One of the biggest benefits has been that Substack forces me to sit down and spend time working through names outside of work, and has been really useful in refining my thought process working through bio names.

Over time I’m getting a better picture of how to start formatting the Substack. The ‘Speculative Series’ catalyst focused pieces that are shorter seem to work better (and I’ve been getting good feedback on them) than what I did in the past which was squeezing these ideas into one paragraph or briefly mentioning them in the monthlies.

Going forward, I’ll probably use those to cover near-term catalyst names in more detail, keep them paywalled for 30 days, and continue publishing my higher-conviction, longer-term ideas freely. I’m also going to experiment with a short daily biotech update through Notes which I have running through an Agent anyways.

TENX Post Mortem

Ultimately a lottery ticket - that fell flat. Most people I’m aware of played it for the upside skew, myself included with a lotto position.

LEVEL topline results were released Aug 10, missing the primary endpoint. TNX-103 only improved 6MWD by 3.5m delta against placebo with no stat sig, whilst KCCQ was essentially the same. Clearly the drug was doing something hemodynamically given there was a 49% reduction in NT-proBNP but no translation into functional benefit.

Now, this part really baffled me, because given the mgmt’s history I would have thought they would have had better enrollment criteria - but LEVEL enrolled a materially less severe and less selected population than HELP ph2. The baseline 6MWD was 320m vs 285m. HELP had also first exposed everyone to IV levosimendan and randomised responders. I had touched on this as a potential problem.

Read more

PepGen was one of the names I mentioned at the start of August one of the other names I’d write a note on. The stock has moved quite a bit since (unfortunately), and I do slightly regret not getting this out sooner, my DMs and Twitter attest to my “appreciation” of PEPG before this actually came out.

Wavestrum who I retweeted has done some very good work on PEPG

The price has moved from trading at almost half the $3.10 RA Capital paid in the last financing at the start of the month; now around $3, after the 6 August safety update removed a meaningful part of the near-term risk. At <$2.00 the valuation seemed very disconnected.

My view around the name is that : the market overreacted to the 5 mg/kg MAD cohort data, the 10 mg/kg dose will hit, and the main safety concern has now been materially derisked.

My base case is a 35–40% mean splice correction, which would rerate the stock towards $5–6 / share.

Catalyst: FREEDOM2 10mg/kg multiple dose data - November 2026

Secondary catalysts: OLE update January, 12.5mg/kg data in H1’27

Positive read PT: $5-6 / share

By reading this you accept you have read and acknowledged the disclaimer.

1. Quick take

First, PGN-EDODM1 is being tested in 6 active patients receiving four doses of 10 mg/kg, with a November readout

Second, prior single ascending dose shows DDR (caveat small n): mean splice correction increased from 12.3% > 29.1% > 53.7% across 5, 10 and 15 mg/kg.

Third, the 5 mg/kg MAD cohort showed bad data and delta vs placebo, but the 7.3% vs 6.8% placebo headline was skewed by a outlier underperformer (and to a lesser extent a positive performing placebo patient), leading to an overreaction and subsequent drop in SP.

Fourth, the main bear case for the 10mg/kg dose was repeated dose safety, but the August safety update has reduced this significantly: a clean 35–40% splice result would move the stock to $5–6, and a fail to $1s.

Read more

I skip a lot of details but try and get my point across with some context.

By reading this you accept you have read and acknowledged the disclaimer.

1. A Foreword

I have held TENX as a small speculative position for some time, and the catalyst is now approaching.:

Catalyst: August 2026 topline

ESC late-breaker: 29 August 2026

The way I view TENX is fairly simple: I think the upcoming LEVEL readout is roughly a coin flip, and at the current valuation that is enough to make the stock attractive as a small speculative position.

If I assume roughly $2/share on failure and something like $45–60/share on a credible success, the current share price implies a probability of success somewhere around the mid-20s. If the true odds are closer to 50%, that is an interesting mismatch.

There is, of course, another perfectly sensible way to play this: wait for the data. If LEVEL works, buy the stock afterwards and remove most of the binary risk, albeit at a much higher price. I recently discussed TENX with a very smart healthcare buy-side analyst, and his preference was effectively this approach. I can understand why. This is an all-or-nothing readout, and I do not think anyone can credibly claim to know whether it will work.

For me, I am comfortable holding roughly a 1–2% position through the readout. I want to stress that this is not a high-conviction clinical call. I simply think the probability/payoff setup is attractive enough to justify taking the risk.

Read more

Into the meat:

CCCC C4 Therapeutics

C4 price action has been a bit poor - falling from around $5 to sitting near $3.50.

The concern came from Johnson & Johnson’s July 23 MonumenTAL-6 results in patients with one to four prior lines of multiple-myeloma treatment.

MonumenTAL directly evaluated two different combination strategies in patients who had received one to four prior lines of therapy:

• Tec-Tal: teclistamab + talquetamab, BCMA bispecific + GPRC5D bispecific

• Tal-Pom: talquetamab + pomalidomide, GPRC5D bispecific + oral IKZF1/3 degrader

Read more

What has changed since my writeup last year:

• Positive randomised Ph2B AD induction data followed by convincing W52 maintenance and deepening data, including Q12W maintenance and rising EASI-100 rates.

  • W16 efficacy is slightly slower / lower than best dupilumab cross-trial benchmarks, but itch is competitive, and responses deepen materially through W24 and W52.

  • No real safety issues except common injection site reactions (which are minor)

• AA moved from PoC readout at W36 to a more relevant W52 profile reaching low dose JAK efficacy band

  • W36 ITT primary missed: corrected/mITT reached nominal significance. W52 SALT <= 20 reached 27.6% at high dose vs 6.7% placebo and continued to deepen.

• Kaposi sarcoma clusters in OX40/OX40L agents, rocatinlimab now discontinued, and 2 cases in Sanofi’s amlitelimab program

• ZENITH AD-1 now registered and recruiting, testing proposed 24 ug/kg Q2W induction regimen in biologic/JAK naive patients 12yo+

• Recent raise (3,532,609 shares of common stock at a public offering price of $92.00 per share) leaves the balance sheet with c$1.1B pro forma cash removing the capital bear case.

Nektar is still remembered as a failed oncology cytokine story. That memory matters because it suppresses willingness to underwrite another Nektar cytokine program. But the current company is economically and clinically different: Rezpeg is an immunoregulatory rather than immunostimulatory IL-2 construct, it is wholly owned, the balance sheet has been recapitalized, and the lead program has produced randomized Phase 2b data in two large dermatology indications

1.1 The market is discounting Nektar still

It’s my view that this not so rosy view of NKTR lies in its past and current ‘issues’. NKTR is well known in the PEGylation and polymer-conjugate drug development space, and in its past had ambitions in oncology.

NKTR at its 2018 peak was worth some $10-11B on enthusiasm for bempegaldesleukin, a PEGylated IL-2 oncology asset partnered with BMS. Bempeg subsequently failed across key late-stage combinations with

Nivolumab, crashing valuation to a range of $150-500m… The collapse destroyed capital and credibility in the company, and has led to an understandable heuristic with some investors to discount the company.

There is a second discount - a legal overhang which will leave some spec funds to wait for resolution before entering. Rezpeg was partnered with Eli Lilly under a 2017 agreement that included $150m upfront to NKTR, with up to $250m in milestones, a 75/25 Lilly/Nektar Phase 2 cost share, and double-digit royalties that could increase with Nektar Phase 3 participation. Lilly later terminated the agreement and returned rights to Nektar, with Nektar suing Lilly, alleging that Lilly and vendors mishandled or misanalysed earlier Rezpeg clinical data in AD and psoriasis making the asset look weaker than it actually was. NKTR also argues that Lilly’s acquisition of Dermira bringing in lebrikizumab for AD created a conflict of interest around Rezpeg development.

The corrected Ph1b AD analysis was later published, and the efficacy direction was replicated in the much larger REZOLVE-AD study. Sep 8 2026 jury trial date is something I see as a source of non dilutive upside for at least the milestone promised as a minimum. (Note: I’m not a legal expert).

My base case is upped to a PT of $145 for NKTR (with zero value assigned to the litigation). The current price is c$68 and is trading at an EV of c$1.3B.

By reading this you accept you have read and acknowledged the disclaimer.

2. Rezpeg (One?) Shot - Treg Reset

2.1 Strengthening the immune brakes

Rezpeg (previously NKTR-368 / LY3471851) is a subcutaneously administered, long-acting, PEGylated form of recombinant human IL-2 designed to preferentially stimulate regulatory T cells (Tregs).

Tregs act as the body’s immune brakes. Effector T cells and natural killer cells drive attack when the body detects a real threat: infection, cancer, or abnormal cells. Tregs do the opposite: they restrain excessive immune activity, support self-tolerance, and help prevent that attack from spilling over into healthy tissue. That is the core problem in autoimmunity: the immune system mistakes “self” for foreign, or fails to shut itself down properly.

Many chronic immune-mediated diseases are a result of this broader ‘failure’ of immune regulation & balance - alopecia areata (immune cells attack hair follicles), atopic dermatitis (chronic inflammatory signalling).

2.2 Alternatives target downstream pathways

Most immunology drugs work to manage this imbalance by blocking one of the inflammatory signals after the immune response is already switched on:

• Dupixent blocks IL-4/IL-13 signalling (key cytokines involved in Type 2 inflammation), a major immune pathway driving atopic dermatitis.

• IL-13 antibodies, like lebrikizumab and tralokinumab, target IL-13 specifically within the same Type 2 inflammatory axis.

• IL-31 receptor antagonists target itch biology more directly via IL-31Ra reducing neuronal activation and inflammatory cascade driving itch.

• JAK inhibitors go broader by interfering with signalling from multiple cytokines inside the cell, lending to their strong efficacy, but also why the class carries more safety baggage.

2.3 Rezpeg strengthens the regulatory compartment, and doesn’t block just one cytokine by ‘attenuating’ the IL-2 signal

Rather than blocking one inflammatory message after it has been sent, Rezpeg is designed to strengthen the regulatory compartment that regulates the immune response. Rezpeg produces an attenuated IL-2 signal impacting CD25-high Tregs impacting Treg expansion (+function).

By expanding Tregs, Rezpeg will (ideally) restore immune balance upstream and regulate several inflammatory programmes at once, including Th1-, Th2- and Th17-driven disease, instead of only blocking one cytokine axis.

2.4 IL-2 can be selective at low exposure

Rezpeg biology works through IL-2, which is a novel MoA in both AD and AA. IL-2 is a key cytokine for T-cell biology. At high enough exposures, IL-2 can stimulate effector T cells and NK cells (hence historic usages in oncology for immune activation). But Tregs are unusually dependent on IL-2 for survival, expansion and suppressive functions. They express high levels of CD25: the alpha chain of the high-affinity IL-2 receptor, which lets them respond to lower or more attenuated IL-2 signalling than many other conventional immune cells.

Now, native IL-2 itself isn’t an ideal chronic autoimmune drug. It has a short half-life, narrow therapeutic window, poor practical dosing characteristics, and as mentioned at high exposures can stimulate the wrong immune cell population.

Rezpeg is an attempt to preserve the aforementioned low-dose IL-2 / Treg biology without broadly activating the immune system. PEGylation (NKTR’s bread and butter) is how they ‘solve this’, by attaching polyethylene glycol chains to recombinant human IL-2, it extends half life, reduces rapid clearance and alters receptor engagement; all in order to sufficiently attenuate the molecule so that CD25-high Tregs receive a sustained signal, whilst conventional effector T cells receive much less stimulation (in a long lasting fashion).

2.5 Demands for an upstream mechanism

Three useful features are demanded for an upstream regulatory therapy: breadth across Th1-, Th2- and Th17-influenced diseases; efficacy that deepens as the regulatory compartment changes; and residual control after dosing stops. The AD and AA datasets provide early support for the first two. The third - remittive or disease-modifying activity - remains to be seen, but we should get more colour with off-treatment datasets in AA and AD released in Q4 2026 and Q1 2027 (expected).

3. AD is Base case: novel Treg and MoA biologic

3.1 What AD is

Atopic dermatitis (AD / eczema) is a chronic, relapsing inflammatory skin disease - best understood as a self reinforcing loop between a broken skin barrier, immune activation, itch and flares.

First, regarding the barrier: in healthy skin the outer epidermis keeps water in and irritants, allergens and microbes out. In AD, the barrier becomes ‘leaky’, driven by genetic factors such as filaggrin loss of function, altered skin lipids, impaired tight junctions, or repeated inflammatory damage. Once the barrier weakens, the skin loses water more easily, and outside triggers penetrate much more easily, resulting in the dry, reactive and prone to flaring characterisation of AD skin.

Second, the immune response. Keratinocytes (predominant cells in epidermis) when damaged release ‘alarm’ signals: e.g. TSLP, IL-25 and IL-35 which push the skin into a Type 2 inflammatory state. Type 2 inflammation is the allergic / atopic arm of immunity, and in AD the key cytokines (chemical messengers) are IL-4 and Il-13. These cytokines drive inflammation, impair barrier repair, reduce antimicrobial defences and perpetuate the itch<>scratch cycle.

AD becomes circular - a weak barrier > inflammation > weakens the barrier further. Itching > scratching > skin break > restarts the cycle. Scratching also increases risk of infection and microbial dysbiosis (esp. Staph aureus colonisation) amplifying inflammation.

Itch is also part of the disease (not just a symptom): IL-31, an itch associated cytokine and neuroimmune signalling between immune cells, keratinocytes, and sensory nerves result in persistent itches despite improvements to inflammation.

AD also evolves over time: acute AD is heavily Type-2 driven, but chronic AD becomes much broader: Th1, Th17, and Th22 pathways can contribute, and tissue resident memory T cells can help lesions recur in the same areas. This is to highlight that despite IL-4 and Il-13 being key AD pathways, they aren’t the only ones. Conceptually a Treg based IL-2 drug (Rezpeg) will work to improve the regulatory side of the immune system rather than one inflammatory output.

3.2 How AD trials are judged

Practical hierarchy is simple: does the drug hit IGA and EASI-75; does it deliver rapid and clinically relevant itch relief; can patients stay on it safely for years; and does response persist beyond induction? Cross-trial comparisons I will use later on, but I want to put in the caveat that they will likely have differences baseline severity, concomitant topicals, endpoint definitions and missing data handling.

For readers less familiar with AD trial endpoints, I’ve put a better explainer here: https://open.substack.com/pub/hanseoulohno/p/how-to-read-atopic-dermatitis-easi?r=3jon20&utm_campaign=post-expanded-share&utm_medium=web

3.3 AD Current Drug Environment (and Limitations)

The AD treatment pathway starts with topicals: moisturisers, topical corticosteroids, topical calcineurin inhibitors, PDE4 inhibitors and topical JAK inhibitors. For mild or localised disease, these can be enough. Nearly all patients will try topical steroids, and some 50% of moderate and 80% of severe patients fail long-term topical maintenance (TDCowen KOLs). Once disease is too widespread, persistent or difficult to control, patients begin moving towards phototherapy, older immunosuppressants or, increasingly, targeted biologics and oral JAK inhibitors.

Within systemic treatment IL-4/IL-13 blockade is the SoC. Dupixent (dupilimab), the most widely prescribed biologic for AD, blocks IL-4Ra and inhibits signalling from both IL-4 and IL-13, the central Type 2 inflammatory axis.

Dupixent remains the SoC due to its combination of meaningful efficacy, chronic use safety profile and prescriber experience. Dupixent alone in 2025 generated over 15B EUR in total sales, up >25% YoY, and Regeneron said the drug had more than 1.4m active patients worldwide in Jan’26 (spanning all indications).

*Despite* those numbers, AD is still unpenetrated, In the US adult population c17m have AD, including c7m with moderate-to-severe disease (+ 10m children with AD). Of this, estimates are that c1.5m are eligible for biologic treatment, but with an estimated only 225k US AD patients on Dupixent in 2025, penetration is implied of around 15%. Sanofi in April 2026 noted that advanced therapy penetration in AD was only just about 18%.

Dupixent also leaves clinical room. Responses are not universal or always deep. In the SOLO 1/2 monotherapy trials, only 51% and 44% of patients receiving Q2W Dupixent achieved EASI-75 at week 16. Among initial responders who continued weekly or biweekly treatment in SOLO-CONTINUE, 71.6% maintained EASI-75 through week 52, leaving c28% that lost this response one year on. AD is an incredibly flare-prone and chronic disease, rather than a one time induction problem…

There are also tolerability gaps. Dupixent’s most recognisable AD-specific issue is ocular toxicity, particularly conjunctivitis. This clearly has not prevented Dupixent becoming dominant, but it creates another switch population alongside patients with inadequate efficacy or secondary loss of response. Cowen’s physician work suggests around 20–25% of Dupixent-treated patients either do not achieve sufficient disease control or tolerate treatment poorly. Applied to its 225k US AD estimate, that implies an illustrative existing switch pool of roughly 45–56k patients (w/o discussion of further mkt expansion).

Lebrikizumab and tralokinumab offer alternative Type 2 biologics by selectively targeting IL-13. They can compete through dosing, access or tolerability, and may be useful after secondary loss of response to Dupixent. But for a true primary non-responder, switching within the same IL-4/IL-13 axis isn’t as attractive as changing to an entire new MoA. Ebglyss (lebrikizumab), generated c$400m in 2025.

At the more potent end of the treatment ladder are the oral JAK inhibitors, principally upadacitinib (Rinvoq) and abrocitinib (Cibinqo). These inhibit intracellular signalling downstream of multiple cytokine receptors, resulting in fast itch relief and deeper responses than biologics. But with a caveat! They’re black boxed for serious infections, malignancy, adverse cardiovascular events, thrombosis and mortality (and more), alongside monitoring requirements. And as such, dermatologists often reserve them for patients who fail, lose response to or cannot tolerate biologic treatment.

And if I was a patient seeing that label I would want that as a last resort:

Rinvoq generated $8.3bn globally in 2025 (across all indications), but that figure spans several large inflammatory indications. Cibinqo is a cleaner AD-specific benchmark and generated $284m. Their AD-specific commercial footprint remains far below Dupixent despite their strong efficacy (due to the label).

Nemolizumab (Nemluvio) is the last option on the list, occupying a different niche: blocking IL-31RA and targeting a neuroimmune pathway tied to itch. Here the positioning is to treat itch dominant disease rather than as a broad replacement for Type 2 control profile. Nemluvio generated around $452m in 2025 across AD and prurigo nodulari. To also reiterate, the efficacy profile of nemluvio is not ‘amazing’ when put into context vs placebo.

Approved (and closest) systemic agents relevant to NKTR benchmarking:

3.4 The upstream competitor class has weakened (especially OX40/OX40L)

The most anticipated ‘new’ MoA beyond conventional cytokine blockade are (were?) OX40 / OX40L agents. The thesis was similar in shape to Rezpeg: act earlier in Tcell activation, reduce inflammatory memory, deepen response over time and potentially use Q12W dosing. The class has since developed a serious safety cloud.

OX40 is a co-stimulatory receptor expressed on activated T cells and binding by OX40L on antigen-presenting cells supports T-cell expansion, survival and inflammatory memory. Rocatinlimab targeted OX40 itself and was designed to reduce OX40-positive pathogenic T cells, whilst Sanofi’s amlitelimab blocks OX40L without directly depleting T cells.

On 3 March 2026, Kyowa Kirin discontinued every rocatinlimab trial after identifying one new confirmed and one suspected Kaposi case, in addition to an earlier confirmed case, and said a mechanistic connection to OX40 modulation could not be excluded. On 28 March 2026, Sanofi disclosed two cutaneous Kaposi cases among 3,778 confirmed amlitelimab-exposed patients; both reportedly had known risk factors. Sanofi is continuing development, and no regulator has adjudicated a class effect.

To put it into context: Kaposi sarcoma is rare with rates in the US around 0.3 per 100,000, making recurrence difficult to dismiss, especially when put into the context that inherited OX40 deficiency has been linked to impaired immune control of HHV-8, the virus driving Kaposi sarcoma. I find it difficult to believe that it will be commercially viable when safer alternatives will come forward (like Rezpeg).

Sanofi had previously assigned amlitelimab more than €5bn in multi-indication peak-sales potential in December 2023, and with a ‘write off’ of the class, it leaves a lot of the ‘novel’ upstream immunology white space available to Rezpeg.

3.5 Pipeline beyond OX40/OX40L

I’m not entirely concerned about the upcoming competitors. I’ve colour coded the table, light blue competitors are following a similar MoA to the traditional Il-4/IL-13 MoA. The market is large enough for more players like longer acting IL-13, or oral STAT6 degraders (which act on the same axis), and doesn’t play into my own mind as a serious challenge to the 2L ask that Rezpeg will take.

The wave of new MoA still lack randomised AD efficacy, and are as yet unproven in safety. Especially examples e.g. ITK inhibitors (Corvus) which carry safety concerns especially in their use in cancer therapies.

3.6 Rezpeg will own 2L, and potentially have a shot at 1L (assuming Ph3 is maintained)

The market is large enough, and with OX40/OX40L being out, I feel confidence that Rezpeg will secure a good share and physician confidence in 2L until new options arrive.

There is good appeal. Rezpeg offers:

• Alternative MoA in scenario of IL-4/Il-13 non response (JAKs have safety baggage)

• Strong safety, thus far safety is largely limited to injection site reactions (minor), and a switch opportunity for secondary failures, and conjunctivitis issues

• Low frequency maintenance potential: Q12W, the other competitor here being Zumilokibart Q3M/Q6M (still on the IL-4/IL-13 axis)

4. Rezpeg in AD

4.1 Timelines

• Oct 2023: REZOLVE-AD Initiated

• 24 Jun 2025: W16 Topline

• 8 Nov 2025: + asthma crossover W24

• 10 Feb 2026: W52 maintenance

• 28 Mar 2026: analysis AAD subgroup

• 8 Jul: 2026: ZENITH AD-1 registered

• Q1 2027: Ph2b off Tx W52 data

• May 2028: AD-1 primary completion estimate

4.2 What REOLVE-AD Phase 2b established

Is Rezpeg active?

