Targeted Therapies Advance Treatment in AML, But Toxicity Challenges Remain

Harry P. Erba, MD, PhD

Harry P. Erba, MD, PhD

Targeted therapies for IDH1/2 and FLT3 mutations have made headway in the treatment of patients with relapsed/refractory acute myeloid leukemia (AML), but several challenges with associated toxicities still exist, said Harry P. Erba, MD, PhD.

Among the oral inhibitors developed to target the mutant forms of IDH1/2 are ivosidenib (Tibsovo) and enasidenib (Idhifa). Early phase trials have shown that both of these agents lead to approximately 25% to 30% of patients achieving a complete remission (CR) or CR with partial hematologic recovery, said Erba, a professor of medicine at Duke University School of Medicine and a member at Duke Cancer Institute. However, differentiation syndrome is a major toxicity that has been observed with these drugs.

“[Differentiation syndrome] is recognized by clinical features; there is no diagnostic test available for it,” said Erba. “The clinician has to be aware that this can happen and that it has been estimated to happen in about 10% to 20% of patients.” He added that the symptoms associated with this condition can sometimes be confused for progressive AML. This makes management that much harder, as steroids should be given immediately after the syndrome is suspected.

As for patients with FLT3-positive relapsed/refractory AML, the first-generation FLT3 inhibitor midostaurin (Rydapt) was approved by the FDA in April 2017 for use in combination with standard cytarabine and daunorubicin induction and cytarabine consolidation based on data from the phase III RATIFY trial. This approval was followed by that of gilteritinib (Xospata) in November 2018. The agent was found to significantly improve overall survival (OS) in this patient population compared with salvage chemotherapy, based on findings from the phase III ADMIRAL trial.¹ On May 30, 2019, the FDA approved a supplemental new drug application to update the label for gilteritinib to include the final OS data from the trial.

A third inhibitor, quizartinib, has also emerged in the landscape in recent years; however, it will have to overcome regulatory adversity in order to join the growing list of agents approved for use in this space. In May 2019, the FDA’s Oncologic Drugs Advisory Committee voted 8-3 against approving the new drug application (NDA) for quizartinib in patients with FLT3-ITD—positive AML.²

Although the phase III QuANTUM-R study showed a statistically significant improvement in OS with quizartinib compared with chemotherapy, the FDA cited concerns with the credibility and generalizability of the trial data. The FDA is now scheduled to make a final decision on the NDA for the second-generation FLT3 inhibitor by August 25, 2019. Erba adds that the potential for immunosuppression is another issue to be cognizant of with this agent.

In an interview during the 2019 OncLive^® State of the Science Summit™ on Hematologic Malignancies, Erba provided insight into how to navigate through the targeted therapies available in the treatment of AML and discussed the work that lies ahead in this space.

OncLive^®: What does the current treatment paradigm look like for relapsed/refractory AML?

Erba: Unfortunately, many patients with AML either do not achieve first remission with chemotherapy, which is the standard of care, or they will relapse afterwards. In fact, about 25% to 30% of patients have primary refractory disease. The majority of patients will relapse. Chemotherapy has been the standard of care for those patients in an effort to utilize allogeneic stem cell transplant to cure their disease. We have learned that certain biologic and clinical features can help predict the likelihood of achieving a remission with chemotherapy, such as the duration of first remission, cytogenetics, age of the patient, and whether or not they have had a prior transplant. The outcomes remain less than optimal.

In my presentation, I spoke about targeted therapies for those patients with AML who have been refractory to or resistant to intensive chemotherapy. Among the drugs we now have available are the oral inhibitors against the mutant forms of IDH1/2 and an inhibitor of the FLT3 activating mutations. In terms of the IDH inhibitors, about 20% of patients will have either an IDH1 or IDH2 mutation. These mutations are gain-of-function mutations, meaning they actually will change the normal function of IDH from creating alpha-ketoglutarate to actually [alpha]-hydroxyglutarate. These mutations lead to a block in differentiation. One of the drugs that we have for IDH1/2 is ivosidenib, which is an oral, selective, and potent inhibitor against IDH1. For IDH2, the oral and potent inhibitor that we have is enasidenib.

