The American Journal of Managed Care
At an AJMC Stakeholder Interchange in Houston, leukemia specialists from academic centers and community practices said menin inhibitors work best in combination, warned that differentiation syndrome can strike months into therapy, and described payer rules that delay access even at flagship centers.
In Houston, Texas, home to one of the nation’s largest leukemia programs, the question of how menin inhibitors reach patients with relapsed or refractory (R/R) acute myeloid leukemia (AML) looks different than it does elsewhere. At a Stakeholder Interchange convened by The American Journal of Managed Care on July 1, 2026, community oncologists in the region said they refer nearly all patients with R/R disease to academic centers, while specialists at those centers said single-agent menin inhibitors, as approved by the FDA, are giving way to combination regimens.
The discussion was moderated by Hagop M. Kantarjian, MD, professor in the Department of Leukemia at The University of Texas MD Anderson Cancer Center. As with other interchanges in the series, this meeting aimed to (1) outline the current therapeutic positioning of menin inhibitors in AML; (2) evaluate the role of community oncology in delivering novel targeted therapies; and (3) define best practices for collaboration and patient comanagement across care settings.
Participants were:
Kantarjian said he doubts practice will stay within the single-agent FDA indications for revumenib (Revuforj; Syndax Pharmaceuticals) and ziftomenib (Komzifti; Kura Oncology), “because the combination therapies in either of these molecular abnormality settings are much better than single-agent therapy.”
DiNardo said MD Anderson typically builds a triplet around the target, with the backbone chosen based on frontline therapy. “We tend not to do just the single-agent–approved target because while it can work, it usually only works in about a third of the patients and often for short durations,” she said. “So, we’re trying to improve on that, and of course trying to get to a transplant.”
Kantarjian highlighted the single-center SAVE trial (NCT05360160) of oral revumenib, decitabine/cedazuridine, and venetoclax in patients with KMT2A-rearranged (KMT2Ar), NPM1-mutated (NPM1m), or NUP98-rearranged AML. Among 42 patients with a median of 2 prior therapies, 60% achieved complete remission (CR) or CR with partial hematologic recovery, and 45% proceeded to transplant.¹ “I, for one, am amazed with the results because these patients are patients who we—not at MD Anderson, but they are generally sent to hospice care,” he said.
Community physicians described a narrower set of options. “In the community, we probably have to do a single agent because I don’t think we have the power to go a dual combination,” Hsu said. Rios said his team at Memorial Hermann gives older patients single agents but questioned conventional limits on combining targeted drugs. “Is there anybody who is taking whatever is the [targeted agent] and combining as long as you don’t have crossover [adverse] effects?”
Comutations raised similar questions. Taveras Alam asked how to approach patients with both NPM1 and FLT3 mutations: “Because I remember it was one of the screening questions for the bleximenib trial.” Kantarjian said he favors “quartet” therapy—a hypomethylating agent (HMA), venetoclax, a menin inhibitor, and a FLT3 inhibitor, with shortened schedules—in relapsed disease. Senapati asked whether alternating agents by cycle might limit cytopenias, but Kornblau cautioned against it. “If you want to breed resistant bacteria or resistant cancer cells in the lab, the way you do it is you either start with a low dose for constant exposure, or you cycle them off and on and off and on to breed resistance,” he said.
Rafaeli described a patient with KMT2Ar AML who reached transplant after salvage with a menin inhibitor combination but relapsed afterward, and asked whether menin inhibitors are being studied as maintenance. Marvin-Peek pointed to data presented at the 2026 American Society of Clinical Oncology Annual Meeting by MD Anderson’s Hannah Goulart, MD. “She did present single-arm data from patients treated with revumenib, and the KMT2A patients had really, really good outcomes compared [with] a historical control of patients previously,” Marvin-Peek said. “They did note some toxicity of it, but overall, patients were able to tolerate it well, thrombocytopenia being the largest.”
In that study of 24 patients, 2-year overall survival was 90%, and thrombocytopenia occurred in 83%.² In SAVE, 63% of patients who underwent transplant resumed revumenib afterward.¹ “So it’s always a little bit of a struggle, like when’s the right time to start?” DiNardo said, citing graft-vs-host disease and engraftment as factors.
On duration, Kantarjian favors 2 years or longer with minimal residual disease monitoring; trials generally use 1 year. Abbas noted that the indication can shape decisions when toxicity arises, saying a clinician might “stop and just wait” in a patient on maintenance, “rather than somebody with active disease, who might want to continue the menin inhibitor as much as possible.”
Differentiation syndrome dominated the safety discussion. Kantarjian said MD Anderson clinicians initially missed cases in which patients with normal-appearing blood counts arrived with fever and a sepsis-like picture. “Early on, we lost a lot of good remissions because we assumed that the patient was resistant, and we lost some patients from differentiation multiorgan failure, particularly patients who were on months 2 or 3 of the menin inhibitor,” he said.
He described patients on ziftomenib who had been in remission for 3 months before presenting with differentiation syndrome, and suggested ziftomenib’s longer half-life might explain why it is seen more often with that agent. Rafaeli asked whether most cases still occur at the time of response; Kantarjian said most occurred early with single-agent therapy and are less common with combinations. Kornblau pressed on the biology: “So what’s your postulated mechanism for them getting differentiation syndrome when they’re in CR? They don’t have a bunch of blasts that are stuck predifferentiation.” DiNardo and Kantarjian suggested residual leukemic cells may still be differentiating, and Senapati proposed that a late event could signal higher relapse risk.
