During a Targeted Oncology live event, Grzegorz S. Nowakowski, MD, discussed the case of a patient treated with tafasitamab plus lenalidomide in the second line for diffuse B-cell lymphoma.
Targeted OncologyTM: What are the options for second-line therapy in this patient with DLBCL?
NOWAKOWSKI: The current NCCN [National Comprehensive Cancer Network] guidelines [for patients who are not candidates for transplant] have gemcitabine [Gemzar] plus oxaliplatin [Eloxatin] plus or minus rituximab as a preferred regimen.1
Polatuzumab vedotin [Polivy] plus bendamustine [Treanda] plus rituximab is also included in the NCCN guidelines. Tafasitamab [Monjuvi] plus lenalidomide [Revlimid], which is another option, is FDA approved for second-line therapy and beyond. A lot of us in the field, in patients who are not willing to go for more intensive regimens [such as] transplant or CAR [chimeric antigen receptor] T-cell therapy, are looking more into these chemotherapy combinations, particularly if the patient progresses after chemotherapy. The idea is it’s going to be a different mode of action. CAR T-cell therapy is [used in the third-line setting] as of now. Again, this may change in the future.
What is the rationale behind the patient receiving this combination?
The FDA granted accelerated approval for the combination of tafasitamab and lenalidomide for relapsed or refractory DLBCL based on [results from] the L-MIND study [NCT02399085].2
Tafasitamab has a cool concept where the antibody cells target CD19, just as in CAR T-cell therapy and loncastuximab tesirine [Zynlonta], which is another recently approved antibody. There were initial developments before studying [CD19] where we felt it could be a good target, but some antibodies didn’t work so well. Now there is this renaissance of interest in CD19-targeting agents such as CAR T-cell therapy, tafasitamab, and loncastuximab.
The [tafasitamab] antibody is engineered to have this enhanced Fc function that increases ADCC [antibody-dependent cellular cytotoxicity], ADCP [antibody-dependent cellular phagocytosis], and cell death. It causes direct cell death because CD19 is important in B-cell receptor signaling and not only in the immune system, but it gives some antisignaling properties as well.3
Lenalidomide has the properties of immune activation and microenvironment function and there are dozens of papers postulating many mechanisms of action for lenalidomide. It’s very pleiotropic, but it does immune activation, and we know from R2 [lenalidomide plus rituximab] and other antibody combinations that it tends to synergize with the antibodies very well. This preclinical idea led to the development of the combination of this naked antibody and lenalidomide in patients with relapsed or refractory DLBCL.
Which trial data supported the approval of tafasitamab/lenalidomide?
L-MIND was a single-arm, phase 2 study [that enrolled patients who had] 1 to 3 prior regimens and who were either relapsing after transplant or were not eligible for transplant. The primary refractory patients were to be excluded, but because of changing definitions, they accrued, to some degree, to the study, and had pretty good results anyway.4
Tafasitamab is an infusion, just like other antibodies. It’s given on days 1, 8, 15, and 22, for 1 to 3 cycles. In cycles 4 to 12, it is given every 2 weeks. Lenalidomide is given at 25 mg daily on days 1 to 25, [just as] in multiple myeloma. This is a different dose [from the R2 regimen], which is 20 mg, but the 25 mg was well tolerated, and this was based on the initial [pilot study]. After 12 cycles of therapy, patients received tafasitamab until disease progression.3,4
Frequently [we are asked] why we would plan on continuing forever. I was involved in the design of the study, and the salvage options for patients were quite limited for those who were not transplant eligible and some of the investigators asked why we would want to stop if it is working. We gave investigators discretion to [decide] whether the patient was benefiting from the treatment and to continue until disease progression. The primary end point of the study was overall response rate [ORR], which has frequently been the most reliable end point for the activity of the combination in this setting because it tends to have less bias in patient selection. The secondary end points were PFS [progression-free survival], duration of response [DOR], overall survival [OS], and so forth.4,5
There were some lenalidomide dose reduction studies where patients were given doses of 25 mg down to 5 mg using step reductions.5
This was a study of the [safety] population, and 81 patients were accrued overall. The median age was 72. The IPI risk score, Ann Arbor stage, and LDH results were typical for refractory DLBCL. Patients with primary refractory disease were supposed to be excluded, but 19 of 81 patients had it and 44 of 81 patients were refractory to prior therapies. Relatively few patients had a prior stem cell transplant and the majority were not eligible for it due to comorbidities, unwillingness to do so, or not responding to salvage therapy. [Not responding] to previous therapies was a major reason [for not getting a transplant].5
How did patients do on the L-MIND trial?
