Nilanjan Ghosh, MD, Levine Cancer Institute, Charlotte, North Carolina, highlights preliminary results from a dose-escalation and safety run-in phase 1b study of glofitamab plus R-CHOP, which was shown to induces high response rates with minimal cytokine release syndrome in patients with relapsed or refractory (R/R) non-Hodgkin lymphoma (NHL), including follicular lymphoma (FL), as well as previously untreated diffuse large B-cell lymphoma (DLBCL). These data were presented at the 2021 American Society of Hematology (ASH) Annual Meeting.

Transcript

Hi, I'm Nilanjan Ghosh from the Levine Cancer Institute in Charlotte, North Carolina. I lead the lymphoma division over there. I'm presenting you from the American Society of Hematology meeting at 2021 from Atlanta, Georgia.

Despite recent advances in relapsed/refractory (R/R) non-Hodgkin lymphoma (NHL), we still have significant unmet needs. Patients with R/R disease have poor outcomes.

That led us to incorporate this novel T-cell engaging bi specific antibody, glofitamab, into well-established chemotherapy background, R-CHOP, which everyone is familiar with, in R/R NHL to be able to assess safety and to have the optimal dose finding. Then apply that to previously untreated diffused large b-cell lymphoma (DLBCL), where, with front-line therapies, we are able to potentially cure at least two-thirds of patients, but one-third of patients still relapse. Even though there are salvage treatments available, there is still an unmet need in that space. If we are able to cure more patients in the frontline, that would be the ultimate goal. 

That led us to incorporate this treatment with the background, knowing that the single agent, glofitamab, has already shown activity in R/R NHL. When you have an active compound, in this case, a bispecific antibody, then adding it to a known chemoimmunotherapy background would hopefully give us new frontiers in treatments for this disease. 

Glofitamab is a T-cell-engaging bispecific antibody with a 2:1 molecular configuration that allows for high-affinity bivalent binding to CD20 on B-cells and monovalent binding to CD3 and t-cells.

In this study, we first started with R/R NHL. Patients who had previously been treated and relapsed or were refractory received 1 cycle of either R-CHOP or obinutuzumab-CHOP. The aim for that was to debulk the tumor and lower the risk of cytokine release syndrome. Then glofitamab was given from cycle 2 onwards. We used a step-up dosing of the glofitamab, again to mitigate the risk of cytokine release syndrome. The target dose of glofitamab was reached from cycle 3. In cycle 2, glofitamab was given on cycle 2, day 8 and 15, and then the target dose reached in cycle 3, day 8.

Ultimately, we found the target dose of glofitamab as 30 mg, which was chosen as the optimal biological dose. The step-up dosing included using 2.5 mg and 10 mg in cycle 2 on day 8 and day 15, respectively. 

Then we assessed for responses based on the end of treatment by the standard Lugano response criteria. We then applied this to the frontline setting in previously untreated diffused large-cell lymphoma (DLCL). We took patients who had ECOG performance even up to 3, so 0 to 3. They received glofitamab, 30 milligrams, after the initial step up in cycle 2. First cycle with R-CHOP, second cycle was R-CHOP, less step-up dosing of glofitamab, and from cycle 3 to 30 mg of glofitamab on day 8. Again, we assessed for the same endpoints of complete response at end of induction, using the investigator-assessed and by the Lugano criteria. We also looked at safety, looking at CRS and other safety events.

The baseline characteristics for patients with R/R NHL showed that there were several patients with high-risk disease as shown by their refractoriness to prior therapy. There were 55 percent of patients who were refractory to any prior therapy, and 48 percent of patients who were refractory to the most immediate prior therapy. We know that refractory patients don't do well, but a significant number of patients in this study had refractory disease. 

When we look at the breakup of histology for R/R NHL, you see that follicular lymphoma (FL), grades 1 to 3a, were the majority of the patients, but in the R/R cohort, we also included transformed FL, which is a more aggressive form and is often treated in similar lines as aggressive B-cell lymphoma. We also included patients with marginal-zone lymphoma (MZL) and mantle-cell lymphoma (MCL), although there were only 1 patient each from those. We had 31 patients in this R/R cohort. In the front-line DLBCL cohort, we had 26 patients. In that group, we looked at the IPI to risk stratify these patients. We found that 46 percent or so of patients were high-intermediate or high-risk IPI. We also had 65 percent of patients who had extra-nodal disease. This highlights the high-risk baseline characteristics for these patients who were enrolled in the study. 