• Yes. All three active regimens beat placebo on mean EASI at Week 16; high and middle doses also separated on key categorical endpoints

Is pharmacology on target?

• Yes. High-dose treatment expanded circulating Tregs up to approximately six-fold and reduced multiple Type 2 biomarkers.

Does response deepen?

• Yes. Placebo-crossover patients improved further between 16 and 24 weeks; new and complete responses accumulated through Week 52. Note: the maintenance population was responder enriched.

Can dosing be reduced?

• Potentially: Q4W and Q12W maintenance preserved EASI-75/90 in many W16 responders. Once a quarter maintenance is massive!

Is it safe?

• Systemic safety clean, and despite ISRs being common, mostly erythema, were usually mild/moderate and rarely caused discontinuation. Underpowered for rare events though.

4.3 REZOLVE-AD Design

REZOLVE-AD enrolled 393 biologic and JAK naive adults with moderate to severe AD. Patients were randomized 3:3:3:2 to 24 ug/kg Q2W, 18 ug/kg Q2W, 24 ug/kg Q4W or placebo for 16 weeks. The primary endpoint was mean percent change in EASI; secondary endpoints included EASI-50/75/90, vIGA-AD 0/1 and itch NRS.

At W16 patients achieving at least EASI-50 were re-randomised within their induction dose to Q4W or Q12W maintenance through Week 52. Patients below EASI-50 entered an active high-dose escape arm; placebo EASI-50 responders could remain on placebo maintenance.

4.4 W16 induction efficacy

All three doses met the primary endpoint with p<0.001. The proposed high-dose regimen produced a 61% mean EASI improvement versus 31% placebo, EASI-75 in 42% versus 17%, EASI-90 in 25% versus 9%, vIGA-AD 0/1 in 20% versus 8%, and itch NRS-4 in 42% versus 16%.

Cross-trial, the skin-clearance endpoints are somewhat below dupilumab at Week 16, while itch is broadly competitive.

4.5 W16 safety and tolerability

The safety was impressive despite not looking like it when you see the % of patients with ISR. 69.7% of patients presented with local ISR but 99.6% of them were mild/moderate, and less than <1% discontinued due to ISRs. ISRs in the vast majority of cases noted by management are not recurrent and are composed of just reddening of the skin (hence the low discontinuation rates).

When pooled, only 1.6% and 3.1% of patients treated with Rezpeg experienced serious and severe TEAEs, respectively. Excluding injection-site reactions (ISRs), overall TEAE rates were highly comparable with placebo. The most common non-ISR TEAEs occurring in more than 5% of Rezpeg-treated patients were eosinophilia (7.8% vs 2.7% with placebo), pyrexia (6.3% vs 2.7%), headache (6.3% vs 4.1%), and arthralgia (5.0% vs 1.4%).

Two treatment-related serious adverse events were reported (severe drug hypersensitivity and moderate tonsillitis), both of which were resolved. Severe treatment-related non-serious events were limited to pyrexia and several local reaction/chest pain combinations, all of which also resolved. Notably, there was no apparent increase in conjunctivitis, oral ulcers, infections (including oral herpes), or major cardiovascular events relative to placebo.

4.6 Crossover replication, W24 depth and pharmacodynamics - placebo crossover patients receiving 24 ug/kg Q2W

Forty-two patients initially assigned to placebo crossed to the high-dose escape regimen. After 16 weeks of active treatment, mean EASI improvement was 60% - almost identical to the 61% blinded high-dose result - with EASI-75 of 50% and vIGA 0/1 of 28%. Among 39 evaluable patients at 24 weeks, mean EASI improvement deepened to 75%, EASI-75 to 62% and vIGA 0/1 to 38%.

4.7 W52 maintenance showed no erosion and a deepening of response

Patients achieving at least EASI-50 at Week 16 were re-randomized to Q4W or Q12W maintenance at the same dose level. In the pooled 24 ug summary, 71% on Q4W and 83% on Q12W maintained EASI-75 at Week 52; EASI-90 maintenance was 80% and 78%; vIGA maintenance was 85% and 63%; itch maintenance was 75% and 77%. There were no new safety concerns through Week 52.

And when compared against peers like dupixent, responses are on par if not better:

And responses are new and deepen:

4.8 Phase 3: ZENITH AD-1

Nektar has announced they’re planning a two study global programme and a BLA objective in 2029. AD-1 is the biologic/JAK-naive study. The second study will focus on the representation of IL-4/IL-13-experienced patients.

For now AD-1 is planned to start on 31 July 2026 with primary completion on 31 May 2028 and study completion in Dec 2028 estimated.

What I’m looking for:

The data we have thus far with the crossover gives me confidence that 6-8 more weeks can lift EASI and IGA outcomes. I’m confident in the success of the Ph3. For modelling I use 65% PoS to be more conservative.

5. Peak sales AD case

5.1 Base Case Model

Again, Rezpeg in my scenario doesn’t need to be BiC to be successful, but rather offer a ‘good enough’ solution to patients for whom IL-4/IL-13 blockade doesn’t work, and I model it accordingly. The data we have leads me to be very confident that Rezpeg will take 2L with some potential usage in 1L - especially with OX40/OX40L being a ‘dead’ class. Add to that once a quarter maintenance, and it’s extremely compelling.

I have lower assumptions, with lower market share (and no inflation modelling) - to remain more conservative.

The framework:

Population: 6.6m US adults w/ moderate-severe criteria + 0.86m adolescents. One quarter assumed to be persistent systemic candidates = 1.89m patients (Sanofi estimates 2.2m for 6yo+)

Penetration and share: Mature advanced-therapy penetration rises to 30%, from a market still below one-fifth penetration today. Rezpeg captures 12% of treated patients: 8 points from a mechanism-switch segment and 4 points from first-line/other switching. Lower base accounting for: APG777/zumilokibart, oral STAT6 degradation and other new mechanisms taking pieces of the 2L pool.

Price. Dupixent’s January 2026 WAC is $4,193.03 per monthly carton, or approximately $50k annual list price. I assume $38k annual net revenue for Rezpeg - a 24-30% discount to list before accounting for the eventual mix of Q4W and Q12W maintenance.

On a side; there are estimates for peak sales for Rezpeg from SS ranging from $4B US to $5.5B worldwide - I’m deliberately below these ranges to show just how cheap Rezpeg is on a valuation perspective.

6. AA Provides Additional Upside Potential

Some readers will remember how apprehensive I was about Nektar’s chances in alopecia areata (AA), it’s difficult to treat, and hair growth is slow etc. I had written it as an optionality, but with the W36 release and W52 extension data I feel that the setup now allows for AA to provide explicit value rather than be treated as a free call option.

At its core, it’s a different proposition to AD, and Rezpeg doesn’t look more efficacious than the strongest JAK regimens, in fact it looks comparable to low dose Olumiant (a JAK), but you should remember that JAKs come with boxed warnings, and here Rezpeg would be the first biologic offering in AA that is safe, with no need for daily oral adherence!

7. AA Introduction and the Space

7.1 What is alopecia areata

Alopecia areata is an autoimmune, non-scarring hair-loss disorder. In healthy anagen hair follicles, local immune privilege limits antigen presentation and inflammatory infiltration. In AA, that privilege collapses: interferon-gamma and IL-15 signalling recruit and activate cytotoxic T cells and NK cells around the follicle, leading to hair shedding and a self-reinforcing inflammatory loop. Treg number or suppressive function may also be impaired, which provides a biological rationale for Rezpeg.

The collapse of immune privilege in the anagen hair follicle during alopecia areata.

Disease ranges from discrete patches to alopecia totalis and universalis. Nearly 7m people in the US will experience AA over a lifetime, and about 700,000 currently live with some form, with severe disease with more than 50% scalp loss estimated in up to roughly 300,000 US patients.

MedComm, Ma et al (2020)

7.2 How would Rezpeg work in AA?

Rezpeg does not directly block IFN-gamma or IL-15. It attempts to restore the regulatory side of the disease. AA tissue studies report reduced or dysfunctional suppressive Treg populations, including marked reductions in FOXP3+CD39+ Tregs around lesional follicles. Separate preclinical work shows that skin Tregs protect the hair-follicle stem-cell niche through high-affinity IL-2 receptor biology.

1. Rezpeg provides a sustained CD25-biased IL-2 signal and expands functional Tregs

2. Tregs suppress antigen-presenting cells, pathogenic CD8 and Th1 activity through CTLA-4, inhibitory cytokines, metabolic control and direct effector restraint.

3. The IFN-gamma/IL-15 amplification loop weakens and follicular immune privilege can recover.

4. Preserved follicles re-enter anagen and visible hair growth follows with a biological lag.

7.2 How AA trials are judged

Hair biology is slow, unfortunately some of you may know after a botched hair cut how long it takes even for your hair to reset. In AA, a drug can show early change in mean SALT while still needing 36-52 weeks to produce clinically meaningful SALT <= 20 response.

7.3 Current systemic therapy is a JAK market

Olumiant, Litfulo and Leqselvi are the dominant therapies in AA, but each is a JAK-pathway inhibitor, again with boxed warnings (as mentioned in AD the label issue) around serious infection, malignancy, cardiovascular events, thrombosis and mortality. These labels also require laboratory and infection monitoring. Olumiant starts at 2 mg daily and can be increased to 4 mg if response is inadequate.

8. Rezpeg in AA

In all honesty, I put much less effort into this than in AD. Ultimately, Rezpeg’s bar here is to be approved, because if there is a ‘safe’ alternative to JAKs, that biologic is going to be prescribed first. Really you just need something ‘good enough’ on hair growth, and easier to live with on safety, monitoring and maintenance versus a JAK which you will need to be on for potentially decades for a non-lethal disease.

Hence this will be a much shorter section than AD, and much more open to scrutiny I imagine.

8.1 Timelines AA

• FDA End-of-Phase 2 feedback : H2 2026

• REZOLVE-AA 24-week off-treatment data: Q4 2026

• Phase 3 initiation: Estimate 1H 2027

• Phase 3 topline: Estimate 2H 2029

8.2 REZOLVE-AA

REZOLVE-AA enrolled 92 JAK- and biologic-naive patients with severe-to-very-severe AA. Patients were randomized 3:3:2 to 24 µg/kg Q2W (n=35), 18 µg/kg Q2W (n=37) or placebo (n=20) for 36 weeks. Mean baseline SALT was 78.5 in the active arms and 76.6 in placebo; median time from onset was approximately seven years. Enrollment was geographically concentrated: 64% Poland, 23% Canada and 13% United States. Patients with SALT >=95 were capped at 25%, making the cohort less severe than several pivotal JAK programmes.

Nektar Therapeutics - REZOLVE-AA Trial Design

8.3 Study violation controversy?

Dec 2025 induction data from REZOLVE-AA was presented but! the study missed statistical significance on Week 36 miTT. Again - I am biased against Nektar here considering I went in thinking that the trial would miss.

Four participants had major baseline eligibility violations. One placebo patient with unstable disease accounted for the 5.5 percentage-point difference between the two placebo estimates. The active-arm mean effect barely changed, while placebo performance fell by half (which would mean the placebo patient achieved a near 100% SALT reduction).

Now there are some more controversies including imputed results from missing data resulting in rescue. A smart person discussing this issue is hannibalspeaks on X. I’m not going to linger on the issue again because my base case is just that Nektar needs to get approved, not whether the Ph2 has ‘dodgy’ results.

Ultimately the results I’ve seen from the Ph2 is that Rezpeg works, but it takes time!

8.4 Interpreting the results

Rezpeg appears slow acting. Separation from placebo emerged gradually, with the majority of patients beginning to show hair growth at or after Week 16. More importantly, the deeper and more clinically relevant response thresholds continued to improve after the original Week 36 endpoint:

From the chart we still see no clear plateau! Right now it’s matching low dose olumiant (which plateaued).

n a cross-trial basis, high-dose Rezpeg’s 27.6% SALT ≤20 response at Week 52 sits in the same numerical range as Olumiant 2mg, which produced approximately 21.2% and 24.4% SALT ≤20 responses across BRAVE-AA1 and BRAVE-AA2 at Week 52.

And: the slower response profile is consistent with the proposed mechanism if Treg expansion, to allow time for cells to traffic into diseased tissue, restrain the follicular cytotoxic-T-cell response, reduce the IFN-γ/IL-15 inflammatory loop, and finally allow preserved follicles to re-enter anagen and produce visible hair.

8.5 Week 52

Rezpeg continued to show deepened response from Week 36 to Week 52.

Thirty-one partial responders entered the 16-week extension: 13 high dose, 14 low dose and four placebo. From Week 36 to 52, 31% and 29% of the active extension cohorts newly achieved SALT <=20 versus none on placebo.

Ninety-four percent completed the extension, and the share with at least 75% SALT improvement increased from 4% at Week 36 to 26% at Week 52. The overall-population Week 52 analysis used the mITT-adapted population and carried forward the primary estimand rules from the Week 36 database lock. It therefore remains linked to the exclusion of the four major eligibility violators and is not a new prespecified randomised primary analysis. Cross-trial, 27.6% SALT <=20 is in the range of Olumiant 2 mg at Week 52 (approximately 21.2-24.4% across BRAVE-AA1/2) but below higher-dose JAK regimens.

8.6 Safety

We already have a good understanding of safety from the larger n cohorts in the AD trials run in Rezpeg, and again many of the issues here lie in ISRs which the vast majority are self resolving mild-moderate erythemas. As long as Rezpeg can maintain safety ahead of black box JAKs this is more of a formality.

8.7 AA Phase 3 - what I’m looking for

As yet, NKTR has not released many details on the final AA Ph3 protocol. But I’m fairly optimistic on the Ph3 outcomes. We saw genuine responses from the drug showing it worked (if you don’t believe me have a look at the corporate presentation for real images).

20 active-treated patients, approximately 28% of the two active arms, discontinued by patient decision during the original treatment period; 15 of those 20 had not reached SALT below 30. By contrast, AE-driven discontinuation was 1.4%. The inference here that slow visible benefit and expectation-setting, rather than acute toxicity, drove much of the attrition. With the Ph2 dataset, patients enrolled will be much less likely to discontinue if they know that response may come back ended, and imputation here shouldn’t lower the Week 52 response rate in the Ph3.

To keep expectations centered, I use a 45% PoS, to reflect p value issues in the Ph2 and just preserve a margin of safety.

Here, a two-arm study with 150 patients per arm has approximately 97% power for 24% versus 8%, around 89% for 22% versus 9%, and only about 69% for 20% versus 10%, before multiplicity and estimand details. The primary risk is effect compression and retention, not mathematical power if the full Phase 2 gap is real - which I believe exists.

9. Rezpeg Peak Sales

Explaining the assumptions:

Population: Approximately 700,000 Americans have some form of AA, moderate-severe prevalence near 0.09% roughly 300,000 people. Assume 60% chronic, diagnosed, in specialist care and willing to use a systemic agent.

Penetration: Assume 40% mature advanced therapy penetration and 35% Rezpeg share, assuming an expansion of the market beyond current JAK use.

Price: $42k assumption.

10. Lilly Litigation: Nearest Catalyst

Unfortunately I’m not a lawyer or a legal expert so I’ll do my best here and talk about it as some potential upside. I’m not modelling anything from this in.

Nektar filed its complaint against Lilly in the Northern District of California on 7 August 2023, alleging breach of contract and breach of the implied covenant of good faith and fair dealing in connection with the Rezpeg collaboration. Lilly filed counterclaims for alleged confidentiality breaches and defamation, then voluntarily dismissed those counterclaims with prejudice in 2025. Lilly’s summary-judgment motion and other pre-trial motions remained pending in the 2025 10-K. The jury trial is currently scheduled for 8 September 2026.

Nektar’s narrative is that Lilly and third-party contractors mishandled or misanalysed the early clinical data and that the Dermira acquisition created a competitive conflict. Lilly disputes the case: their position being that the calculation error originated with a contract research organisation.

Now, as a non legal expert - Lilly has admitted that there was an error in the data calculation, there’s been a year+ delay in potential revenue from this, and the loss of the milestone payments. I’d expect at least a minimum of the $250m milestone to be settled, and much more considering the development years lost (and the Ph2 datasets showing Rezpeg’s efficacy).

11. Adding it all together and Tying it Up

Considering the adjusted EV of Nektar is around $1.5B (FD market cap is at $2.5B) even with my grossly conservative modelling and PoS I’m getting 112% upside. No comments left, I rest my case.

Rezpeg does not need to own every second-line patient or beat the strongest JAK on raw efficacy; it needs to become a safe credible mechanism-change biologic in AD and the first commercially useful biologic in AA.

The existing data that we have paints NKTR to be reasonably cheap, and this is whilst I assign no value to the Lilly litigation, T1D, asthma or the rest of the pipeline.

AD is a mix of visible inflammatory skill signs, body-area involvement, itch, sleep disruption, QoL burden and chronic relapse risk. A good AD drug thus needs to improve both what the physician can see and what the patient actually feels.

AD trials are trying to answer four relevant questions:

• How bad does the skin look?

• How much of the body is affected?

• How much does the patient itch?

• How much does the disease affect daily life?

The endpoints are designed around these questions. The main physician assessed endpoints are EASI and IGA / vIGA-AD. The most important patient-relevant symptom endpoint is usually Peak Pruritus Numeric Rating Scale (PP-NRS) / itch NRS.

See the main endpoints in the graphic below:

In moderate-to-severe AD trials, patients are typically only eligible for study enrollment when they possess enough disease burden at baseline. E.g. Dupixent’s FDA review entry criteria included for adults: BSA > 10%, IGA>=3, EASI>=16, and itch NRS>3. Patients need to present with visible eczema over a meaningful body area, moderate / severe physician rated disease, and meaningful itch: these drugs are not designed for mild eczema, and are testing whether the drug can move inflammed & symptomatic patients towards controlled, clear or near-clear disease.

In actual trials, FDA precedent is still centered on w/16 skin endpoints. In practice, the two most important registration grade benchmarks are EASI-75 and IGA 0/1. They’re also the cleanest cross-trial shorthand for us market participants. So it’s well worth understanding these two a bit more:

EASI (Eczema Area and Severity Index) measures how severe the eczema looks and how much body area is affected. Doctors score signs such as redness, thickness, scratching, and lichenification across body areas. “How much does visible eczema improve?”

EASI-50 = 50% improvement from baseline, EASI-75 = 75% improvement from baseline, EASI-90 = 90% improvement from baseline, EASI-100 = Complete clearance by EASI.

IGA (Investigator Global Assessment) ; think of this like the ‘clear or almost clear’ endpoint. This is a ‘simpler’ physician global score, usually running from 0 to 4: clear, almost clear, mild, moderate, severe. In trials, the key readout is usually the percentage of patients who reach IGA/vIGA-AD 0 or 1, often with at least a 2-grade improvement from baseline.

“Does the doctor think the patient’s skin is now clear or almost clear”.

The third major endpoint is itch. This is usually captured by Peak Pruritus NRS or itch NRS, a patient-reported 0–10 scale. A 4pt improvement is commonly used as a clinically meaningful response threshold in moderate-to-severe. Itch is often what patients care about most - it drives scratching, sleep loss, distress and treatment switching. It’s also a reason why JAK inhibitors remain very relevant despite the safety baggage in that patients can derive fast and meaningful symptom relief.

Supporting endpoints: BSA tells you how much body surface area is affected. DLQI captures dermatology-related quality of life. POEM is patient-reported eczema burden. ADCT captures patient-perceived disease control.

For FDA-relevant purposes, the key week-16 skin endpoints have historically been IGA/vIGA-AD 0/1 and EASI-75. Itch NRS, safety and longer-term durability then inform on whether the product profile is commercially attractive.

How does this translate into what the FDA-relevant endpoints look at:

• IGA / vIGA-AD 0/1 = Shows clear or almost clear skin

• EASI-75 = Major improvement in eczema severity

• Itch NRS Improvement = Shows patient symptom benefit

• Safety = Benefit is worth the risk

• Durability / maintenance = Shows whether response persists after induction

In order, what I check is whether EASI-75 was hit (or IGA), itch, then safety, and whether this is a durable benefit.

Now looking into this a bit more you can expand these endpoints if you are trying to evaluate as a market participant:

• Skin efficacy = Clear placebo-adjusted benefit on EASI-75 and IGA

• Depth of response = Good EASI-90 ! Not just EASI-75

• Patient benefit = Meaningful itch improvement, ideally early

• Usability = Clean safety, convenient dosing, durable maintenance

  ---

Several things to consider: First, AD trials can differ meaningfully in baseline severity, background topical steroid use, rescue-medication rules, geography, and placebo response. These differences can make one drug appear better than another in cross-trial comparisons, even when the comparison is not clean.

Second, IGA is an imperfect endpoint. It is relevant for regulators because it captures clear or almost-clear skin, but it is also coarse. A patient can improve substantially and still fail IGA 0/1 if they do not get close enough to clearance.

Third, itch cannot be ignored. Itch is often the symptom patients care about most, and it is frequently the reason they seek treatment or switch therapies. This is commercially important, and it helps explain why JAK inhibitors remain used despite boxed warnings: they can provide fast, meaningful symptom relief.

AD is a chronic disease, so the durability of response matters. We should also assess whether the response is maintained, whether dosing can be stretched, whether patients still need background topical steroids, and whether safety remains clean with longer exposure.

Abivax (Best in Disease)

At the start of June ABVX dropped significantly on what I and many had viewed as an unsubstantiated cancer signal; in part due to poor explanation. Comments from management indicated they would drop part 2 data in the coming weeks to alleviate these fears, which provided a good catalyst, considering that I believed ABVX was ultimately a safe drug. The efficacy that we have seen from obe, and totality of patient hours on the drug, I see a derisked and best in class asset.

The efficacy from ABVX ABTECT maintenance is best in disease!

Obe delivered 48% endoscopic remission (50mg) vs 10% placebo. Rinvoq, the strongest approved JAK benchmark in UC shows 24-26%, Xeljanz 15-17%. Cross-trial caveats apply but this is an insane(!) delta difference. This is before mentioning class warnings for JAKs.

And on the safety concerns. Management walked through the slides incredibly well, and the reaction from the market has been positive (up some 40%). In short - the cancer risk is in my view, non-existent.

