Moleculin Biotech has reported a preliminary 37% composite complete remission rate among 62 evaluable patients in Part A of its pivotal Phase 2/3 MIRACLE trial of Annamycin plus cytarabine for relapsed or refractory acute myeloid leukemia.
The July 31 update also showed a 24% complete remission rate across the blinded study population. Among 30 evaluable patients who had previously received a venetoclax-based first-line regimen, the complete remission and composite complete remission rates were 23% and 37%, respectively.
Those figures are encouraging because acute myeloid leukemia that progresses after venetoclax treatment frequently has aggressive disease biology and a poor prognosis. They do not, however, represent response rates produced solely by Annamycin. The blinded analysis combines patients receiving one of two Annamycin doses with participants receiving cytarabine and placebo.
An earlier prespecified analysis of the first 45 patients provided a more direct, although still preliminary, comparison. Complete remission occurred in 43% and 36% of patients receiving the two Annamycin doses, compared with 12% in the control group. Composite complete remission rates were 50%, 57% and 29%, respectively.
The trial had enrolled 74 of its planned 90 Part A participants when Moleculin announced the update. The company expects to treat the final Part A patient in September 2026 and release comprehensive unblinded results between December 2026 and February 2027.
What does the 37% blinded remission rate in the MIRACLE trial actually represent?
The 37% figure is the composite complete remission rate across all three randomized treatment groups among 62 evaluable patients.
Approximately 23 patients had therefore achieved either a conventional complete remission or a complete remission with incomplete blood-count recovery, subject to rounding of the reported percentage. Approximately 15 achieved the stricter complete-remission definition.
A conventional complete remission in acute myeloid leukemia generally requires bone marrow blasts to fall below 5%, the absence of circulating blasts and extramedullary disease, and recovery of neutrophil and platelet counts beyond specified thresholds. A complete remission with incomplete recovery indicates that the leukemia has cleared to a similar morphological standard, but one or more blood-count requirements have not been met.
MIRACLE’s composite endpoint includes complete remission and complete remission with incomplete hematologic recovery. It does not include morphologic leukemia-free state, a broader response category that can increase reported composite response rates in some acute myeloid leukemia studies.
The present analysis remains blinded. Moleculin, investigators and patients do not know which of the additional participants received Annamycin or control treatment. The pooled result can show whether remission activity across the study remains within a broadly encouraging range, but it cannot measure Annamycin’s incremental benefit over cytarabine.
The company has emphasized that the blinded composite rate has remained between approximately 37% and 40% across analyses involving 30, 45 and 62 evaluable patients. That consistency reduces the likelihood that the earliest pooled signal disappeared as enrollment expanded.
It does not eliminate the possibility that the difference between the experimental and control arms will narrow. Only the complete unblinded analysis can show how the additional responses are distributed.
Why is the venetoclax-failed subgroup particularly important in relapsed acute myeloid leukemia?
Venetoclax combined with a hypomethylating medicine has become an important first-line option for older adults and patients considered unsuitable for intensive induction chemotherapy.
The combination can produce substantial remission rates, but the outlook becomes considerably worse when the leukemia proves resistant or returns. Venetoclax failure frequently coincides with adverse genetic features, clonal evolution and alternative survival mechanisms that allow malignant cells to escape BCL-2 inhibition.
Moleculin reported that 30 of the 62 evaluable MIRACLE participants had previously received a venetoclax-based regimen. Seven achieved complete remission, producing a rounded rate of 23%, while approximately 11 achieved a composite complete remission, producing the reported 37% rate.
The similarity between the venetoclax subgroup and the complete pooled population suggests that prior venetoclax exposure has not yet eliminated the possibility of remission within MIRACLE. That is a clinically relevant observation because previous treatment can sometimes select a more resistant leukemia population.
Moleculin compared the result with a retrospective study in which only three of 24 patients receiving salvage treatment after frontline venetoclax and hypomethylating therapy achieved complete remission or complete remission with incomplete recovery. Median survival for all 41 patients in that report was 2.4 months.
That historical comparison requires caution. The earlier study came from a single institution, included a small number of patients and used several different salvage treatments. Its population also contained many older adults with adverse-risk disease, and 17 patients did not receive salvage therapy at all.
MIRACLE enrolls patients who can receive high-dose cytarabine and meet clinical-trial eligibility requirements. They may be fitter than the broader population represented by the 2.4-month survival estimate.
