MaaT Pharma plans to request a re-examination of the negative Committee for Medicinal Products for Human Use opinion on MaaT013, also known as Xervyteg, for adults with acute graft-versus-host disease involving the gastrointestinal tract after prior treatments have failed. The European regulatory dispute now centres on whether the clinical benefit observed in the ARES study can be attributed sufficiently to MaaT013 despite the use of concomitant treatments in a severely ill population.
Why the MaaT013 re-examination will focus on causal attribution rather than unmet need
The negative opinion does not erase the clinical response and survival outcomes reported in ARES, nor does it suggest that the need for additional third-line treatments is in doubt. Instead, it exposes a narrower but decisive regulatory problem: a single-arm study can show that patients improved after receiving a therapy, but it may not prove with enough certainty that the therapy caused the improvement. In acute graft-versus-host disease, that distinction becomes especially difficult because patients frequently receive several immunosuppressive agents, anti-infective treatments, nutritional support and other interventions at the same time.
MaaT Pharma intends to ask for different rapporteurs and a Scientific Advisory Group involving haematology specialists. That could help the Committee for Medicinal Products for Human Use assess whether conventional expectations around treatment attribution are realistic in a third-line setting where clinical practice is complex and therapeutic options are heterogeneous. However, the re-examination cannot become a second clinical development programme. European Medicines Agency procedure limits the review to the evidence available when the original opinion was adopted, which means MaaT Pharma must win through a stronger interpretation of the existing dossier rather than by adding new trial results.
This makes the appeal scientifically meaningful but procedurally constrained. The French biotechnology developer can argue that the totality of evidence is coherent across ARES, earlier studies and early access use, but regulators must still decide whether coherence is enough to support conditional marketing authorisation. Severe disease and limited alternatives may justify greater tolerance for uncertainty, yet they do not remove the requirement to establish a positive and reasonably attributable benefit-risk profile.
What the ARES response and survival results reveal, and why they still leave a regulatory gap
ARES enrolled 66 adults with gastrointestinal acute graft-versus-host disease who had failed corticosteroids and were refractory to ruxolitinib. The study reported a gastrointestinal overall response rate of 62% at Day 28, including a 38% complete response rate, while gastrointestinal response remained 47% at Day 56 and 44% at three months. One-year overall survival reached 54%, and patients who responded by Day 28 had substantially better survival than non-responders.
Those results are clinically notable because the enrolled population was heavily treated and predominantly had severe disease. The depth of response also matters. A response profile dominated by complete and very good partial responses is more persuasive than a result driven mainly by marginal improvements, particularly when gastrointestinal involvement can lead to profound diarrhoea, bleeding, infection, malnutrition and organ failure. The persistence of responses beyond Day 28 further reduces the possibility that the primary endpoint captured only a brief fluctuation in symptoms.
The regulatory gap remains the absence of a concurrent control group. ARES was open-label and single-arm, so differences in supportive care, background immunosuppression, infection management, centre experience and patient selection cannot be fully separated from the apparent effect of MaaT013. The strong association between early response and survival is important, but it does not independently establish that MaaT013 generated the response. Patients with less biologically aggressive disease may have been more likely to respond and survive regardless of the investigational treatment.
This is the core tension in the dossier. The efficacy signal appears larger and more durable than clinicians would generally expect in this setting, yet the trial architecture leaves room for alternative explanations. Conditional approval can accommodate residual uncertainty when an unmet need is substantial, but the regulator must still be convinced that the remaining uncertainty is manageable through post-authorisation evidence rather than fundamental to the initial efficacy claim.
Why CHRONOS strengthens clinical context without replacing randomized evidence
The CHRONOS retrospective study gives MaaT Pharma an important contemporary benchmark. It examined 59 adults treated at European transplant centres after corticosteroid and ruxolitinib failure, excluding microbiome-based therapy. Patients received a range of third-line approaches, including anti-tumour necrosis factor therapies, extracorporeal photopheresis and vedolizumab. Gastrointestinal response was 37% at Day 28, fell to 22% at Day 56, and one-year overall survival was 29%, with median survival of 86 days.
Compared descriptively, ARES produced higher early response, greater response persistence and better one-year survival. That comparison supports the argument that MaaT013 may be doing more than reflecting improvements in modern supportive care. CHRONOS also demonstrates why a single standard comparator would have been difficult to select, since real-world third-line treatment is fragmented rather than dominated by one accepted regimen.
However, CHRONOS cannot function as a synthetic control in the same way as a prospectively designed and carefully matched external comparator. It was retrospective, treatment choices varied, and the populations were not randomized or formally matched. Differences in eligibility, disease severity, treatment timing, supportive care, centre practices and follow-up can create an apparent treatment advantage even when headline criteria look similar. The study was also sponsored to contextualise ARES, which does not invalidate its findings but increases the importance of transparent methods and independent scrutiny.
CHRONOS therefore strengthens the unmet-need argument more decisively than it settles the efficacy argument. It shows that patients progressing beyond ruxolitinib face poor outcomes and inconsistent treatment pathways. It also makes the ARES results harder to dismiss as ordinary third-line performance. What it cannot do is close the causal gap identified by the Committee for Medicinal Products for Human Use.
How MaaT013 could alter third-line care after corticosteroids and ruxolitinib failure
Acute graft-versus-host disease is initially treated with systemic corticosteroids, while ruxolitinib has become the principal approved option when corticosteroids do not provide an adequate response. Once ruxolitinib fails, clinicians enter a less standardised environment in which treatment choice depends on organ involvement, infection risk, prior exposure, centre experience and the patient’s rapidly changing condition. The absence of an established European third-line therapy is therefore both a clinical problem and a trial-design problem.