Malignancies excluding NMSC remained within the cited UC background range (normal). Case level details also make it seem very unlikely to be related to obe. MPN patient had pre-existing thrombocythaemia and platelets of 909,000/µL at screening. And the prostate cancer case had only 3.4 months of obe exposure. NMSC cases 3/4 hada thiopurine / azathioprine exposure (known risk factor), with 2/4 having prior skin cancer history.

ABVX announced a raise for $600m, which is reasonable, considering the cost to flesh out a team for commercialisation. My own base case is that the company is holding a best in disease asset, it’s cheap, and it’s a likely BO candidate for JPM 2027.

BO for the asset should be >$200 (and this is before accounting for Crohn’s as an opportunity).

The other picks

Read more

The stock is now around $72.50 after the ABTECT maintenance readout (which hit at one point >$150 on headline), with a view of “great efficacy, but we now have a potential cancer overhang.”

The way I see it is: this was a narrative-control failure, not obvious evidence that obefazimod is a cancer-causing drug.

By reading this you accept you have read and acknowledged the disclaimer.

The headline efficacy was genuinely exceptional. In ABTECT maintenance, clinical remission at Week 44 was 50.8% on 25 mg and 51.3% on 50 mg versus 10.4% on placebo; endoscopic remission was 41.5% and 47.7% versus 9.9% on placebo.

If you only looked at efficacy, this was exactly the data people wanted. This is why I think management expected the stock to go up, maybe materially. They released quickly, messaged “strong efficacy, favorable safety, no new safety signals,” and probably thought the drug had just been significantly de-risked. Bu, yeah. No. The safety table was not framed properly.

A few points and colour we have now:

The issue is that ABVX put prostate cancer, breast cancer, and colonic dysplasia under “malignancies other than NMSC,” all in the 50 mg arm. That looks terrible at first glance - and like dose response. But the first important point is semantic: colonic dysplasia is not invasive cancer. NCI defines dysplasia as abnormal cells that are “not cancer,” although it may sometimes become cancer depending on severity and context. So this is not really “three internal cancers.”

It’s more accurately two common organ cancers + one UC-relevant premalignant/surveillance lesion + an NMSC imbalance. That still needs explaining. But it is a very different fact pattern from “obe caused three cancers.”

Case level colour introduces some more information against causality:

• Prostate case was a 53-year-old male, 367 days on treatment, refractory to six prior advanced therapies, with a six-year disease history

• Breast case was a 65-year-old female, 206 days on treatment, failed corticosteroids only

• Colonic dysplasia case was a 49-year-old male, 369 days on treatment

Management color also said the relevant event timing ranged from day 49 to day 311, with no temporal clustering.

There may be two clocks floating around here: total treatment exposure versus maintenance-period timing, but either way the logic is the same: these are not cancers appearing on a biologically sensible de novo drug-induction timeline.

The breast cancer case is the easiest to rationalize: 65-year-old woman, 206 days on drug, and SEER puts the median age at breast cancer diagnosis at 64, and iirc was her first mammography.

The prostate cancer case is a bit less clean because 53 is young versus SEER median prostate diagnosis age of 68, but prostate cancer is the most common cancer category in men, often screening-detected, and a diagnosis within roughly a year or on the company’s alternate clock as early as weeks/months is not consistent with straightforward de novo carcinogenesis. Six prior advanced therapies isn’t a low-risk patient.

(I see comments about a carry forward of a prostate cancer case but I would imagine they would not have been included and thus there were 2 cases of prostate cancer, with one excluded.)

• NMSC, the mean age of NMSC cases was 62 vs 42 for the overall trial population, 2/4 50 mg NMSC patients were deemed not or unlikely related, and one of the remaining two had prior skin-cancer history.

End of the day there very well could be a case that high dose accelerates malignancies, but I’m more inclined towards noise, especially for the non-NMSC cases.

In totality I can understand why management didn’t clock it as a cancer signal. We have many datasets with ABVX:

In the Ph2b open label maintenance study, n=217 on 50mg once daily, with n=164 completing 2 years, ABVX reported no safety signals. On this it’s already 328 patient years of fexposure before counting partial year exposure from dropouts.

And as of its Nov’22 safety cut n=1074 patients had received obe, including n=209 on 50mg for at least 1 year with no opportunistic infection or malignancy signal detected.

Study 108 adds more comfort that after 2-4 years of prior 50mg open label treatment n=130 patients deescalated to 25mg, 104/130 completed 144weeks, patients had up to 7 years of total exposure and no new safety signals were observed.

And mechanistically, despite being a bit of an unknown obe doesn’t look like a pro-cancer immunosuppressant from what I can see. It’s an oral small molecule that binds the CBC 80/20 cap-binding complex upregulating miR-124: which reduces inflammatory mediators including IL-6, TNF-alpha, and CCL2/MCP-1.

It’s pretty different from a JAK inhibitor or broad immune suppression story. The vague MoA we have doesn’t naturally explaain rapid induction of cancer over a few months to a year.

In fact, miR-124 is often discussed in colorectal biology as tumor-suppressive.

miR-124, a potential therapeutic target in colorectal cancer

Anyhow what has management said in private conversations after yesterday’s debacle.

They will in the coming weeks provide safety package, which originally would have come at UEG. They’ll publish a more comprehensive safety analysis on the data we just received.

And the ABTECT maintenance Part 2 for induction non-responder cohort. There should be roughly n=500 additional non-responder patients with around n=150 for 50mg for 44 weeks. Look for safety here. If we see a signal the stock is going to struggle.

They confirmed they won’t dilute (yet). Thankfully, it would have been incredibly destructive. They have around 4 months to raise I would say, and a cash runway until Q4 2027. They’ll try their best to clean up the safety story, and let the stock recover, and try better to finance from strength. (They will need $500m-$1B more capital for a proper commercial buildout).

Arguably the price now makes it more attractive to BP than prior, but now pretty capped looking towards $150 for BO. Precedent favours no black box as the likely outcome, assuming overall safety package is free. And the KOL who mentioned AdComm on the conference call… I’m not commenting.

Not a bad opportunity at $70 if you think the market will be convinced by safety data coming out and a new dataset on non-responders in the next couple weeks (with CD for free almost). Personally, I think this really was a case of bad luck, and management failing to prepare properly for questions they should have expected, and I’m happy to be long from the $70s.

Moving off that good news:
Cytokinetics CYTK

Cytokinetics finally read out its Phase 3 ACACIA-HCM trial in non-obstructive HCM, and the stock is now up around 18% this month.

I was positive going into the trial. My view was that higher drug exposure should translate into better outcomes and increase the likelihood of hitting the primary endpoint. That is broadly what happened.

I will keep this section short, as I covered most of my thoughts in the post, but the key point is this: Cytokinetics now has the first drug with a credible path across the entire HCM spectrum. They also reported their first sales for aficamten in obstructive HCM, which marks the beginning of what I think can become a very large franchise.

This reaffirms my base case assumption that aficamten will become a multi-billion-dollar annual product, potentially reaching $5 billion-plus at peak. Cardiologists I speak to continue to view aficamten favourably versus Camzyos, particularly around drug-drug interactions, titration, manageability, safety, and overall ease of use.

The broader point is that aficamten is not just another HCM drug. It is one of only two clearly efficacious agents in the HCM space after beta-blockers failed to show meaningful efficacy in the head-to-head study last year. I would expect a significant share of new starts and treatment-naïve patients to move towards aficamten over time, and we should begin to see this reflected in sales.

My target remains $100 for CYTK.

C4 Therapeutics CCCC

You may also have noted that there’s a new stock write up on the substack. Written in collaboration with , we think that C4 is a great oncology name on the market.

C4 is developing its lead asset Cemsidomide to be a backbone for refractory relapsing multiple myeloma. Our view is that cemsidomide has the potential to be best-in-class, with a cleaner safety profile than competing degraders from BMS. Its ability to pair with other agents also makes C4 an attractive strategic asset and a credible M&A candidate.

Our SOTP PT is $7.90 using what we view as conservative assumptions. The full write-up goes into more detail.

The first catalyst we highlighted in the post came through this weekend, with BMS presenting positive data at ASCO last Friday. C4 is now up around 25% over the past week on the news.

The relevance is: BMS is also developing a next-generation IKZF1/3 degrader, which works through a similar mechanism to C4’s lead asset, cemsidomide. Positive BMS data help validate the class and reinforce the idea that this mechanism continues to have meaningful clinical utility in multiple myeloma.

Importantly, the BMS data continues to highlight a key area where C4 may be differentiated. Neutropenia remains a central toxicity concern for the class (61.1% here - C4 will! do better), and this is where cemsi’s safety profile could become a major advantage.

Next catalyst is EHA on 12 June where we should see updated Ph1 cemsi + dex data.

Broader Commentary

May was mixed and selective. XBI was up around 4% for the month, but, as usual, still lagged the Nasdaq and S&P 500.

Read more

C4 Therapeutics is a multiple myeloma company approaching several inflection points in one of the most validated mechanisms in haematology. For more than two decades, IKZF1/3 degradation has anchored myeloma treatment through Revlimid and Pomalyst, yet the market largely assumes Bristol Myers Squibb (BMS) will continue dominating the next generation of degraders through iberdomide and mezigdomide.

We believe that assumption is wrong. C4’s lead asset, Cemsidomide, is beginning to look differentiated where it matters most in modern relapsed/refractory myeloma: activity in post-BCMA patients, dose maintainability, and combinability with T-cell redirecting therapies. Cemsi’s Phase 1 data was generated in one of the most heavily pretreated degrader cohorts (median 7 prior lines, 75% post-BCMA; full comparator is on p. 9).

Our thesis is straightforward. Over the next 12–24 months, C4 will generate two datasets: the ongoing MOMENTUM study evaluating Cemsidomide plus dexamethasone in 4L+ multiple myeloma, and the Elrexfio combination study evaluating Cemsidomide with a BCMA T-cell engager in earlier-line disease. At the same time, Bristol Myers is expected to report additional iberdomide and mezigdomide data at ASCO (May 29-June 2), which could further validate IKZF1/3 degradation as the next-generation myeloma backbone, and highlight C4 as a sympathy play.

If MOMENTUM confirms the Phase 1 efficacy and durability profile, we believe the standalone asset is materially undervalued today. If the Elrexfio study shows Cemsidomide can safely deepen and extend T-cell engager responses, the opportunity expands beyond late-line salvage therapy into becoming a differentiated oral backbone partner for the broader myeloma ecosystem and will become a key M&A target for multiple myeloma players - not just undervalued.

At $3.80/sh and 249m fully diluted shares outstanding, C4 trades at an implied fully diluted market capitalization of roughly $946m. Our base-case SOTP yields a $7.90 PT (~108% upside) using conservative assumptions: 10% peak share in both 4L+ and 2L+ settings, 50% PoS for the standalone 4L+ opportunity, 33% PoS for earlier-line combinations, and zero value assigned to the broader internal pipeline outside Cemsidomide.

By Sanya and Han

By reading this you accept you have read and acknowledged the disclaimer.

Read On For More!

Subscribe now

The $16 Billion Vacuum

Since Bristol-Myers Squibb (BMS) acquisition of Celgene in 2019 for ~$74b, the market has long believed that IKZF1/3 degraders, known also as immunomodulatory drugs (IMiDs), would be permanently monopolised by BMS, with little to no room for competition. The belief held undeniably true for 6 years, with IKZF1/3 degraders (specifically Celgene’s Revlimid and Pomalyst) yielding more than $16 billion per annum to BMS at peak. The situation has since changed, as both degraders are now generic and are rapidly losing revenue. BMS aims to substitute some of the lost revenue through the introduction of two novel IKZF1/3-degrading CELMoDs (iberdomide and mezigdomide; a third CELMoD, golcadomide, is positioned in B-cell lymphoma, not multiple myeloma).

Enter C4 Therapeutics (C4), a targeted protein degradation company. C4’s key asset is Cemsidomide (Cemsi’), a novel IKZF1/3 “Monofunctional Degradative Agent Connector” (MonoDAC). Cemsi’ is a (potentially) best-in-class, most tolerable IKZF1/3 degrader rivalling novel CELMoDs in development which aim to replace now off-patent first-generation degraders (IMiD class). Functionally, the MoA of all three is the same: bind cereblon, recruit IKZF1 and IKZF3 for degradation, and the myeloma cell dies.

That’s a lot of nasty words… We’ll break it down.

Multiple Myeloma: An Introduction

Multiple myeloma (MM) is an incurable but increasingly treatable blood cancer of plasma cells in bone marrow. Under normal conditions, plasma cells are part of the immune system, helping the body to fight infections by producing antibodies (Ig / immunoglobulins).

In MM, however, this process breaks down… A single clone of abnormal plasma cells begins to grow uncontrollably inside the bone marrow. These malignant plasma cells crowd out ‘normal’ blood-forming cells, producing large amounts of an abnormal antibody protein, or a light chain, known as M protein. This abnormal protein can be detected in the blood or urine and is one of the key markers used to diagnose and track the disease.

The M Protein & M Spike - In a healthy person, antibody proteins are more evenly distributed. In multiple myeloma, the abnormal clone produces a large amount of one monoclonal protein, creating a visible ‘spike’:

The presence of abnormal plasma cells (and crowding out of normal cells), disruption of normal antibody production, and deposition of light chain in end organs results in the ultimate manifestations of myeloma which can include anemia, kidney failure, infections, and bone lesions.

Clinically, major signs of symptomatic myeloma are summarised by CRAB:

• C: High calcium levels (hyperCalcemia)

• R: Kidney problems (Renal impairment)

• A: Anaemia

• B: Bone damage or lesions

A patient is generally considered to have active or symptomatic MM when ≥10% of the cells in the bone marrow are cancerous plasma cells + at least one myeloma-defining event. That event could be a CRAB feature or a biomarker suggesting imminent organ damage.

About 190,000 Americans live with MM today, with ~36,000 new diagnoses and ~11,000 deaths projected for 2026 (SEER). Five-year relative survival has reached 62% (2015-2021), nearly double the figure from two decades ago. MM is becoming a long-tail disease: patients live longer, relapse more, and cycle through more lines of therapy. The pool grows every year.

Myeloma Keeps Relapsing…

MM should be best thought of as a long-term relapsing cancer. Patients often go through several rounds, or ‘lines’ of treatment over the course of the disease.

The current treatment backbone is built around combinations of drugs, and patients are typically treated with a mix of:

• IMiDs (e.g. Revlimid, Pomalyst)

• Proteasome inhibitors (e,g, Velcade, Kyprolis)

• Anti-CD38 antibodies (e.g. Darzalex, Sarclisa)

• Steroids (e.g. Dexamethasone)

Guggenheim Securities LLC

However, patients may initially respond well to a specific treatment regimen, sometimes for years, but the cancer will often return (relapses) or stop responding to treatment (becomes refractory). And despite the number of potential treatments we could use, patients can be more and more refractory.

Relapsed / refractory MM (RRMM) describes disease that continues to progress after prior therapy , inadequate response on treatment, progression within about 60 days after therapy in a patient who had achieved remission, or primary refractory from the beginning. Labels here matter to help describe which drug classes are no longer likely to work for patients.

• Triple-class refractory is disease refractory to at least three major categories of drugs:

  • IMiDs (immunomodulatory drugs)

  • Proteasome inhibitors

  • Anti-CD38 antibodies

• Penta-refractory MM is even harder to treat and is refractory to:

  • Two IMiDs

  • Two proteasome inhibitors

  • One anti-CD38 antibody

A reader will understand that the clinical problem has become more severe. Later-line myeloma isn’t just ‘more myeloma’, but biologically and clinically different. The disease is genetically heterogeneous and has evolved under treatment pressure; drug-sensitive clones killed, whilst resistant subclones survive and expand.

At the same time this is happening, the patient is becoming more fragile, accumulating cytopenias, infections, marrow reserve limitations, frailty, and toxicities from prior therapies…

Where a drug is used matters almost as much as whether it works. A frontline backbone, maintenance therapy, 4L salvage option, or post-CAR-T therapy are all being used in very different patients with very different expectations for efficacy, durability, tolerability, and commercial opportunity.

The modern late-line patient is now very difficult to treat… They aren’t ‘just’ a post-Revlimid, or post-Pomalyst patient. They’re now post-IMiD, post-proteosome inhibitor, post-CD38, post-BCMA, and sometimes post-GPRC5D…

CAR-Ts and bispecific antibodies have helped… but, they’ve not come close to solving the problem, carrying limitations of their own. Keeping it more simple, CAR-Ts can provide very deep responses (BCMA 73-98% ORR), but come with manufacturing issues, access limitations, patient fitness requirements, monitoring and safety concerns. Bispecifics are more accessible by nature of being off the shelf, but concerns around response depth (BCMA bispecific 71-74% ORR) durability (9-17m), infections and T-cell exposure exist.

From this, we can now see a paradox. MM has more treatment options than ever, but the modern refractory patient is also more heavily pretreated than ever. The more patients that live through each line of therapy, the more the field needs drugs that can still work after everything else.

It will be important to remember that the modern contemporary 4L+ or ‘salvage’ population is materially different from older late-line studies as BCMA CAR-T, BCMA bispecifics, and BCMA ADCs begin to form part of the treatment landscape.

IMiDs are a Treatment Backbone in MM

IMiDs are used across R/R MM (Highlighted in orange) - C4 Therapeutics

Over the last 30 years, multiple myeloma has been transformed from a treatment landscape that had only high-dose steroids & chemotherapy to now a multiplex of quadruplet or triplet regimens based on novel drug classes, cellular therapies, and bispecifics, that are often administered sequentially until progression or intolerance. The current treatment paradigm for myeloma is complex and ever changing, but has recurring motifs.

A core, persistent motif is that IMiDs are a regular backbone of treatment, whether induction or maintenance, transplant-eligible or ineligible. Many first-line regimens, whether quadruplet or triplet, will have Revlimid as a component. Once a patient frankly progresses on Revlimid, the regimen is often changed to a Pomalyst-based one (both IMiDs).

And We Still Believe the Next Backbone Will Still Be a IKZF1/3 Degradation

At first, it may seem strange that the next backbone could still come from an IMiD like mechanism, since we just mentioned that patients are already becoming refractory to IMiDs.

But that misses the point!

The IMiD mechanism remains one of the most validated mechanisms in MM. Revlimid and Pomalyst became foundational by binding cereblon and directing it to degrade IKZF1 and IKZF3; transcription factors that myeloma cells depend on to survive. When IKZF1/3 are degraded, downstream survival programs (e.g. IRF4 & MYC) are disrupted, myeloma cells die and immune activity improves.

These 1st generation drugs were not optimised from first principles, and were not originally designed with full knowledge of cereblon mediated IKZF1/3 degradation. Despite this limitation, they still are incredibly successful drugs due to the power of the underlying degradation mechanism.

IMiDs Degrade IKZF1/3

To highlight that prior IMiD resistance may not be as limiting as it first appears: Pomalyst demonstrated that patients refractory to Revlimid could still respond to a more potent cereblon degrader, despite both drugs belonging to the same IMiD class. Pomalyst binds cereblon more strongly and drives more refined IKZF1/3 degradation, establishing an important precedent: improving cereblon engagement and degradation kinetics can restore efficacy even after older IMiDs fail. That same principle now underpins next-generation degraders such as mezigdomide, iberdomide, and cemsidomide, all of which are being developed specifically for patients who would otherwise be considered IMiD-refractory.

So follows, the idea which is simple: keep the validated MoA, but build a more potent, more selective, more durable, and more dose optimised drug. A better designed degrader will in our view retain activity after old IMiD resistance, overcome partial cereblon downregulation, and serve as a new oral backbone in R/R MM.

That is exactly what BMS is attempting with iberdomide and mezigdomide, which we view less as a reason to avoid the space and more as external validation that large pharma still believes IKZF1/3 degradation will remain a core myeloma backbone for years to come; be it as a degrader, or also acting in combination with immune-based myeloma regimes.

The question is therefore not whether IKZF1/3 degradation will continue to work, but rather if Cemsi is differentiated enough? (Hint: We wouldn’t have written this if we didn’t think so!)

Cemsi: C4’s Differentiated Bet

C4’s lead asset, Cemsi, is an oral IKZF1/3 degrader improving on the old IMiD mechanism and competing with next-generation CELMoDs. We believe differentiation lies in three areas: more optimized pharmacology, activity in a more modern refractory population, and better dose maintainability / combinability:

First: Cemsi is more potent and pharmacologically optimised. Preclinical work suggests Cemsi binds cereblon far tighter than pomalyst and drives faster, deeper IKZF1/3 degradation, even as the older IMiDs can lose activity as the disease adapts and cereblon biology changes.

AACR 2021 - Cemsi has more activity than Pomalyst (Various Assays In Vitro)

Lonian et al., Cancer Res 2022 - Cemsi vs Mezig shows longer plasma concentration and IKZF3 Degradation

NJEM 2023 - Cemsi Begins IKZF3 Degradation in <3h

Secondly: Cemsi has been tested in a much more refractory and commercially relevant patient population than the BMS CELMoD datasets - and shown comparable (if not better) results. In C4’s Ph1, patients had a median 7 prior lines, with 75% prior BCMA exposure and 29% prior CAR-T plus T-cell engager exposure. By comparison, mezi’s earlier studies had only 12–30% prior BCMA exposure, while iberdomide’s expansion study excluded prior BCMA therapy altogether. Cemsi’s efficacy is being generated in patients much closer to today’s real-world post-BCMA late-line myeloma population, and these should be kept in mind when ‘reviewing’ the ORR numbers.

C4 Ph1 Dose Escalation Trial

Third: Cemsi is more dose-maintainable and combinable than mezigdomide.

The dosing point is important. IKZF1/3 degradation kills myeloma cells, but IKZF1 also plays a role in neutrophil (white-blood cell) maturation. If the IKZF1/3 degrader is too continuously suppressive, patients can develop neutropenia and infections resulting in dose interruptions, reductions, and discontinuations. So, balancing a potent molecule to deliver enough target degradation to kill myeloma whilst allowing the bone marrow to recover is ideal. (Read more below).

