CME
Physicians: Maximum of 1.00 AMA PRA Category 1 Credit™
Released: February 26, 2021
Expiration: February 25, 2022
John M. Burke, MD:
Parsaclisib is a next-generation PI3K inhibitor that is highly selective for δ vs α/β/γ isoforms and isdesigned to minimize risk of hepatotoxicity.8 In early studies, parsaclisib appeared active in patients with R/R MZL and other B-cell lymphomas.9
At ASH 2020, interim results from the open-label phase II CITADEL-204 trial of parsaclisib for patients with R/R MZL were presented.10 In this trial, patients with no previous exposure to BTK inhibitors or PI3K inhibitors were treated with parsaclisib 20 mg/day for 8 weeks followed by either 20 mg once weekly (weekly group, n = 28) or 2.5 mg once daily (daily group, n = 72). Of note, the study originally included an ibrutinib‑experienced cohort of 10 patients that was terminated due to slow enrollment.
The primary endpoint was ORR, with secondary endpoints including the CR rate, duration of response, PFS, OS, and safety. A key aim with this agent and trial was to minimize or avoid some of the long‑term toxicities that emerge with continuous PI3K inhibitor therapy.
John M. Burke, MD:
During the trial, daily dosing of parsaclisib at 2.5 mg/day was chosen as the preferred regimen. Among the 72 patients who received this regimen, approximately one third each had MZL of nodal, extranodal, or splenic subtypes. Most (74%) had received previous chemotherapy, and fewer than 15% had undergone previous surgery or radiation. About half of all patients were refractory to their last systemic therapy. Of note, the median duration of parsaclisib therapy was 9 months in the daily group, and the median follow-up at this analysis was 14.9 months.
John M. Burke, MD:
In the overall study population, the primary endpoint, ORR, was 57.0% by independent review committee assessment, including CRs in 6% of patients. Results were similar in the daily treatment subgroup: The ORR was 56.9%, with CRs in 5.6%. Of note, every patient treated on this study achieved at least stable disease.
John M. Burke, MD:
Responses were seen in all MZL subtypes, ranging from 51.6% in nodal MZL to 55.9% in extranodal MZL to 62.9% in splenic MZL. Most responses were PRs, with only 2 CRs reported in each subtype. The median time to first response was 8.1 weeks. Of importance, by the first assessment, 67% of responders had already achieved a CR or PR.
John M. Burke, MD:
In the overall study population, the median duration of response was 12 months and the median PFS was 19 months. In the daily-dosing group, the median duration of response and median PFS were not reached.
John M. Burke, MD:
Toxicities associated with parsaclisib in the daily treatment group included diarrhea in 49% of the patients (grade 3 in 14%). Overall, 14% of patients discontinued treatment due to diarrhea or colitis. Of note, in the daily group, the median onset of grade ≥ 3 diarrhea was 5.1 months, and improvement to grade ≤ 2 took approximately 2 weeks.
In addition, grade 3/4 neutropenia occurred in 9% of patients (11% in the daily group). Grade 3/4 elevated transaminases were not observed frequently on this study. Serious AEs in the daily group included pneumonia in 8% and colitis in 7%. In the daily group, more than one half of patients had to interrupt treatment or reduce the dose due to toxicity, and 35% discontinued due to toxicity, which is a considerable amount.
John M. Burke, MD:
The results reported from CITADEL-204 show that treatment with parsaclisib led to rapid and durable responses in patients with R/R MZL, with an ORR of 57% in the daily dosing group and in the overall population. The median time to response was 8.1 weeks. However, as with many PI3K inhibitors, the toxicity was significant, as evidenced by the fact that 35% of patients had to discontinue the drug due to toxicity, even when the dose was reduced from 20 to 2.5 mg daily. I think these data show that parsaclisib may be a potential new option in this setting. Of importance, there was a relative absence of elevated transaminases as a serious problem. Diarrhea was relatively common at 49%, but grade 3 in only 14% of patients. Of note, efficacy was similar between patients with nodal, extranodal, and splenic MZL.
Jeff P. Sharman, MD:
In the setting of follicular lymphoma (FL), 3 PI3K inhibitors are already approved by the FDA: idelalisib, duvelisib, and copanlisib. Copanlisib differs in that it is given intravenously weekly instead of orally. Among the oral therapies, diarrhea has remained problematic, even with lower dosages, as discussed. In general, this drug class is associated with approximately a 12‑month response. It was nice to see lower rates of transaminase elevation, which can be quite frightening with some of the PI3K inhibitors and tend to occur early, often within the first 2-3 months of therapy.
Let’s return to a question from earlier.