The trial is not powered to prove efficacy specifically within the 30-patient venetoclax subgroup. The subgroup analysis is descriptive, blinded and based on a small denominator. It identifies a potentially valuable signal rather than establishing a new post-venetoclax treatment standard.
How is the randomized Phase 2/3 MIRACLE trial designed to test Annamycin?
MIRACLE is a multinational, randomized, double-blind, placebo-controlled and adaptive Phase 2/3 study enrolling adults with acute myeloid leukemia that is refractory to, or has relapsed after, one previous induction regimen.
Part A is designed to select the preferred Annamycin dose. Patients are randomized among 190 milligrams per square metre of Annamycin plus high-dose cytarabine, 230 milligrams per square metre plus high-dose cytarabine, or placebo plus high-dose cytarabine.
Annamycin or placebo is administered for three consecutive days. High-dose cytarabine is given for five consecutive days, with remission assessed after one treatment cycle.
The initial randomization is approximately 1:1:1. Part A is expected to include 90 participants, although the design allowed the dose-selection stage to use between 75 and 90 patients.
Dose selection cannot be based on the highest response percentage alone. Investigators must consider complete remission, composite remission, serious adverse events, treatment-related mortality, pharmacokinetics and the tolerability of each dose.
Once an optimal dose is selected, Part B is expected to enroll approximately 222 additional patients. They will be randomized 1:1 to receive the selected Annamycin regimen plus cytarabine or placebo plus cytarabine.
Relevant patients from the chosen Part A dose group and the concurrent control group are intended to contribute to the final pivotal analysis. Complete remission after one cycle is the principal efficacy endpoint, while overall survival is an important secondary outcome.
The adaptive structure can save time and participants by moving directly from dose selection into confirmatory testing. It also makes the integrity of dose-selection rules, statistical adjustments and control of false-positive findings especially important.
How encouraging were the first unblinded results from 45 MIRACLE participants?
The prespecified June analysis included 14 patients receiving 190 milligrams per square metre of Annamycin, 14 receiving 230 milligrams per square metre and 17 receiving cytarabine plus placebo.
Six patients in the lower-dose Annamycin group achieved complete remission, producing a rate of 43%. Five patients in the higher-dose group achieved complete remission, producing a rate of 36%. Two control participants achieved complete remission, producing a rate of 12%.
Composite complete remission occurred in seven patients receiving the lower dose, eight receiving the higher dose and five receiving control treatment. The corresponding rates were 50%, 57% and 29%.
These differences numerically favoured both Annamycin regimens. The absolute complete-remission advantages over control were approximately 31 percentage points for the lower dose and 24 percentage points for the higher dose.
The results did not reach statistical significance at the early interim analysis. Each experimental group contained only 14 patients, meaning the response percentages could change considerably when a few additional outcomes are added.
Baseline imbalances can also affect results in a small randomized population. Molecular risk, duration of the first remission, previous venetoclax exposure, age and whether the disease was primarily refractory or had relapsed after an initial response can all influence the likelihood of successful salvage treatment.
The two Annamycin doses did not produce a simple dose-response pattern. The lower dose generated the higher strict complete-remission rate, while the higher dose generated the higher composite rate. That makes the complete 90-patient dataset particularly important for selecting the regimen that should enter Part B.
The independent data-monitoring committee recommended continuing both experimental groups. Its decision supported further investigation but should not be interpreted as a finding that either dose has already demonstrated efficacy.
How could liposomal Annamycin overcome resistance to older anthracyclines?
Annamycin, also known as naxtarubicin, is an investigational anthracycline formulated within liposomes.
Like other anthracyclines, it is intended to enter cancer cells, insert itself between DNA base pairs and interfere with topoisomerase II. The resulting DNA damage disrupts replication and repair and can cause rapidly dividing leukemia cells to die.
One proposed distinction is reduced susceptibility to P-glycoprotein-mediated drug resistance. P-glycoprotein is an efflux transporter that can pump certain chemotherapy medicines out of malignant cells, reducing the intracellular concentration needed to damage the cancer.
Laboratory studies have indicated that Annamycin retention and activity are less affected by P-glycoprotein than doxorubicin. This could be useful in acute myeloid leukemia that has survived previous anthracycline-containing treatment.
The liposomal delivery system may also alter how the medicine is distributed among tissues. Encapsulation can change drug exposure and potentially limit delivery to organs where toxicity is undesirable.
