Longest Follow-Up Maintains Ibrutinib Benefit in CLL/SLL Across Treatment Settings
April 13, 2020 05:00pm
By Lisa Astor
During a Targeted Oncology live case-based peer perspectives presentation, Anthony Mato, MD, MSCE, explained to a group of physicians the diagnostic workup and treatment considerations and decisions he makes when seeing a patient with chronic lymphocytic leukemia in the clinic.
Anthony Mato, MD, MSCE
During aTargeted Oncologylive case-based peer perspectives presentation, Anthony Mato, MD, MSCE, explained to a group of physicians the diagnostic workup and treatment considerations and decisions he makes when seeing a patient with chronic lymphocytic leukemia (CLL) in the clinic. Mato, the director of the Chronic Lymphocytic Leukemia Program at Memorial Sloan Kettering Cancer Center, discussed treatment options in relation to a case scenario of a patient with newly diagnosed CLL.
A 75-year-old woman reported symptoms of weight loss and fullness in her left upper abdominal quadrant. Her past medical history included hypertension and hypercholesterolemia, both medically controlled.
A physical exam was notable for moderate axillary lymphadenopathy, and her spleen was palpable about 4 cm below costal margin. Otherwise, she was well appearing and continued her daily activities. Her Coombs test was negative.
Laboratory findings were remarkable for a white blood cell count of 48,000/μL with 73% lymphocytes; hemoglobin, 9.1 g/dL; platelets, 125 × 103/µL; absolute neutrophil count, 1800/µL (within normal limits); lactate dehydrogenase, 250 U/L; and β2-macroglobulin, 4.2 µg/L.
Flow cytometry indicated CD5-positive, CD19-positive, and CD23-positive disease. Fluorescence in situ hybridization (FISH) testing revealed an 11q deletion (del[11q]), and molecular analysis indicated IGHV unmutated and TP53 wild-type disease. Bone marrow biopsy revealed diffuse infiltration by CLL.
What do you test for when you see a patient with newly diagnosed CLL who needs treatment?
Prognostic testing is something that is a moving target in CLL. The way I like to think of it is that the tests are only as good as the patient populations for whom they are studied. For example, del(17p) may have less significance when we are talking about targeted agents.
I actually think there is still value in assessing for a del(17p) or a p53 mutation, even if you are giving a patient ibrutinib, because those are the patients who tend to have slightly inferior outcomes compared with patients who do not have del(17p). Of course, there are not a lot of frontline data for ibrutinib in patients with del(17p) based on the RESONATE-2 trial [NCT01722487], which included patients who were 65 years and older who did not have del(17p).
RegardingIGHVstatus, you could argue maybe less importance, but even for patients who are receiving a targeted agent like ibrutinib, theIGHVstatus actually helps predict how quickly the treatment-related lymphocytosis will resolve. Patients [with a mutation] can have a lymphocytosis that can last up to a year; patients [with no mutation], less than that, probably 6 months as the median time to resolution.
In our practice, we do test FISH, we do next-generation sequencing, and we do check forIGHVstatus. And you could also argue if you are going to perform a bone marrow biopsy, then you would want to perform karyotyping to assess for a complex karyotype, all of which are probably important prognostically.
In terms of del(11q), this is where you have to think about the patient population. If you are thinking about giving a targeted agent to patients, del(11q) probably does not have any meaning anymore. Ibrutinib completely overcomes that. It also completely overcomes theIGHVstatus, and it almost completely overcomes the del(17p), although I would argue there is still some room for improvement in those patients.
How would these characteristics factor into your decision for a first-line treatment regimen?
In terms of our patient, when we want to look at prognostic factors [such as] the presence of del(11q) versus the absence of del(11q)…probably [the patients with] del(11q) do a little better with ibrutinib than the [patients without] del(11q), and thenIGHVstatus does not seem to have any prognostic significance for patients receiving ibrutinib as a monotherapy.
What data would factor into your decision of which frontline regimen to give to this patient?
[To start with, there is] the CLL11 trial, which was conducted by the German CLL study group and presented back in 2015. It was a direct comparison between the anti-CD20 glycoengineered monoclonal antibody obinutuzumab [Gazyva] in combination with chlorambucil as compared with chlorambucil as the primary analysis and then a secondary comparison with rituximab [Rituxan] plus chlorambucil.1The big questions here, at least from a regulatory perspective, were “Could you beat chlorambucil?”not a hard thing to do—and then from a scientific perspective, “Does the glycoengineered antibody give you any advantage over rituximab?”
This is a frontline trial, not necessarily in elderly patients, but it turned out they were mostly elderly because the patients had to have significant comorbidities to participate in the study, so they were usually older patients with comorbid conditions. It was a large trialprimary comparison, 238 versus 118 patients. Not a major surprise; the antibody plus chlorambucil was better than chlorambucil in terms of progression-free survival [PFS] and time to next therapy, and also overall survival [OS] was positive here. Across the board this became a new standard of care.
I think more interestingly, when the comparison was made with obinutuzumab/chlorambucil versus rituximab/chlorambucilhere we are talking about a 663-patient comparison, pretty long-term follow-up now—obinutuzumab was better than rituximab in terms of PFS and, with longer-term follow-up, OS.
When thinking about using a monoclonal antibody, if that is the treatment strategy in the frontline setting, one should probably choose obinutuzumab for patients over rituximab. [I personally do not think] chlorambucil adds anything to the regimen. So when I prescribe this regimen, if I am thinking about giving obinutuzumab, I just give it as a monotherapy. I think the chlorambucil adds toxicity, nothing more, and probably secondary leukemia somewhere along the line if the patient lives long enough after the treatment and is young enough when they receive it.
The PFS for obinutuzumab/chlorambucil versus rituximab/chlorambucil showed a major difference between the curves. The hazard ratio was 0.49, and thePvalue was <.0001. This was a positive trial, and again this influenced the standard of care. The OS data with longer follow-up do appear to remain significant, so maybe they built a better antibody.
The next trial that I will mention is the RESONATE-2 trial. You start to see this theme with these CLL trials, in that the comparison here is ibrutinib compared with chlorambucil.2This is a frontline trial; patients had to be 65 years or older to participate. There are not a lot of data for ibrutinib in a younger patient population based on this initial approval. Patients could not have del(17p) and participate in this trial, just to give a sense of the patient population. Again not a big surprise, ibrutinib was overwhelmingly better in terms of PFS compared with chlorambucil. In addition, it was also superior in terms of OS.
We are starting to see this theme across all the CLL studies in the frontline setting, that there appear to be OS advantages, particularly with the targeted agents…. Chlorambucil monotherapy is a regimen that was introduced in the United States in the 1950s and probably is not the standard of care anywhere in the United States any longer.
The next trial that I will mention is the ECOG-E1912 study. This data set, I think, is helping to confirm what, for most oncologists, is the standard of care already. This is a frontline trial that did a 2:1 comparison of FCR [fludarabine, cyclophosphamide, and rituximab], which is standard of care to many providers, compared with ibrutinib plus rituximab.3These patients had to be less than 70 years and could not have a del(17p) to participate in this trial; the primary comparison was PFS. FCR was given as a standard dosing planned for 6 cycles of therapythe way it was done in all the previous studies—and ibrutinib and rituximab were given with ibrutinib as a monotherapy for the first month, the combination with rituximab for 6 cycles, and then beyond cycle 7, ibrutinib as a monotherapy with a treat-to-progression strategy. This is standard ibrutinib plus rituximab for cycles 2 through 7.
The data were presented at the 2018 American Society of Hematology [ASH] Annual Meeting as part of the late-breaking abstracts. Maybe it is a surprise that ibrutinib/rituximab was superior in terms of PFS compared with FCR. But the big surprise was that OS with ibrutinib/rituximab was better than FCR across all patients who participated in the trial. When they looked at subgroup analyses, that advantage was maintained whether the patient wasIGHVmutated or unmutated but was statistically significant only in the patients [with no mutation]. For young patients [with no mutation] who are fit, in my mind ibrutinib is now the standard of care. For patients [with a mutation], I think you can still make an argument for, or at least have a discussion about, chemoimmunotherapy, specifically FCR, but I also think it is very reasonable to provide ibrutinib to these patients.
Now, the bigger question that comes out of these data is “Does rituximab need to be combined with ibrutinib to have this result?” I would say probably not because the next trial that was presented at ASH and subsequently published in theNew England Journal of Medicinewas the ALLIANCE trial, which compared BR [bendamustine and rituximab]a very popular favorite—to ibrutinib/rituximab or ibrutinib as a monotherapy.4The primary endpoint here was a comparison of the targeted agent with BR in terms of PFS. The primary endpoint was met, as ibrutinib-based therapy was better than BR in terms of PFS. Here they did allow patients to participate who had del(17p), so [the design was] a little different from the ECOG data.
One of the findings that I think is important from these data is that rituximab added nothing to ibrutinib. Whether a patient got ibrutinib/rituximab or ibrutinib by itself, the results were completely superimposable. Extrapolating that back to the ECOG data, I do not think there is a lot of evidence for adding rituximab to ibrutinib. If an oncologist is going to give it as a first-line therapy and is considering ibrutinib or ibrutinib/rituximab, they should probably just give ibrutinib and treat to progression. Two-year PFS for BR was 74%, 87% for ibrutinib, and 88% for ibrutinib/rituximab. Again, no difference. The OS in this trial, unlike that of the ECOG data, was not significant.
The rate of grade 3 to 5 nonhematologic AEs [adverse events] was similar [across the arms of the trial], but there were certainly more grade 3 to 5 hematologic AEs for chemotherapy.
To make things even more confusing, I will also talk about another trial that was an oral presentation at ASH called the iLLUMINATE trial, in which ibrutinib plus obinutuzumab was compared with the original CLL11 regimen obinutuzumab/chlorambucil.5This, again, is a treat-to-progression strategy. Ibrutinib/obinutuzumab was better than obinutuzumab/chlorambucil, which makes one wonder whether the CLL11 regimen should still be considered a category 1 recommendation because now we have a regimen that is better in terms of PFS but not in terms of OS.
People are probably wondering, “Do you need this anti-CD20 with ibrutinib?” This trial did not include an ibrutinib monotherapy control, so we do not know that answer. I think the jury is out. I am pretty comfortable saying that if an oncologist is thinking about using rituximab with ibrutinib, they do not need to. I cannot confidently say how much more the obinutuzumab is adding to ibrutinib here. One thing I can say is that we see higher rates of complete response [CR] with obinutuzumab/ibrutinib than we would have seen with ibrutinib as a monotherapy.
It is not a big surprise that this patient was started on ibrutinib.
The patient was started on ibrutinib (Imbruvica) 420 mg daily.
Is there a role for obinutuzumab plus chlorambucil anymore?
I think chlorambucil is out. The people from MD Anderson Cancer Center would argue that for a patient who is young, fit, with no mutation, non-del(17p), probably trisomy 12 or del(13q), you should at least discuss it because in the long-term follow-up data, there is a theoretical tail on the curve that may result in cure for a subset of patients. But it is a small number, and it is hard to argue against the data from ECOG-E1912. I think if FCR is not gone yet, it is probably close to being gone. Of course, there are attempts to keep resuscitating chemotherapy.
Do you consider the risk of atrial fibrillation in such a patient? What else would be a concern when giving ibrutinib?
It has not changed my standard of care, but it obviously makes me a little more anxious. I feel as if I give the patient more caution when we are starting. It adds risk, but I do not think it adds a significant risk.
The risk of atrial fibrillation on ibrutinib is, just to give a ballpark, between 5% and 10%. Whether it exacerbates underlying atrial fibrillation is hard to know. I can tell you that, at my practice, if the patient is well managed, I treat as if it were not there.
I am also comfortable giving ibrutinib to patients who are taking anticoagulants, as long as they are not on warfarin. Warfarin is where all the earlier eventsaspirin, too—happened. Subsequently, all the clinical trials allowed for anticoagulation, except for warfarin. I do not have great experience with ibrutinib plus warfarin. If I feel ibrutinib is still the standard of care, and if the patient is not bleeding or having bleeding issues on whatever anticoagulant they are on, then I will give them ibrutinib, and part of the informed consent will say they may have a higher risk of bleeding. I feel pretty comfortable starting ibrutinib in a patient who has atrial fibrillation and in a patient who is anticoagulated.
Would giving lower-dose ibrutinib help minimize the bleeding risk?
As far as I call tell, none of the ibrutinib toxicities are dose dependent. I am not sure the bleeding risk is lower on a lower dose; it has not really been studied. [Results from] the lower-dose studiesand they are not really studies; a couple of them are retrospective analyses—suggest maybe outcomes are similar. I would not start at a lower dose to minimize a toxicity that has not yet happened. We just do not know whether the bleeding risk or atrial fibrillation risk will be affected by the lower dose.
Certainly, if I have somebody who is on ibrutinib and they start bleeding and then I take them off and I decide to rechallenge them, I usually do rechallenge them at a lower dose, but it is not proved that I did it based on any evidence.
What if the patient has any source of minimal bleeding or bruising? Do you stop the drug?
Probably not. You have to work it up, but microscopic hematuria is not enough to stop ibrutinib.
What are the next-line treatment options if ibrutinib is stopped? What factors influence your selection after ibrutinib?
There are not a lot of sequencing data coming from these trials. Once a patient is censored, we do not know what happens to them. I think the best prospective data available post ibrutinib are for venetoclax [Venclexta], and I would also argue that maybe there is a rationale for using chemotherapy, particularly in patients who are great candidates for it in the first place. If the patient isIGHVmutated, it is probably reasonable to consider.
We are anticipating positive data from the regimen of obinutuzumab plus venetoclax. This was a frontline trial that was presented at ASH in 2017.6Keep in mind that this was a relatively small trial that looked at 2 different strategies for obinutuzumab dosing, either before venetoclax to minimize TLS [tumor lysis syndrome] risk or after venetoclax ramp-up. There were 32 frontline patients presented here, so not like the 600-patient trials that I previously mentioned, and the encouraging thing to come from these data was that the response rate to obinutuzumab/venetoclax was 100%. Unlike the data that we saw from RESONATE-2, [the results showed] a very high proportion of patients who are in CR, 72%. Across the board, for the very small numbers of patients with del(17p), del(11q), and unmutatedIGHV, for example, that CR rate and that response rate were maintained. So obinutuzumab/venetoclax looks like a very active regimen.
Just to add to it, there is a high proportion of patients who are minimal residual disease [MRD] undetectable across the board. But whether they had a CR or a partial response, the MRD status is very negative in a high proportion of patients. Again, very small numbers.
Twelve-month PFS was 100%, 15-month was 94%, and 18-month was 91%. Only 3 progression events were noted. Two patients developed Richter transformation, and 1 developed progression of disease. We do not have venetoclax available as a first therapy anyway at the moment, and I would not change my standard of care based on 32 patients.
The toxicity profile for venetoclax is largely hematologic: neutropenia, thrombocytopenia, a little bit of gastrointestinal toxicity, and then that theoretical risk of TLS, which, if you follow the ramp-up, does not seem to be a major issue. But those data led to the CLL14 trial [NCT02242942], which has not yet been reported but will be some time soon. It is looking at that same regimen of obinutuzumab/venetoclax with obinutuzumab/chlorambucil in a randomized comparison. We do not yet officially know which one won, but I would bet on this one for sure. The appealing thing about this approach is that it is a 12-month regimen and then everyone stops.
One of the limitations, if we want to talk about sequencing, is that although in my mind there are very strong data for using venetoclax after ibrutinib discontinuation, I do not have any data to say that ibrutinib works after venetoclax discontinuation. If one is looking for levels of evidence, and none of these regimens are curative, I am still sequencing ibrutinib first over venetoclax because there was a prospective trial specifically done in ibrutinib failures where the response rate to venetoclax after ibrutinib failure was 65% to 70% and the median PFS was about 2 years.7We do not have that information for the opposite sequence. We need to do that trial.