In general, the phase I studies of these drugs have showed pretty similar results, with about 25% to 30% of patients achieving either a CR or CR with partial hematologic recovery. These remissions are durable in the sense that their median duration is over 6 months in both cases. What has been remarkable about both drugs is that they can lead to transfusion independence even if the patient doesn't achieve CR with clearance of the blasts. We have learned, however, that these remissions can take some time; the median time to a CR is about 3 to 4 months with both of the drugs.

We have also learned about the major toxicity we have seen with these drugs, which is differentiation syndrome, as you might expect; this is often, but not always, associated with leukocytosis. The syndrome can occur as early as 1 week and as late as months after starting the drugs; it is characterized by fever, hypoxia, dyspnea, weight gain, hypotension, and pulmonary infiltrates. As soon as the syndrome is suspected, patients need to be treated with steroids. We use dexamethasone 10 mg intravenously twice daily and continue that for at least 3 days before considering tapering. These patients are typically best monitored in the hospital if that occurs.

The challenge that we have is that for those symptoms, there's no perfect diagnostic test. Those symptoms can also be very similar to disease progression. Therefore, a patient having progressive leukocytosis from leukemia and coming in with pneumonia can have all of those same [symptoms] as differentiation syndrome. In my own practice, if I see the counts going up as if the patient is responding, and especially if I see more differentiated granule sites, that can be very helpful.

Secondly, the response to dexamethasone itself is important to check. If the patient symptomatically gets better within 1 to 3 days, that's likely going to be differentiation syndrome. We manage these patients with hydroxyurea for the high white blood cell count. The response rates are only seen in about 25% to 30% of patients, and yet these drugs will suppress [alpha-]hydroxyglutarate levels in almost every patient. Therefore, why doesn't everyone respond? We don't know for certain.

The one thing that really appears to correlate with likelihood of response is the mutational complexity of the blast; the more mutations, the less likely it appears that the patient will respond. On top of that, the specific mutations matter. When FLT3 and RAS co-occur with IDH, that is often associated with a lower chance of response. However, clearly, patients in both studies were able to move on to transplant. Older patients can become transfusion-independent, and at least one-third of patients who are transfusion-independent at baseline become independent. There is clearly a benefit, and what we are learning now is how to incorporate the single agents into initial therapy with intensive chemotherapy.

Could you expand on the progress that has been made with FLT3 inhibitors?

As you know, midostaurin is FDA approved based on data from the RATIFY trial, where midostaurin was added to standard induction and high-dose cytarabine consolidation, followed by maintenance midostaurin, versus placebo. In that study, there was no difference in response rates, but the survival was longer with midostaurin. At 4 years, 51% of patients were still alive if they received midostaurin, compared with 44% of patients who received placebo. Based on that, and without a change in early mortality with the addition of midostaurin, the drug was approved. Midostaurin is considered a first-generation FLT3 inhibitor, which means it's not very selective. It's a Type I inhibitor, meaning it inhibits FLT3-ITD mutations.

Other drugs have been in development that have been rationally designed toward more specificity in FLT3. One that has now been FDA approved for patients with relapsed/refractory AML with either a FLT3-ITD or FLT3-TKD mutation is gilteritinib. That was approved late in 2018 based on the results of the ADMIRAL trial, which was a phase III study randomizing patients 2:1 to receive either oral gilteritinib or chemotherapy; it was investigator's choice for the chemotherapy. The co-primary endpoint was met; the response rate was doubled with gilteritinib compared with chemotherapy and the survival was also prolonged. The FDA approved this agent based on the response rates, but at the 2019 AACR Annual Meeting, Alexander Perl, MD, of the University of Pennsylvania, presented the OS data from the phase III study. Even when you censor survival at the time of transplant, there is a benefit with gilteritinib over chemotherapy.