Cytarabine provides a complementary mechanism. It is converted within cells into an active metabolite that interferes with DNA synthesis. Combining an anthracycline with cytarabine follows a well-established acute myeloid leukemia treatment principle, although Annamycin is intended to improve the efficacy and safety characteristics of the anthracycline component.
The biological rationale is credible, but resistance after venetoclax is complex. Some leukemia cells develop alterations involving TP53, RAS signalling, FLT3, MCL-1 dependence, mitochondrial metabolism and other pathways. Avoiding one drug-efflux mechanism may not overcome every form of resistant disease.
MIRACLE must therefore demonstrate a clinical advantage rather than relying on the formulation’s laboratory properties.
Has Moleculin proved that Annamycin does not cause anthracycline-related heart damage?
Moleculin reported that the MIRACLE trial continues to show no evidence of cardiotoxicity based on available ejection-fraction measurements and adverse-event reporting.
This is potentially important because conventional anthracyclines can produce cumulative and sometimes irreversible heart-muscle damage. Risk increases with total lifetime exposure and can be influenced by age, existing cardiovascular disease, chest radiation and concurrent treatment.
A less cardiotoxic anthracycline could allow clinicians to treat patients who have previously received substantial chemotherapy without adding the same degree of cumulative cardiac risk. It might also permit additional treatment cycles if continued dosing proves useful.
The available evidence does not justify describing Annamycin as definitively non-cardiotoxic. MIRACLE’s current population is still small, treatment involves a limited number of doses and follow-up may be too short to detect delayed heart dysfunction.
Eligibility criteria also reduce the trial’s exposure to patients at greatest cardiac risk. The study excludes certain patients with impaired cardiac function, previous mediastinal radiation or cumulative anthracycline exposure above the protocol threshold.
Cardiac surveillance should extend beyond left ventricular ejection fraction. Global longitudinal strain, electrocardiograms, cardiac biomarkers, symptomatic heart-failure events and longer follow-up can identify damage that a single ejection-fraction measurement might miss.
The correct conclusion is that no cardiotoxicity signal has emerged so far. Demonstrating a meaningful cardiac-safety advantage will require a larger exposed population, transparent cardiac-monitoring data and sufficient follow-up to detect late effects.
What safety risks remain even if Annamycin avoids detectable cardiotoxicity?
Avoiding heart damage would address only one part of the safety burden associated with intensive salvage treatment.
High-dose cytarabine can cause profound reductions in white blood cells, platelets and red blood cells. Those effects can lead to severe infection, sepsis, bleeding, transfusion requirements and extended hospitalization.
Cytarabine can also produce neurological toxicity, particularly in older adults and patients with impaired kidney function. Cerebellar effects may present as difficulty walking, impaired coordination, slurred speech or changes in mental status.
Ocular toxicity is another recognized risk, which is why preventive corticosteroid eye drops may be used during high-dose treatment. Liver abnormalities, gastrointestinal symptoms, rash and inflammatory complications can also occur.
Annamycin remains a cytotoxic medicine. Even if its tissue distribution reduces cardiac exposure, it may contribute to bone-marrow suppression, infection, mucosal injury, nausea and other chemotherapy-related events.
The July 31 update did not provide a detailed table of adverse events, serious adverse events, dose interruptions or early deaths for the 62 evaluable participants. Those data will be essential when the complete Part A results are released.
Particular attention should be given to 30-day and 60-day mortality. A salvage regimen can produce more remissions while still failing to improve outcomes if treatment-related complications offset the antileukemic benefit.
Why is complete remission after one treatment cycle only the beginning of the evidence?
Achieving remission can relieve cytopenias, reduce leukemia-related complications and create an opportunity for potentially curative allogeneic stem-cell transplantation.
A higher complete-remission rate is therefore clinically meaningful. It is also an earlier endpoint than overall survival, allowing a pivotal programme to produce evidence without waiting several years for every survival outcome.
Not every remission carries the same value. A response lasting only a few weeks may provide limited clinical benefit, while a deeper remission without measurable residual disease may support successful transplantation or prolonged disease control.
The next analyses need to report duration of remission, measurable residual disease status, event-free survival, overall survival and the proportion of responding patients proceeding to transplant.
Blood-count recovery also matters. A patient classified as having complete remission with incomplete recovery may remain vulnerable to infection and bleeding despite substantial leukemia clearance.
