Harry Erba, MD, PhD: This has been a great discussion of what’s available. Let’s talk about what’s coming and what we’ve learned at ASH [American Society of Hematology Annual Meeting & Exposition]. I’m going to start with Naval. You’ve had a lot of experience looking at checkpoint inhibitors. Why don’t you talk about some of the things we learned about at ASH?

Naval Daver, MD: For immune checkpoint inhibitors specifically, there are a number of trials ongoing at our site [The University of Texas MD Anderson Cancer Center], as well as others represented from other sites. To start with, the ones that are more traditional immune checkpoints for T cells are the PD-1 [programmed cell death protein 1], CTLA4 inhibitors. Our group has been in combinations of those with different HMAs [hypomethylating agents], as well as frontline with induction chemotherapy in maintenance and other settings. The most data where we’re seeing encouraging activity are with HMA in combination with the PD-1 inhibitors, nivolumab or ipilimumab, both in AML [acute myeloid leukemia] and MDS [myelodysplastic syndrome].

At this time, in the relapsed setting with the HMA-nivolumab, we’re seeing about 35% response rates. There are immune itises that occur, in about 20% of the patients, and this is where a learning curve and experience are critical when using these drugs because they are manageable if identified early. If not, they can progress. Pneumonitis and colitis can be quite severe. The key with this whole field, without going into details of each of the abstracts—Dr [Ivana] Gojo presented an abstract from [Johns Hopkins University] looking at HMA plus PEMBRO [pembrolizumab] frontline relapsed. It had very similar data to the relapsed: 30% or so response, improved survival in the early salvage. Dr [Joshua] Zeidner presented some with HDAC [histone deacetylase]/PEMBRO [pembrolizumab]. The key is that all of us showed it using different technologies: TCR [T-cell receptor repertoire]—based analysis, we’ve done RNA sequencing as well as CyTOF [single-cell mass cytometry]. Others have done immune markers using infiltration by flow cytometry, which is pretty easy, looking at CD3.

There is a subset of AML, 20% to 30% where these drugs are giving very high response rates, and these are the immune-activated subsets. They have either high T-cell infiltration to begin with, or they have the RNA signature showing upregulation in interferon-gamma pathways, or they have a TCR clone that you can track. This is how this field has to move forward just as we did with IDH/FLT3. Using this for everybody with a 20% toxicity and a 35% response is not viable, and that’s where it’s going to go. I think now to the very positive checkpoint-based therapy.

Harry Erba, MD, PhD: I’ll just make the point for fair balance. The randomized phase II study of DURVA [durvalumab] with azacitidine versus azacitidine did not show increases in remissions or survival, but probably because we need to, as you’re saying, have the biomarker that identifies the right population. It shouldn’t be the end of investigation; it should be the beginning.

Naval Daver, MD: Yeah, and that’s the other thing. For PD-1, PD-L1 [programmed death-ligand 1], I thought they were quite similar. I’ve been talking to a lot of our immune colleagues. We’re fortunate to have a very powerful immunotherapy platform at [The University of Texas] MD Anderson [Cancer Center]. Just 2 weeks ago, when I asked them, they said absolutely, there’s a huge difference. If you look at the market share, it’s obvious: NIVO [nivolumab]—PEMBRO have 5 to 6 times more than DURVA [durvalumab]–ATEZO [atezolizumab], and there are published data. There are a couple of caveats. It should be biomarker driven—PD-1 and PD-L1, although in general people feel they’re similar, are quite different. Of course, there are other checkpoints such as TIM-3 that were presented here at ASH. It seems to be a more specific AML/MDS checkpoint. That’s the point: usually we’ve had different pathways than solid tumors for molecular, and that may be the same for the immune.

Going along that line, the CD47, 1 of the major highlights of ASH this year, this is a macrophage checkpoint developed very academically by Stanford colleagues Ravindra Majeti and others. They showed a frontline study of azacitidine—CD47 in both AML and MDS. The data in about 48 patients showed very good response rates: 65% CR [complete response]; CR in older AML unfit for chemotherapy and in MDS, using MDS response criteria: about 90% response. More so than the responses to me, the tolerability is quite impressive. The 60-day mortality was 0. They had 1 of 61 patients come off study with median follow-up of 8 months.

They then showed data in the TP53 subset, which is very small in number. But 7 of 9 or 8 of 10 patients had responded, and with follow-up, only 1 had relapsed, which is exciting. Safety is very nice. Of course, it opens the door that all of us are thinking about: can this be combined with HMA—VEN [venetoclax] and other things like that. One thing we do see with that drug’s adverse effect is it does have early anemia, and that’s an on-target clearance of old red blood cells because, apparently, those are the only cells that express CD47. It’s in about the first 10 days that patients get a big drop in hemoglobin. If you can get through that, then it’s well tolerated. Especially for TP53, which we can discuss more, this was very exciting.

Mark Levis, MD, PhD: That’s the neat thing about this concept—this doesn’t uncover a block against the T cell, which is what these other agents are doing. These are macrophage. We figured out how to kill a TP53 cell: you eat it. OK, you just swallow the thing.

Rami Komrokji, MD: That’s a good point because, in MDS, what we started with is that the T cells are not that functional. The quality of the T cells existing are not good. It seems particularly with TP53, it evades the immune system, whether with transplant or anything else, that there is this evasion of the immune system’s abilities.

Mark Levis, MD, PhD: Just swallow that sucker.

Harry Erba, MD, PhD: To follow up on your points about checkpoint inhibitors, the randomized phase II study should give us a pause about bringing this into our routine practice at this point. I know community oncologists, and I’m sure they’re better at handling checkpoint inhibitors than I am, right?

Naval Daver, MD: They are. We see that a lot.

Harry Erba, MD, PhD: I think it should give us pause until we know which group of patients would actually benefit, not only from that immune environment standpoint of who’s going to respond. I’ll point out that the SWOG 1612 study was closed because of a higher mortality. It wasn’t statistically significant, but it had a higher mortality in our NIVO [nivolumab] arm. Some of these appear to be because of NIVO [nivolumab] toxicities, so we need to have some caution there.

Transcript Edited for Clarity