The ORR for this combination was quite high at greater than 60%, which is comparable with what we see in CAR T-cell therapy or intensive chemotherapy. So this was quite significant and impressive at the time the [results were] published. The CR [complete response] rate was even more impressive at 43%. Again, this was in patients who were relapsed or refractory, not transplant eligible, or those relapsing after transplant, so a 43% CR rate is high.5,6
As clinicians, we care about the DOR, too. So if you are a regulator, say at the FDA, you only worry about response rates because it’s less about patient selection, but clinicians like responses to be durable. The median PFS was 12.1 months.5 The median PFS doesn’t fully reflect the activity of this regimen because it plateaus just after the median. CAR T-cell therapy data look very similar, too. For a relatively well-tolerated combination, these were very impressive results at the time of presentation. The median OS was not reached and, as with the PFS results, the OS also plateaued. So these were very impressive results in terms of DOR.
The patients in CR were primarily driving this benefit, but even the patients in PR [partial response] had [an approximately] 30% sustained response.6 The treatment was active in the patients treated both with 1 prior or 2 or more prior lines of therapy. Responses, particularly the CR rates, were somewhat higher in the patients who were on second-line treatment. This would be the patients who were not eligible for transplant.
Do you feel comfortable using this regimen in patients with GCB [germinal center B-cell–like] subtypes because they were underrepresented in the study?
There was a whole debate about it. We believe that the combination of the antibodies and lenalidomide works well in GCB subtypes as well. It is a little bit different with single agents because the data showed response rates and activity were better in ABC [activated B-cell] or non–GCB subtypes of DLBCL, but in combination, there appeared to be less of a differential by cell of origin.
But in the [forest plot] analysis, both subtypes benefited. There was a trend toward a little bit of a high response rate in patients with the ABC subtype, but overall, the response rate was high in patients with GCB patients as well. I believe it was approximately 45% to 50% in both subtypes.
What about the R2 regimen? Do you prefer not to use it in GCB subtypes?
Yes, I prefer not to use it in GCB subtypes. [Results of] the ECOG-ACRIN E1412 study [NCT01856192] were recently published in the Journal of Clinical Oncology and I was a PI [principal investigator] in it.7 This study was looking at all-comers, so it was the only randomized frontline phase 2 study, where lenalidomide was added to R-CHOP. This one was cell-of-origin agnostic, so they could have the GCB or ABC subtype. There was [approximately] a 12% difference in PFS in this study and a favorable hazard ratio.
Another study, the ROBUST study [NCT02285062], was focused on patients with the ABC subtype.8 It didn’t show a difference using different lenalidomide scheduled doses, though there were other patient selection issues in the study. As a single agent, lenalidomide is more active in the ABC subtype and I use it myself in clinical practice more in ABC or non–GCB subtypes. In combination with the antibodies, or even chemotherapy, this may not be necessarily true. Because most of these patients are already exposed to rituximab, I think based on the R2 study [results], they didn’t see much of a differential based on cell of origin, which is a little bit disappointing, because we were hoping we could [use it to] select the high responders, but that didn’t pan out. REMARC [NCT01122472] was a study done by a French group that used lenalidomide maintenance after R-CHOP but didn’t track the cell of origin.
In fact, the GCB subtype tended to benefit more, and an idea was that maybe some microenvironment influences played a role…. In my clinical practice, in non–GCB subtypes, I use a single agent, but for combination of the antibodies, the activity seems to be agnostic to cell of origin.