The trial must additionally show whether responses are distributed across molecular subtypes. Activity in TP53-mutated, adverse-risk or venetoclax-resistant leukemia would be particularly relevant, but small subgroup analyses can easily produce unstable estimates.
MIRACLE’s single-cycle endpoint creates a clean comparison across treatment groups. It does not establish that one cycle provides sufficient disease control or that later treatment and transplant outcomes will favour Annamycin.
Which patients are represented by the MIRACLE trial and which patients are not?
The trial enrolls adults between 18 and 80 years of age with relapsed or refractory acute myeloid leukemia after one previous induction regimen and an Eastern Cooperative Oncology Group performance status of zero to two.
It is specifically testing second-line remission induction. The findings should not automatically be extended to patients who have received several salvage treatments, newly diagnosed patients or people too frail for high-dose cytarabine.
Patients with acute promyelocytic leukemia are excluded because that disease has a distinct biology and treatment pathway. The trial also excludes certain patients with central nervous system leukemia, significant cardiac impairment or excessive previous anthracycline exposure.
Some patients with FLT3-mutated leukemia are excluded in regions where the targeted medicine gilteritinib is available. Other patients with actionable mutations may have targeted treatment options involving IDH1, IDH2 or additional molecular pathways.
MIRACLE is therefore not attempting to replace every second-line acute myeloid leukemia treatment. Its clearest potential role may be among patients fit for intensive salvage therapy who lack a more appropriate mutation-directed option.
The increasing proportion of participants previously treated with venetoclax makes the population relevant to contemporary practice. Its requirement for high-dose cytarabine nevertheless limits applicability to many older or medically vulnerable people who receive venetoclax precisely because they cannot tolerate intensive chemotherapy.
What must the full 90-patient Part A readout establish before MIRACLE advances?
The complete analysis must confirm that the numerical advantage seen among the first 45 participants persists as the second half of Part A is added.
Investigators need to identify how the additional responses are distributed across the 190-milligram, 230-milligram and control groups. The current pooled 37% rate cannot answer that question.
Dose selection should reflect a consistent benefit-risk judgment. If one dose produces slightly more remissions but substantially greater infection, marrow suppression or treatment-related mortality, the lower dose may offer the stronger pivotal profile.
The analysis should report confidence intervals and the statistical rules applied to the interim examinations. Repeated access to accumulating data can increase the risk of a chance-positive conclusion unless the protocol controls for those looks appropriately.
Detailed patient characteristics will help determine whether the treatment groups were balanced. Prior venetoclax exposure, molecular risk, first-remission duration, age and refractory versus relapsed disease deserve particular attention.
Cardiac findings should be reported with actual measurements, follow-up duration and the number of evaluable patients rather than a general statement that no cardiotoxicity was observed.
The optimal dose will then face the more consequential test in the approximately 222-patient randomized Part B. That stage must reproduce the remission advantage in a larger population and show whether the responses translate into durable disease control and improved survival.
Could Annamycin become an important treatment after venetoclax failure?
The emerging evidence gives Annamycin a credible opportunity. Both experimental doses numerically outperformed randomized control treatment in the first 45 MIRACLE patients, and the pooled remission signal has remained relatively stable as the evaluable population expanded to 62.
The venetoclax subgroup is especially interesting. A 37% composite remission rate in this population would be clinically meaningful if the eventual unblinded results show that the responses are concentrated in the Annamycin groups rather than the cytarabine control group.
The present update cannot demonstrate that. It pools active and control participants, covers only 30 venetoclax-exposed patients and provides no mature survival or response-duration analysis.
The cardiac findings are similarly promising but preliminary. No detected cardiotoxicity is different from proof that the medicine carries no cardiac risk, particularly after limited dosing in a selected clinical-trial population.
Moleculin expects the complete Part A results between December 2026 and February 2027. That readout should provide the first sufficiently developed assessment of the dose decision, response consistency and safety profile before the pivotal expansion begins.
Annamycin has received Fast Track and Orphan Drug designations from the United States Food and Drug Administration for relapsed or refractory acute myeloid leukemia, as well as orphan designation from the European Medicines Agency. Those designations support development but do not indicate that efficacy has been established or approval is likely.
The 37% blinded remission rate keeps the MIRACLE programme moving forward. The decisive question is no longer whether the study contains remissions, but whether the final unblinded comparison shows that Annamycin produced enough additional, durable and safely achieved remissions to change second-line acute myeloid leukemia care.