Both labels carry boxed warnings for differentiation syndrome.³,⁴ The ziftomenib label directs clinicians to interrupt therapy when the syndrome is suspected; the revumenib label calls for interruption if severe signs persist beyond 48 hours, after starting corticosteroids.³,⁴ Kantarjian said MD Anderson always stops the menin inhibitor, starts high-dose steroids with a taper, and restarts only after all symptoms resolve. DiNardo described one exception. “There are some people in the clinic, toward the end of the first cycle, that the white count is rising, they have some diffuse myalgias, maybe, or a little bit of a creatinine spike, and in those situations, I’ve done steroids without stopping.”
Revumenib’s reputation for QTc prolongation drew skepticism. Rios said pharmacists now weigh heavily in treatment decisions. “If a pharmacist says there’s a drug interaction here, it’s very hard to go back and say, ‘Well, I’m going to try it anyway,’ ” he said, noting that the field moved past similar QTc concerns in acute promyelocytic leukemia. Kantarjian said automated electrocardiogram readings often overstate prolongation and that manual reads by cardiologists frequently show normal intervals. His approach is to correct potassium and magnesium, avoid QTc-prolonging antibiotics, and switch to isavuconazole when an azole is needed.
The adult revumenib dose is 270 mg twice daily, reduced to 160 mg twice daily with strong CYP3A4 inhibitors; ziftomenib is dosed at 600 mg once daily.³,⁴ DiNardo said thrombocytopenia can limit revumenib after transplant. She described a patient with NUP98-rearranged AML, now past her third transplant, whose platelets fall below 50,000 each time revumenib is restarted. “And so, we’ve gone from the 270 dose to the 160 to the 110,” she said. Abbas said he has a similar patient.
Even at MD Anderson, repeat molecular testing at relapse takes about a week, and Kornblau said it should be faster. “You can isolate DNA in a few hours, half a day, you can do (polymerase chain reaction) PCR in a couple hours, and run it on gel,” he said. “They could, with automation, have some of these things out within 24 hours.” Kantarjian said samples are batched twice weekly, so weekends add a delay. Rios and Taveras Alam said results at Memorial Hermann take about a week and a half, requiring patients to wait in the hospital.
DiNardo cited a German analysis of more than 2200 patients showing that time from diagnosis to treatment did not affect survival in clinically stable patients.⁵ “If you can’t wait, you can’t wait. But I think it provides another rationale that you can,” she said. Kantarjian called the question controversial for proliferative disease.
MD Anderson starts menin inhibitors in the hospital and keeps patients approximately 14 days; Rios said Memorial Hermann patients typically stay approximately 3 weeks. “It depends on reliability, the social aspect, the liability, etc,” Abbas said of discharge timing. DiNardo said proximity matters most: “I feel like [differentiation syndrome] tends to happen week 3 and 4, so we discharge them on week 2, and then they’re going to come back with DS. So, I feel comfortable discharging, but they have to be close by that first level.” MD Anderson sees discharged patients 3 times a week in a fast-track clinic. Rios credited nurse navigators with follow-up and said Memorial Hermann now always holds a bed open for the acute leukemia program, after persuading administrators.
Some participants identified payer requirements as the greatest barrier to timely menin inhibitor access. Kantarjian described a patient with KMT2A-rearranged acute lymphoblastic leukemia who achieved minimal residual disease–negative remission and was headed to transplant; 3 weeks after he requested revumenib, it had not arrived because the insurer required active disease and single-agent use. Cervoni-Curet acknowledged limited experience in that setting, saying of patients with B-cell acute lymphoblastic leukemia, “I haven’t prescribed a lot of B-ALL patients, but I would.”
Others said wording often decides the outcome. Senapati said he obtained approval for a patient with a fusion outside the label by describing it differently. “It’s a MEIS1-HOXA9–dependent leukemia, that’s what I put as, and they approved it,” he said. “But otherwise, I actually reach out to the company, and they can sometimes help with the wording.”
Rafaeli said UTHealth’s pharmacist obtains each insurer’s written criteria in advance. “There’s the aspect of the policy of the insurance itself. So, if we learn what the policy for the insurance is, which we can ask for, we can easily justify it,” she said. “They have these policies. They exist. We just need to ask for it.” DiNardo described the frustration of shifting rules: “Every time you have the peer-to-peer, you’re talking to a different company that has different rules, and they’re like, ‘This is what it says.’ And you’re like, ‘Well, you didn’t tell me that.’” She added that documenting minimal residual disease as persistent disease may help. Kornblau suggested attaching the FDA label language directly to requests.
Texas Oncology physicians said they rarely manage R/R AML themselves. Arani said that when patients progress, his practice calls academic colleagues, and “more often than not, they end up in clinical trial.” He added, “For us in Houston, we’re blessed…they’re 20, 25 miles away from us.”
Hsu said duplicating the academic workup makes little sense. “It makes no sense for us to…send it to NeoGenomics, have them do all the testing, which is going to take 12 days, whereas send it to tertiary, downtown medical center, just repeating another bone marrow,” he said. “I feel bad wanting to have 2 separate marrows done.”
Rafaeli offered a supporting example. “We recently had a patient who we shared with a community provider, had a bone marrow at the community provider 2 weeks before being admitted to us, and we got our peripheral blood [next-generation sequencing; NGS] quicker than he got his bone marrow NGS.” Rios said acute leukemia cases across the Memorial Hermann system are sent to its main Texas Medical Center campus.
That model shapes the meeting’s recommendations. Rather than build specialty pharmacy and molecular testing capacity in community practices, the group favored efficient referral pathways for testing and menin inhibitor initiation, standardized readiness for differentiation syndrome across academic, community, and infusion teams, and practical prior authorization guidance tailored to each payer.
References