MaaT013 takes a different approach from additional systemic immunosuppression. The pooled-donor, full-ecosystem microbiome therapy is administered by enema in hospital and is intended to restore microbial diversity and immune balance in a gut damaged by transplantation, antibiotics, inflammation and repeated treatment. Should that mechanism translate reliably into clinical benefit, MaaT013 could add a treatment modality that acts through the intestinal ecosystem rather than simply intensifying broad immune suppression.
That distinction could matter for patients with gastrointestinal disease, but adoption would not be automatic. Transplant teams would need confidence in donor screening, pathogen controls, batch consistency, product handling and administration across medically fragile patients. Clinicians would also need practical guidance on how MaaT013 should be sequenced with corticosteroids, ruxolitinib and other rescue therapies. Ironically, the concomitant treatments that complicate the regulatory assessment are likely to remain part of real-world use, making clear treatment protocols and post-authorisation evidence essential.
The route of administration may also influence uptake. An off-the-shelf hospital product can be operationally attractive in an acute setting, but enema administration may be challenging in patients with severe intestinal inflammation, bleeding or poor tolerance. A marketing authorisation would answer the benefit-risk question at a population level, not eliminate these patient-level implementation decisions.
What the negative CHMP opinion reveals about regulatory expectations for microbiome therapies
MaaT013 is not a conventional small molecule with a single defined active ingredient. It is a high-diversity biological ecosystem derived from pooled donors and manufactured as a standardised therapeutic product. That complexity is central to its scientific rationale, but it also raises regulatory questions around identity, potency, consistency, contamination control and the relationship between product composition and clinical effect.
For microbiome developers, the MaaT013 review shows that manufacturing sophistication and biological plausibility cannot compensate for uncertainty in treatment attribution. Regulators may accept that a complex ecosystem cannot be reduced to one organism or metabolite, but they still need a reproducible product and interpretable clinical evidence. A positive re-examination would suggest that a large effect size, durability, survival outcomes, external context and severe unmet need can collectively support approval even without randomisation. A second negative opinion would signal that future microbiome programmes may need stronger concurrent controls or prospectively defined external-control strategies.
The decision may also shape how developers select endpoints. Day 28 overall response is clinically established in acute graft-versus-host disease, but survival, response durability, corticosteroid exposure, infection burden and hospital utilisation can strengthen the relevance of that endpoint. MaaT013 reported encouraging results across several of these dimensions, yet the regulator’s concern indicates that multiple favourable endpoints do not automatically resolve confounding when they arise from the same uncontrolled study.
This is not a rejection of the microbiome as a therapeutic field. It is a reminder that platform novelty often increases, rather than reduces, the need for disciplined evidence generation. The more biologically complex a product is, the more clearly the clinical programme must separate product effect from background care and natural disease variation.
Why continued early access helps clinical learning but cannot decide the European appeal
MaaT013 remains available to eligible patients through an early access programme that has treated more than 300 patients across 13 countries since 2019. Continuity matters because the negative Committee for Medicinal Products for Human Use opinion does not immediately remove an option being used in a population with few alternatives. It also allows transplant centres to retain operational experience with product delivery and patient management.
Early access experience can provide useful information on safety, feasibility and outcomes outside a tightly controlled study. A broad geographic footprint may reveal whether results are reproducible across centres with different clinical practices. However, such programmes are vulnerable to the same interpretive limitations as other non-randomized evidence, including treatment selection, incomplete data capture, variable concomitant therapy and the absence of a comparable untreated group.
More importantly, new early access observations cannot be introduced to cure the current re-examination problem. They may support future regulatory discussions, confirmatory work or post-authorisation commitments, but the appeal must be decided using the existing evidence base. Continued access therefore protects clinical continuity while leaving the central marketing authorisation question unresolved.
What the September 2026 decision could mean for MaaT Pharma and the wider microbiome field
A revised positive opinion would reopen the path towards European conditional marketing authorisation and could establish MaaT013 as a new third-line option for adults with gastrointestinal acute graft-versus-host disease. It would also validate a development strategy built around a single-arm pivotal study, contextual real-world evidence and early access experience in a rare and life-threatening setting. Commercial preparation and treatment-centre education would then become immediate priorities, alongside any confirmatory evidence required by regulators.
A final negative opinion would be more consequential than a simple delay. MaaT Pharma would need to determine whether the most credible route forward is a controlled European study, a redesigned global programme, or further engagement with regulators around a narrower population or different evidence framework. The setback would also affect how industry observers assess the read-through from MaaT013 to oral candidate MaaT033 and the broader microbiome platform, although each programme will ultimately depend on its own formulation, indication and trial design.
In my assessment, MaaT Pharma has a clinically credible argument but a difficult regulatory one. The ARES effect size, response depth, durability and survival compare favourably with contemporary third-line outcomes, and the unmet need is undeniable. Yet the Committee for Medicinal Products for Human Use has identified a weakness that expert testimony alone may not fully solve: the pivotal study was not designed to isolate the contribution of MaaT013 from complex background therapy.
The September review will therefore test how much uncertainty European regulators are prepared to accept when a severe disease, fragmented standard of care and promising survival signal collide with an uncontrolled trial. Clinicians will watch whether the Scientific Advisory Group considers the ARES results sufficiently persuasive for conditional approval. Microbiome developers will watch something broader, whether regulatory flexibility for rare and lethal diseases can extend to complex ecosystem therapies without lowering the evidentiary threshold beyond what regulators consider scientifically defensible.
