Jorge E. Cortes, MD: We have [therapies] that are targeted, but many patients don’t have a specific target. They don’t have FLT3 or IDH. What do we use for salvage [therapy]? I want to divide them into the fit and the unfit. What’s the standard [therapy]? Let’s start with fit patients. What do we do for a fit patient who has relapsed/refractory [disease] and doesn’t have FLT3 or IDH? Margie, what’s your approach? Do you use CPX-351, for example?

Margaret T. Kasner, MD: I’m going to fall back on my clinical trial answer. If that isn’t an option, I encourage people to think about that if they can, because we don’t have a good standard. Do I use CPX-351? I do, but more rarely than I use something like FLAG [fludarabine, cytarabine, GCSF]–idarubicin–venetoclax if they haven’t seen venetoclax in the past. If the thing they already got was more like a CPX-351– or 7+3-type agent, I might pick a different combination. I’m not a huge fan of azacitidine- venetoclax in the relapsed setting, especially for a very fit patient who could move to transplant; the data aren’t excellent in that setting. I try to stick with a more aggressive approach. There are a lot of factors. Are they refractory vs did they have a long time when they didn’t relapse? In that case, for relapsed patients I like a high-dose cytarabine-based regimen if they haven’t seen that in a long time.

There isn’t a good standard or a great answer for these patients. If they’re fit and their cardiovascular status is good and they can get some more midostaurin or idarubicin, and some high-dose cytarabine, that will get them in remission to get to transplant. The question is, are your transplanters into the transplant at nadir vs in remission and transplant after that? That’s a conversation I have with transplanters in terms of the next step, because for fit patients transplant is my answer.

Jorge E. Cortes, MD: What about gemtuzumab? Gemtuzumab came, went, and came back. One of the approved indications is for a single agent for relapsed patients. Hetty, where do you put gemtuzumab in your algorithm? Do you use it at all in this context?

Hetty E. Carraway, MD: I’ve had a number of young patients relapse lately, even 1 with core-binding factor [acute myeloid] leukemia who relapsed a year and a half after induction 7+3, with GO [gemtuzumab, ozogamicin] in the induction and consolidation. We were thinking about what to do for this patient as we reinduce him because he should respond to 7+3? Do we add GO [gemtuzumab, ozogamicin] in the setting? We’re eager to get him to transplant. We worry about VOD [veno-occlusive disease]. He had GO [gemtuzumab, ozogamicin] before, a year and a half ago, but we’re eager and hopeful that we’re moving to CR2 [second complete remission] and to transplant. I don’t know the perfect answer to that. It can be different for different folks. For him, I don’t want to do anything to compromise this patient’s ability to get to transplant. He’ll likely have a nice response to 7+3, so I may forego GO [gemtuzumab, ozogamicin] in that setting.

[On the other hand] there are patients who have relapsed leukemia who are not responding to therapy and who are relapsing less than 1 year. GO [gemtuzumab, ozogamicin] is a reasonable therapy that’s well tolerated and can be given in the outpatient setting. For many of those reasons, we can use that as an agent for our patients. We try to use clinical trials. I’m a fan of FLAG [fludarabine, cytarabine, GCSF]–idarubicin–venetoclax, MEC [mitoxantrone, etoposide, cytarabine], and high-dose cytarabine. Any of those options are reasonable, but we typically land in the clinical trial arena. The problem I have with GO [gemtuzumab, ozogamicin] is that I don’t use it if we’re looking to obtain a CR2 and get them to transplant, because I worry that the CR [complete remission] won’t be long enough in duration to get them to that transplant. That’s where we land with GO [gemtuzumab, ozogamicin].

Jorge E. Cortes, MD: Both you and Margie mentioned FLAG [fludarabine, cytarabine, GCSF]–venetoclax. Those data look very good. It’s a single-arm [regimen], but it looks very attractive. It will be interesting to see what evolves with that. Other combinations will be more intensive, including venetoclax. Brian, what about the unfit patient who doesn’t have a specific mutation? There’s FLT3 and IDH. They’ve probably already gone through azacitidine-venetoclax. You’re going to tell me clinical trials, but I don’t have clinical trials. What do I do?

Brian Andrew Jonas, MD, PhD: Run for the hills, unfortunately. We’re in a very tough situation there. [The University of Texas] MD Anderson [Cancer Center] was the first to show its experience with this population with dismal outcomes. There was another abstracted at ASH [American Society of Hematology Annual Meeting]—I believe it was the Mayo [Clinic] experience—and it was the same story. We’re looking at median survivals of 2 to 3 months in these patients who are HMA [hypomethylating agent]–venetoclax failures. It’s depressing. The answer is clinical trials because once they fail that treatment, unless they have a mutation, we don’t have a lot to offer. Sometimes I’ll try gemtuzumab in those patients, and I’ve been able to get some mileage out of it in that setting. I’ve tried cladribine-based treatments, largely based on the MD Anderson combination studies with low-dose cladribine and low-dose cytarabine. In those situations, the results aren’t very appealing. If I don’t have a trial available, I’ll discuss those with patients as potential option. But a frank discussion with them about the odds they’re facing is almost as important as which treatment you’re going to choose, so that expectations aren’t unrealistic.

Jorge E. Cortes, MD: You mentioned gemtuzumab, and that was initially approved. That’s probably the most data we have. Not much has been published after azacitidine-venetoclax, so it’s hard to tell how much it will change that dismal outcome, but it’s something to consider. Something we didn’t talk about in the unfit patient as frontline [therapy] is low-dose araC [cytarabine]–glasdegib, which is an approved combination. It’s gone to the rearview mirror with the azacitidine-venetoclax data. In some patients where you’re worried about myelosuppression, that’s where I’ve used it. There’s no data, but would you consider it a salvage [therapy] after azacitidine-venetoclax? Anybody want to give it a shot?

Hetty E. Carraway, MD: We tend to use low-dose araC [cytarabine]. We’ve had patients who’ve tolerated that well. When I’ve run out of options, I’m not shy of using low-dose araC [cytarabine]. I’m not eager to add glasdegib. I’m not sure it adds more to the likelihood of response for patients primarily because the control arm in that study had poor outcomes with single-agent low-dose araC [cytarabine]. That’s where we’re on the fence. Will I use low doses of cytarabine? Certainly, especially if patients have had limited exposure to it.

Transcript edited for clarity.