Jeff P. Sharman, MD:
At ASH 2020, final efficacy and safety results from the phase IIb UNITY‑NHL trial were presented.11 This trial evaluated monotherapy or combination therapy with umbralisib, a novel inhibitor of PI3Kδ and casein kinase-1ε, in multiple disease-specific cohorts and treatment combinations; this analysis focused on patients with previously treated, R/R indolent NHL (N = 208). Previous results from UNITY-NHL suggested umbralisib is active with manageable safety in this setting.12 As discussed, umbralisib has improved selectivity and is generally associated with fewer autoimmune-like toxicities compared with other PI3Kδ inhibitors.13,14
In this trial, patients received umbralisib 800 mg once daily until progression or unacceptable toxicity. The primary endpoint of this analysis was ORR, with secondary endpoints including duration of response, PFS, time to response, and safety.
Jeff P. Sharman, MD:
UNITY-NHL included 69 patients with MZL, 117 with FL, and 22 with small lymphocytic lymphoma. Virtually all patients had had previous CD20-targeted therapy, with a significant amount having received previous chemoimmunotherapy therapy as well (91%). The median age in the study was 66 years, with a slight male predominance. Approximately three quarters of patients had stage III/IV disease. The median number of previous therapies was 2, with a range of 1-10.
The median duration of follow-up was quite long in this study (27.5-29.3 months), longer than that observed in studies of idelalisib and duvelisib. At a median follow-up of more than 2 years, approximately 30% of patients remained on therapy, indicating that there is a subgroup of patients who can derive sustained benefit from this drug.
Jeff P. Sharman, MD:
Here, again, the most common AE was diarrhea, which afflicted nearly 60% of patients; however, only 10% of these events were grade 3, and no cases of grade 4 diarrhea were seen. Unlike studies of other PI3K inhibitors, minimal serious transaminase elevation was seen—approximately 20% overall, with grade 3 elevations occuring relatively infrequently (~ 7%). Neutropenia was also relatively mild, at 16% overall with approximately 11% being grade 3/4.
Discontinuations due to transaminase elevation and grade 3 diarrhea occurred in approximately 3% of patients each, which is less common than what has been observed with other PI3K inhibitors. Grade 3/4 opportunistic infections were not common (3.4%) and grade 3/4 pneumonia or pneumonitis was infrequently seen (< 2%).
Jeff P. Sharman, MD:
In the indolent NHL cohort the ORR (per independent review committee) was 45% to 50%, with an additional one third of patients achieving stable disease in each disease cohort. The disease-control rate was 81.3% across the cohorts.
Jeff P. Sharman, MD:
The ORR was similar in each disease cohort: 49.3% in MZL, 45.3% in FL, and 50.0% in SLL. In MZL specifically, the highest response rate was observed in patients with nodal disease (60% vs 45% for mucosa-associated lymphoid tissue and splenic subtypes). Overall, the majority of patients had some reduction in their burden of disease.
Jeff P. Sharman, MD:
The duration of response was greatest in the MZL cohort, where the median was not reached, vs 11.1 months in FL and 18.3 months in SLL. Likewise, median PFS was not reached in the MZL cohort vs 10.6 months for FL and 20.9 months for SLL.
Jeff P. Sharman, MD
The PI3Kδ inhibitor umbralisib demonstrated clinical activity in patients with R/R indolent NHL, with an ORR of 47.1% and a 16.0% CR rate in patients with MZL and no patients progressing to date. Moreover, umbralisib is associated with an acceptable safety profile and a low rate of discontinuations due to AEs. Notably, umbralisib was approved by the FDA in February 2021 for patients with relapsed/refractory MZL who received ≥ 1 prior anti-CD20-based regimen and patients with relapsed/refractory FL who received ≥ 3 prior lines of systemic therapy.
John M. Burke, MD:
I think the study provides further demonstration that PI3K inhibitors have meaningful activity in not only patients with FL but also patients with MZL. The question for healthcare professionals will be how to incorporate these agents into their treatment arsenal for patients with this disease. The BTK inhibitor ibrutinib is approved by the FDA for patients with MZL who require systemic therapy and have received at least 1 previous anti-CD20–based therapy, with response rates at approximately 50%, similar to results with umbralisib in this study.15
As mentioned, the results of this study have led to the FDA approval of umbralisib not only in FL, but also in MZL. This is the first approval of a PI3K inhibitor in MZL, so it is great to have formal recognition of the efficacy of this drug category in MZL. In terms of sequencing, I think the choice between ibrutinib and umbralisib will come down to an individual discussion with each patient, with potential risks and benefits and patient characteristics taken into account. I envision sequencing one of these drugs after the other; that is, in patients previously treated with ibrutinib, I will try umbralisib next, and vice versa. It is also important to recognize that umbralisib appears to offer a manageable toxicity profile such that some patients may be able to receive this drug for a longer period of time than with idelalisib and duvelisib, although healthcare professionals should remain aware that umbralisib is associated with known toxicities of the PI3K inhibitor class of drugs.
