RESTORE-PV: phlebotomy reduction with an uncontrolled-completer limitation

September 9 disclosure / reviewed October 4

Initial RESTORE-PV results were disclosed September 9, followed by a September 10 oral presentation at SOHO. This History event concerns DISC-3405 in polycythemia vera, not bitopertin/APOLLO or selcodebart/RALLY-MF. The original poster was recovered and its tables visually checked; the cutoff is July 13, 2026. [1] [2]

Biology / TMPRSS6 and iron restriction

DISC-3405 is a humanized antibody against TMPRSS6 that increases endogenous hepcidin and restricts circulating iron available for erythropoiesis. Analysis: this acts on a resource needed for excess red-cell production, rather than demonstrating elimination of the underlying clonal disorder. Reduced phlebotomy burden is clinically relevant, but is not itself proof of lower thrombosis or transformation risk. [1] [2]

Prior healthy-volunteer studies supplied pharmacological proof of mechanism. The patient poster reports dose-proportional exposure, increased hepcidin and lower iron, with mean hematocrit below 45%. Numerical half-life and exposure-response thresholds are not established here. Analysis: biochemical changes should be read alongside anemia, symptoms and the practical burden of treatment, because iron restriction can create a tradeoff rather than an unqualified benefit.

Design, treatment and disposition

NCT06985147 is a 52-week, multicenter open-label study without a randomized concurrent control. Forty adults enrolled, twenty per cohort; twenty in A and eighteen in B had been dosed at cutoff. Observation lasted 4–12 weeks; dose escalation was followed by 300 mg subcutaneous maintenance every two weeks in A or every four weeks in B. [1] [2]

Thirteen A participants had completed 26 weeks and nine had completed the first maintenance period. The poster identifies two voluntary withdrawals unrelated to AEs. The remaining difference between enrolled and completer counts must not all be called dropout. The cohorts also had unequal average time on study, making a crude safety-rate comparison especially vulnerable to exposure imbalance.

Efficacy / matched observation periods

For the thirteen Week-26 completers, phlebotomy events fell from 4.0 before dosing to 0.6 after dosing over matched 26-week periods. The paired analysis is within-patient, not against placebo. If one episode meeting phlebotomy eligibility but not actually treated is included, the post-dose mean becomes 0.7 and the difference −3.3 (95% CI −4.5 to −2.1). [2]

Statistical interpretation / selection and attribution

Analysis: regression to the mean, changing background therapy, assessment intensity and clinical decisions about phlebotomy can affect a before/after comparison. A small paired p-value quantifies compatibility with a null model under that design; it does not remove those alternative explanations. The eligibility-event sensitivity analysis is useful because it exposes how procedure decisions affect the endpoint. [2]

The 61.5% and 77.8% figures answer different interval/subset questions and must not be portrayed as an improving response rate in the same full cohort. Mean hematocrit below a threshold does not show every patient remained below it. A controlled study with prespecified intervention criteria, intention-to-treat handling and long-term thrombotic outcomes would strengthen causal and clinical interpretation.

Safety / arm-level treated populations

The poster provides the actual treated denominators and separates related from all-cause events. One Grade-2 anemia event in B resolved after dose reduction. Deaths were not separately tabulated; no related SAE and no TEAE-driven discontinuation do not mean no deaths or no all-cause SAEs. The reported 82.4% Grade-1 proportion refers to events, not the percentage of patients with mild events. [2]

Analysis: monitoring hematologic effects is integral to a drug intended to restrict iron. Cohort B’s shorter exposure means its lower total AE percentage cannot establish a safer schedule. Rare-event reassurance remains limited by 38 exposed participants and immature follow-up; later comparisons should include cumulative exposure and consistent ascertainment.

Capital and management / next evidence

Disc reported $717.7 million of cash, equivalents and marketable securities at June 30 and projected runway into 2029; Q2 R&D was $46.9 million. The liquidity figure is historical and R&D expense is not operating cash burn. No fully diluted share count or financing probability is inferred. The program is licensed from Mabwell; product rights and development obligations must be assessed separately from clinical activity. [1] [3]

Analysis: the early patient signal supports further testing and a clearer dose/schedule decision, not a registration-level efficacy claim. Delivering more mature A/B data and a credible pivotal design is the next management test. The end-2026 update guidance remains prospective; this September event is complete even though the underlying trial and evidence development continue.

Sources / References

1. Disc · Initial results · September 9, 2026
2. SOHO 2026 original poster · Data cutoff July 13, 2026
3. Disc · Q2 results · July 30, 2026
4. ClinicalTrials.gov · RESTORE-PV