2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Javier L. Munoz, MD, discusses the nuances of choosing treatment options for patients with CLL in the frontline and relapsed/refractory settings, as well as novel therapeutic strategies on the horizon.
NEXT-GENERATION NONCOVALENT
Next-generation noncovalent BTK inhibitors and chimeric antigen receptor (CAR) T-cell therapy are currently under investigation in chronic lymphocytic leukemia (CLL). Treatment selection variables include age, comorbidities, concomitant medications, high-risk features, performance status, patient preference, dosing schedule, and duration of therapy, according to Javier L. Munoz, MD.
“CLL is a rapidly evolving field, and the follow-up of some novel therapies may be too short to fully visualize long-term toxicities or survival benefits,” said Munoz. “Ultimately, one size does not fit all when choosing therapies for patients with CLL.”
For example, during the 2020 ASH Annual Meeting and Exposition, results of the phase 1/2 BRUIN trial (NCT03740529), which were presented virtually, demonstrated a 63% objective response rate with the investigational BTK inhibitor LOXO-305 in patients with CLL and small lymphocytic lymphoma (SLL).1
Additionally, at the virtual meeting, findings from the ongoing phase 1/2 TRANSCEND CLL 004 trial were presented, revealing that the CAR T-cell therapy lisocabtagene maraleucel demonstrated high rates of undetectable minimal residual disease (MRD) with rapid and durable responses among patients with high-risk relapsed/refractory CLL or SLL.2
In an interview with OncLive®, Munoz, director of the lymphoma program at Mayo Clinic, discussed the nuances of choosing treatment options for patients with CLL in the frontline and relapsed/refractory settings, as well as novel therapeutic strategies on the horizon.
Munoz: First, not all patients with CLL require immediate therapy at the time of diagnosis. CLL is a very heterogeneous disease. Second, the goal of therapy is palliative for most patients with CLL, with the exception of allogeneic hematopoietic cell transplant for select patients. The National Comprehensive Cancer Network mentions multiple therapeutic options for patients with newly diagnosed CLL. We are fortunate to have several nonchemotherapy options available in the frontline setting, including the BCL-2 inhibitor venetoclax [Venclexta] and BTK inhibitors ibrutinib [Imbruvica] and acalabrutinib [Calquence], plus or minus monoclonal antibodies.
Multiple variables help us make treatment decisions, including age, comorbidities, presence or absence of high-risk features, performance status, dosing, concomitant medications, and desire of finite vs continuous therapy. We put all the options on the table, but ultimately, the patient is the final decision maker based on preference and goals of care.
How do the options change when a patient has relapsed/refractory disease? How can we optimize available treatments in the relapsed/refractory setting?
Similarly, asymptomatic relapse does not necessarily require immediate therapy, particularly because patients with CLL are not cured with conventional therapy. Several treatment options are available, and there is no single standard treatment regimen for all patients with relapsed or refractory CLL.
In order to make a decision, it is important to know which f irst-line therapy was chosen—the depth and duration of response the patient derived to such therapy. In general, we make the decision based on intrinsic characteristics of the patient [age, comorbidities, performance status], the disease [presence or absence of high-risk features, disease tempo], and the treatment [dosing, concomitant medications, oral vs intravenous (IV), fixed duration vs continuous].
In patients who have received more than 2 systemic therapies, we also have PI3K inhibitors [eg, idelalisib (Zydelig), duvelisib (Copiktra)] available in addition to other targeted agents such as BTK inhibitors [eg, ibrutinib, acalabrutinib] and BCL-2 inhibitors [eg, venetoclax].
Patients with high-risk features [eg, deletion 17p (del (17p)] deletion and/or TP53 mutations) do not respond to chemoimmunotherapy or relapse shortly thereafter. In general, I try very hard not to prescribe chemotherapy. That said, FCR [fludarabine, cyclophosphamide, and rituximab (Rituxan)] could still be considered for a young or fit patient with mutated IGHV who lacks high-risk features and desires fixed-duration therapy.
Targeted therapies have been compared with chemotherapy in randomized trials [eg, RESONATE-2, NCT01722487; iLLUMINATE, NCT02264574; E1912, NCT02048813; Alliance A041202, NCT01886872; ELEVATE-TN, NCT02475681] favoring the non-chemotherapy arms. If a patient strongly values having a treatment-free interval, venetoclax plus obinutuzumab [Gazyva] is a very reasonable nonchemotherapy option, with the downsides of tumor lysis monitoring and requiring infusions with a monoclonal antibody.
We try hard to look at biologic age instead of chronologic age because it is very important to consider comorbidities and performance status when making decisions. CLL is usually a disease affecting elderly individuals; thus, some trials have been designed to capture younger patients who may tolerate more aggressive therapies.
For example, the ECOG-ACRIN E1912 trial [ibrutinib plus rituximab vs FCR] enrolled younger adults with CLL, whereas the RESONATE-2 trial enrolled older adults with CLL, so we have guidance for each scenario.
