John M. Burke, MD, reviews the treatment options he considers when treating a patient with chronic lymphocytic leukemia and the factors that sway his clinical decisions.
John M. Burke, MD
At aTargeted Oncologylive case-based peer perspectives event, John M. Burke, MD, reviewed with a group of physicians the treatment options he considers when treating a patient with chronic lymphocytic leukemia (CLL) and the factors that sway his clinical decisions. Burke, the associate chair of the Hematology Research Program, US Oncology Network, Rocky Mountain Cancer Centers, discussed these treatment choices on the basis of 2 case scenarios of patients with CLL, one who is newly diagnosed and another with progressive disease.
A 75-year-old woman reported symptoms of weight loss and fullness in her left upper quadrant. She had a medical history of prior hypertension and hypercholesterolemia, both of which were medically controlled.
On physical exam, moderate axillary lymphadenopathy was noted and the spleen was found to be palpable about 4 cm below the costal margin. Otherwise, she appeared well and continued her daily activities.
Her laboratory results were remarkable for a white blood cell (WBC) count of 48,000/μL and 73% lymphocytes; a hemoglobin (Hb) level of 9.1 g/dL; platelets, 125,000/μL; absolute neutrophil count (ANC), 1800/μL (within normal limits; WNL); lactate dehydrogenase (LDH), 250 U/L; and β2microglobulin of 4.2 µg/L. Flow cytometry noted CD5-positive, CD19-positive, and CD23-positive cells. Also, fluorescence in situ hybridization (FISH) testing revealed deletion 11q (del[11q]).
Molecular analysis showed that she wasIGHVunmutated andTP53wild-type. A bone marrow biopsy was performed which showed diffuse infiltration by CLL.
Please describe the patient’s case at presentation.
This is a 75-year-old woman who presented with weight loss and left upper quadrant fullness. Her past medical history was notable for past hypertension and hyperlipidemia. On exam, she has axillary lymphadenopathy and papillary palpable spleen 4 cm below the ribs. She is otherwise well-appearing and is a functional, active, otherwise reasonably healthy woman.
Her laboratory results show that she has a white blood cell count of 48,000, made up of mostly lymphocytes. Her hemoglobin is 9.1, and she has mild thrombocytopenia. β2microglobulin is 4.2 µg/L and flow cytometry shows a monoclonal B-cell population with typical CLL profiling. Her FISH shows a del(11q) and her test forIGHVmutational status comes back as unmutated. A molecular test shows she does not have aTP53mutation (it is wild-type). The bone marrow biopsy shows diffuse infiltration by CLL.
What is your usual approach for patients with newly diagnosed CLL?
I don’t test bone marrow in every patient with CLL. I do it only if the clinical trial requires it, and in my experience, fewer clinical trials are now requiring it. I have patients who I’m putting on treatment without doing a marrow and just running the tests listed here. This patient meets requirements to start treatment: She has symptomatic splenomegaly and anemia.
Based on the National Comprehensive Cancer Network (NCCN) recommendations, what is the best course of treatment for this patient?
The NCCN has a treatment pathway for patients without deletion 17p [del(17p)], with a whole separate list for patients with del(17p). They break it down by mutational status and comorbidities. Appropriate courses are recommended for patients considered frail, those patients aged ≥65 years or younger patients with significant comorbidities, and young patients without comorbidities.1
The CLL11 trial compared chlorambucil plus obinutuzumab (Gazyva) with chlorambucil plus rituximab, and it put obinutuzumab on the map as frontline therapy.2These data are several years old. Progression-free survival (PFS) was better with the obinutuzumab-based combination. Obinutuzumab was better than rituximab at improving PFS. When these data were initially shown, there was no difference in overall survival (OS), but that changed with a presentation at the 23rd European Hematology Association Congress last year.3A small difference in OS emerged with obinutuzumab when compared with rituximab.
I’ve seen a lot of [clinicians] use single-agent rituximab to treat patients with CLL. It was never the NCCN standard, but I think a lot of doctors in practice used to do it. The standard dose would be 375 mg/m2for the first dose, ramped up to 500 mg/m2thereafter.
With a median follow-up of 59.4 months, the median PFS with obinutuzumab was 28.9 months versus 15.7 months with rituximab (HR, 0.49; 95% CI, 0.41-0.58;P<.0001). Time to next treatment was longer and the OS was better with obinutuzumab.
Would you consider ibrutinib (Imbruvica) alone or in combination with another agent to treat this patient?
The RESONATE-2 study is not a [wildly] popular study in the United States [in which patients were randomized to either chlorambucil or ibrutinib]. Clearly, the PFS was markedly better with ibrutinib than it was with chlorambucil. The results were published in theNew England Journal of Medicine4 years ago and contributed to FDA approval of ibrutinib for all patients with CLL.4
A subgroup analysis of RESONATE-2 was relevant to our patient, who had del(11q) and unmutatedIGHV. If you look at the Kaplan-Meier curves for PFS, ibrutinib was always better than chlorambucil. The del(11q) did not have a major impact on prognosis when ibrutinib was used. The same was true ofIGHVmutational status. Mutated or unmutatedIGHVdid not really matter all that much for patient outcomes, with ibrutinib. With chlorambucil, however, patient outcomes were worse with unmutatedIGHV. No matter what the patient’s status was, ibrutinib was better than chlorambucil.
At the 61st American Society of Hematology (ASH) Annual Meeting, a major study presented was ALLIANCE A041202, which has since been published in theNew England Journal of Medicine(December 2018). This is among the first studies to compare ibrutinib with chemotherapy as initial therapy. Patients had to be aged ≥65 years with untreated CLL. They were randomized to either bendamustine plus rituximab (BR), ibrutinib alone, or ibrutinib plus rituximab. The primary endpoint was PFS. The conclusion was that in all comers, an ibrutinib-containing regimen improved PFS compared with BR as additional therapy.5
The secondary conclusion that you could perhaps get from this is that adding rituximab did not do much. The ibrutinibrituximab curve and the ibrutinib-alone curve are superimposable in terms of PFS (HR, 1.00; 95% CI, 0.62-1.62;P= .49). It does not help to add rituximab.
In the BR arm, the 2-year PFS estimate was 74%, compared with 87% with ibrutinib alone and 88% with ibrutinib plus rituximab. No difference exists between those last two. For OS, no significant difference existed between the ibrutinib arms and the BR arm. So, it is not the end of the world if you give BR to start because you can give ibrutinib later and you do not lose any OS.
Case 1 (continued):
The patient was started on ibrutinib 420 mg daily.
What are some toxicities observed with the use of ibrutinib?
Grade ≥3 hematologic adverse events (AEs) were higher with BR, but grade ≥3 nonhematologic AEs were higher with ibrutinib. Think about diarrhea, atrial fibrillation, infections, and hypertension. People were also on the ibrutinib longer.
Could any other ibrutinib combinations be considered?
Another important study presented at ASH was iLLUMINATE, another randomized trial comparing ibrutinib with chemoimmunotherapy. The control arm this time was obinutuzumab plus chlorambucil based on the CLL11 trial and the investigation arm was obinutuzumab plus ibrutinib. The key point is that chlorambucil was continued for the usual 6 cycles, which is how they did it in CLL11, whereas the ibrutinib oral drug was continued foreveruntil progression or unacceptable toxicity. There was a little imbalance on the duration of treatment but the obinutuzumab in both arms was for 6 cycles.6
Ibrutinib plus obinutuzumab showed a dramatic improvement in PFS versus chlorambucil plus obinutuzumab. It is notable to observe the timing of when the Kaplan-Meier curve started to separate. In the first 6 to 9 months, when everybody in the cohort is getting treatment, the curves are superimposable. Remember, the chlorambucil stops at 6 months, and pretty soon thereafter is where the separation starts. Part of the advantage is that you have a better drug with ibrutinib, but another advantage is that you can do longer treatment with ibrutinib and chemotherapy. A feature of ibrutinib that makes it better than other agents is that it can be continued and will not wipe out the bone marrow.
In a subgroup analysis, the high-risk patientswhich included all patients with del(17p), del(11q),TP53mutations, or unmutatedIGHVwere grouped together and [the investigators said], “Let’s see how that turns out.” Sure enough, the outcomes are hugely different. And patients with bulky disease, those with tumors ≥5 cm, do slightly worse than patients with tumors that are <5 cm. I am not sure if that makes a big difference in my treatment decision for a patient, but it is an interesting observation: Patients with big bulky lymph nodes ≥5 cm are likely to have a less-favorable outcomes than those with nonbulky disease.
A summary of the outcomes of iLLUMINATE shows that the median PFS was not reached with the ibrutinib regimen. The 30-month PFS was 79% versus 31% with the control. The most common toxicities in both arms included thrombocytopenia and neutropenia, and treatment-related deaths were rare.
How do you feel about ibrutinib as initial therapy? Would you use it in combination or as a single agent?
This patient was started on ibrutinib and I think that was reasonable. Obinutuzumab has FDA approval now with ibrutinib for frontline therapy, but rituximab does not. You can also just go with ibrutinib alone. It is a judgment call whether to add the obinutuzumab or not.
Some scary data came out in the days when ibrutinib was used as a last resort, from the University of Texas MD Anderson Cancer Center. They showed that if you stopped ibrutinib, the median survival was about 3 months.7If we stopped it, the patient was in a bad place. I do not think that is really true in this day and age, because we use ibrutinib much earlier; we do not see these explosive progressions.
How do you manage patients who are taking blood thinners at the same time as Bruton’s tyrosine kinase (BTK) inhibitors?
I had a patient with Waldenstrom macroglobulinemia who was too frail for chemotherapy, so I put her on acalabrutinib (Calquence) although she needed rivaroxaban (Xarelto) for recent deep vein thrombosis. We were 1 month away from taking her off of the Xarelto and she had a retroperitoneal hemorrhage. She made it through, but it was not fun. I had another patient on ibrutinib and warfarin for years and she was on death’s doorstep. She actually signed up for hospice but then decided to let me start her on ibrutinib. I did, and now here she is 4 years later doing fine on ibrutinib and warfarin most of the time.
What are the next-line treatment options?
If this patient progressed on ibrutinib, then venetoclax (Venclexta) would be an option. Other options would be a CD-20 antibody [similar to rituximab] or idelalisib (Zydelig).
The CLL14 study may give us another treatment option in the future.8It’s a really important study that [will receive a lot of attention]. It mimics the iLLUMINATE study in that the control arm is chlorambucil plus obinutuzumab, but in this case the experimental arm was venetoclax plus obinutuzumab. The design is slightly different, too. Obinutuzumab was used for 6 months in both arms. In the control arm, they did something [I think was innovative]: They gave the chlorambucil for 12 months instead of 6 months, not just stopping the chlorambucil at 6 months like they did in iLLUMINATE. In the experimental arm, they also continued venetoclax for 12 months and then shut it off. The time on therapy was the same for both arms.
Nobody has seen the results of this study. The top-line press release from last year states that the study met its primary endpoint: Venetoclax plus obinutuzumab improves PFS compared with chlorambucil plus obinutuzumab.9It is kind of the same outcome as iLLUMINATE, and so we’ll see [what happens]. This study did not get presented at ASH although the press release occurred a month before the conference, and the data at this time have still not been presented to the public, to my knowledge. I have heard they are impressive. We will probably get to see these data soon and it is conceivable that [this combination] might become an option for initial therapy.
How would you choose between ibrutinib or venetoclax as initial therapy? What factors would you think about?
If the patient were on an anticoagulation medication, you could go with venetoclax. If somebody just wanted 1 year of therapy, you could give them the venetoclax option. Another factor to consider is that very few data exist on the effectiveness of ibrutinib after venetoclax. Are you losing efficacy that way? This is uncertain. On the [flip side, though], data show that venetoclax after ibrutinib works well.
Does the greater effect of venetoclax on minimal residual disease (MRD) in CLL indicate that it is a more effective drug than ibrutinib?
In the MURANO study, the rate of MRD in relapsed CLL was about two-thirds (62.4%) with venetoclax and rituximab [versus 13.3% with BR].10We have not seen what the MRD rate in [the CLL14 trial] is yet. We know from iLLUMINATE that the MRD rate with ibrutinib plus obinutuzumab in the frontline was about 25% to 30%. We know that degree of MRD correlates with long-term outcomes, from what we have seen in CLL. But I do not know that comparing this drug with another drug [in terms of MRD negativity] is going to lead to better outcomes. If you would achieve MRD with venetoclax, is it better than not achieving MRD? We do not know that venetoclax will lead to longer PFS than ibrutinib, just because more people achieve MRD.
In a myeloma study, venetoclax is being studied in patients with translocation t(11;40) [NCT03314181]. More deaths occurred in venetoclax-treated patients. Apparently venetoclax looks very effective but is linked to more deaths from infections. Just achieving MRD is not the goal. That was the take-home point from the study. You do not want to achieve just MRD; you have to do more than that.
What other options are becoming available for CLL?
Acalabrutinib, another future player in CLL, is currently approved for mantle cell lymphoma. It has shown really good activity in CLL. At ASH, we saw data of the 3-year follow up of a phase I/II study in which almost everybody got acalabrutinib at the current standard dose of 100 mg twice daily. The data from the presentation showed an overall response rate [ORR] of 97%, which I found very impressive. Almost all of these were partial responses, with a few complete responsesonly about 5% CRs, as with ibrutinib.
AEs with acalabrutinib include diarrhea, headaches, contusion, upper respiratory tract infection, weight increase, arthralgia, nausea, and cough. Patients get more headaches with acalabrutinib than with ibrutinib. I warn patients about this toxicity, but it does get better with time.