John M. Burke, MD:
Mosunetuzumab is a novel bispecific antibody that crosslinks both CD3 and CD20, thereby redirecting T-cells to target malignant B-cells. Bispecific antibodies have produced exciting results in lymphoid malignancies as well as in multiple myeloma and other diseases. Like CAR T-cell therapies, they can cause both cytokine release syndrome (CRS) and neurotoxicity, although generally at a lower rate than observed with CAR T-cell therapy.
At ASH 2020, data from the FL cohort of a phase I/Ib dose-escalation study of mosunetuzumab in patients with R/R B-cell NHL were presented (N = 62).16 Patients had to have received at least 2 previous systemic therapies and had an Eastern Cooperative Oncology Group performance status of 0 or 1. Mosunetuzumab was administered in 21-day cycles. Treatment in cycle 1 was given as step-up dosing on Days 1, 8, and 15: 0.4/1.0/2.8 mg to 1.0/2.0/13.5 mg. In cycles 2-8, patients received their cycle 1, Day 15 dose (2.8-13.5 mg) on Day 1. The step-up dosing in this study was used to try to minimize CRS. The primary endpoints were safety and tolerability, determination of the MTD and dose-limiting toxicities, recommended phase II dose, and responses, with secondary endpoints including survival and quality of life.
John M. Burke, MD:
The median age in this study was relatively young at 59 years. A sizeable percentage of patients had disease double refractory to both an alkylating agent and a CD20-targeted therapy (61.3%), and many had experienced early progression of disease within 24 months of frontline therapy (46.8%). Patients in this study had received a median of 3 previous therapies.
John M. Burke, MD:
The investigator‑assessed ORR for mosunetuzumab in patients with R/R FL was 67.7%. In those patients who were refractory to the last therapy, the ORR was 65.2%, and in those with progression within 24 months, the ORR was 75.9%.
Of interest, more than one half of patients (51.6%) achieved a CR across all dose levels, which is a very deep response in this difficult-to-treat population. Also, at a median follow-up of 18.4 months after the first response, the median duration of response was 20.4 months, which is really quite good in this population. Overall, responses were consistent across the high-risk groups.
John M. Burke, MD:
Notable AEs included CRS (occurring in 17.7%) and hypophosphatemia (21.0%). Grade ≥ 3 AEs were seen in 67.7% of patients, but only 6.5% discontinued due to treatment-related AEs, and only 1 grade 5 AE occurred.
Neutropenia was relatively common (22.6%), but this is not surprising in a heavily pretreated population; febrile neutropenia was quite rare. All 13 cases of hypophosphatemia resolved, mostly in cycle 1.
John M. Burke, MD:
In this study, all CRS events were grade 1/2, occurred during the first cycle, and were resolved without tocilizumab, ICU admission, or vasopressors. The median duration of CRS was 2 days (range: 1-8). In addition, the ORR did not appear to differ greatly whether the patient had a CRS event or not (45.0% with CRS vs 52.9% with no CRS). Evidence of CRS included pyrexia in 17.7%, chills in 9.7%, headache in 4.8%, and tachycardia in 3.2%.
Jeff P. Sharman, MD:
In this early study, mosunetuzumab produced high and consistent responses in heavily pretreated patients with FL, including in those in high-risk subgroups (disease progression within 24 months and double-refractory disease). Assessment of higher dose levels is ongoing, but to date, the safety profile is acceptable, with low rates of grade 2 events and no grade ≥ 3 CRS.
Overall, mosunetuzumab appears to be quite active in this population, and I anticipate that we will see FDA approval of this therapy. Pivotal studies are being launched that will look at the combination of mosunetuzumab/lenalidomide vs lenalidomide/rituximab in the second‑line setting for FL.17
John M. Burke, MD:
Mosunetuzumab clearly has activity in patients with relapsed FL. For me, the highlights of this study are the high CR rate of approximately 50% in a population of heavily pretreated patients, many of whom were refractory to their previous therapy.
Another important feature of bispecific antibodies, as Dr. Sharman mentioned, is the risk of CRS and neurotoxicity, which can be similar to that seen with CAR T‑cell therapy. However, in this study, CRS rates were relatively low, and all were grade 1/2. Compared with other bispecific antibody drugs, mosetuzumab appears to have a favorable toxicity profile that, in theory, could be managed by most oncologists.