Cemsi’s efficacy looks even more impressive once you adjust for the patient population. With ORR that compares well against mezigdomide’s and iberdomide’s ORR, whilst being tested in a much tougher, more modern R/R MM population.

The read through for us isn’t that “Cemsi matches BMS CELMoDs”; but rather that Cemsi is producing comparable-to-better responses in post-BCMA, multi-refractory patients that now reflect the modern patient population. It’s this data that leads us to believe Cemsi is BiC.

C4 Offers Safer and Gentler Degradation

On-target Neutropenia is a class-wide side-effect of IKZF1/3 degraders. It is both manageable and widely expected, with physicians having a variety of countermeasures to treat it if it occurs. Nonetheless, less neutropenia is always better, as high rates frequently lead to both dose reductions and sometimes, drug discontinuation. In this regard, we believe Cemsi is a much more tolerable and safer degrader than both of the next-generation CELMoDs which we believe will be a key differentiation for late-line treatments.

A key reason may be C4’s alternative 14-on / 14-off dosing schedule, which gives neutrophils more time to recover than conventional 21-on / 7-off IMiD-style dosing. Whilst fewer dosing days would normally be expected to reduce efficacy, Cemsi appears to offset this through stronger potency (see above) and better dose tuning. In practice, improved tolerability may actually increase effective drug exposure over time by reducing dose reductions, interruptions, and discontinuations, allowing patients to remain on therapy longer and at intended doses; which on the surface may have seemed counterintuitive.

Importantly, cemsidomide-associated adverse events appear to cluster early, primarily during the first two treatment cycles, with toxicity rates declining meaningfully thereafter. We view that pattern as encouraging because it suggests patients may stabilize on therapy over time, potentially contributing to the low dose-reduction burden and absence of drug-attributed discontinuations seen thus far. We have not seen data from Mezigdomide thus far showing the same degree of apparent tolerability stabilization, which we attribute to reflect its more intensive dosing schedule and reduced neutrophil recovery time between cycles.

The detailed comparison panel is provided below:

Cemsi is clearly the most tolerated option, with

a) lowest amount of dose reductions;

b) no drug-attributed discontinuations and

c) lower Grade 3+ neutropenia than mezigdomide (57% vs 71%); iberdomide retains the lowest at 42%, though with materially lower ORR

We note that when compared to Iberdomide, Cemsi looks less differentiated on safety alone. However, it is important to remember that Iber’ did not produce impressive ORRs and both Cemsi and Mezig are undeniably better options from an efficacy perspective than Iber, making such comparison incorrect.

We also want to highlight that these safety results are in the context of Cemsi enrolling the most frail and weak population. These numbers are important for our long term view that Cemsi will derive value not just from efficacy, but generating responses that are sustainable and combinable with future immune therapies - which cannot be done as easily with a drug that is not safe!

4L+ Cemsi+Dex MOMENTUM

MOMENTUM is the near-term registration leg. The study evaluates Cemsi + dexamethasone (a steroid control partner) in 4L+ relapsed/refractory multiple myeloma, meaning it is not testing the broader “combination backbone” thesis with a novel immune therapy yet. Instead, it asks a simpler question: can Cemsi, with only steroid support, reproduce the Phase 1 efficacy, durability, and tolerability profile in a larger registration-oriented 4L+ dataset?

First patient was dosed in Q1’2026 and initial data is expected in H2’2027. Ph1 already showed cemsi works. For MOMENTUM, we think >40% overall response rate and ≥6 months median duration of response is the base-case success threshold. That would be clinically meaningful, competitive with mezigdomide, and supportive of an accelerated approval discussion in 4L+.

A true blue sky readout would be a repeat of the Ph1 high dose profile: ~50%+ overall response rate, >8–9 months medium duration of response.

Pfizer’s Elrexfio: First In Degrader BCMA Combination

The next part of the story is the test for the combination thesis, starting with Pfizer. In Dec 2025, C4 announced a free drug collaboration to evaluate Cemsi with Pfizer’s BCMA bispecific antibody Elrexfio (elranatamab). C4 will have the first non-large pharma IKZF1/3 degrader in the clinic with a BCMA T-Cell engager in Q2’2026. To make it a bit more exciting, Pfizer has a relevant myeloma asset, and no internal IKZF1/3 degrader franchise, and an incentive to make Elrexfio combinations more competitive (M&A?).

Mechanistically, pairing the two makes sense, T-cell engagers recruit patient immune cells to kill myeloma cells, but depth and durability are limited by T-cell fitness, exhaustion, infection risk, and ongoing cytopenias. Cemsi degrades IKZF1/3, kills myeloma cells directly, and activates immune effector biology: promoting T-cell/NK-cell activation and inflammatory cytokines (e.g. IL-2 and IFN-gamma). Preclinical work done by C4 also showed that Cemsidomide pretreatment enhanced Tecvayli-mediated T-cell dependent cytotoxicity.

That’s a whole lot of words... In English: Elrexfio pulls the immune trigger, and Cemsi can also make the T cells better ammunition.

An open label Ph1b in 2L R/R myeloma with n=30-50 will be conducted. C4 will test Cemsi 50ug, 75ug, and 100ug QD on a 14/14 schedule, with dexamethasone weekly through cycle 4 and Elrexfio dosed per label. Primary objective is safety and tolerability, with secondary endpoints including preliminary efficacy and immunologic activity. Cemsi will likely start in cycle 2 to reduce overlapping hematologic toxicity during Elrexfio step-up dosing.

We’re expecting (and hoping) to see deepened responses, and extended durability without breaking the bispecific dosing schedule. Elrexfio already produces a 58% ORR in BCMA naive patients,and 33% in BCMA exposed patients. The combination we expect should be trying to hit a 90%+ ORR, and extending duration of the response through better T-cell activation and persistence. If Cemsi hits this, C4 can replicate the template as an immune-therapy partner, if they aren’t immediately bought out (they will be bought out.).

We expect data in mid 2027. This will be one of the largest inflection points for C4’s valuation.

The Wider Template (and M&A?)

Pfizer is the first example; and if it can be replicated, then IKZF1/3 degraders will in our view become the preferred partner for T-cell redirecting therapies across MM. It’s a simple template: pair a cleaner oral degrader with immune therapies that don’t own an internal degrader.

Preclinical data suggests IMiDs/CELMoDs increase IL-2 transcription and CD8+ T-cell expansion, whilst early clinical data from JNJ Talvey + pomalidomide combination show higher response depth and durability than from TCE alone. Multiple trials are already testing iberdomide or mezigdomide with elranatamab, Tecvayli, Talvey, cevostamab, and etentamig.

C4 - Cemsi induces T-Cell Activation

That activity validates the mechanism, but BMS owning iberdomide and mezigdomide is also a strategic problem. Other MM players have assets with varying MoAs but lack an IKZF1/3 degrader, and combinations are already where the field is going. A clean external degrader is scarce.

C4 isn’t a copy of ‘BMS’s’ degraders, rather its Ph1 profile suggests it’s safer, gentler, and also more efficacious than BMS’s offering. For combinations, that may be the whole game! A backbone drug that will combine clearly, will be a backbone, whilst the ones that can’t combine clearly, will be salvage drugs. A 14/14 schedule, lower dose reduction burden, and immune activation profile makes C4 much more attractive as a backbone, and an attractive best fit partner for TCE regimens.

The first step is BCMA and Elrexfio, and the next is broader anti myeloma combinations. C4 is planning to disclose an additional Ph1b in mid 2026 with other anti myeloma agents. This will tell us whether C4 is building a single Pfizer trial, or a franchise architecture. We hope you’re as excited by what this means. It will mean that the upside won’t just be Cemsi + dexamethasone in 4L+, but as a supplier of a differentiated and better IKZF1/3 backbone to companies that are in MM, but lack the mechanism.

Even as the field moves towards T-cell redirection therapies, a combination is still needed to boost efficacy!

Sum Of The Parts: Turning Our Thinking into Numbers

We modelled the “base case” Price Target (PT) of $7.90 for C4, versus the spot price of $3.80. The model is risk-adjusted and the price target of $7.90 is assigned under present-day assumptions. Naturally, as C4 provides more data, our PoS and Market Share assumptions will change, shifting the PT.

a) Cemsi’ WAC-discount adjusted peak sales are exceeding of $3.4b (ex. probability, time and other adjustments)

b) C4’s strong balance sheet position with cash into Q1 2028, as well as cash expected from warrant exercises

c) $60m value assigned to Roche and Biogen partnerships (20m and 40m respectively) as we believe both have potential, albeit Roche $20m figure is merely an upfront payment carried out by Roche

d) C4’s pipeline/platform beyond Cemsi’ is valued at zero.

e) We believe a peak-sales based approach rather than a detailed revenue build is more appropriate for the company at this stage as it is unclear just how much Cemsi will differentiate (albeit it is evident it likely will). Once more data is generated by the group, we will update the modelling. We see 10% Market Share as a likely conservative number.

We also make a number of assumptions behind our model beyond those explicitly stated in the opus below are:

a) standardized scaling DR (yielding 14% at today’s fully diluted mcap versus 18.5% on common-only); note that as spot price increases our DR decreases, increasing PT (contact us for full derivations).

b) full exercise of all warrants by holders (with corresponding dilution and cash impacts, as we view this as reasonably likely and reflect uncertainty via PoS impact for both 2L-4L and 4L+)

C4 Peak Sales Build

Cemsidomide Only

4L+ Peak Sales Build (Cemsi + dex)

2L-4L Peak Sales Build (Cemsi combos)

Beyond the Base Case

A careful reader noticed numerous highly-conservative assumptions within the model, with fundamental ones being a low, 10% market share across 2L and beyond, and ⅓ PoS for 2-4L setting. We realize that these might be too conservative, with the sensitivity table below representing different assumptions and corresponding price targets that yields. Interestingly, we believe that even under a 33% PoS and a slim 5% market share, C4 remains undervalued by >25%, and that is assuming full warrant exercise.

What We’re Watching

What Could Go Wrong?

There are numerous ways that C4 can fail. Headline reasons include low cash reserves sufficient only until Q1 2028 (excluding warrant proceeds, which we see as likely and which will provide C4 with >$224m in additional financing); low credibility of management due to numerous pipeline failures and simply, fierce competition from BMS. We also believe the risk of ORR deflation in further Cemsi’ data is not as high as in traditional oncology programs (but is still present). As said before: we believe that Cemsi is unlikely to see reduced ORRs even with increases in sample number with two key reasons supporting our conclusion: a) Cemsi’, unlike Mezi’ is not seeing >25% rates of dose-reductions due to better tolerability, partially invalidating the risk of dose-associated reduction in ORRs and b) Cemsi’s current data are in unusually frail population (median of seven prior lines). MOMENTUM, however, is expected to enrol a less heavily pretreated 4L+ cohort (median prior lines likely closer to the contemporary 4L+ population than Ph1’s median 7-line setting) further reducing risk of reduced ORR. There is also some potential to see ORRs improve due to healthier baselines, we are not pricing that in, however.

The 14-on / 14-off dosing schedule remains one of the more important bear arguments. Fewer dosing days could theoretically reduce efficacy because the drug spends less time degrading IKZF1/3. However, we do not think this should be framed as a clear efficacy handicap. Cemsi appears to offset the shorter schedule through greater potency, durable target degradation, and better dose tuning. More importantly, the longer off-period appears to reduce neutropenia-driven dose interruptions, reductions, and discontinuations. In practice, that may mean patients receive more consistent effective exposure over time, even with fewer scheduled dosing days. The risk is that larger studies show the schedule is not enough to maintain efficacy or safety; the upside is that 14/14 proves to be the better optimized balance.

From the clinical standpoint, we are looking for a ≥40% Cemsidomide ORR in 4L+ setting for AA path to materialize as well as ≥6mo mDOR (median duration of response), which would make Accelerated Approval likely, as well as also be clinically meaningful compared to mezigdomide (equivalent efficacy). The bar IS high, which by itself creates inherent risk to Cemsi’s 4L+ aspirations. The bar for Elrexfio + Cemsi’ combo is likely at around 80-90% ORR, again, high. We want to re-iterate that we see Cemsi as likely to succeed in both of the settings outlined.

Of course, the risks outlined above are substantial. We account for all via our PoS (50% for 4L+ and 33% for 2L-4L) as well as low Market Capture expectations (at this time)

Net Net

Multiple myeloma has been a degrader story for twenty years. From Revlimid to Pomalyst, and now mezigdomide and iberdomide, the class has anchored every meaningful regimen we have seen, whether the partner is a proteasome inhibitor, a CD38 antibody or a BCMA bispecific. Neither we nor leading KOLs see that changing in the next 10+ years. The MM market is also big enough for all three next-gen degraders to coexist, and Cemsi doesn’t have to beat mezigdomide to be a share-taker. We reflect this in our model, whereas 10% share base case bakes in efficacy equivalent to Mezig’ (≥) and likely better safety (median 7 prior lines in Cemsi’s 4L+ cohort, the most heavily pretreated tested against any next-gen degrader), so further upside if Cemsi’ efficacy outperforms is not priced in by neither our model / market.

Interestingly, the next inflection point for C4’s story may be imminent, with new data for mezigdomide and iberdomide coming this weekend at ASCO. We think Bristol Myers will create a lot of noise around the presentations. Good data from BMS degraders is likely to attract renewed attention to the space, highlight C4’s scarcity value as the only real competition, and underline C4’s combination opportunity / 4L+ value. We plan to update the safety, efficacy and our model after this event. Even if ASCO does not yield sufficient attention to C4, key readouts in 2027 remain as key catalysts regardless.

We want to highlight that our model is a stress test with low embedded expectations. Under stress-tested assumptions as highlighted in the Sum of The Parts, the math still gets us to a $7.90 PT against a $3.80 spot. Any efficacy edge over mezigdomide is unmodeled/free upside in either our model or the market price. Taken together, we think the current share price of C4 represents an attractive entry point and positive long-term expected value.

Thank you for reading and reaching the end! We hope you appreciate C4 as much as we do!

Appendix: Holders

A lot of information dropped yesterday. The long awaited topline readout from aficamten in nHCM in ACACIA, and the first quarter of sales since aficamten has been approved in oHCM. Cytokinetics also announced a offering of $650m.

Now that most of the regulatory hurdles are past us, this really is going to be a show-me story. In my opinion, this is going to follow a trajectory similar to BBIO’s own stock, as sales start to flow in.

By reading this you accept you have read and acknowledged the disclaimer.

In summary:

• ACACIA-HCM Ph3 met dual primary endpoints

• MAPLE-HCM sNDA accepted, PDUFA Nov. 14, 2026

• MYQORZO Q1 launch: >275 HCPs, ~680 pts

• $4.8M Q1 MYQORZO revenue, ~$1.1B cash

• Proposed public offering announced of $650m
  

ACACIA-HCM

The past writeup I had on ACACIA-HCM

ACACIA-HCM met both dual primary endpoints. W/36 (aficamten vs placebo):

• KCCQ-CSS improved by 11.4 points vs 8.4 points | delta = +3 points (p=0.021)

• pVO2 increased by 0.64 mL/kg/min vs -0.03 mL/kg/min | delta = +0.67 (p=0.003)

There’s a clear! improvement in both metrics.

One particularly encouraging detail was the washout data (+ rapid early improvement + sustained benefit through treatment). After discontinuation of aficamten, KCCQ scores declined back toward placebo levels. Now nHCM is very messy, but this is very clear evidence for the FDA that the observed benefit is drug related.

So why the muted reaction? There’s 2 points I can think of; and not because expectations were too high. The options market was implying about a 30% probability of success going into the readout.

The first issue was the headline KCCQ delta at Week 36, which came in at ‘only’ +3 points. But: most of the study, the separation versus placebo looked closer to ~5–7 points based on the longitudinal curve.

That likely explains the strong move in Edgewise (+21%), a competitor company also developing a next-generation cardiac myosin inhibitor. The interpretation appears to be that aficamten demonstrated efficacy, but the delta leaves room for a ‘safer competitor’.

In practice though, cardiologists are unlikely to evaluate aficamten based purely on the w/36 snapshot. It’s not hard to look at the full curve, and see that the overall efficacy profile appears stronger than the headline +3 point that will appear on the label.

Safety was the second issue, but I think the concern is significantly overblown and largely anticipated.

No new safety signals were reported. The key issue was reduced LVEF, which is already a well-understood on-target effect of cardiac myosin inhibition. The relevant question is therefore not whether LVEF reductions occur; we already know they do, but whether they are predictable, reversible, and manageable.

In ACACIA:

• 10% of aficamten patients developed LVEF <50% (vs 1% placebo)

• 3% required treatment interruption due to LVEF <40%

• 2 patients experienced serious heart failure adverse events associated with reduced LVEF

Management responded that:

• the majority of LVEF reductions were asymptomatic

• EF recovery generally occurred within days to 1 week after interruption

• patients could often restart at lower doses and continue therapy

This was also an aggressive titration strategy in nHCM, dosing toward maximum tolerated exposure rather than the minimum effective dosing strategy often pursued in oHCM.

Most patients from ACACIA have now rolled into the FOREST-HCM open-label extension, which should provide important longer-term safety and durability data. We should also get substantially more detail at upcoming medical meetings.

Overall, I think market is underappreciating that this is the first drug to hit in nHCM, and overemphasising the w/36 delta. Being able to be a drug that encompasses the entire HCM spectrum will make the sell so much easier.

We get more details on efficacy and safety in follow up conferences and in FOREST. This should be a easy approval. Well done Blum on excellent trial design again!

Sales Q1

Moving Forward

This builds on my prior post around aficamten potentially expanding across the full HCM spectrum.

MYQORZO launched on January 27; Q1 is a half quarter of sales. Even so, the early launch metrics look strong, and management said: ‘most of these metrics exceed our launch expectations’. (Trust in Blum!)

More than 275 HCPs prescribed MYQORZO during Q1, with over 50% coming from high-volume HCM prescribers. Through April, that number had already expanded to more than 425 HCPs. Cytokinetics also reported:

• c1,400 REMS-certified HCPs

• c680 patients by Q1 end

• already c1,100 patients through April

  • 70% of dispensed patients on paid prescriptions

• conversion to paid therapy averaging <2 weeks

That April patient growth is notable in showing acceleration. They added roughly 420 patients in about a month after exiting Q1 at 680.

Aficamten appears to be penetrating the specialist HCM community quickly. Cytokinetics noted that high-volume HCM prescribers historically account for ~80% of HCM prescriptions, and the company detailed >90% of these physicians during Q1. That strategy makes sense: dominate the key specialists first, then broaden into the wider cardiology community later.

Prescribers averaged 2.4 prescriptions per HCP, while Cytokinetics estimates MYQORZO exited Q1 with >30% new-to-brand share. For a second entrant competing against an entrenched incumbent in Camzyos, that is a strong early signal.

The broader commercial setup also continues improving.

Once aficamten is approved across both obstructive and non-obstructive HCM, prescribing should become simpler across the entire HCM spectrum. There are currently no approved disease-specific therapies in nHCM, and in real-world practice differentiating perfectly between oHCM and nHCM is not always straightforward. A single therapy usable across both settings lowers prescribing friction meaningfully.

Operationally, aficamten also is much easier to use than Camzyos:

• fewer DDI concerns,

• simpler REMS requirements,

• more flexible echo timing,

• and cleaner pharmacy/distribution logistics.

Administrative simplicity can drive adoption just as much as small efficacy differences.

I continue to think aficamten can become a $5B drug over time. If execution continues, Cytokinetics looks much closer to getting to that future $100 stock price than what the market is pricing today. But, that’ll take time, and sales to show up!

Grace Therapeutics GRCE

PDUFA was on the 23rd of April, and as much as I’d have hoped it would have passed with approval from the FDA, GRCE received a CRL from the administration on Chemistry, Manufacturing, and Controls related issues. The details we have so far is that it’s a Class 2 CRL, which will be at least a 6-8 month review from the FDA, after fixing the issues identified in the letter. I’ve been checking the FDA CRL database, and have yet to find it released for viewing which is a bit annoying so we’re only working with what GRCE has specified.

My case was that mortality would not be a concern with the FDA, considering the pathway, and it seems that is holding through. The drug itself is ‘approveable’ by the FDA but the issues from the CDMO and manufacturing is the pain:

What happens going forward and what are the issues then in plain? Leachable coming from manufacturing equipment, infusion bags, needles, tubing etc is the main issue; the commercial CDMO dataset is incomplete; they don’t have enough commercial real world data is what I’m guessing. They are asking for baseline + longitudinal data from the actual system that will be used in the market.

The baseline work can be done in 2-3 months, and fast work in 6 months. The issue and risk here is if anything quantifies above thresholds, which will require reformulation, material changes, and/or additional tox works, which makes this timeline much much longer. (We need the CRL letter and CDMO details to determine this!).

And because of this leachables dataset being lacking, the FDA couldn’t complete the nonclinical tox assessment. Here I’m less concerned, alcohol levels (<2%) are within known limits, there aren’t novel excipients, and can run in parallel.

Now, the second main issue, which is that the CDMO has cGMP deficiencies. They have facility / process compliance issues ! What the hell? Are they using Catalent’s cousin ? I can’t comment without knowing what CDMO this is, but this means that the FDA needs to inspect the facility AGAIN, book a new inspection slot and get their act ready again.

What does this mean? It’s a 1+ year to get this all fixed to PDUFA again. Considering the chairman of GRCE runs one of the largest private CDMOs in America I’m shocked this is an issue that even came up… What I’ll expect is the Type A meeting goes ahead, and they raise 3-6 months within this duration. We also need to wait for next financial report to understand the dilution that will be expected. For the time being I’m out of this and will look maybe again later down the line if it’s at a significant discount.

Ultimately the drug is still the same, it works, is approveable, and is still capable of generating $150m p.a. peak sales, the manufacturing is definitely resolveable, but the timelines are pushed back long.

And very quickly on the second large catalyst Cytokinetics:

It isn’t happening in April! As you may have guessed by the fact it’s now May…

I think it hits, but I’ve been wrong and blindsided by the CMC for GRCE, so we’ll see how nHCM goes in May. Good luck!

