NMN in ITP: A Phase 1/2 Signal That Should Rewrite How We Design Rare Disease Trials

Picture the scene: a hematologist reviewing a platelet count in an ITP patient who has already failed corticosteroids, rituximab, and a thrombopoietin receptor agonist. The count sits at 18 × 10³/μL. The next option on the list is a splenectomy. What sits between that patient and the operating room, right now, is a very short list of orally administered agents — and that list just got more interesting. A phase 1/2 trial published in Nature Medicine has now placed low-dose oral nicotinamide mononucleotide (NMN) on that list, reporting a clean safety profile and a platelet response signal in immune thrombocytopenia patients. The clinical trials community should read this carefully — not because NMN is the next blockbuster, but because of what this trial’s architecture reveals about what we’re still getting wrong in rare disease development.

ITP is not an obscure orphan condition. Annual incidence in the United States runs at 6.1 per 100,000 persons, with the highest burden falling on children aged 0–4 years (8.1 per 100,000) and adults over 65 (13.7 per 100,000). Despite that patient volume, the approved armamentarium is narrow: corticosteroids, IVIg, rituximab, and thrombopoietin receptor agonists like romiplostim (Nplate) and eltrombopag (Promacta). When Nplate received expanded approval for earlier use in adult ITP, it was treated as a milestone. It was also a reminder that the disease has been waiting decades for mechanisms that address immune dysregulation rather than just compensating for platelet destruction.

NMN enters that vacuum through a metabolic door that most hematology sponsors haven’t bothered to knock on — and the biology, once you examine it, raises a legitimate question about how long that door has been sitting open.

The Biology They Weren’t Looking For

Here is the counterintuitive reading of this trial: the mechanism of action is not novel because NMN is a new molecule. NMN supplementation has been studied in healthy humans since at least 2019, when a single-dose safety study in 10 healthy men tested 100 mg, 250 mg, and 500 mg doses and found no clinically significant adverse events at any level. A subsequent 12-week study confirmed that 250 mg/day was safe and well-tolerated in healthy subjects, with measurable increases in whole-blood NAD+ levels. The safety profile was already established. What was missing was the disease-specific rationale — and that rationale, it turns out, has been sitting in the immunology literature, largely ignored by ITP drug developers.

NAD+ levels are decreased in the plasma and CD4+ T cells of ITP patients, and in murine ITP models, restoring cytosolic NAD+ in those CD4+ T cells promotes regulatory T cell (Treg) differentiation. Tregs are the immunological brake that ITP pathophysiology releases — their suppression of autoreactive B and T cells is precisely what breaks down when platelet autoantibody production accelerates. The phase 1/2 trial isn’t proposing a mechanism from thin air. It is testing whether a well-characterized metabolic substrate can restore a known immune defect at a dose low enough to avoid the toxicity concerns that have historically kept NAD+ precursors out of serious oncology and immunology pipelines.

That biological coherence matters enormously for what comes next in the regulatory pathway — because the FDA’s appetite for biomarker-driven endpoints in rare disease trials has grown substantially since the agency’s Breakthrough Therapy designation program formalized the expectation that early mechanistic signals can anchor an accelerated development program. The FDA’s Breakthrough Therapy guidance explicitly describes the designation as available for drugs where preliminary clinical evidence indicates substantial improvement over available therapy on a clinically significant endpoint — or a pharmacodynamic biomarker that is reasonably likely to predict clinical benefit. NAD+ restoration in CD4+ T cells, directly upstream of Treg reconstitution, is exactly the kind of mechanistic anchor that argument is built on.

But the design of the phase 1/2 trial itself is where sponsors running comparable rare disease programs should pay the closest attention.

What the Trial Design Exposes

The standard efficacy threshold in ITP clinical trials — a platelet count of ≥ 50 × 10³/μL for two-thirds of eight or more weekly measurements during the last 12 weeks of a 24-week blinded treatment period — was established by precedent from the romiplostim and eltrombopag development programs. That threshold was designed for thrombopoiesis-stimulating agents. It measures the output of platelet production, not the upstream immune dysregulation that causes platelet destruction in the first place. For an immune-modulating mechanism like NMN, anchoring the primary endpoint to that same platelet count threshold creates a structural problem: you are measuring the downstream consequence of a process you are intervening in at the source, with a lag that the 24-week window may not fully capture.