When it comes to other agents, fixed-duration venetoclax plus obinutuzumab improved progression-free survival [PFS] compared with chlorambucil plus obinutuzumab; thus, venetoclax [CLL14 trial; NCT02242942] is particularly reasonable for older patients with cardiovascular comorbidities [eg, history of uncontrolled atrial fibrillation or serious bleeding] that may make you favor BCL-2 instead of BTK inhibition. In patients with high-risk features, I would particularly avoid chemotherapy, though. Patients with high- or low-risk features can benefit from targeted agents [eg, ibrutinib, acalabrutinib, or venetoclax) plus or minus monoclonal antibodies, and the rationale to choose one particular agent is multifactorial.
The phase 3 ECOG-ACRIN E1912 trial randomized 529 young adults, with a median age of 57 years, with previously untreated CLL and without 17p deletion, to receive ibrutinib plus rituximab vs 6 cycles of FCR. The study showed superior PFS [89% vs 73%] and superior overall survival [99% vs 92%] at 3 years, despite the fact that patients assigned to FCR were allowed to subsequently receive ibrutinib.
I have indeed used ibrutinib plus rituximab in young patients with CLL, with encouraging results. Extrapolating data from other randomized trials that did not show benefit from adding rituximab to ibrutinib [eg, ALLIANCE A041202 trial of bendamustine plus rituximab vs ibrutinib vs ibrutinib plus rituximab in older individuals], some experts prescribe singleagent BTK inhibitors without the monoclonal antibodies.
I have to admit that, particularly during the coronavirus disease 2019 [COVID-19] pandemic, I have favored oral therapies, bypassing the burden of IV administration, in order to keep patients at home and decrease their exposure to COVID19. Of note, the E1912 trial did not have a single-agent ibrutinib arm. Nevertheless, the decision to add a monoclonal antibody to a BTK inhibitor must always be individualized.
Both acalabrutinib and ibrutinib are oral therapies that are FDA approved in CLL and prescribed continuously until disease progression. That said, we have longer-term data regarding the efficacy/toxicity of ibrutinib. The choice between these therapies is made based upon patient preference and comorbidities while taking into consideration the most serious toxicities with each agent and the predicted tolerability in each particular patient. Acalabrutinib is prescribed twice a day, whereas ibrutinib is prescribed once a day.
The absorption of ibrutinib or zanubrutinib [Brukinsa] does not seem to be impaired by proton pump inhibitors, but zanubrutinib is not FDA approved for CLL at this point. When you think about BTK inhibitors, toxicities that come to mind include infections, hypertension, atrial fibrillation, hemorrhage, and arthralgia. Extrapolating data from the ASPEN trial [NCT03053440; ibrutinib vs zanubrutinib] in Waldenström macroglobulinemia, it is possible that the differences between BTK inhibitors may be mainly reflected in the safety profile instead of efficacy. That said, cross-trial comparisons are misleading; thus, we need to wait for head-to-head comparison data [eg, ibrutinib vs acalabrutinib] to guide us regarding the differences among BTK inhibitors in patients with CLL.
Before changing therapy due to BTK intolerance, we need to perform dose adjustment following recommendations from the package insert and consider prescribing symptom-targeted measures in order to avoid early discontinuation. The biology of disease progression while taking a BTK inhibitor is different from intolerance to a BTK inhibitor. If a patient is progressing on a covalent BTK inhibitor, I would be hesitant to switch therapy to another BTK inhibitor outside of a clinical trial, particularly when we have other options available, such as BCL-2 and PI3K inhibitors that have a different mechanism of action and a different set of toxicities.
That said, if a patient is intolerant to a BTK inhibitor, then we could potentially try a different agent from the same family in the future. For example, acalabrutinib monotherapy in patients with CLL who were intolerant to ibrutinib showed that 72% of the ibrutinib-related adverse effects did not recur.3 Furthermore, some interesting clinical trials have evaluated the use of BTK degraders, highly selective noncovalent BTK inhibitors in previously treated CLL. Some of these trials allow patients to enroll despite previous exposure to covalent BTK inhibitors. The activity of covalent BTK inhibitors is markedly reduced in the presence of BTK cysteine binding site mutations; hence, using a noncovalent BTK inhibitor that inhibits both wild-type and mutated BTK seems to be a reasonable approach to explore in clinical trials.1
Unmet needs include Richter’s transformation and patients with 17p deletion and TP53 mutation that relapse after targeted agents.
Clinical trials are evaluating the role of MRD in timelimited regimens. We always encourage patients to enroll in studies testing investigational agents because it is the only way to improve upon our current standard of care. One of the downsides of single-agent BTK inhibition would be that few patients achieve CR [complete response] or even MRD negativity; hence, we welcome combinatorial approaches [eg, BTK inhibitors plus BCL-2 inhibitors plus monoclonal antibodies] in the clinical trial setting.
CAR T-cell therapies are under investigation in multiple lymphoproliferative disorders and also may have a future role in CLL, alone [eg, ZUMA-8] or in combination with targeted agents [eg, TRANSCEND CLL 004].4 We are seeing deep, lasting remissions with novel agents; hence, we are very enthusiastic about the therapeutic landscape in CLL.
Disclosures: Munoz reports relationships with Pharmacyclics/ Janssen, AstraZeneca, BeiGene, Kite/Gilead and Juno/Bristol Myers Squibb/Celgene.
Related Content: