An Open-Label Phase II Study of Buparlisib (BKM120) in Patients with Relapsed and Refractory Diffuse Large B-Cell Lymphoma (DLBCL), Mantle Cell Lymphoma (MCL) and Follicular Lymphoma (FL)

Background: Patients (pts) with relapsed or refractory (R/R) non-Hodgkin's lymphoma (NHL), including DLBCL, MCL, and FL, have poor outcomes and few therapy options. Activation of the phosphatidylinositol 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) pathway plays an important role in NHL pathogenesis. The relative expression of aberrant PI3K isoforms can change over the course of disease: for example in MCL, PI3Kd (frequently overexpressed in B-cell malignancies) is more common initially, while PI3Kα plays a more significant role in disease progression and has been identified as an escape mechanism to selective PI3Kd inhibition (Iyengar et al. 2013). Therefore, an agent that targets multiple PI3K isoforms might elicit more durable responses, particularly in later stages of disease. The oral pan-class I PI3K inhibitor buparlisib (BKM120) showed activity in DLBCL cells in vitro (decreased phosphorylation of downstream PI3K signaling effectors, reduced cell proliferation and survival (Zang et al. 2014)) and has shown clinical activity in solid tumors. This study evaluated the safety and efficacy of buparlisib in pts with R/R NHL.

Methods: In this multicenter, parallel-arm, open-label, global Phase II study (NCT01693614), pts with DLBCL, MCL, or FL received buparlisib (100 mg QD) continuously until progression, intolerance, or withdrawal of consent. Key inclusion criteria were R/R disease following ≥1 prior therapy, ≥1 measurable lesion per local assessment using standard criteria (Cheson et al. 2007), and Eastern Cooperative Oncology Group (ECOG) performance status ≤2. Pts with DLBCL were eligible if they had received or were ineligible for autologous stem cell transplant (ASCT). The primary endpoint was overall response rate (ORR). Secondary endpoints were progression-free survival (PFS), duration of response, overall survival, safety, and tolerability. PI3K pathway activation status, an exploratory endpoint, was assessed by mutation analysis (PIK3CA, PTEN) and immunohistochemistry (PTEN loss) in archival tumor tissue samples. Results for the 6-month primary efficacy and safety analysis of buparlisib in pts with DLBCL and FL were previously presented (Younes et al. ASH 2014). Here, we report results for pts enrolled in the MCL cohort.

Results: At this 6-month primary analysis (data cut-off, February 25, 2015), 22 pts with MCL were enrolled. The median age was 68.5 years (range 47–79 years). The median number of prior therapies was 2 (range 1–6), which included intensive combination chemoimmunotherapy regimens, radiotherapy, and/or ASCT.

The median duration of exposure was 20.6 weeks (range 1.7–54.1 weeks). Overall, 17 (77.3%) pts discontinued study treatment for the following reasons: 9 (40.9%) adverse events (AEs); 5 (22.7%) disease progressions, 2 (9.1%) protocol deviations, and 1 (4.5%) physician decision. Four pts died during the study: 1 due general deterioration and 3 due to underlying disease (with 1 on-treatment death due to disease progression). The most common AEs regardless of causality were hyperglycemia (n=10 [45.5%]), and fatigue, anxiety, depression, and weight decrease (n=7 [31.8%] each). The most frequent grade 3/4 AEs were asthenia, confusional state, leukopenia, and febrile neutropenia (n=2 [9.1%] each). Abnormal laboratory values grade 3/4 included elevated levels of fasting glucose (18.2%), aspartate transaminase and alanine transaminase (4.5% each), and neutropenia (18.2%).

ORR was 22.7% (95% CI, 7.8–45.4), with 1 (4.5%) complete response and 4 (18.2%) partial responses (Figure 1). Thirteen (59.1%) pts had stable disease. The disease control rate was 81.8% (95% CI, 59.7–94.8). With 7 PFS events, the median PFS was 11.3 months (95% CI, 3.8–not estimable).The estimated PFS rate at 6 months was 68.6% (95% CI, 39.8–85.7). Of the 11 pts with evaluable samples, none showed PI3K pathway activation at baseline.

Conclusions: Buparlisib treatment was moderately well tolerated, had an acceptable safety profile, and elicited sustained reduction in tumor burden in pts with R/R MCL with a median PFS of 11.3 months. Despite the limited ORR (22.7%), the observed disease control with targeted inhibition of all 4 PI3K isoforms in this cohort of pts with MCL is encouraging, and warrants further exploration of buparlisib in pts with NHL. Updated results for the DLBCL and FL cohorts will be presented at this meeting.