My Treatment Approach: Acute Myeloid Leukemia (AML) - Episode 9
An overview of treatment options in patients who present with IDH1/IDH2 mutations and how to manage treatment-related adverse events.
Richard Stone, MD: Let’s move on to the next part of this case. We gave him the 7+3 [chemotherapy regimen]. He had some mucositis and febrile neutropenia, but he got through it. He went into remission. He was still MRD [minimal residual disease] positive at the time of remission. He received 2 cycles of intermediate-dose cytarabine. We recommended a reduced-intensity allogeneic stem cell transplant because of his age. That’s the way to go for anybody who’s older.
He could have gotten an allogeneic transplant, but he didn’t want one. In lieu of that, we were going to give him oral azacitidine, even though it’s a bit of a dicey thing. But while we were waiting to get him the oral azacitidine, he relapsed clinically with fatigue, aches and pains, and gum bleeding. His white blood cell count was a little higher. He was still anemic. He was thrombocytopenic. He was neutropenic. He had circulating myeloblasts. His bone marrow biopsy was hypercellular. He had the same 3 mutations that he had before.
We were faced with a decision about what to do. This is a 67-year-old man who had already been through 7+3 roughly half a year earlier. Were we going to give him intensive chemotherapy? Were we going to give him something else? We chose the approved agent for this situation, enasidenib, as a single agent. This goes back to the data that Eytan Stein, [MD,] published about 5 years ago with a large phase 1 trial using enasidenib in people with IDH2 mutations who relapsed or were refractory to their initial therapy. The response rate wasn’t fantastic, but it was encouraging in about 30% overall. About half of the responses were complete remissions [CRs]. It’s unknown whether adding something at this stage to an IDH inhibitor, like azacitidine, instead of IDH2 inhibitor enasidenib alone is beneficial. It’s unknown whether adding chemotherapy to enasidenib at this stage is beneficial. We elected to give him enasidenib alone. Eunice, would you have done anything differently?
Eunice Wang, MD: It doesn’t say what his ECOG performance status was at the time of relapse.
Richard Stone, MD: It’s still good.
Eunice Wang, MD: It was still good. In this situation, one certainly could do enasidenib as a single agent. Given there have been some data looking at enasidenib and azacitidine, as well as ivosidenib and venetoclax plus or minus azacitidine, we know that venetoclax-azacitidine works very well for the treatment of IDH2-mutated disease in the upfront setting. The alternative would be whether we’d go with enasidenib or try venetoclax-azacitidine if he were a bit fitter and still robust in this situation, given the sensitivity of IDH2-mutated disease to a venetoclax-based regimen. Alternatively, we’d ideally love to do a triplet trial of enasidenib and venetoclax. That would potentially be a very active combination based on what we’ve seen with the IDH1 inhibitors.
Richard Stone, MD: Enasidenib, venetoclax, and azacitidine might be the way to go here, even though there aren’t vast data for that. There are anecdotal reports. But that’s certainly the way we’re going with this case. Azacitidine-venetoclax would be reasonable, as you pointed out, although the data for azacitidine-venetoclax in the relapsed setting aren’t robust. There was a paper from Memorial Sloan Kettering Cancer Center describing azacitidine-venetoclax results by Maximilian Stahl, [MD,] and others that showed that azacitidine-venetoclax had a role as a relapse agent, particularly in patients with this type of biology. But we went with enasidenib—the approved agent in this situation—as a single agent, 100 mg daily. The indirect bilirubin went up. Is that something we should have expected?
Eunice Wang, MD: Yes. The incidence of indirect hyperbilirubinemia in patients getting enasidenib is about 8%. In our practice, because it’s asymptomatic and expected, we typically don’t hold or change our dosing for hyperbilirubinemia.
Richard Stone, MD: Exactly. There’s some discussion in the package insert about that. But by and large, it’s benign and indirect, and people can get through it just fine. It isn’t associated with hemolysis. The patient was taking these pills every day. On day 28, he still had peripheral blasts. His ANC [absolute neutrophil count] was 1100 per mm3. His red blood cell transfusion needs had disappeared. But 7 days later, he had some clinical problems. He started to have dyspnea on exertion and mild lower extremity swelling. He had bibasilar crackles on exam and bilateral infiltrates on chest x-ray. What do you think is going on, Eunice?
Eunice Wang, MD: Given the timing of this and the monotherapy with enasidenib, I’d suspect differentiation syndrome. Differentiation syndrome can occur in up to 7% or 9% of patients who get treated with enasidenib. The important thing is to distinguish it from other underlying comorbidities. This could be pneumonia, congestive heart failure, or differentiation syndrome.