The word has always been that you get less bleeding and seemingly less atrial fibrillation with acalabrutinib. A randomized trial is comparing acalabrutinib with ibrutinib in relapse CLL although we do not yet know the results [NCT02477696]. My current opinion is that I am agnostic right now about whether acalabrutinib causes less bleeding.
A 62-year-old man who was diagnosed with CLL 6 years ago reported increasing fatigue on routine follow-up. At diagnosis, flow cytometry was done which showed a monoclonal population of CD5-positive, CD19-positive, and CD23-positive cells. FISH testing had revealed del(17p) and molecular analysis had showedIGHVmutated,TP53wild-type disease. He was initially treated with ibrutinib for 4 years and achieved stable partial remission.
On physical exam he now presented with cervical lymphadenopathy, about 4 cm, and his spleen was found to be palpable about 5 cm below the left costal margin. Laboratory results were remarkable for WBC count of 153,000/μL with 73% lymphocytes (absolute lymphocyte count, 112,000/μL); Hb, 9.1 g/dL; platelets, 125,000/μL; ANC, 1800/μL (WNL), LDH, 250 U/L, and β2microglobulin of 4.2 µg/L.
Do you do additional molecular testing for patients who progressed on ibrutinib?
No. To me, it matters [only] if you are going to start a second-line BTK inhibitor that has a different binding site on the BTK and you want to know if they have a mutation. For venetoclax or idelalisib, it does not really matter if they have aBTKmutation. The treatment options are venetoclax, a PI3K inhibitor, duvelisib (Copiktra), idelalisib, chemotherapy with a del(17p) patient, or a targeted therapy.
Would this patient be a good candidate for venetoclax in the second line?
I will refer to the original phase II trial that got venetoclax approved as a CLL regimen in patients with del(17p). They did the usual ramp-up phase and the primary endpoint was ORR. There were 107 patients in the cohort and the ORR was 85%. The most common toxicities shown here were neutropenia (40%), infections (20%), anemia (18%), and thrombocytopenia (15%).
It is important to understand the risk stratification for tumor lysis syndrome (TLS). I think everybody knows you are supposed to hospitalize a high-risk patient who is receiving their first dose of venetoclax.High-risk disease is defined as the presence of any node ≥10 cm, or a node ≥5 cm plus a lymphocyte count ≥25 x 109/L.
If you are going use venetoclax in a high-risk patient, you hydrate them. What I have done is put them in the hospital at night and start them on hydration. I take TLS labs early the next morning, TLS labs at noon, and TLS labs again the following morning. Then, they can go home the following morning if there is no TLS. It ends up being about a 48-hour in-patient stay.
The MURANO trial included patients who had previously received chemotherapy then relapsed.10It was looking at BR versus venetoclax plus rituximab. Prior bendamustine was allowed if the duration of response was at least 2 years. Stratification was done by del(17p). No one here had received prior ibrutinib alone as an initial therapy; it was all prior chemotherapy.
The subgroup analysis worth noting is in the patients with del(17p), but basically every subgroup here demonstrated improvement with venetoclax. Venetoclax was worse in no patient subgroup, including patients with del(17p).
Which PI3K inhibitors could be used in this setting?
The DUO study, a randomized trial of duvelisib versus ofatumumab with the primary endpoint of PFS, gave duvelisib its spot as new kid on the block last year. Duvelisib is a PI3K inhibitor just like idelalisib. The PFS curve favored duvelisib and there was a significant improvement in [investigator-assessed] PFS: 17.6 months versus 9.7 months [P<.0001]. The ORR was better and most patients experienced a good reduction in lymph node burden. They did not use rituximab in this trial. It was single-agent duvelisib. Unlike idelalisib, which in CLL is approved to be given with rituximab, duvelisib is approved as a single agent with no CD-20 antibody required.
Toxicities, as far as I can tell, seem similar to those of idelalisib. Grade ≥3 AEs include neutropenia, which occurs in about 30% of patients; diarrhea, which occurs in about 15%; pneumonitis [14% vs 1% in the control arm]; and colitis [12% vs 1%]. These are pretty similar to PI3K toxicities seen with idelalisib. AEs of special interest include rash, pneumonitis, transaminitis, colitis, pneumonia, diarrhea, and neutropenia. You are supposed to use prophylaxis these days because of the risk of pneumocystis pneumonia.
I had one woman refuse Bactrim when she got pneumocystis pneumonia. It was a shame because [the duvelisib] worked like gangbusters in a CLL that was refractory to ibrutinib and refractory to venetoclax. I put her on idelalisib and rituximab and she had her best response yet.