Read more

Given that, I don’t think it makes much sense to introduce new ideas right now. Instead, I’ll focus on two upcoming April catalysts where I’m already positioned long: Cytokinetics’ Phase 3 nHCM readout and Grace’s GTx-104 PDUFA decision.

I’ll keep this one relatively brief and outside the paywall, as I don’t think the content is especially substantial this month.

I’m long two readouts into April, and I think that’s enough to tackle this month. I’ve already extensively written into both of them: GRCE and CYTK.

Grace Therapeutics $GRCE

Written extensively about this. There’s been a lot of traction in regards to knowledge around the company now. There are strong votes of confidence and there’s a bit of bafflement at how the company was trading so cheap, and the improvement in SP is reflecting this. The PDUFA date is set for April 23. On approval I’d expect it to tick up to $6 handily. I’m not expecting a raise; the company has enough cash to get to cash flow positive.

We’ll see how it goes on the day. Main bear issuesare as so far: mortality on the dataset, which most clinicians aren’t concerned about. At TDCowen, although there was no disclosure on the mortality by grade, Kohli attributed the higher mortalities to discrepencies in grade, which would make sense condiering the greater Grade V. Now on CMC concerns, micelle formulations are commonly used, and it’s well understood and it’s not a complicated product ‘really’ from a handling perspective.

I think it passes PDUFA at 90% PoS, and commercially it’s a very straightforward sale into neuro centers.

Cytokinetics $CYTK

Now, there’s been no actual confirmation that the readout is in April, but justding by the timelines of enrolment, it’s pointing towards April.

If it hits, it’s a $90-100 stock easy. The nHCM market aficamten has already been approved in makes me feel comfortable in a $40 backstop minimum.

Considering the main 1L treatment and only alternative are beta blockers which Cytokinetics showed last year in a H2H (vs metoprolol) are ineffective and can actually be damaging I’m confident that oHCM will provide a strong backstop. What alternatives are there?

On REMS: Despite concerns that REMS could slow uptake, cardiologists I’ve spoken with don’t view it as a meaningful barrier. Camzyos has already established the framework, meaning aficamten will launch into a pre-educated market, which will ease adoption.

In practice, resource constraints at community centers may actually work in Cytokinetics’ favor, with I’d imagine most if not all of treatment-naive patients more starting on aficamten. This is combined with its improved PK profile and fewer DDI, which is a greater impact than the REMS.

On the upcoming trial, I have a suspicion that BMY botched up recruitment, nHCM is pretty hard to diagnose and looking at the data, I believe some of the patients weren’t actually nHCM which would explain the placebo outperforming in their trial. Coupled with the fact that their drug was unable to reach peak sustained exposure... It was actually surprising how close considering these challenges they were to hitting stat sign which shows that the MoA works!

And despite Blum’s criticisms, his team and he are extremely experienced in trial design. The trial is a game of p values, but higher systemic exposure and a better trial design makes me feel pretty confident at least one will hit if not both primary endpoints.

If successful, Cytokinetics would be the first company to span the full HCM spectrum with a truly efficacious therapy.

Read more

Recommendation: [Long]

Current Target Upside

Price: 62.00 90.00 45.16%

Last time I discussed CYTK, the target was $55 into the PDUFA. The stock has since moved from the low $30s to $62, and I think it’s long overdue a recap on what’s happened and what will happen now that Camzyos got approved. Next major catalyst is the ACACIA-nHCM trial which reads out in Q2 I’d expect there will be a +/- 30% stock price moving event.

This update is split into two parts:

A short recap of what has happened and a focus on oHCM (Past Note)

A deeper PoS dive into ACACIA in nHCM (This Note)

By reading this you accept you have read and acknowledged the disclaimer.

Foreword

I don’t think the market fully appreciates that ACACIA may have a better chance of success than many expect.

At first glance, the precedent looks bad… ODYSSEY was the first Ph3 trial of a cardiac myosin inhibitor in nHCM and it missed both co-primary endpoints. The natural takeaway is that myosin inhibition simply doesn’t work in nHCM, and therefore aficamten; another myosin inhibitor should fail as well.

I think that conclusion is too simplistic.

ODYSSEY tells us a lot about why the trial failed, and those lessons actually suggest ACACIA has a reasonable path to success.

ACACIA should read out before the upcoming earnings, and based on enrollment completion I would expect results sometime in April.

Importantly, my base case for CYTK does not rely solely on this readout. As discussed in the previous post on oHCM, the oHCM franchise alone already provides a strong commercial foundation. That gives the company a meaningful backstop even if ACACIA were to miss. But on nHCM specifically, I’m constructive heading into the data.

Refresher on aficamten & mavacamten

Aficamten is a next-generation cardiac myosin inhibitor. It’s considered next-generation because it improves on the first drug in the class, mavacamten (Camzyos) from $BMY. The key improvement is in its pharmacokinetics profile. Aficamten has a much shorter half-life (c1.5 days vs c7 days for mavacamten), which allows physicians to titrate dosing more precisely and respond more quickly to changes in cardiac function.

Both drugs are now approved for obstructive HCM (oHCM) where myosin inhibition has shown clear clinical benefit. However, neither drug is currently approved for non-obstructive HCM.

Refresher on nHCM

Hypertrophic cardiomyopathy exists in two main forms: obstructive (oHCM) and non-obstructive (nHCM).

Rough estimates suggest nHCM represents about 30–35% of all HCM cases.

In oHCM, the thickened septum narrows the left ventricular outflow tract (LVOT), which physically obstructs blood flow leaving the heart. This mechanical obstruction drives many of the symptoms, including reduced cardiac output, elevated filling pressures, and impaired exercise tolerance.

In nHCM, patients still have hypertrophy and the same underlying sarcomere mutations, but there is no LVOT obstruction. Because there is no clear mechanical blockage to relieve, treatment responses are usually measured through patient-reported symptoms and exercise capacity rather than changes in outflow gradients.

Both forms ultimately stem from the same core biological problem: excessive sarcomere cross-bridge formation and hypercontractility. In theory, that means both should be addressable through cardiac myosin inhibition.

However, there are currently no disease-specific therapies approved for nHCM. Treatment today largely focuses on symptom control, improving diastolic function, preventing arrhythmias, and general heart-failure management.

If aficamten works in nHCM, it will be the first therapy addressing the entire HCM spectrum, and in my view become a >$5B franchise drug.

Before getting into the meat of it just a metrics legend to help provide context to the measurements:

NYHA Functional Class: (Both ACACIA & ODYSSEY enrolled Classes II-III as inclusion)

• I : No symptoms with normal activity

• II : Mild symptoms with ordinary activity

• III : Marked limitation ; symptoms with minimal activity

• IV : Symptoms at rest

KCCQ-CSS (Kansas City Cardiomyopathy Questionnaire = Clinical Summary Score):

• A scale from 0-100: 0 being worst possible symptoms 100 being no symptoms (c5 point improvement = clinically meaningful change

  • 75–100 Mild symptoms

  • 50–74 Moderate symptoms

  • 25–49 Severe symptoms

  • <25 Very severe symptoms

LVEF (Left Ventricular Ejection Fraction): How much blood the heart pumps out with each beat

• 55–70% Normal

• 50–55% Mildly reduced

• 40–49% Moderately reduced

• <40% Systolic heart failure

Peak VO₂ (ml oxygen / kg body weight / minute)

• Maximum oxygen consumption during exercise testing. Measures cardiopulmonary fitness and exercise capacity.

• Typical ranges: (Typical meaningful improvement = 1.0ml/kg/min)

• >35 Highly fit

• 25–35 Normal

• 15–25 Reduced capacity

• <15 Severe impairment

NT-proBNP: A biomarker released when the heart muscle is under stress.

• Typical ranges:

• <125 pg/mL Normal

• 125–300 Mild elevation

• 300–900 Moderate elevation

• >900 Significant cardiac stress

Introduction to CYTK’s ACACIA

ACACIA is the Phase 3 trial evaluating aficamten in adults with symptomatic non-obstructive HCM.

The study was originally planned to enroll around 420 patients based on the HFSA 2023 trial design, but enrollment has overenrolled to 516 patients (500 planned).

The trial design also has changed slightly.

Initially the study used KCCQ-CSS as the sole primary endpoint, but it was later modified to include dual co-primary endpoints:

• KCCQ-CSS (symptom and quality-of-life score)

• Peak VO₂ (exercise capacity)

The primary analysis evaluates change from baseline at Week 36.

The larger sample size allows the trial to maintain approximately 90% statistical power even with two primary endpoints. Because the endpoints are split, the statistical significance threshold is 0.025 per endpoint, instead of the traditional 0.05.

Technically, hitting one endpoint with strong supportive trends on the other would constitute a positive result, but I suspect that the market and FDA will look for both endpoints to be hit, or at least one positive endpoint and a positive ‘near miss’ / positive trend.

ODYSSEY: Mavacamten (Camzyos) by $BMY missed both co-primary endpoints in nHCM

Fortunately, since ODYSSEY has effectively the same dosing and targets the same MoA, we can use it to bridge to ACACIA and better figure out if ACACIA has a shot on goal?

Context (it’s asking the same question as aficamten in ACACIA)

Can mavacamten improve symptoms and exercise capacity in non-obstructive hypertrophic cardiomyopathy (nHCM)? The biological logic was straightforward and is the same logic underlying aficamten. Reducing sarcomere hypercontractility -> less forceful contraction -> better cardiac efficiency, and ultimately helps patients feel better and exercise better.

That logic had already worked in obstructive HCM, where myosin inhibition clearly improves hemodynamics and symptoms. ODYSSEY tested whether that same mechanism would translate into non-obstructive HCM.

What patients were enrolled

The ODYSSEY population was moderately symptomatic, but with advanced and long-standing disease. Roughly 70% were NYHA Class II and 30% were NYHA Class III, with a mean KCCQ-CSS of about 57, which indicates meaningful symptom burden. Mean LVEF was 65.8%, so systolic function was preserved. Mean peak VO₂ was 18.2 ml/kg/min, showing impaired exercise capacity, and median NT-proBNP was about 917, consistent with significant cardiac stress.

Overall: patients enrolled had hearts that still pumped well at rest, but were clearly symptomatic and exercise-limited.

Endpoints and what to hit

The trial required improvements in coprimary endpoints of Peak VO2, & KCCQ-CSS better than placebo.

Patients received titrated doses:

% Patients received vs Dose (mg)

• 3% 1mg

• 6% 2.5mg

• 31% 5mg

• 28% 10mg

• 22% 15mg

Note: 10-15mg was the high dose, and about half of patients reached it.

What actually happened then?

Mavacamten narrowly missed statistical significance, but improved both co-primary endpoints numerically, and there were also directional improvements in biomarkers and imaging parameters.

New England Journal of Medicine

Peak VO2 change +0.52 vs +0.05, difference = 0.47, p = 0.07

KCCQ +13.1 vs +10.4, difference +2.7, p = 0.06

Improvement on both measured primary endpoints! But not stat sig.

ODYSSEY: Why the miss then if it showed efficacy?

1. The placebo problem… (nothing aficamten can fix if bad luck strikes again)

Placebo patients improved +10.4 points on KCCQ-CSS, which is unusually large. In most heart-failure settings, about 5 points is already considered clinically meaningful, meaning placebo in ODYSSEY improved by roughly double the meaningful threshold…

Several factors likely contributed. Patients in global Phase 3 trials often receive better background care, including medication optimization, arrhythmia management, and closer specialist follow-up. KCCQ is also subjective, making it sensitive to increased monitoring and patient expectations. Meanwhile peak VO₂ testing is effort-dependent, and patients often perform better at later visits simply due to familiarity with the exercise protocol. However, all of these issues will confound ACACIA… Fingers crossed.

2. Safety Effects Reduced Drug Exposure (aficamten can fix this!)

Mavacamten interruptions occurred in 25.7% of patients versus 7.6% on placebo, due to reductions in LVEF below 50% from systolic dysfunctions. In ODYSSEY these interruptions lasted about four weeks, meaning patients lost effective exposure during a trial attempting to measure sustained benefit...

3. nHCM may be a harder biological target (aficamten better exposure may fix this?)

Unfortunately the biology itself may be less favorable in nHCM than in obstructive disease.

In oHCM, hypercontractility directly drives left ventricular outflow tract obstruction, so reducing contractility has a clear therapeutic effect. But in nHCM, symptoms often arise from a broader mix of mechanisms including diastolic dysfunction, myocardial ischemia, ventricular stiffness, and impaired energetics.

Myosin inhibition may only address one component of the disease, BUT better exposure may allow it to overcome this barrier.

Further ODYSSEY Data (Post-hoc analyses)

Post-hoc analysis support that drug exposure determines efficacy (intuitively this makes sense). In ODYSSEY, patients with baseline LVEF ≥65% achieved >2x higher mavacamten concentrations by Week 48 and experienced fewer treatment interruption; of this higher exposure group we saw much greater efficacy:

ODYSSEY high-exposure subgroup

• KCCQ-CSS: +c5.1 placebo-adjusted

• Peak VO2: +c0.75 placebo-adjusted

For a refresher, the overall ODYSSEY trial results were smaller:

• KCCQ-CSS: +2.7

• Peak VO2: +0.47

Pulling it all Together in CYTK’s ACACIA

Are there meaningful differences in ACACIA that would result in a different outcome to ODYSSEY?

I think yes. One of the biggest issues in ODYSSEY was exposure instability, due to dose interruption.

Aficamten behaves differently. Unlike mavacamten, where treatment is typically temporarily discontinued once LVEF falls below 50%, aficamten dosing generally down-titrates first, with interruption typically reserved for LVEF <40%. Cytokinetics has also noted that full treatment interruptions occur extremely rarely.

ACACIA is leveraging essentially the same titration framework used in prior aficamten studies, meaning the dosing strategy has already been stress-tested in both oHCM and earlier nHCM cohorts, as we see in the 96-week FOREST (n=34) extension from the REDWOOD-HCM cohort.

A total of 34 patients were followed for 96 weeks. After dose titration, most patients were able to reach and maintain the highest therapeutic doses, with the final dose distribution:

• 58.8% on 20 mg

• 17.6% on 15 mg

• 11.8% on 10 mg

• 11.8% on 5 mg

76% of patients sustained high-dose therapy over a long term.

Safety events related to excessive myosin inhibition were limited. LVEF <50% occurred in 4 patients (5.4 events per 100 patient-years) and was reversible with dose reduction or short interruption, with patients able to safely resume therapy. Biomarkers improved alongside symptoms, with NT-proBNP falling c753 pg/mL (c70%) and hs-cTnI declining significantly by week 96. Cardiac function remained stable, with LVEF normalizing from c70% to c64% and remaining stable throughout treatment.

FOREST gives me confidence that aficamten will be able to maintain much more consistent therapeutic exposure in ACACIA than in ODYSSEY.

“The numbers, Mason!” - Doing a Estimate of PoS

NOTE: THIS ENTIRE THING IS COMPLETE GUESSTIMATION I COOKED UP TO JUSTIFY MY OWN BIAS THAT THIS WILL HIT AT LEAST ONE PRIMARY ENDPOINT

If ACACIA can improve exposure, and exposure can increase treatment size, with more patients on maximal dosing:

Exposure is higher for aficamten & should translate to a better response

Pretty poor man’s exposure > pharmacodynamic > clinical response framework

Myosin inhibitors produce concentration-dependent reductions in contractility:

PD(E) = Emax x E / (EC50 + E)

Higher sustained doses = higher pharmacodynamic effect

FOREST showed 76% @ 15-20mg (59% on 20mg), LVEF <50% events rare (5.4/100pt-years). ODYSSEY c40% <5mg, c50% >10-15mg. Mean dose = c7.8mg. Vs FOREST mean dose = c16.3mg.

Efficacy uplift = PD(E2) / PD(E1) = [E2 / (EC50 + E2)] / [E1 / (EC50 + E1)]

For an EC50 range from 5-15mg, you get 1.3-1.5x greater pharmacodynamic effect, and if system is further from saturation 1.6x.

If you take it at face value you get efficacy expectations from the direct uplift of:

• KCCQ-CSS = c4pts

• Peak VO2 = c0.65mL/kgmin

This is working on the basis of an increased effective exposure which aficamten is well placed to achieve.

Accounting for placebo

The placebo is a case of bad luck for ODYSSEY, but to keep it fare and just keep the harsh placebo improvement we’ll base it similarly. ODYSSEY placebo patients improved +10.4 KCCQ points from a baseline of ~57. Because KCCQ is bounded 0–100, placebo response scales with baseline severity.

So:

Placebo ΔKCCQ = cλ × (100 − baseline KCCQ)

From ODYSSEY:

λ = 10.4 / (100 − 57) = c0.24

For an ACACIA-like population (baseline KCCQ = c57–60):

Expected placebo improvement = c9.5–10.5 points; so if aficamten produces c14–15 point total improvement, the placebo-adjusted effect is c4–5 points.

Statistical power in ACACIA

Enrolled: 516 patients (c258/arm)

Assuming:

SD = 15

α = 0.025

SE = SD × sqrt(2 / n)

SE = c1.32

Using ACACIA assumption of ΔKCCQ ≈ 4.0–4.5 points, then:

• Z = 4.0 / 1.32 ≈ 3.03

• Z = 4.5 / 1.32 ≈ 3.41

With alpha = 0.025, the critical value is about 2.24, so KCCQ power is roughly 79–88%.

Peak VO2 power

Peak VO₂ variability in cardiopulmonary exercise testing studies typically shows standard deviations around 3 mL/kg/min

SE = 3 × sqrt(2 / 258)

SE ≈ 0.264

With dpVO₂ ≈ 0.60–0.65

Z = 0.60 / 0.264 ≈ 2.27

Z = 0.65 / 0.264 ≈ 2.46

Which implies roughly:

51–59% power for the peak VO₂ endpoint.

And putting it all together

If the endpoints were independent, the probability of both hitting would simply be KCCQ power × peak VO₂ power, or roughly 0.835 × 0.555 = 46%. But this understates true joint probability, because KCCQ and peak VO₂ reflect the same underlying physiology. Correlating the approximation, P(A ∩ B) ≈ p1p2 + ρ × sqrt[p1(1−p1)p2(1−p2)], and assuming a modest positive correlation of ρ ≈ 0.25–0.45, the probability of both endpoints hitting rises to roughly 51–55%.

Roughly speaking:

c50–55% chance both endpoints hit

c65–75% chance at least one endpoint hits with a positive trend on the other

ODYSSEY showed that myosin inhibition can move the right endpoints in nHCM, but the trial narrowly missed statistical significance due to large placebo response and unstable drug exposure. Aficamten’s dosing profile appears meaningfully different: in FOREST most patients sustained high doses with few LVEF-related interruptions, suggesting more consistent exposure. If exposure truly drives efficacy, even modest improvements over ODYSSEY’s treatment effects could be enough for ACACIA to achieve statistical separation.

Based on (very) rough guesswork and ACACIA’s larger sample size, the probability of both endpoints hitting may sit around c50–55%, with a c65–75% chance that at least one endpoint hits with a positive trend on the other. From the current price of $62, I see c$90 (c45% upside) if ACACIA reads out positively, while a clear miss would likely push the stock back toward the $40s, with longer-term value still supported by the oHCM franchise around c$80.

We’ll see how it goes in Q2!

Date: [04/03/2026]

Recommendation: [Long]

Current Target Upside

Price: 60.72 90.00 +c45%

By reading this you accept you have read and acknowledged the disclaimer.

Last time I discussed CYTK, the target was $55 into the PDUFA. The stock has since moved from the low $30s to $60s, and I think it’s long overdue a recap on what’s happened and what will happen now that Camzyos got approved. Next major catalyst is the ACACIA-nHCM trial which reads out in Q2 and I’d expect the event to be a +/- 30% stock price moving event. Regardless, the current oHCM approval gives a fairly nice backstop to a miss.

This update will be split into two parts:

A short recap of what has happened and a focus on oHCM (this note)

A deeper PoS dive into ACACIA in nHCM (next note)

ESC 2025 (September)

We got two reads, that helped understand the positioning for aficamten:

1. Aficamten won h2h against metoprolol in oHCM (MAPLE).

Metoprolol (a beta blocker) remains standard first-line therapy in obstructive HCM, despite limited evidence that it meaningfully improves exercise capacity. It reduces heart rate and symptoms in some patients, but it does not directly address the underlying hypercontractility driving LVOT obstruction.

MAPLE was important because it tested CYTK’s aficamten directly against SoC in symptomatic oHCM (n=175, 24 weeks).

MAPLE showed (aficamten vs metoprolol):

• Peak VO₂: +2.3 mL/kg/min vs metoprolol (p<0.0001)

• KCCQ-CSS: +6.9 points

• NYHA improvement: 51% vs 26% (NNT ≈ 4)

• LVOT gradient reductions: −30 mmHg (rest), −35 mmHg (Valsalva)

• LVEF <50%: 1.1%

Mean peak oxygen consumption (VO₂) increased with aficamten and declined with metoprolol over 24 weeks in symptomatic oHCM.

Consistent benefit of aficamten versus metoprolol, with treatment effects favoring aficamten across age, sex, baseline functional class, and gradient strata

Importantly, mean peak oxygen consumption increased with aficamten and declined with metoprolol over 24 weeks. Metoprolol reduced exercise capacity by −1.2 mL/kg/min and did not meaningfully improve gradients. Aficamten improved both symptoms and hemodynamics.

The treatment effect was consistent across age, sex, baseline functional class, and gradient strata; it was broadly reproducible.

Aficamten demonstrated superiority to the current first-line therapy. And proved beta-blockers don’t really work!

2. Mavacamten (Camzyos) by $BMY missed both co-primary endpoints in nHCM (ODYSSEY)

BMY’s ODYSSEY trial in non-obstructive HCM (nHCM) missed both co-primary endpoints. On the surface, that reads as a categorical failure for the class in nHCM.

The reality is more nuanced.

The trial narrowly missed statistical significance on both endpoints:

• KCCQ-CSS: p = 0.06

• Peak VO₂: p = 0.07

These weren’t big misses. Directionally, the data favored mavacamten.