How does an anti-CD19 antibody downregulate the CD19 receptor?
There is limited information, but they did a study looking at the CD19 expression after tafasitamab exposure in chronic lymphocytic leukemia and [there was no impact] and in DLBCL as well. The CD19 expression is just a part of the story because you worry that a part of the CD19 molecule could be mutated and then the CAR T-cell agents would not bind or that part of the molecule could be missed because of alternative splicing or losing one of the exons because of the evolutionary pressure of the treatment. We did whole exome and RNA sequencing and saw no abnormalities within the CD19 cells. It appears to be expressed after tafasitamab exposure, and there are no point mutations, exon deletions, or other changes that would affect the integrity of CD19, to the best of our knowledge.
Of course, the best data would come from clinical evidence if we note that CAR T-cell therapy is working. In this study, only 1 patient proceeded with CAR T-cell therapy and had good clinical benefit and was in remission last time I saw the data. So it appears that in anecdotal experiences CAR T-cell therapy will still work in those patients.
The opposite is true, too. There is a huge interest now in this combination and [whether] it will be active in post– CAR T-cell relapses. Lenalidomide as a single agent is frequently used in this setting. How active will this combination be in post–CAR T-cell relapse? We know that lenalidomide is active. A lot of patients with CAR T-cell relapses will still have CD19, so we believe that is also an option, but more data will be needed.
Do patients tolerate the 25-mg lenalidomide dose in combination with tafasitamab, or is the dose modified often?
[Approximately] 30% of patients will have to drop to 20 mg, particularly with subsequent cycles. The nice thing for lenalidomide is that you can use the growth factor support because it is primarily neutropenia that causes some of the dose reductions. Studies are different from real life, so in the real world we always have some patients who are already cytopenic from the previous therapy. I usually support them with a growth factor, and sometimes I start my patients at 20 mg. The dosing intensity of lenalidomide seems to be important, though.
I wouldn’t very liberally decrease it because there appears to be some dose relation to the response, at least as a single agent in a refractory setting in DLBCL in contrast to follicular [lymphoma], but somewhere from 15 mg to 20 mg is the golden spot for response.
The 25 mg was used in those studies as a single agent, so, about one-third of patients did require dose reductions. If you use this combination, you follow the lenalidomide package inserts, and if you need to reduce because of creatinine clearance, you reduce the lenalidomide or if you see significant neutropenia despite the growth factor used, then you can reduce on a subsequent cycle to 20 mg, or interrupt and reduce to 20 mg.
Does patient preference weigh into the decision to choose finite therapy vs therapy until progression of disease in the second-line setting?
Yes, it comes down to the patient’s preference. I don’t practice in the community, so I don’t have more experience with this. We have this policy at Mayo Clinic that [any clinician] from around the world can call us at any time for advice about their patients. So, routinely, we are getting quite a few phone calls from those who are responsible for patients with lymphoma, or for any other disease type from outside, and practitioners call asking what to do.
I am always surprised by how many patients do not want to proceed with CAR T-cell therapy or stem cells or even clinical trials, which we often have here, because of the preference of being near the local center. Travel is not always possible and some patients want to stay where they are, which is a very reasonable option.
Are there trials comparing this with transplant or something lenalidomide alone?
We did 2 things to differentiate this from lenalidomide alone. A study called RE-MIND [NCT04150328] with close matching of the patients with real-world data showed that the combination was definitely much more active than lenalidomide alone. [We knew this] but wanted to double-check in a very close-matched cohort. A confirmatory study for this is [the frontMIND study (NCT04824092), which is a frontline study that compares] R-CHOP as standard therapy vs R2-CHOP plus tafasitamab.
I am the principal investigator globally for this study, and one of the reasons why we designed it this way was there was some activity already from randomized phase 2 studies using lenalidomide. It was safe and effective and also the doublet was already approved, so it was logical to move it forward.