This is the trial design trap that rare disease sponsors fall into repeatedly — borrowing endpoints from the predecessor therapy because those endpoints passed regulatory review before, rather than building the endpoint around the intervention’s mechanism.

A phase 1/2 design in this context has genuine strategic advantages that a conventional Phase 2 cannot replicate. By combining dose-finding and early efficacy signal detection in a single protocol, the investigator team can prospectively anchor biomarker co-primary endpoints — NAD+ levels, Treg frequency, autoantibody titers — alongside the platelet count threshold, without the regulatory cost of running two sequential studies. That biomarker ladder matters not just for scientific credibility but for the Breakthrough Therapy dossier. When a sponsor walks into a Type B meeting with dose-exposure-response data across a pharmacodynamic biomarker that is mechanistically upstream of the clinical endpoint, the conversation about accelerated approval is categorically different from a conversation anchored to platelet counts alone.

The NMN trial’s low-dose oral design adds a layer that the thrombopoietin receptor agonist programs couldn’t offer: a tolerability profile clean enough to contemplate combination therapy. Romiplostim carries injection-site reactions and a bone marrow reticulin formation signal that limits long-term use in younger patients. Eltrombopag carries hepatotoxicity warnings and drug-food interactions significant enough to require fasted administration. An oral agent at doses well below those associated with any safety signal in healthy subjects is a combination partner that a sponsor running a refractory ITP program could legitimately propose to the FDA without a lengthy toxicology justification.

The Regulatory Blind Spot Sponsors Must Fix

Here is what concerns me about where this trial goes next. The ITP drug development space has a well-documented tendency to treat Phase 2 completion as a credentialing exercise — a box to check before the real Phase 3 begins — rather than as the moment when the mechanistic argument for accelerated approval either gets made or gets abandoned. If the sponsors or academic investigators behind this NMN program treat the phase 1/2 data as a proof-of-concept that simply justifies a larger, longer, more conventional randomized controlled trial, they will spend three to five years and significant capital arriving at a dataset that tells them roughly what the phase 1/2 signal already suggested.

The FDA’s own guidance on Breakthrough Therapy designation points at a different path. The agency has stated that preliminary clinical evidence of substantial improvement — not Phase 3 confirmation — is sufficient to initiate the intensive FDA guidance that comes with the designation. For a rare disease with an annual U.S. incidence of 6.1 per 100,000 and a patient population that runs out of approved options before age 60, the ethical and regulatory calculus both push toward the accelerated path.

What that path requires, practically, is a biomarker validation package built in parallel with the clinical development plan — not appended to it after Phase 3 enrollment closes. NAD+ levels in CD4+ T cells are not yet a validated surrogate endpoint. Treg frequency is not yet a validated surrogate endpoint. Making them validated surrogate endpoints requires prospective data collection, reproducible assay methodology, and an FDA engagement strategy that begins at the End-of-Phase-1 meeting, not after a Phase 2 database lock. That is the operational discipline this program needs to demonstrate to move from a signal in Nature Medicine to a New Drug Application that the agency can actually review on an accelerated timeline.

The biology is coherent. The safety data, stretching back to 2019 in healthy subjects, is cleaner than most novel mechanisms reach Phase 2 with. The patient need is documented and acute. What ITP patients cannot afford is a development program designed around the regulatory comfort of the investigators rather than the urgency of the disease — and right now, that choice is still being made.

References

1. Nature Medicine — “Low-dose oral nicotinamide mononucleotide for immune thrombocytopenia: a phase 1/2 trial”
2. PubMed — ITP annual incidence in the United States: 6.1 per 100,000 persons
3. PubMed — Single oral administration of NMN at 100 mg, 250 mg, and 500 mg: safety study in 10 healthy men
4. PMC — NMN 250 mg/day for 12 weeks: safety, tolerability, and NAD+ elevation in healthy subjects
5. PubMed — NAD+ levels decreased in plasma and CD4+ T cells of ITP patients; NAD+ precursor supplementation and Treg differentiation in murine ITP model
6. FDA — Breakthrough Therapy Designation: guidance and eligibility criteria
7. AJMC — ITP durable platelet response endpoint definition (≥ 50 × 10³/μL threshold) and treatment landscape