John M. Burke, MD:
Returning to CAR T-cell therapy, ZUMA‑5 was a single-arm phase II trial of the CD19-targeted CAR T‑cell product axicabtagene ciloleucel for patients with R/R indolent NHL (N = 146).18 In this trial, patients with grade 1-3a FL or MZL (nodal or extranodal) and at least 2 previous lines of therapy (including an anti-CD20 antibody) were treated with lymphodepleting fludarabine and cyclophosphamide for 5 days before axicabtagene ciloleucel 2 x 106 cell/kg. The primary endpoint was ORR, with secondary endpoints including the CR rate, ORR per investigator, duration of response, PFS, OS, safety, and CAR T-cell and cytokine levels.
The median number of days from leukapheresis to delivery of the product to the study site was 17 days.
John M. Burke, MD:
Patients with FL had a median age of 60 years and those with MZL had a median age of 66 years. The median number of previous therapies in each group was 3. Most patients (85% to 90%) had stage III/IV disease, nearly one half had a Follicular Lymphoma International Prognostic Index score of at least 3, and approximately 50% had high tumor bulk. More than 50% of patients had progressed within 24 months of their first anti-CD20 antibody treatment.
John M. Burke, MD:
In the FL group, the ORR with axicabtagene ciloleucel was 94% and the CR rate was 80%. In the MZL group, the ORR was 85% and the CR rate was 60%. The median time to first response was 1 month. Among 25 patients with FL who initially achieved a PR, one half converted to a CR after a median of 2.2 additional months.
All subgroups responded to CAR T‑cell therapy, including those with 4 or more previous therapies, those who had previous autologous stem cell transplant, those who were refractory to their previous therapy, or those who had experienced progression of disease within 24 months of their initial CD20-targeted therapy.
John M. Burke, MD
The median duration of response in the FL group was not reached, with 77% of patients maintaining a response at 1 year. The median duration of response in the MZL group was 10.6 months. In both populations, those patients who achieved a CR had a more favorable duration of response than those who achieved only a PR.
John M. Burke, MD:
At 12 months, 77.5% of the patients with FL were progression free, and the median PFS was not yet reached. Median PFS in the MZL group was 11.8 months. Median OS was not reached in either group, and the 12-month overall OS rate was 92.9% at the time of the report.
John M. Burke, MD:
In ZUMA-5, grade ≥ 3 AEs occurred in 86% of patients, including cytopenias in 70% and infections in 16%. Three patients died while on study, with 1 death thought to be related to axicabtagene ciloleucel. Other notable treatment-emergent AEs of any grade included pyrexia (84%), neutropenia (64%), hypotension (49%), and headaches (45%).
John M. Burke, MD:
CRS was seen in 82% of patients; more than 90% had pyrexia and more than 40% experienced hypotension. Most CRS was grade 1/2, but grade ≥ 3 CRS occurred in 7% of patients. Forty‑nine percent of patients required tocilizumab, and 17% required corticosteroids. The median time to onset of CRS was 4 days, with a median duration of 6 days (range: 1-27). All cases of CRS were resolved, except for 1 patient who died with CRS (due to organ failure) before resolution.
John M. Burke, MD:
Neurologic AEs of any grade occurred in 60% of patients and were grade ≥ 3 in 19%. Patients most commonly experienced tremor (52%) and confusion (40%). Thirty‑six percent of these cases were treated with corticosteroids and 6% with tocilizumab. The median time to onset of neurologic AE was 7 days, with a median duration of 14 days.
John M. Burke, MD:
Results from ZUMA-5 show that single-agent axicabtagene ciloleucel can produce high response rates in patients with R/R indolent NHL, including subgroups with high-risk characteristics. The ORR was 92%, with CRs of 80% in the FL cohort and 60% in the MZL cohort. Moreover, responses were durable, with 64% of patients in the FL cohort and 50% in the MZL cohort continuing to respond at the time of data analysis. Axicabtagene ciloleucel had a manageable safety profile. Fewer grade ≥ 3 neurologic AEs occurred in FL vs MZL.
Jeff P. Sharman, MD:
As a drug class, CAR T‑cell therapy has transformed the management of relapsed large‑cell lymphoma, acute lymphoblastic leukemia, and mantle cell lymphoma and has been approved by the FDA in all 3 settings. It stands to reason that this therapy should work well in indolent lymphomas as well—and perhaps even better because there is more time for patients to receive therapy. In the diseases for which CAR T-cell therapy is approved, we have growing optimism that this treatment could potentially be curative for some patients, and I wonder if some patients in ZUMA-5 may likewise continue to improve with time without relapse or with long-term undetectable MRD.
The main challenge in implementing CAR T-cell therapies is access; even in diseases like large‑cell lymphoma, a surprisingly small portion of potentially eligible patients are receiving this treatment. This may be because many potentially eligible patients do not have access to these therapies, whether due to geography or the difficulty of staying for a prolonged period at the treatment center. Nevertheless, I have great hope for the use of CAR T-cell therapy in indolent lymphomas.
Back