We do exactly what was done here. Because of the IDH inhibitor, we were very suspicious that it could be differentiation syndrome. We’d start dexamethasone 10 mg [twice a day], but we’d concomitantly start antibiotics, admit the patient, potentially do an echocardiogram to make sure there are no cardiac issues, check troponins, etc. We’d do a multitiered approach, which includes active therapy for potential differentiation syndrome. It would be a little atypical in that the peripheral blasts are 15%. The ANC is a little better. The question is, on day 35, is the white blood cell count now much greater or going up? Is there evidence of neutrophilia? If that’s the case, then in addition to dexamethasone, we might start hydroxyurea to cytoreduce the leukocytosis in addition to the measures listed here.
Richard Stone, MD: Absolutely. Is there any role for stopping the IDH2 inhibitor?
Eunice Wang, MD: If there are very severe symptoms, we hold. But I have to remind myself all the time that even though the temptation is to hold, enasidenib has a long half-life. If you hold the drug and wait for the washout, it isn’t going to be 4 or 5 hours. It could be a week or two. It’s important, the first measure is to start the steroids. When there are severe symptoms, we stop the enasidenib, but that doesn’t magically help everything. We have to always remind ourselves that it takes a while for that drug to clear, particularly if they’ve been on it for about 5 weeks.
Richard Stone, MD: After six 28-day cycles, he was in morphologic CR. MRD flow showed 2% blasts, so he wasn’t in a measurable residual disease negative state. He still had a bit of an IDH2 mutation and has a DNMT3A mutation.
Eunice Wang, MD: Six 28-day cycles would be 6 months. How long do you wait before you start saying, “Maybe they should have responded?” Because we all give azacitidine. We all give 4 to 6 cycles of azacitidine alone for MDS [myelodysplastic syndrome]. But for AML [acute myeloid leukemia], with venetoclax-azacitidine, we expect something within a couple of months. Sometimes it’s hard for us to sit tight on these patients. Would you wait 6 months? Or would you start to get nervous around 3 or 4 months?
Richard Stone, MD: There’s probably some value of stable disease in these patients. It’s a little different from a chemotherapy regimen. This is a slower remission. They can have peripheral blasts for a long time. They can be stable and not progress. It’s like the enasidenib is suppressing things rather than killing things. We wait a while. We didn’t kill all the cells here, or even a lot of them. We killed many of them, but there’s still residual disease. He still has a high chance of relapse after transplant because he has MRD going into it.
Eunice Wang, MD: Would you transplant him now? Or would you give it another couple of months?
Richard Stone, MD: I’d do it now. Six months is a long time.
Eunice Wang, MD: I think so, too. I would be impatient.
Richard Stone, MD: He’s going to develop another clone or have a formal relapse, and then the transplanters won’t take him. It isn’t an ideal situation because he’s going to have a reduced-intensity transplant because of his age, so we’re going to be relying on graft-vs-leukemia effect to work here. There might be a case for enrolling him in a clinical trial post-transplant with an IDH2 inhibitor if there’s one available, or at least azacitidine and venetoclax or something, to try to reduce his risk of relapse post-transplant, which is going to be high.
Eunice Wang, MD: Do you continue to do the next-generation sequencing [NGS] at all of these subsequent time points? You did it again here after 6 cycles, and it still is the DNMT3A, IDH2, and RUNX1. Why would you continue to repeat? Sometimes I get the question, “Why are you continuing to repeat this very expensive next-generation sequencing in these people who are on targeted therapy?” What are you looking for? Are you looking for RAS mutations that might predict resistance? Why would you continue to do next-generation sequencing on these patients?
Richard Stone, MD: That’s a great question. I wonder if it’s cost effective, but it’s biologically interesting. It may have some clinical impact. If you look at the data on residual disease, there’s the excellent paper by the HOVON group that was published a few years ago—the first author was Mojca Jongen-Lavrencic, [MD, PhD,]—where they looked at both flow cytometric and NGS-based detection of MRD, and they were complementary. You don’t know before you send this that the MRD flow is going to be positive.
In a patient like this, because both are positive for MRD, that would suggest a very high risk for relapse, at least without transplant. It would probably be a significant risk with transplant as well. It gives some prognostic information. I’m going to transplant this patient no matter what, but this gives me a little more insight into what’s going on. You’re right, there may be another mutation—a FLT3 mutation—that could crop up that would make me think differently about what’s going on, but it’s certainly a valid question.
Transcript edited for clarity.