Post-hoc analyses presented suggest that exposure-response relationships and dosing constraints played a role. ODYSSEY had treatment interruptions and dose limitations tied to LVEF thresholds and pharmacokinetic considerations. In nHCM, without a LVOT gradient to guide dosing results in titration being a safety-balancing act. I believe this limited effective exposure in a meaningful subset of patients.

My takeaway is not that cardiac myosin inhibition fails in nHCM. Rather, ODYSSEY suggests the mechanism has activity, but dosing flexibility will matter more in this population. Aficamten’s shorter half-life and more flexible titration framework will give it a cleaner shot on goal to allow for better exposure.

I’ll go deeper into ACACIA and probability of success in the next note; but I think it’s higher than people might give it credit for.

The Label & Usage

First I’ll go through the differences in the two labels and usage, and it should help paint a picture of where aficamten stands out.

Myqorzo (Aficamten) (CYTK)

Camzyos (Mavacamten) (BMY)

Aficamten was approved by the FDA on December 26 (yay!), for oHCM. Camzyos the current incumbent is approved for NYHA Class II–III symptomatic oHCM. Myqorzo’s indication does not restrict by NYHA class, making it broader on paper. Whether payers apply the same breadth in practice will remain to be seen, but from a regulatory standpoint the labeled population is wider.

Both agents carry boxed warnings for systolic dysfunction and require REMS enrollment with serial echocardiographic (echo) monitoring.

Echo Monitoring Requirement (Mandatory):

Camzyos

Baseline Echo > Week 4 > Week 8 > Week 12 (4 Mandatory)

Myqorzo

Baseline Echo > Week 4 > Week 12 (3 Mandatory in Smooth Titration)

Week 8 (only if dose adjustment occurs)

Strictly by label, there’s one fewer required echo during initiation.

A second echo difference shows up in interruption mechanics; Camzyos requires mandatory interruption if LVEF falls below 50%, which can trigger re-baselining and retitration cycles (and therefore echos). Myqorzo allows dose reduction in the 40–50% range and mandates interruption only below 40% or with heart failure symptoms. Fewer forced stop-start cycles could in theory translate to c2 fewer early echos in practice depending on workflow.

LVEF Thresholds (Operational Impact)

LVEF threshold = left ventricular ejection fraction cutoff at which therapy must be dose-reduced or interrupted to prevent systolic dysfunction.

Camzyos

LVEF <50% > Mandatory interruption > Retitration cycle

Myqorzo

LVEF 40–50% > Dose reduction

Interrupt only if <40% or HF symptoms

With Camzyos, if LVEF drops below 50% doses must be interrupted, re-imaged, and re-escalated. With Myqorzo, physicians can adjust dose and continue therapy unless EF falls below 40%.

DDI Monitoring

DDI monitoring = managing dose and safety when drug–drug interactions (typically via CYP metabolism) alter exposure. Think when you take lots of drugs at once. (Example DDI: Don’t mix Tylenol with alcohol!)

Myqorzo

Primarily metabolized via CYP3A4

Avoid or adjust with strong CYP3A4 inhibitors/inducers

LVEF reassessment recommended 2–8 weeks after certain interacting drugs are started or stopped

Camzyos

Metabolized via CYP2C19 and CYP3A4

Multiple formal contraindications (moderate/strong CYP2C19 inhibitors; strong CYP3A4 inhibitors/inducers)

Detailed DDI adjustment algorithms embedded in prescribing guidance

In plain English, both drugs will still require monitoring. But camzyos carry broader metabolic constraints and a more complex interaction framework, which adds a lot of complexity for prescribing friction.

In a cardiology population frequently on polypharmacy, fewer metabolic variables simplify prescribing logistics and specialty pharmacy coordination

oHCM Opportunity is Validated by Camzyos’s Launch

My numbers have changed since my last write-up, largely because of the added data from Camzyos’s launch. Camzyos is annualizing >$1B, and consensus expectsc$1.3B in 2026 and c$1.6B+ in 2027, at only c15–18% penetration in oHCM (the market is nowhere close to saturation). Similarly to transthyretin (TTR) cardiomyopathy, we’re seeing an expansion of TAM now that efficacious pharmacologic options are available (as proven by MAPLE, beta-blockers don’t really work). Couple this with increasing diagnosis rates, and the CMI class itself expanding.

Cardiovascular launches are typically slower and more linear than oncology. Luckily aficamten isn’t launching into a cold market, camzyos has laid a lot of the education groundwork around the MoA, and demonstrated demand. Beta blockers still remain first-line despite weak evidence on exercise capacity, and MAPLE should lay the groundwork for 1L use, and myosin inhibitors taking market majority.

Initially I’d expect aficamten’s opportunity primarily in new patient starts, rather than large-scale switching. I would expect only a minority of stable Camzyos patients to switch (<30%), given inertia and clinical stability.

In lower-resource and community settings, aficamten’s operational simplicity should make it the more straightforward first choice. But I do not see a structural reason why that advantage would be confined to those settings. Even in large academic centers where Camzyos is entrenched, physicians still care about workflow efficiency and treatment continuity. If aficamten proves smoother to manage in practice, it should compete effectively for new starts across both community and major HCM centers.

A fairly basic revenue model just to guesstimate the peak sales. 2032 2.5x peak sales you’ll get around 115m fully diluted EV, so around $80 a share for the oHCM franchise alone? Play with that how you like.

It’s early for aficamten, now that the drug has launched, and management expect that >50% of CMI naive patients will be on aficamten by YE. Let’s see how sales are for Q1 in May.

I think aficamten will be a market leader. Add on the upcoming ACACIA and a success and you have a $5b/year franchise.

PART 1 RECAP (THIS POST).

PART 2 ON ACACIA UPCOMING.

Anyhow, this post I’ll spend a bit of time just recapping and going over a couple new names. I’m planning on next article to go into PoS for CYTK’s ACACIA so I’ll be abit more brief on CYTK’s side. It’s very very catalyst rich over the next couple months. Also trying a new structure out to see if this makes it clearer to read.

By reading this you accept you have read and acknowledged the disclaimer.

Read more

Date: [05/01/2025]

Recommendation: [Long]

By reading this you accept you have read and acknowledged the disclaimer.

This will cover everything that I feel is relevant so far to anyone wanting to understand the opportunity around Grace so it will be long. If you don’t want to read everything, there is a shorter collection before the full writeup. Key points on the PDUFA and bear arguments are addressed near the bottom of the writeup after introducing the disease and need for Grace’s solution. I have adjusted the PT since my last writeup, and fixed discrepencies in previous writeups.

Summary

Grace Therapeutics (Grace) is a microcap biotech company focusing on the treatment of rare diseases, trading well under its assets’ fair value. The lead program is GTx-104, and although there are two others (GTX-102 & GTX-101), GTx-104 is the company’s most important drug asset, and Grace should be evaluated as a single-asset company.

GTx-104 is a new intravenous (IV) formulation of a previously approved drug nimodipine which is used to treat aneurysmal subarachnoid hemorrhage (aSAH). Contextually nimodipine is the only FDA-approved medication able to improve aSAH patient outcomes and is de-facto mandated for ALL PATIENTS suspected to have aSAH (Class I recommendation with level A evidence per AHA/ASA).

Problematically, despite nimodipine being the gold standard for treatment for >40 years, the oral form possesses numerous issues such as low absorption into the body, variable levels in the blood, and difficulties in being administered to critically ill patients. GTx-104 solves these issues by presenting an IV formulation, allowing for 100% bioavailability. GTx-104 has been granted Orphan Drug Designation by the FDA, providing 7 years of marketing exclusivity post-launch in the US.

In its recent Phase 3 study GTx-104 demonstrated significant reductions in clinically significant hypotension, improved dose intensity, and better overall patient outcomes compared to the standard oral nimodipine treatment, hitting all primary endpoints. At the same time, Grace secured additional funding through a PIPE, bringing in c$15.8 million, increasing fully diluted share count (if 100% exercised) from 14.3m to 22.2m. The cash runway is expected to last through the regulatory milestones and until cash flow positivity.

Grace possesses a derisked asset that will bring the first innovation to the disease in >40 years, and is extremely likely (PoS >90%) to pass its PDUFA and launch in H2’26 into a market that has already demonstrated a strong willingness to use and pay at premium for ease-of-use products.

Trading at current levels, Grace is a >3x multibagger opportunity.

Timelines:

Feb 10 2025: GTx-104 reported data on its STRIVE-ON Phase 3 trial

Jun 25 2025: NDA Submission to the FDA

Aug 27 2025: FDA accepts NDA Submission

Apr 23 2026: PDUFA decision target date

Previous Writeups:
Nov 2024: Comprehensive (Pre Ph3 Trial) - Same as this but older (and worse)

Jun 2025: Update (Post Ph3 and Pre NDA Submission) - Ph3 Details (I have better thoughts here at the bottom on the Phase 3)

How we get to my $11.10 PT (The Thesis in Short)

For people who don’t want to read an essay, see this section.

1. GTx-104 makes sense as an innovation to standard of care (SoC)

• First meaningful advance in aSAH vasospasm management in >40 years; IV reformulation of nimodipine, a proven drug with known efficacy, which makes it an ‘easier’ approval than drugs that hold a new MoA.

• Reduces clinically significant hypotension (the goal of the drug) vs oral SoC: 19% relative reduction in Phase 3.

• Material dosing advantage:

  • 54% of GTx-104 patients received >95% of planned dose vs 8% with oral nimodipine in the Phase 3 trial.

  • Dose control matters in acute brain injury management, with oral absorption being unreliable due to gastric emptying, ileus, and interruptions.

• Operational advantages:

  • No waking patients q4h, fewer nursing interventions.

  • Usable in dysphagic patients; avoids NG tube variability.

  • IV is operationally easier to administer.

• Phase 3 signals fewer ICU days, which alone could justify a meaningful price premium (ICU day cost >> drug cost).

• I hold a 90% success of the FDA approving the drug in April.

2. Sales penetration is achievable, and is compelling for hospitals and reimbursers

• The market is highly concentrated, with c250 hospitals treating c50% of all aSAH cases, a concentrated sales force can achieve early and fast penetration. Grace targets 15-20 sales reps.

• In STRIVE-ON, 30 hospitals were enrolled, with >100 expressing interest; strong early pull through is already being shown + the AHA/ASA are already preparing to include IV management guidelines. Neurologists I communicate with, agree with my own view that GTx-104 will be widely adopted.

• The reimbursement backdrop is incredibly favourable as well:

  • aSAH is among the most expensive diagnosis related groups (avg cost >$500k per case)

  • Oral nimodipine is cheap ($50-100/day), but the market already tolerates premiums, with Nymalize being priced at c8-10x capsules and possessing a 40% penetration, despite the aforementioned GI side effects and limitations.

• You get optionality in the TAM!

  • Literature based claims (which I still use in my PT) hint at around aSAH 40,000 patients.

  • Claims data (ClearView) indicate that there are actually c70k eligible survivors annually.

  • And of this, the primary market of dysphagic patients 50-70% would make the IV appropriate population, where this would hold the first line closer to c40k versus the 20-30k commonly cited.

3. Financing, cash, and the backers work well

• The company has passed its dilution overhang: through conversations with management and my own models, the current cash is enough to get the company to cash flow positive.

• And the price you’re paying right now is not that much higher than the institutionals (inc. ADAR1) who partipated in a PIPE @$3.395.

4. Add This Up:

I take conservative assumptions that GTx-104 will match worse peers’ penetration at 45%, whilst priced at the same levels (despite the ability to price well above this) in a 40k aSAH patient pool.

Base Case Price Target: $11.10

At the base case PT, Grace would carry an equity value of approximately $313M, implying an EV of c$280–285M after adjusting for net cash. This compares to base-case peak U.S. sales of c$125M, or roughly 2.3× peak sales, which is conservative for a hospital-based, high-margin product with regulatory exclusivity and limited commercial risk.

At the current price of $3.66, Grace trades at a fully diluted market cap of approximately $103M and an enterprise value of c$65–70M. The market is discounting approval and execution risk far more heavily than the underlying data and balance sheet justify. The stock is trading at half of risk-adjusted peak sales, despite a funded runway through cash-flow breakeven and a single, well-de-risked asset.

In this context the upside is a c3x from current levels, driven by fundamental normalisation. Given the limited remaining development risk, strong regulatory precedent, clean financial runway, and incredibly strong feedback from potential adopters, you have an extremely asymmetric risk-reward profile.

This is before accounting for upside from an underestimated patient population and what I model as a more conservative pricing than is likely ($600-$1000/day range potential) which would introduce a far greater upside than one I present, to well over 5x (see table).

Grace is incredibly undervalued.

For people who want to get into the work, enjoy. If you’ve read the original it will rhyme a lot. See below:

Background to Aneurysmal Subarachnoid Hemorrhage (aSAH)

What is aSAH?

aSAH is a life-threatening type of stroke that occurs when a bulge in a blood vessel (aneurysm) bursts; either spontaneously or secondary to a head injury. It results in bleeding in the area between the brain and its protective membranes, known as the subarachnoid space; the burst creates a rush of blood that can increase pressure in the brain, damage surrounding tissue, and lead to severe complications. To help visualise the impact, imagine popping a balloon inside a carefully maintained terrarium.

aSAH key aspects:

• Ruptured aneurysms: Weak spots in blood vessels are the main cause of aSAH.

• Vasospasm: Blood from the rupture irritates nearby vessels, causing spasms and narrowings, restricting blood flow, causing further damage.

• Delayed cerebral ischemia (DCI): This complication reduces oxygen supply to the brain; potentially leading to long-term deficits.

Graphic from Physiopedia

Who does aSAH Affect and what’s the Impact?

aSAH most commonly affects patients in their 50s and 60s, with women at modestly higher risk than men. U.S. incidence is typically cited at c20,000–50,000 cases annually based on epidemiologic studies (with higher figures of c70,000 suggested by claims-based analyses (Clearview)). However, despite modern neurocritical care, outcomes remain poor:

• Mortality: Prehospital mortality rates have been reported to be 22-26%, with hospital inpatient rates around 13-14% in the US (19-20% globally).

• Long-term disability: A third of survivors face permanent disabilities, like memory issues or movement difficulties. Most deficits tend to improve, but c50% of patients tend to suffer from cognitive impairments for up to a year.

• Cost: aSAH care is among the most expensive DRG categories in the US. Average total per-patient costs exceed $500,000, reflecting high ICU utilization, prolonged length of stay, and frequent complications. Acute hospitalization alone averages c$370,000 per case, rising to c$530,000 in patients who develop delayed cerebral ischemia (DCI), before accounting for long-term rehabilitation, outpatient care, or lost productivity.

aSAH represents a high-mortality, high-cost condition. Of the c70% of patients who survive, outcomes bifurcate: roughly half experience persistent functional & neurological limitations, whilst the remainder achieve a more favorable functional recovery.

How is aSAH Treated?

Surviving aSAH requires quick and effective treatment typically involving three steps:

• Securing the aneurysm: The ruptured aneurysm is treated either via microsurgical clipping or endovascular coiling to prevent rebleeding.

• Managing brain pressure: Elevated intracranial pressure is addressed through measures such as external ventricular drainage to relieve hydrocephalus and reduce secondary brain injury.

• Preventing secondary complications: This is where nimodipine plays a critical role, administered to reduce the risk of cerebral vasospasm and delayed cerebral ischemia (DCI), preserving cerebral blood flow and improving neurological outcomes.

aSAH is Bad…

Despite a well-established treatment paradigm, aSAH remains the deadliest form of stroke, with persistently high mortality and morbidity driven by several well-characterized complications:

• Rebleed risks: Ruptured aneurysms can rebleed, typically within the first 24h if the aneurysm isn’t secured, with an attached 70% mortality rate, often with severe neurological injury among survivors.

• Hydrocephalus: After ruptures, blood can block normal flows of cerebrospinal fluids (CSF), resulting in a buildup of fluids in the brain and an increase in intracranial pressures (hydrocephalus; 20-30% of cases). This pressure if untreated can deteriorate conditions, and result in death.

• Vasospasm and Delayed Cerebral Ischemia (DCI): Approximately 60-70% of aSAH patients develop cerebral vasospasm, and c40% of these progress to DCI. Reduced cerebral perfusion and oxygen delivery are strongly associated with infarction, long-term neurological deficits, and increased mortality.

• Secondary Brain swellings and inflammation: The presence of blood from the initial aneurysm also triggers an inflammatory response, further complicating treatment with brain swelling and tissue damage.

All in all, it’s fairly bad, and 30 days after treatment, from the effects above, almost 80% of patients have either died or sustained neurological injuries.

The Drug Nimodipine (Why?)

What is Nimodipine and How does it Work?
Nimodipine is a type of drug known as a calcium channel blocker, a class of drugs designed to relax blood vessels and improve blood flow (vasoconstriction). Under normal conditions, calcium entry into these cells promotes constriction / tightening the blood vessels. After an aSAH, this process can become dysregulated, resulting in vasospasm; pathologic constriction of cerebral arteries that reduces blood flow and oxygen delivery to brain tissue.

By blocking calcium entry into vascular smooth muscle, nimodipine regulates this vasoconstrictive signaling, promotes cerebral vasodilation, and helps preserve cerebral blood flow. In doing so, it reduces the risk of vasospasm-related ischemia and downstream neuronal injury.

In practical terms, nimodipine helps prevent secondary injury in a brain already compromised by hemorrhage, serving as a safeguard against more delayed neurological deterioration.

Usage in aSAH Treatment
Nimodipine is the only FDA-approved drug specifically indicated for improving outcomes in aSAH patients. Its role is so critical that its use is recommended for every single patient suspected of aSAH by the American Heart Association (AHA) and American Stroke Association (ASA). This guidance is based on the strongest level of evidence available (Class I recommendation with Level A evidence).

So why is nimodipine so special, surely there’s a range of alternative calcium blockers that can be used? Nimodipine is unique in its ability to target blood vessels in the brain. Unlike most other calcium channel blockers, which act systemically, nimodipine is sufficiently lipophilic to cross the blood–brain barrier; a protective filter that regulates what substances can enter the brain from the bloodstream.

The drug is typically administered within 96 hours of the hemorrhage and continues for 21 days, with 60mg (two 30mg capsules) to be delivered orally every 4 hours. The primary purpose is to prevent vasospasm and the resulting DCI, 60% and 40% incidence respectively in aSAH patients. This administration is aimed to start at least within 96 hours of aSAH, and in the case of the patient being unconscious, the typical suggestion is that the capsule contents should be extracted and delivered by a nasogastric tube, and rinsed down the tube with a saline solution.

The Evidence
Nimodipine has effectively been rubber-stamped as the gold standard for aSAH treatment for over three decades now, but what’s the evidence behind this reasoning? Nimodipine was first approved in 1988 following clinical trials in the US, France, and the UK.

See below the FDA label data, showing the results of multiple randomised, double blind, placebo-controlled trials (i.e. nimodipine versus no drug), with nimodipine typically administered every 4 hours for 21 days: (Note Hunt and Hess grades, I-V is the range, I being the least severe patients, V being the most severe cases)

US Study (1983): Severe deficits due to vasospasm significantly reduced (1 vs 8)

French Study (1984): Significant reduction in all spasm-related deficits and severe deficits (2 vs 10)

UK Study (1989): Grades I-III. Nimodipine significantly improved good recovery (199 vs 169) and reduced severe disability (12 vs 31). Mortality trends favoured nimodipine (43 vs. 60)

Canadian Study (1984): Grades III-V (Severe cases). Nimodipine (90mg) reduced spasm-related delayed ischemic deficits (11% vs. 31%).

Pooled Analysis (1984 & 1992): Grades IV-V (Severe cases). Combined data from UK and Canadian severe trials. Nimodipine improved good recovery (25.3% vs. 10.9%) and reduced severe disability and vegetative survival.

And a meta-analysis result sorted based on GRADE certainty confirming nimodipine as the standard of care, and as the drug that has displayed the best outcomes in improving neurological function for aSAH patients.

A Problem with Oral Nimodipine Currently

1. Poor Absorption and Variability

By far the largest problem is nimodipine has poor absorption and variable blood levels. When you take an oral medication, not all of it makes it into your bloodstream. Oral nimodipine’s bioavailability (a measure of how much of the drug actually reaches your system and is available to do its job) sits at around 7-13%. This is very low.

This matters: it presents a large inefficiency when you have a 100mg pill but only around 10mg is actually doing the work (and even this is an estimate). Then also imagine that you want to control dosages when you’re doing guess work when administering the pill, hoping that you get a stable level of the drug absorbed and that levels won’t be low or fluctuate.

Bioavailability sits so low because the drug has to pass through the stomach and liver, where much of the drug is broken down before it even enters the bloodstream (first-pass metabolism).

Even when nimodipine gets into the bloodstream, its levels tend to fluctuate. This is because the oral formulation breaks down quickly, and the body absorbs it inconsistently.

So why does this fluctuation actually matter? Imagine trying to maintain a steady level of oxygen in a tropical fish tank, but the supply keeps going up and down unpredictably. You’ll find it incredibly difficult to keep the fish alive. For nimodipine, when levels are too low, the drug is unable to effectively control the problem of vasospasm, and when levels are too high it can cause dangerous drops in blood pressure (hypotension). In the setting of acute brain injury where cerebral perfusion is already fragile, either extreme can meaningfully worsen outcomes.

2. Difficulties in Administration

Many aSAH patients are in a critical state (around half) - unconscious, intubated, semi-conscious, or mechanically ventilated. Swallowing pills is not an option for these patients, to work around this healthcare providers have to do this in a pretty crude and unmeasured fashion. They extract from the capsules using a needle the nimodipine and deliver the drug through a tube inserted into the stomach (nasogastric tube). The alternative is crushing the capsule which is equally pretty crude.

These methods are inherently imprecise. Drug loss during preparation is common, liquids can adhere to tubing, and incomplete delivery is frequent. As a result, patients may receive inconsistent or unintended doses (compounding the already significant variability in oral absorption). The process is also time consuming and there’s a risk of drug loss during preparation.

3. Altered Physiology in Critical Patients

Administration challenges are compounded by the altered physiology of critically ill aSAH patients. Reduced consciousness, autonomic dysfunction, sedation, and neurogenic injury can significantly impair gastrointestinal function.