However, the biggest [issue we had when] presenting this concept to some regulatory authorities was that we were a little bit naive in the past, thinking that adding 1 drug at a time is going to move the bar a whole lot. R-CHOP already has 5 different compounds, so I think the sixth one probably is not going to move the bar a whole lot. There are some studies that failed, I think, 1 drug at a time. So the ambitious plan here is to add a doublet. But the study is designed to capture very high-risk patients, [meaning] IPI 3 and above. It’s looking at the highest-risk population and is adding doublet on top of R-CHOP. There are some study centers in the United States that are in the process of either opening or even have it open currently.
Could tafasitamab/lenalidomide be moved to the first-line setting with more targeted agents as chemotherapies are eliminated?
Yes. There is a pilot study led by my colleague Dr [Jason] Westin at [The University of Texas MD Anderson Cancer Center]. He is basically pioneering the so-called smart-start, or smart-stop now, where he is adding exactly this combination to R-CHOP. The question is: Can he strip some of the chemotherapy agents [such as anthracyclines]?
[The patient] tried to shorten and then to remove different cytotoxic drugs with the idea that maybe over time he can develop a chemotherapy-free regimen. [Results of] the initial pilot study have shown this combination plus ibrutinib [Imbruvica] is producing high response rates. He still added chemotherapy later because he was worried that he may miss the possibility of curing the patient, but after initial feasibility, he is slowly stripping chemotherapy. We may get there one day.
What are the similarities and differences of loncastuximab tesirine and tafasitamab?
I think cross-study comparisons are usually difficult. I am very cautious always when comparing different study results because the patient population is not always the same. I happen to be involved with the FDA in different reviews and I do believe that the response rate is what tends to reflect the most activity and is less dependent on patient selection, though not completely.
The ORR of loncastuximab is [approximately] 50% or very close to that. The DOR appears to be a little bit shorter, but this could be due to patient selection, so it looks very encouraging. It has a little bit of a different adverse event [AE] profile. At this point it doesn’t have as strong a follow-up as this study, so we don’t know if the same very encouraging plateaus in responding patients will be seen with it.
Maybe it’s going to happen, but it is more of a traditional cytotoxic therapy that is directed like polatuzumab. It works more on the immune microenvironment in immune activation. There is this renaissance of CD19 targeting and for CAR T-cell therapies, all the approved products target CD19, and now loncastuximab and tafasitamab.
I usually tell the industry to not develop any more agents targeting CD19. We have enough. There are some other good targets, too. Some of the CAR T-cell therapies are targeting different molecules on the surface.
How many of these patients on the L-MIND trial stopped therapy early? What is the safety profile of combination lenalidomide and tafasitamab?
The primary reason for stopping therapy early was disease progression because some patients just didn’t respond. The toxicities were primarily hematologic, which is consistent with what you would see with lenalidomide. Nonhematologic AEs [included] fatigue and diarrhea, but nothing striking or unusual. Discontinuation of combination [therapy due to] AEs was seen in 12% of the patients [n = 10/81].5
A comparison of the AEs of combination therapy vs monotherapy showed the hematologic and other toxicities were driven by lenalidomide. Tafasitamab alone had [an approximate] 27% ORR and when combined with lenalidomide the response rate doubles, so there’s a true synergy between those drugs.
The monotherapies are quite well tolerated. Some patients can develop neutropenia, as was seen in the monotherapy trials, but overall the toxicity is minimal for the antibody alone.
What is the rapidity of the response for this regimen? Who wouldn’t be eligible for it?
The first evaluation was done after 2 cycles of therapy, so within 8 weeks the response was right there. The response is quite brisk. If I had any concern about putting [a patient] on lenalidomide, it would be for reasons such as it can cause some rashes as seen previously with lenalidomide combinations, so with previous hypersensitivity, I probably would not [use it].
If patients have very rapidly progressive symptoms, I may stabilize them with radiation or some other treatment first, maybe hydroxysteroids, rituximab, or something such as that just to remove the disease burden before I start this combination. I expected that the responses would be dipping over time, but the responses were brisk and happened after 2 cycles of therapy.