Moreover, in the study, patients were allowed to continue gilteritinib after transplant, and the survival of patients after transplant was better if they started on the drug versus not. The questions for the future are focused on if we can do better than less than half of patients having complete responses. Is there a role of combining gilteritinib with chemotherapy in the relapsed/refractory setting to get a higher response rate and potentially deeper remissions? Of course, what about using these second-generation, more specific drugs upfront? Midostaurin is approved for use in combination with chemotherapy, and now there are phase I data on combining gilteritinib with chemotherapy that look promising.

The second FLT3 inhibitor that will hopefully be approved soon by the FDA is quizartinib, based on the QuANTUM-R study. This may be another drug available for us. Will physicians be tempted to use quizartinib, if it becomes available, in place of midostaurin? It's an interesting thing to consider, but I'm not quite there yet. We have a large, international phase III study, RATIFY, showing an improvement in survival. Although we would all like to see a better improvement in survival by adding a drug in the upfront setting, we don't know that if substituting a more selective FLT3 inhibitor for midostaurin will actually do that. It's been suggested that maybe the benefit of midostaurin upfront is that it's a less selective therapy, so it's inhibiting other tyrosine kinases. Caution is warranted until we have the results of second-generation FLT3 inhibitors with chemotherapy in the frontline setting. QuANTUM-R is the furthest study along.

Are there any issues with quizartinib that need to be addressed?

OS was the primary endpoint of the QuANTUM-R study. OS was statistically improved in the study, but the trouble is that people aren't impressed by the amount of improvement observed. Median survival was 27 weeks with quizartinib and only 20 weeks with chemotherapy, not a big difference. I like to consider it in terms of how many patients are still alive 1 year later; median survivals are hard to grasp. We saw similar numbers there actually: 27% were still alive at 1 year with quizartinib and 20% were still alive with chemotherapy. This is statistically significant, but is it clinically significant?

The other issue with quizartinib has been the myelosuppression, or the potential for myelosuppression. One of the issues is that when you look at the response rates with quizartinib, they aren't as robust as [those seen in] the ADMIRAL trial. In fact, the only thing that has been reported so far is the CR rates, which were doubled with quizartinib. That could be due to the myelosuppression prohibiting count recovery, which is something we have to deal with. On the other hand, a significant number of patients proceeded to transplant after that, which is a goal of therapy. In fact, more patients proceeded to transplant in QuANTUM-R than in ADMIRAL. You can't compare these head-to-head, but it gives you a signal that there may be a benefit.

Another issue with quizartinib is a history of QT prolongation. When this drug was first developed in the phase I study, researchers very quickly ramped up to high doses of quizartinib. However, other studies showed that at very low doses, quizartinib just shut down FLT3. Doses that were 20 times higher than what you need were actually being tested in the phase I study and led to QT prolongation. Several subsequent studies used lower doses. Therefore, there might be some concern with the drug in terms of toxicity.

Finally, the question arises of is a need for a second-generation FLT3 inhibitor in terms of the mechanism in which these drugs work. As I mentioned, midostaurin and gilteritinib are both Type I inhibitors, so they bind to the ATP binding site and shut down the protein. Quizartinib is a type-II inhibitor that stabilizes the inactive confirmation and is only active against the ITD form. In the studies of quizartinib as a single agent, patients with a FLT3-TKD mutation have been known to relapse. Some have speculated that this is a potential mechanism of resistance. However, just because a FLT-TKD mutation is there doesn't mean that it is driving the relapse.

My feeling is that a phase III study was done, and a survival benefit was seen, so let's get the drug approved by the FDA and learn how to use it. Once it is approved, we will learn more about where these drugs fit in.

References

1. U.S. FDA Approves Supplemental New Drug Application Adding Overall Survival Data for XOSPATA® (gilteritinib). Astellas Pharma. Published May 30, 2019. https://prn.to/2HKD05m. Accessed May 30, 2019.
2. FDA Briefing Document Oncologic Drugs Advisory Committee (ODAC) Meeting May 14, 2019 NDA 212166 Quizartinib Applicant: Daiichi-Sankyo, Inc. https://www.fda.gov/media/124896/download. Accessed May 14, 2019.