This can result in delayed gastric emptying, which results in the drug taking longer to reach the intestines (where it is absorbed) resulting in inconsistent dosing timings, reduced absorption.

4. Side Effects

Although nimodipine is generally well tolerated, unstable plasma concentrations increase the risk of adverse events. The most clinically significant is hypotension (4.4%). Other side effects include edema, headache and nausea (1%).

5. Quality of Life

Nimodipine must be administered every four hours for 21 days, creating a substantial logistical burden for hospitals and care teams. Frequent dosing increases nursing workload and complicates care coordination in already resource-intensive neuro-ICU settings.

For patients who survive the acute phase and transition out of the ICU, adherence becomes a challenge. Maintaining a strict q4h dosing schedule after discharge is difficult and increases the risk of missed or delayed doses.

To put this also into some extra context (very anecdotal), but the patients that aren’t unresponsive come out of the hospital feeling absolutely brutalised mentally. Imagine being a patient who just suffered a hemorrhage in the brain, you’re feeling pretty battered and on drugs already.

For a minimum of 2 weeks you have a pretty horrific regime. Patients have hourly neurological assessments to check for early signs of vasospasm and changes in motor function. They are then kept euvolemic (properly hydrated) to maintain blood flow to the brain (can be up to 250mL/h). This results in frequent urination. Add to this now nimodipine every four hours administered for 21 days.

This is a borderline Russian sleep experiment, on people with brain damage. It’s an understatement to say they come out completely mentally destroyed and delirious, not counting any complications.

The Alternatives to Oral Nimodipine Currently?

There’s a lot of problems, and hospitals really shouldn’t be performing guess work with a life saving drug on patients that possess a condition with an extremely high mortality %, and the quality of life of both hospital staff and patients should be increased as much as possible. You’ve already suffered fatalities from misinjection of nimodipine into the IV tube rather than the nasogastric tube.

So a couple of other formulations have been developed.

1. Nimotop Intravenous Infusion (Europe Only)

Bayer’s Nimotop is an IV formulation of nimodipine available in Europe. Designed for patients who cannot tolerate oral administration. However, the formulation includes solvents such as ethanol (24%) and polyethylene glycol (PEG) (17%) which has resulted in relatively low uptake in the EU, and the FDA preventing its approval.

Complications with this ethanol content can cause complications especially in patients with liver or kidney impairments, alcohol sensitive women, and pregnant women. This ethanol exposure can also result in central nervous system depression and cardiac arrhythmias, whilst PEG can cause lactic acidosis especially in patients with impaired kidney function, as well as infusion-related toxicities.

2. Nymalize Oral Solution

Nymalize is an oral solution of nimodipine approved in the USA (2013). Its approval was based on a single Phase 1 pharmacokinetic (PK) bridging study, conducted in the aftermath of fatal medication errors involving inadvertent IV administration of oral nimodipine formulations. The regulatory requirement was limited: Nymalize was just required to demonstrate that it delivered nimodipine into systemic circulation at exposure levels comparable to the existing capsule formulation.

It’s supplied as a ready-to-use liquid formulation in the form of pre-filled oral syringes and bottles. They are designed to remove the problem of extracting nimodipine from the capsules to allow easier dosing in patients with swallowing difficulties. These syringes are incompatible with IV tubes to prevent the previously aforementioned fatalities.

However, while Nymalize improves administration safety and convenience, it does not address the core pharmacokinetic limitations of oral nimodipine. The drug remains subject to extensive first-pass metabolism and continues to exhibit highly variable systemic exposure. As a result, Nymalize does not meaningfully improve bioavailability or enable more precise control of therapeutic drug levels; it functions primarily as a safety enhancement and workflow improvement rather than a pharmacologic advance. Anecdotally, as well diarrhea is a common side effect…

Despite these limitations, Nymalize is priced at approximately $500 per day, compared with $25–50 per day for generic nimodipine capsules, showing payer willingness to pay a substantial premium for safety and ease of use, without resolving fundamental exposure and variability issues.

3. NicaPlant (Ph2 Completed) (& a quick discussion on Nicardipine)

NicaPlant (BIT Pharma) is a localized, modified-release nicardipine implant placed intraoperatively around the basal cerebral arteries during microsurgical aneurysm clipping. In an initial Phase 2a study (n=14), the implant was safe, achieved therapeutic cerebrospinal fluid (CSF) nicardipine concentrations without meaningful systemic exposure, and was associated with lower rates of moderate–severe angiographic vasospasm (c10% vs c50% in controls). These findings were supported by a 2024 Phase 2b randomized trial (n=41), which demonstrated significantly reduced rates of moderate–severe vasospasm (20% vs 58%), lower need for rescue therapy (10% vs 58%), fewer new cerebral infarcts (10% vs 32%), and numerically higher favorable functional outcomes at 52 weeks (84% vs 67%, not stat sig), with comparable adverse event rates.

However, NicaPlant’s applicability is structurally constrained: it can only be used in patients undergoing open surgical clipping, which today accounts for roughly 1/3 of the aSAH population. Even within this subset, adoption would likely be well below 100% due to surgeon training requirements, anatomical and exposure limitations, added operative complexity, and uncertainty pending Phase 3 outcomes and reimbursement clarity. Importantly, even in treated patients, NicaPlant would not obviate the need for systemic nimodipine, which remains guideline-mandated.

While Phase 3 discussions with regulators have been reported, no public trial start date has been disclosed.

Nicardipine

I’ve also seen IV nicardipine being discussed as an option, and it feels apt to tackle this whilst on NicaPlant. Nicardipine does not compete with nimodipine’s role in aSAH care. As highlighted by the meta-analysis referenced above, nimodipine remains the only therapy shown to improve neurological outcomes and reduce delayed cerebral ischemia (DCI). Intravenous nicardipine is used primarily for blood-pressure control in the neuro-ICU and is not indicated for DCI prevention, while intrathecal or local delivery of nicardipine (or nimodipine) remains investigational and non-standard.

Some High Level DD from Reddit

Jokes aside. My own informal discussions with practicing neurologists have been largely consistent and have informed my own view: if approved, IV nimodipine would be adopted in routine practice, and there is a clear demand for an IV formulation in aSAH management.

GTX-104, Grace’s Solution

The clinical picture is clear: nimodipine remains the only approved therapy for aSAH, yet its oral formulation is poorly suited to the realities of critical care. Existing alternatives address administration safety or niche use cases but do not resolve the fundamental limitations of oral dosing, namely variable exposure and unreliable delivery. Also, there are currently no currently progressing late-stage programs pursuing novel pharmacologic therapies to replace SoC in aSAH, and systemic alternative compounds evaluated to date have failed to demonstrate superiority or even parity with nimodipine in improving outcomes.

Grace Therapeutics Introduction and Brief History

First, an introduction to Grace. In October 2024, the company was formerly known as Acasti Pharma, Acasti Pharma having acquired Grace Therapeutics in August 2021, a month before Grace launched its pivotal PK study for GTX-104 in September 2021. Grace prior to this had spent a year developing GTX-104 to search for an optimal formulation for nimodipine (which is highly insoluble). Grace had completed Ph1 studies, and required extra capital, whilst Acasti’s lead programme was gone, and held $60m of cash on its balance sheet. Match made in heaven?

GTx-104 completed its pivotal PK bridging study in December 2022, demonstrating bioequivalent exposure to oral nimodipine with meaningfully more consistent plasma levels. In April 2023, Grace finalized the Phase 3 STRIVE-ON trial design with the FDA under a streamlined 505(b)(2) regulatory pathway, with safety as the primary endpoint leveraging nimodipine’s established efficacy. Patient enrollment began in October 2023 and completed in Fall 2024, ahead of typical industry timelines. Topline Phase 3 results were reported on February 10, 2025, supporting the safety and tolerability of IV nimodipine. The company submitted a New Drug Application (NDA) in June 2025, which the FDA accepted for review on August 22, 2025, assigning a PDUFA target date of April 23, 2026.

Alongside clinical execution, Grace undertook several corporate actions to support U.S. commercialization and regulatory engagement. The company delisted from the TSX Venture Exchange in March 2023, redomiciled to the United States, and relisted on Nasdaq. It raised $7.3 million in a private placement in September 2023, appointed new CEO Prashant Kohli and CFO Robert J. DelAversano, and simultaneously with STRIVE-ON topline results announced a PIPE financing of up to $30 million in February 2025 to support the NDA review and pre-launch activities.

In addition to GTx-104, Grace has disclosed two earlier-stage programs, GTx-101 and GTx-102. Both programs remain paused, with resources prioritized toward regulatory approval and potential commercialization of GTx-104. The company has indicated that advancement of these assets is contingent on the outcome of the GTx-104 NDA and broader capital allocation decisions post-PDUFA.

GTX-104 Introduction

GTX-104 is a room temperature stable IV formulation of nimodipine, designed to overcome the limitations of oral nimodipine. It uses micelle-based technology, these being nano-sized structures formed by molecules with hydrophilic (water-attracting) heads and hydrophobic (water-repelling) tails. Nimodipine, being a hydrophobic molecule, is then able to be encapsulated by these micelles in the core, allowing it to dissolve in an aqueous environment ready for IV administration.

Grace Therapeutics

GTx-104 does not alter nimodipine’s mechanism of action or clinical role; rather, it optimises delivery. This allows GTX-104 to deliver nimodipine in a stable, solvent-free IV form, ensuring consistent drug levels, improved safety, and greater efficacy for ill patients. Oral nimodipine suffers from poor absorption, high variability in blood concentration, and administration challenges. An IV delivery resolves these issues by bypassing the gastrointestinal system and directly delivering the drug to the bloodstream.

Data before the Phase 3 Results

GTX-104 Pharmacokinetic Trials

Grace conducted two pivotal pharmacokinetic (PK) trials for GTX-104. PK studies aim to investigate how a drug moves through the body, including how it is absorbed, distributed, metabolised, and excreted. Here the goal was to compare how the drug behaves in the body - and to demonstrate that the IV formulation delivered nimodipine to the bloodstream in a manner that was comparable (bioequivalent) to the oral capsules whilst solving the absorption and variability issues of the oral form.

PK Studies:

1. GTX-104-001 (Undisclosed Date / Prior to GTX-104-002):

• Design: A 4-part, single-center, randomized, safety dose-escalation and crossover study.

• Participants: 104 healthy male and female volunteers.

• Objective: Evaluate the safety and PK profile of GTX-104 at different dose levels.

• Details: The study tested escalating doses of GTX-104 to determine the optimal dose with minimal side effects and consistent drug levels.

2. GTX-104-002 (18 May 2022):

• Design: A single-center, randomized, two-period crossover study.

• Participants: 58 healthy male and female volunteers.

• Objective: Compare the bioavailability and safety of GTX-104 at steady state with that of oral nimodipine capsules.

• Details: Participants alternated between GTX-104 (IV) and oral nimodipine:

  • GTX-104: A 30-minute infusion of 4 mg every four hours combined with a continuous infusion of 0.15 mg/hour for 72 hours.

  • Oral Nimodipine: 60 mg capsules every four hours for 72 hours.

The results from the PK studies were extremely positive, summarised:

1. Improved Bioavailability:

   • GTX-104 achieved 100% bioavailability, meaning the full dose was delivered directly into the bloodstream.

   • In contrast, oral nimodipine showed 7.2% bioavailability.

2. Reduced Blood-Level Variability:

   • GTX-104 provided consistent plasma concentrations of nimodipine, with 20% fluctuations around the mean concentration over a 24-hour period.

   • Oral nimodipine exhibited extreme variability, with peak concentrations 3x the mean and trough concentrations dropping to 50% of the mean.

3. Peak (Cmax) and Overall Exposure (AUC):

   • GTX-104 demonstrated a Cmax (maximum plasma concentration) of 24.8 ng/mL, comparable to oral nimodipine’s 25.2 ng/mL.

   • For AUC (area under the curve, measuring overall drug exposure), GTX-104 achieved equivalence to oral nimodipine within the FDA-required range of 80-125% confidence interval:

     • GTX-104: 239.1 ng·h/mL

     • Oral nimodipine: 235.5 ng·h/mL

4. Safety and Hypotension:

   • GTX-104 showed fewer episodes of hypotension (clinically significant drops in blood pressure) compared to oral nimodipine:

     • GTX-104 8.5%, oral nimodipine 12.3%.

   • No serious adverse events were reported in the GTX-104 group.

5. Faster Onset and Predictability:

   • GTX-104 reached therapeutic levels within 30 minutes of infusion.

   • Oral nimodipine showed delayed and unpredictable absorption, particularly in scenarios involving delayed gastric emptying.

To illustrate this with a graph, displaying the mean concentration (ng/mL) for GTX-104 vs oral nimodipine.

The PK studies demonstrated that GTX-104 solved the absorption inefficiencies of oral capsules (100% bioavailability), provided consistent and predictable blood levels, removing the problem of variability, and lowered the risk of adverse events (hypotension).

The STRIVE-ON Phase 3 Results; The Good, The Criticisms, My Take

Safety and ToleRability of GTX-104 (Nimodipine Injection for IV Infusion) ComparEd with Oral Nimodipine in Patients with aSAH (STRIVE-ON) was a critical Ph3 clinical study designed to evaluate the safety and tolerability of GTX-104 in patients with aSAH, building on the successful and positive PK data from earlier trials, with the aim to confirm its safety profile versus oral nimodipine in a real world hospital setting. Topline results were released on Feb 10 2025. The NDA was submitted April 9, 2025, accepted without issue, and assigned a PDUFA date of April 23, 2026.

The Study Design:

Participants:

• 100 patients diagnosed with aSAH enrolled at 25 U.S. hospitals (as a note these were high volume aSAH hospitals. 70% of SAH patients are treated in 370 hospitals).

• Patients randomised in a 1:1 ratio to receive either GTX-104 or oral nimodipine.

Treatments:

• GTX-104 Group:

  • Continuous IV infusion of 0.15 mg/hour.

  • A 30-minute IV bolus of 4 mg every 4 hours.

• Oral Nimodipine Group:

  • Two 30 mg capsules (60 mg total) every 4 hours.

Duration:

• Both groups received treatment for up to 21 days or until discharge.

• Follow-up assessments occurred at 30 days (+/- 7 days) and 90 days (+/- 10 days).

Primary Endpoint:

• Incidence of clinically significant hypotension, defined as:

  • A drop in systolic blood pressure by more than 20 points or diastolic blood pressure by more than 10 points.

  • Absolute diastolic pressure ≤ 100 mmHg.

Secondary Endpoints:

• Duration and number of hypotension episodes (length and recurrence of blood pressure drops).

• Incidence and severity of adverse events (AEs) (all safety concerns are tracked).

• Incidence of delayed cerebral ischemia (DCI).

• Measures of quality of life, such as hospital discharge disposition and patient independence (Follow ups at 30 and 90 days to assess recovery).

• Suicidal ideation using the Columbia-Suicide Severity Rating Scale (C-SSRS) (Suicidal thoughts of patients).

Patients were included from Hess Grades I to V (the entire range), and aSAH diagnosis confirmed via CT scan and angiography. Patients were excluded if they received <12 doses of oral nimodipine prior to randomisation, and those with cardiovascular events, liver failure, or contraindications to nimodipine.

The primary objective was to confirm that GTx-104 does not increase the frequency or severity of clinically significant hypotension compared with oral nimodipine. It was designed to be a safety-focused assessment consistent with the 505(b)(2) regulatory pathway, given that nimodipine’s efficacy in aSAH is already well established.

STRIVE-ON Results

STRIVE-ON met its primary endpoint, demonstrating that treatment with IV nimodipine (GTx-104) did not increase the incidence of clinically significant hypotension compared with oral nimodipine. Clinically significant hypotension occurred in 28% of patients in the GTx-104 arm versus 35% in the oral nimodipine arm, corresponding to a 19% relative reduction.

Treatment with GTx-104 resulted in substantially improved dose delivery. Fifty-four percent of patients receiving GTx-104 achieved ≥95% of the planned cumulative dose, compared with 8% of patients treated with oral nimodipine, reflecting greater consistency of drug administration with IV delivery.

Functional outcome measures favored GTx-104 at 90 days. Good neurological outcomes (mRS 0–2) were observed in 74% of patients in the GTx-104 arm versus 48% in the oral nimodipine arm. Health-related quality-of-life assessments were directionally consistent, with higher overall scores and improvements in mobility, self-care, and daily activities; pain scores were similar between treatment groups.

Healthcare utilization metrics differed between arms. Patients treated with GTx-104 experienced fewer ICU days (approximately 1.5 fewer on average), fewer ventilator days (approximately 5 fewer), and lower ICU readmission rates (4% vs. 8%). Additional measures of care intensity, including duration of ventilation, sedation use, and time spent comatose, were lower in the GTx-104 group.

Mortality was numerically higher in the GTx-104 arm (8 deaths) compared with the oral nimodipine arm (4 deaths). The GTx-104 group included a higher proportion of patients with Hunt–Hess Grade V disease (8% vs. 2%). No deaths were attributed to GTx-104 or to infusion-related hypotension. Overall adverse event rates were similar between treatment groups, and the rate of serious adverse events per patient was identical.

My Interpretation of STRIVE-ON Results

What STRIVE-ON Was - and Was Not - Designed to Prove

STRIVE-ON was designed to answer: ‘is IV nimodipine as safe as oral nimodipine when administered in real-world hospital settings?’. The framing is consistent with a 505(b)(2) development strategy, where the efficacy of nimodipine is already well established, and the regulatory burden for approval here is to ensure that a new formulation does not introduce additional safety risk.

The trial was powered for its primary safety endpoint: clinically significant hypotension. This is the most relevant concern for IV nimodipine. It was not intended to demonstrate superiority to SoC, or to reassess efficacy endpoints such as mortality or functional outcomes.

On the Primary Endpoint, IV Nimodipine Was as Safe as Oral

STRIVE-ON met its primary endpoint, demonstrating that IV nimodipine did not increase the frequency or severity of clinically significant hypotension compared with oral nimodipine. Hypotension occurred numerically less often in the GTx-104 arm (28% vs. 35%).

From a regulatory and clinical standpoint, this directly addresses the primary historical barrier to IV nimodipine. Because nimodipine is a potent vasodilator, IV administration has long raised concern for precipitating systemic hypotension and compromising cerebral perfusion in aSAH patients. STRIVE-ON shows that controlled IV delivery does not exacerbate this risk relative to oral dosing.

This safety context is important when interpreting downstream outcomes. The landmark British Aneurysm Nimodipine Trial established that nimodipine’s clinical benefit is driven primarily by improvements in neurological outcomes, not by large or consistent reductions in mortality. In that trial, poor neurological outcomes were reduced by approximately 40%, largely through reductions in severe disability and vegetative states, whilst mortality was not significantly different between groups (33% placebo vs. 29% nimodipine).

STRIVE-ON follows the same pattern. Although not powered for efficacy, GTx-104 demonstrated higher rates of favourable functional outcomes (mRS 0–2) and improved quality-of-life measures, without a definitive mortality signal. The consistency between historical oral nimodipine data and STRIVE-ON supports the interpretation that IV nimodipine preserves nimodipine’s established benefit profile for improving functional recovery without necessarily altering mortality, while maintaining an acceptable safety margin.

Mortality: Consistent With Natural History (A Blemish on a Good Dataset)

The tug-of-war between bulls and bears lies in the numerical imbalance in mortality, with 8 deaths in the GTx-104 arm vs. 4 deaths in the oral nimodipine arm. This does warrant heightened scrutiny and introduces some doubt into what is otherwise a good result.

Extra context to evaluate the death imbalance:

1. Baseline severity was imbalanced. The GTx-104 arm enrolled a higher proportion of Grade V patients (4v1). Grade V represents the most severe end of the disease spectrum.

2. Absolute mortality (16%) in the GTx-104 arm is consistent with historical expectations. Overall mortality with nimodipine-treated aSAH patients in historical cohorts and on the oral nimodipine label is typically in the 20% range, depending on severity mix. The mortality observed in the GTx-104 arm falls within this range and is consistent with the known natural history of aSAH rather than suggesting excess risk.

3. No deaths were attributed to GTx-104 and all deaths were marked as being due to the severity of the underlying disease. To determine this, the clinical assessors will have considered whether there is reasonable evidence that the investigational drug contributed to death, including known pharmacology and clinical context. (The FDA will still likely conduct their own assessment).

4. STRIVE-ON was not powered to evaluate mortality. With approximately 100 patients total and low event counts, numerical mortality imbalances of this magnitude are well within expected statistical variability, particularly when baseline severity is uneven. Importantly, functional outcomes moved in the favorable direction despite the mortality imbalance, which makes it harder to explain a hidden harm signal.

Notably, death stratification by Hunt–Hess grade has not been publicly disclosed. In the absence of patient-level data, I use natural-history benchmarks to contextualize the results, and ‘generate’ expected deaths from the initial patient mix.

When mortality is modelled using a Poisson framework anchored to the generated ‘expected’ deaths (λ≈6.4 for GTx-104 & λ≈5.7 for oral), the probability of observing a ≥4-death imbalance between arms is roughly 20%. The observed 8 vs 4 split is only about one standard deviation from expectation. In a c100-patient study with low event counts, differences of this magnitude are common and are compatible with random variation rather than a treatment-related mortality signal. This is meant to be illustrative rather than anything solid, given the lack of disclosed patient-level mortality stratification. In safety-focused trials not powered for mortality, FDA evaluates whether deaths suggest a new or formulation-specific risk, rather than whether numerical imbalances exist per se.

It is important to contextualize mortality in aSAH more broadly. Deaths are largely attributed to primary hemorrhage effects (55%), rebleeding (17%), or medical complications (15%). Vasospasm / delayed cerebral ischemia (DCI) is not a primary predictor of mortality. I would also like to take this opportunity to once again highlight that the role of nimodipine is to specifically improve functional outcomes, and not reduce mortality.

A materially different interpretation would be warranted if deaths clustered early after infusion initiation, were temporally associated with hypotensive events (contradicted by the positive topline), or occurred disproportionately in lower-severity (good-grade) patients. Based on the disclosures to date, there is no indication of such patterns, and deaths have been described as attributable to the severity and complications of the underlying disease.