Most importantly, wanted to talk about the results in terms of efficacy. When we look at the R/R NHL cohort, we see that this regimen of glofitamab plus R-CHOP is very effective, with an overall response rate (ORR) of 90 percent, with a complete metabolic response of 81 percent. That's really high in this R/R population. When we look at the ORR in the indolent cohorts, that's including FL and MZL, we see an overall response rate of 92 percent and a complete response rate of 83 percent. When we look at the transformed lymphomas and the MCL, we see an ORR of 86 percent and a complete response rate of 71 percent. Still, very high responses in R/R disease. The important thing is the median duration of response was not reached, indicating that the responses were durable.

When we now move on to the previously untreated DLBCL, the ORR was really high at 100 percent over here. The numbers are small because, at the time of data cutoff, only 9 out of 26 patients had reached the end of treatment assessment. The complete response rate (CRR) was 89 percent, and 1 patient had a partial response (PR). Everybody responded in this cohort. We also note that this is very manageable from a safety perspective.

If we focus at R/R NHL, adverse events (AEs) were seen in almost all patients, but in terms of related to glofitamab, 84 percent of patients had AEs. If we look at higher grade AEs, 90 percent of them had grade 3 or higher Aes, with 65 percent of the AEs related to glofitamab.

There was 1 fatal AE in the relapsed/refractory cohort, which was unrelated to glofitamab. It was an unfortunate death from COVID-19 pneumonia. When we look at ICANS (immune effector cell-associated neurotoxicity syndrome), it's very low. Only 1 patient, so 3 percent of patients had ICANS, which is an important adverse event with immune effector cells. 

The adverse events leading to glofitamab discontinuation were seen in 3 patients. Three patients had to discontinue glofitamab because of an AE. That translates to about 10 percent of patients. About 29 percent of patients had an AE leading to glofitamab modification or dose interruption. Neutropenia was seen. Again, this is with a chemoimmunotherapy backbone, so sometimes it may be hard to tease out whether the neutropenia is related to chemotherapy or to glofitamab. But, 77 percent of patients had neutropenia grade 3 or above. Febrile neutropenia was seen in 19 percent of patients. Tumor flare was not seen in any patients. 

When we contrast this to the front-line DLBCL, we see a significant decrease in the number of adverse events. We do think that in R/R patients, the number of adverse events are usually higher. In the frontline setting, grade 3 or higher AEs were seen in 58 percent, and only 15 percent were considered to be related to glofitamab. That is a significant difference from the R/R setting, where it was more like 65 percent related to glofitamab in terms of grade 3 or higher. 

There was 1 patient who had a fatal AE, but it was not related to glofitamab. In fact, in that patient, glofitamab had not yet been started. It happened in cycle 1. If you remember, cycle 1 was only R-CHOP, so this adverse event happened at that time, but because the study captures everything in totality, this is included for transparency. There were no AEs which lead to discontinuation of glofitamab in the frontline setting. There was no ICANs in the frontline setting. Even if we look at neutropenia, it was lower at grade 3 or higher being only 42 percent in the frontline setting.

Overall, the safety profile, even including infections, 7.7 percent, and only 3.8 percent grade 3 or higher, and no tumor flare. The safety profile looks really good in frontline DLBCL, certainly very manageable as well in the R/R setting. 

Finally, we are all interested with these bispecific antibodies to look at cytokine release syndrome. We know that this is an important AE of special interest currently. In the R/R NHL cohort, the cytokine release syndrome events were mostly low-grade. 

If you look at all patients who got cytokine release syndrome, it was seen in 17 patients, so 55 percent. Ten of those 17 were low-grade, so that's 32 percent. There were 3 patients with grade 3 or higher, and 4 patients with grade 2. That was the breakup in terms of R/R, in terms of serious AEs of CRS. It was 8 patients, so 26 percent. And 8 patients needed tocilizumab use for the R/R.

As I mentioned previously, the side effect profile was much better in the frontline setting. In the safety run-in phase, we saw only 1 patient who had cytokine release syndrome, so it's a really low number, 3.8 percent, and that patient had grade 1. There was no grade 2, grade 3, or higher CRS. There was no serious CRS, and there was no tocilizumab use. This is another very important thing because our primary interest is to look at this regimen in the frontline setting.

Now, we can conclude that glofitamab plus R-CHOP has excellent efficacy. We saw both in R/R and in the frontline setting, and can certainly be combined with R-CHOP.

One other important thing was that R-CHOP dose intensity was maintained in all patients. CRS was low. In R/R where it occurred, mostly was low-grade CRS. And ICANS was nearly absent in the frontline setting certainly. There were no events of ICANS at all.

This would mean that this combination is suitable for outpatient setting, at least in the front-line setting, and that's what we're looking at in the next step. Thank you.