The Scenarios and How I model the PDUFA & Usage

My base case is FDA approval of GTx-104 (90%), with principal uncertainty on labelling tone rather than approvability. STRIVE-ON was appropriately designed for its pathway, met its primary safety endpoint, and addressed historical concerns that have limited IV nimodipine usage (clinically significant hypotension), without investigators identifying a formulation-specific safety signal. Whilst numerical mortality imbalance may introduce FDA scrutiny, I believe it is consistent with disease severity, natural history, and the limitations of a safety-focused trial not powered for mortality.

From a practical standpoint, I view labeling risk as having limited impact on real-world ICU utilisation. In neuro-ICU settings, nimodipine is already administered under continuous hemodynamic monitoring, and treatment decisions are individualized at the unit and patient level, with the label serving as guidance rather than a rigid protocol.

In practice, I expect GTX-104 to most likely be used during the highest-risk initial 7–14 day period following aneurysmal subarachnoid hemorrhage, and as SoC in dysphagic, intubated, or hemodynamically unstable patients, with transition to oral as patients stabilise. In patients who remain intubated or unable to tolerate oral dosing, IV therapy could be continued for the full 21-day course. Neuro-ICU use and protocols will remain likely fairly individualised per unit/patient, with the label informing monitoring and dose-adjustment considerations (particularly around blood pressure) while allowing clinicians discretion based on patient status. At this stage, specific blood-pressure thresholds or dose-modification guidance remains uncertain to me and will depend on final label language.

Importantly, GTx-104 should be an easy sell to hospital formularies and pharmacy & therapeutics committees. The elimination of Q4 oral dosing, often interrupted and administratively burdensome offers a clear nursing workflow benefit, while more reliable drug delivery and potential reductions in ICU length of stay (where costs routinely exceed $10,000 per day) provide a compelling economic rationale for adoption alongside clinical convenience.

My Take on PDUFA Scenarios

Clean Approval (50%)

The FDA approves GTX-104 under the 505(b)(2) pathway with standard safety language largely aligned to oral nimodipine. The label emphasizes IV use in monitored inpatient settings without material restrictions beyond routine neuro-ICU practice. It reflects acceptance that STRIVE-ON met its primary safety endpoint, did not increase hypotension risk, and did not identify a formulation-specific safety signal. Mortality is acknowledged as part of aSAH’s natural history but is not limiting. Broad hospital adoption with minimal commercial friction. Base case peak sales remain, cash will be enough to reach cash flow positivity.

Approval With Restrictive Labeling (40%)

The FDA grants approval but includes more conservative labeling referencing the numerical mortality imbalance and the complexities of IV use in critically ill patients. While causality is not assigned, the label may note that mortality differences were observed and that STRIVE-ON was not powered for survival outcomes. Emphasis on blood pressure monitoring and ICU use would be heightened, but a boxed warning or REMS is unlikely. Commercial adoption remains strong, but more backended with more adoption anxiety, and more dependent on institutional comfort and operational benefits. Base case peak sales remain, may require another dilution round to reach cash flow positivity depending on adoption anxiety.

CRL – Non-Clinical / Administrative (5%)

CMC, formulation, or labeling-clarification issues rather than concerns with the clinical data. Potential requests could involve micelle consistency, IV stability, sterility, or post-marketing surveillance commitments. Another year delay and dilution necessary.

CRL – Clinical (Mortality Concern) (5%)

Clinical CRL driven by mortality would require the FDA to view the imbalance as unresolved by disease severity, natural history, or trial design. This would likely need evidence of a mechanistically plausible safety signal, such as infusion-linked hypotension or excess deaths in lower-severity patients, none of which have been disclosed. Given nimodipine precedent and STRIVE-ON’s design, this remains a low-probability scenario. Would send the stock to EV / negative EV.

So how does Grace Make Money?

At this point in the writeup, it should be clear why I believe GTx-104 represents a meaningful upgrade to the current standard of care in aSAH. It addresses the core limitations of oral nimodipine: unreliable absorption, dose interruptions, and administration challenges, while fitting naturally into existing neuro-ICU workflows.

Assuming a positive PDUFA decision in April 2026, a US commercial launch in Q3–Q4 2026 is reasonable.

GTx-104 will monetise into a highly concentrated, high-acuity hospital market where pricing is driven less by drug cost and more by operational reliability and ICU economics. aSAH is among the most expensive DRGs in U.S. hospitals, with average total episode costs often exceeding $500,000 per patient. Against this backdrop, drug cost is not the binding constraint; reliability of delivery and avoidance of complications are.

This dynamic is already visible in the market. Generic oral nimodipine costs roughly $50–100 per day, yet Nymalize, an oral liquid formulation that does not resolve absorption variability or first-pass metabolism, is priced near $500 per day and has achieved approximately 30–40% penetration. That precedent demonstrates payer and hospital willingness to pay a meaningful premium for formulation and operational advantages alone.

GTx-104 builds directly on this precedent. Unlike Nymalize, it addresses the core limitations of oral nimodipine, and I will comp a lot of the following estimates to Nymalize’s success in the market also.

Getting Peak Sales Figures

Some assumptions to get peak sales:

1. Market Size: c40,000 aSAH patients / year USA (per literature; and materially below claims-based estimates (c70k))

2. Pricing Scenario (Pricing: Benchmarked to Nymalize (c$500/day) with no clinical benefits)

   1. Benchmarking to Nymalize also offers upside for pricing. $500 allows us to model ‘conservatively'.

3. Duration of treatment: 14 days (Despite full course being 21 days)

4. Market Penetration

   1. 35% penetration, benchmarked to Nymalize adoption

   2. 45% penetration, reflecting broader ICU use in dysphagic, intubated, or unstable patients

   3. Upside case tied to higher claims-based incidence

Risk adjusted for 90% FDA PoS and a 2.5× EV/sales multiple (in line with small, hospital-focused specialty pharma) EV Bridge:

Commercial Execution Is Highly Focused

Commercial execution risk is mitigated by market concentration. Roughly 250 hospitals account for c50% of U.S. aSAH cases. Grace plans a lean commercial model with c15–20 representatives covering these centers.

Importantly, demand signals already exist. STRIVE-ON enrolled 30 hospitals, and management has indicated that over 100 additional centers expressed interest in participation. While commercial discussions cannot occur pre-approval, this level of investigator engagement supports post-approval uptake.

At the guideline level, the American Heart Association/American Stroke Association has already begun acknowledging the potential role of IV nimodipine, further supporting institutional receptivity.

Cash Runway & Financial Position (Continuing with the Base Case)

On a fully diluted, pro forma basis, Grace enters 2026 with approximately 28.2 million shares outstanding, inclusive of the February 11 pre-funded units and associated warrants.

From a liquidity standpoint, Grace begins 2026 in a solid financial position. Pro forma cash is approximately $36 million, comprising $16.9 million in cash as of September 30, 2025, roughly $4.0 million from prior warrant exercises, and $15.0 million from the first tranche of the February 2026 private placement. Under the base-case assumption of a clean FDA approval for GTx-104, the second tranche of the financing, linked to warrant exercise, adds a further $15 million, bringing total available cash to approximately $51 million.

This capital base is sufficient to carry the company through approval and into commercial execution without requiring additional financing (below for my justifications), as my own discussions with the company have confirmed. You’re getting the option to invest with low dilution risk, at a price point not too much higher than the institutional PIPE buy-in @ $3.395 / share.

Operating Burn & Getting to Cash Flow Positivity

Grace’s cost structure and cash runway are straightforward and manageable. In 2026, OPEX remains controlled as development activity winds down post-STRIVE-ON and commercial infrastructure is built selectively ahead of launch. I model total OPEX of c$12–13M, driven by c$4M in R&D and $7–8M in SG&A, resulting in a similar level of cash burn given the absence of revenue.

In 2027, the first commercial year, SG&A increases to support onboarding & education (scale limited due to concentrated market). Total OPEX rises to c$18–20M, while revenue ramps to c$30–40M. Given gross margins of approximately 95% (small molecule COGS), I see cash burn compressing to c$5–8M, even at partial penetration.

By 2028, I model incremental revenue largely dropping through and adoption approaching steady state, revenue @ $70–90M range, while OPEX to c$22–24M, which will get Grace to cash flow positivity (c$35–50M annually).

From a liquidity perspective, Grace enters 2026 with approximately $36M of pro forma cash, rising to c$51M upon exercise of the approval-linked warrant tranche. Against cumulative pre-breakeven cash usage of approximately $20M across 2026–2027, the company reaches cash-flow breakeven with c$30M of cash remaining. Importantly, warrant dilution is already reflected in the fully diluted share count, implying no incremental equity financing is required for the base case.

Other:

Protection

GTX-104 has received orphan drug designation from the FDA, granting it seven years of marketing exclusivity in the US post-approval. During this period, the FDA will not approve another application for the same drug in the same indication, regardless of patents. The designation also extends to ten years in Europe.

The patent portfolio extends well beyond this orphan drug designation period, with the earliest patent for GTX-104 expiring in 2037, and in September 2025 the U.S. Patent and Trademark Office granted a new patent), which covers the IV dosing regimen used in the Phase 3 STRIVE-ON safety trial for GTx-104. This dosing regimen patent extends protection for the product into 2043.

On GTx-101 / GTx-102

Grace has additional preclinical programs (GTx-101 and GTx-102), but these are not material to the investment thesis. GTx-104 is the sole value driver and the focus of both capital allocation and near-term execution.

Summing it all Together

Getting the PT: With a 280-285m EV and 30m net cash at breakeven. EV is $313M, or $11.10 per share fully diluted.

At this level, Grace would carry an equity value of approximately $313M, implying an EV of c$280–285M after adjusting for net cash. This compares to base-case peak U.S. sales of c$125M, or roughly 2.3× peak sales, which is conservative for a hospital-based, high-margin product with regulatory exclusivity and limited commercial risk.

At the current price of $3.66, Grace trades at a fully diluted market cap of approximately $103M and an enterprise value of c$65–70M. The market is discounting approval and execution risk far more heavily than the underlying data and balance sheet justify. The stock is trading at half of risk-adjusted peak sales, despite a funded runway through cash-flow breakeven and a single, well-de-risked asset.

In this context the upside is a c3x from current levels, driven by fundamental normalisation. Given the limited remaining development risk, strong regulatory precedent, and clean financial runway you have an extremely asymmetric risk-reward profile.

PT: $11.10

Thank you for making it to the end. I appreciate it’s a long read. Feel free to reach out!

I’d just like to first of thank all of you for reading (especially a couple who I’ve engaged with a lot; especially for being a strong motivator in the substack world. Otherwise I’d just have stayed and continued writing on Twitter/X only).

I never really thought I’d put much effort into Substack but I think I’ve been motivated just to continue doing this, because it genuinely helps me to improve, talk to more people, and also is fun as something to do when I have time!

Recapping on the writeups I’ve done so far since inception this year:

Grace Therapeutics GRCE | Dec 2024
Pitched: $4.20 | Initial: $3.37 (-20%) | PT: $12

The thesis remains intact (and confirmed by Ph3 results), I plan to push out an update in January after JPM and distributing it on some more channels to get the word out before. But the Phase 3 itself achieved all goals, heard discourse around the mortality % but looking at it, it seems still incredibly undervalued. I still think the $12 PT holds, and is just beaten by undercoverage. You have extremely strong shops in the stock underwriting the PIPEs including ADAR1. It feels like another commercialisation story similar to my strongest conviction holdings BBIO and CYTK, both of which have run up incredibly this year.

Essa Pharma ESSA | Jun 2025 - Jul 2025 (Closed)
Initial: $1.72| Currently: $1.88(+9%) | PT: $2.05-2.10

Essa was a valuation gap for a company that was way too far in liquidating. It ended up being acquired for $1.91. Closed before the final just to get the cash out first. The thesis was always a liquidation for a dead bio. Quick and short, actually faster than I expected for the thesis to materialise.

Cytokinetics CYTK | Aug 2025
Initial: $34 | Currently: $64.18(+88%) | PT: $55 (need to update)

CYTK in the 30s was too good of a R/R even if Blum was CEO. The stock had been battered after rumours of acquisitions. The PT really needs to be revised, and I’ll put an update out next month, now that Aficamten was approved (last week!). Trials also show that Africamten does outperform BMY’s Camzyos drug. Will continue the saga, and see how commercialisation goes.

Nektar NKTR | Aug 2025
Initial: $29.86| Currently: $43.94 (+47%) | PT: $55 (Just noticed it was the same as CYTK)

Bit more tricky. Its results in Atopic Derm have been incredibly promising, as well as the data coming out of Alopecia Areata (despite naysayers - the Ph2 works if you’re treating it as a PoC trial). The main issue now is figuring out the financing overhang for funding two Ph3 trials…

Sarepta SRPT | Oct 2025 - Nov 2025 (Closed)
Initial: $18.50 | Closed: $18.67 (0%) | PT: $30

This was silly of me. Honestly, I still think the company has a very good shot at hitting $30. I’m also invested in Arrowhead who have a couple assets tied with Sarepta, ARRW seems to negate a lot of the risks with SRPT and provides more of the upside. Honestly I don’t understand why I ever touched a company mired with regulatory issues. After the earnings call I lost any confidence in the thesis of a revenue driving re-rating. Too much of a headache when the thesis was tied too closely to earnings; when there are far better options elsewhere in biotech. It’s also made my head a lot clearer to stick to the science.

This has definitely been a year of two halves, with the first half being absolutely miserable to be a biotech investor, and even good data struggled to get rewarded. But towards the second half bio increasingly appeared to be an alternative to the crowded AI, and the rebound in biotech has felt healthier and more rational. You’re getting companies rewarded and punished for good and bad data respectively. Risk appetite has seemingly returned.

On a personal level, this year has been a learning process in how (and what) to publish. I will become more selective about what I put out publicly, focusing on fewer ideas but higher-quality work, and being more deliberate about when something is actually worth writing up for 2026 onwards. I’ll continue sharing my longer-form DD openly and for free, while using the monthlies to be more transparent with paid subscribers about positioning, changes, and the names I’m actively moving in and out of.

I’ve cleared out the portfolio sitting at 30% cash, and will stop doing active work until the New Years. There are a few names I’m already quite confident will turn into full writeups early next year, to post after JPM, and I’m looking forward to getting those out when I have the time to finish writing.

Performance-wise, I’m finishing the year at c45% YTD (with a couple of days still to go). Not the flashiest finish compared to some, but I’m happy as long as it’s green, and especially happy given how difficult the first half of the year was. The late-year boost from ABVX buyout rumours didn’t hurt either, particularly as it’s a name I’ve consistently flagged in the monthlies as a VERY likely JPM takeout candidate.

More than anything, I want to say thank you to everyone who’s read, subscribed, commented, challenged my views, or reached out privately throughout the year. The engagement, especially from people I’ve interacted with regularly, has been genuinely motivating and a big reason I’ve continued putting the work in.

Wishing everyone a Merry Christmas and a great holiday period. Looking forward to a busy, hopefully constructive start to 2025. As always, feel free to reach out on Twitter DMs or here anytime.

As is tradition in the monthlies, a photo from a day in this month.

By reading this you accept you have read and acknowledged the disclaimer.

Another quick update to overall theme and making this one a bit more snappy because FDA shenanigans are absolutely ridiculous.

FDA regulatory volatility and staffing gaps are hindering science. Over 200 industry leader signed a letter urging a restoration of predictability and gold standard science in the drug review process. Need I say more:

• Brain drain and exodus are resulting in knowledge gaps, you have non area experts conducting reviews resulting in missed timelines, and difficulties in getting alignments on requirements

• Padzur’s appointment in CDER is a boon but we’ll see how that goes. There was a very chummy photo published in November but time will tell…

  

• You’re also now getting the FDA starting to forgo traditional AdComs in favour of ad hoc internal ‘expert panels’ - closed door and have been convened on policy issues without public input or recorded votes that AdComs come with… Interesting choice of transparency considering how CRLs has been made public. We saw it in Biohavenn when there was an AdCom planned for troriluzole and was cancelled last minute.

• Not commenting on vaccines. I’ll leave 2 links:

  • https://www.statnews.com/2025/11/29/covid-vaccine-deaths-fda-memo-vinay-prasad/

  • https://www.nbcnews.com/health/health-news/fda-covid-vaccines-children-deaths-rcna246437

  • (Prasad noting FAERs data for Covid causing 10 deaths - bad science)

• Then of course Prasad with his leaked internal email: https://x.com/ZacharyBrennan/status/1993293016958996844?s=20

  • Not even going to comment.

1.Nektar Therapeutics NKTR *Update thoughts on upcoming AA readout* $65.21

Alopecia Areata (AA) has been a notoriously difficult disease to treat is how I will open this… Even though Nektar’s Rezpeg has shown strong results in atopic dermatitis I don’t think this will translate to AA… which is why I’m trimming ahead of REZOLVE-AA readout in December.

In AD, Rezpeg behaves exactly as a systemic IL-2/Treg modulator should: clear, dose-linked Treg expansion in its Phase 1b study, broad reductions across inflammatory proteins, and a statistically meaningful EASI/itch signal in REZOLVE-AD. Even in SLE (ISLAND), where Treg biology is relevant but the disease itself is noisy and heterogeneous, the drug delivered clean pharmacology: strong Treg proliferation, yet only modest clinical gains and an inverted dose–response curve as higher exposures brought more cytokine-like adverse events. That profile makes sense in in a disease where dialing down systemic T-cell activation can quickly alter symptoms driven by Th2 imbalance and the itch–scratch cycle, but not really in AA…

AA is on the other side, the pathology anchored by a concentrated, interferon-y/IL-15/JAK-driven cytotoxic attack maintained by CD8⁺ NKG2D⁺ T cells around the hair bulb. This is much closer to a fixed cytototoxic lesion than a diffuse inflaammatory condition. Once established in the follicule, this population behaves like a resident memory compartment, sustained locally and relatively insulated from broad systemic shifts, which would explain why:

• AA skin can transmit disease to immunocompetent mice with intact Tregs

• IFN-y/IL-15 blockade prevents onset but doesn’t reverse entrenched lesions

• JAK inhibitors remain the only consistent induction therapy

Against that backdrop, the idea that an *upstream* Treg-biased IL-2 agonist can dismantle an entrenched TRM-like CD8⁺ network is a weak biological prior.

Prior IL-2 efforts in AA also argue against Rezpeg. In early AA studies, boosting circulating Tregs do not meaningfully alter the scalp’s immune architecture, hair follicles continued to show high interferon-y signatures, persistent CD8 clustering, and minimal restoration of immune privilege despite clean peripheral pharmacodynamics.

Profiling reinforce this: AA lesions show a qualitative regulatory defect, with epigenetically unstable follicular Tregs and a chemokine environment that favours effector retention over regulatory entry.

Even aggressive Treg-expansion strategies in animals, including IL-2 complex systems that drive large Foxp3+ increases directly in lesional skin, have failed to dismantle the interferon-y/IL-15–driven CD8 TRM network. And JAK inhibitor studies provide a useful counterpoint, regrowth occurs even when Tregs fall locally, highlighting that the real control node is the effector circuit, not the regulatory compartment.

Evidence would seem to show that simple Treg quantity is not the limiting factor once the TRM-like CD8⁺ population is locked in.

The only recent IL-2 ’successes’ in AA depend on conditions that Rezpeg does not recreate: highly localized IL-2 delivery paired with chemokine guidance to physically recruit Tregs into the scalp, or IL-2 as maintenance after a JAK inhibitor has already cleared the disease. Neither support systemic IL-2 monotherapy as an induction therapy.

Mind you, I have spotted some discourse around Rezpeg working in specific AA subgroups, e.g. patients with shorter duration, lower SALT burden, or less deeply entrenched TRM like architecture. Biologically, this has surface plausibility: early AA demonstrates more dynamic perifollicular inflammation, more migration-prone CD8 populations, and a slightly more permissive chemokine profile that could be more responsive to a systemic regulatory shift. However, REZOLVE-AA is not designed to test or exploit any of these potential niches. The study does not meaningfully enrich for early disease (<12 months), patchy or rapidly cycling disease, or low interferon-signature skin. Nor does it stratify by known resistance anchors such as IL-15 axis intensity, baseline TRM density, or lesional vs non-lesional Treg gradients.

As a result, even if a small biologically tractable subgroup exists, the trial’s heterogeneity will dilute that signal, and any effect size is unlikely to survive the noise floor of a mixed-severity, mixed-duration AA population. Historically, when IL-2–pathway agents have shown tiny pockets of response, they emerge only after heavy biomarker enrichment or post-hoc molecular clustering, neither of which is available here. In practice, this means REZOLVE-AA is a pure systemic monotherapy test in an unstratified, entrenched AA population, which is precisely where the biological prior is weakest.

This doesn’t mean I don’t like the stock, just that I am resizing ahead of the trial.

Read more

“Q4 Elevidys infusion volumes would be flat or lower than Q3.” - Mgmt

If my entire thesis is driving based off of sales this feels like an extremely strong indicator of a thesis break even if elevidys sales were better than consensus expected.

My proxy was Roche’s sales in Europe but if they don’t translate to the US in growth I’m not sure how to handle that because it changes my view on the stock for the short-medium term.

2. Confirmatory trial failed

some details on Essence; “ESSENCE, its global, Phase 3 randomized, double-blind, placebo-controlled study evaluating the efficacy and safety of AMONDYS 45 (casimersen) and VYONDYS 53 (golodirsen) compared to placebo in 225 patients, ages 6-13 years old, with Duchenne muscular dystrophy (Duchenne) amenable to exon 45 or 53 skipping.”

The exon skippers were always dubious in efficacy; in the trial missed on primary endpoints - which was not unexpected? PMOs aren’t the be all end all of SRPT, nor a future driver? The trends were good, and excluding Covid year actually showed stat significance.

But I don’t want to be married to a stock with regulatory issues, and the investment now changes from the science and revenue story to fighting with the FDA on getting exon skippers to stay on market when they plan to file with FDA (with this data).

I was wrong and I really should have spotted that this was more of a red flag than not, and stayed a mile away. Lesson learned. If anything changes I’ll come back, but I don’t expect I’ll enter in size for a very long time.