Phase 2 Study Evaluating the Efficacy and Safety of Parsaclisib in Patients with Relapsed or Refractory Marginal Zone Lymphoma (CITADEL-204)

Background: Marginal zone lymphoma (MZL) is a clinically heterogeneous, indolent, non-Hodgkin lymphoma (NHL) with 3 subtypes classified as extranodal, nodal, and splenic MZL. Parsaclisib, a potent, highly-selective, next-generation phosphatidylinositol 3 kinase (PI3K) δ inhibitor, has shown promising response rates in a phase 1/2 study in patients with previously treated B-cell malignancies. CITADEL-204 (NCT03144674) is a multicenter, open-label phase 2 study of parsaclisib in patients with relapsed or refractory (R/R) MZL. Here, we report preliminary results for patients who were not previously treated with a Bruton’s tyrosine kinase inhibitor (BTKi-naïve).

Methods: Eligible patients were ≥18 years of age with histologically confirmed MZL who received ≥1 prior line of systemic therapy, including anti-CD20 therapy, but were BTKi-naïve. Patients had documented disease progression or lack of response to the most recent regimen, had radiologically measurable lymphadenopathy or extranodal lymphoid malignancy (or histologically confirmed bone marrow infiltration in cases of splenic MZL), and an Eastern Cooperative Oncology Group performance status (ECOG PS) ≤2. Prophylaxis for Pneumocystis jirovecii pneumonia (PJP) was required.

Patients were allocated to receive parsaclisib 20 mg once daily (QD) for 8 weeks followed by either 20 mg once weekly (weekly-dosing group [WG]) or 2.5 mg QD (daily-dosing group [DG]). Objective response rate (ORR) was the primary endpoint; complete response rate (CRR), duration of response (DOR), progression-free survival (PFS), overall survival (OS), and safety and tolerability were secondary endpoints. All radiology-based endpoints were assessed by an independent review committee (IRC). Safety was evaluated in all treated patients, and efficacy was evaluated in all treated patients who had a follow-up of ≥9 weeks, had at least 1 post-baseline assessment, or discontinued study participation prematurely.

Results: From December 2017 to January 17, 2020 (data cut-off), 99 patients (WG, n = 28; DG, n = 71) were treated. Median age was 71 years; 52.5% of the patients were male, 94.9% had an ECOG PS ≤1, and 31.3%, 33.3%, and 35.4% had nodal, extranodal, and splenic MZL, respectively. The median number of prior systemic therapies was 2. At cut-off, 40 (40.4%) patients had discontinued treatment, including 20 (20.2%) for disease progression. The median parsaclisib exposure (range) was 7.5 months (0.4−22.4).

At the data cut-off, 94 patients were evaluable for response, including 66 in DG (Table). Median (range) follow-up for the efficacy evaluable population was 11.1 months (1.2−25.0) overall and 9.5 months (1.2−25.0) in DG. The ORR (95% confidence interval [CI]) was 54.3% (43.7−64.6) overall and 57.6% (44.8−69.7) in DG; the ORR (95% CI) was 48.3% (29.4−67.5), 50.0% (31.9−68.1), and 63.6% (45.1−79.6) for patients with nodal, extranodal, and splenic MZL, respectively (Table). The median time to response was 8 weeks. The median DOR (95% CI) was 9.3 months (6.2−not evaluable [NE]) among all responders and 9.4 months (6.0-NE) in DG. The median PFS (95% CI) was 13.8 months (8.8−NE) overall and 11.5 months (8.3-NE) in DG.

Among the 99 treated patients, the most common treatment-emergent AEs (TEAEs) were diarrhea (36.4% of patients), cough (18.2%), and rash (14.1%). The most common TEAEs grade ≥3 were neutropenia and diarrhea (8.1%, each). The most common serious TEAEs were febrile neutropenia and pneumonia (5.1%, each). TEAEs leading to dose interruption or dose reduction occurred in 47.5% and 14.1% of patients, respectively. TEAEs leading to discontinuation occurred in 15.2% of patients; the most common events were diarrhea (5.1%) and colitis (3.0%). TEAEs leading to death occurred in 4 patients, with 2 events, febrile neutropenia (n = 1) and sepsis (n = 1), deemed treatment related. New or worsening grade ≥3 laboratory test values of clinical interest included increase in alanine/aspartate amino transferase (2.0%/1.0% of patients), and decrease in neutrophil count (13.1%), platelet count (3.0%), and hemoglobin (3.0%).

Conclusion: BTKi-naïve patients with R/R MZL achieved rapid and durable responses with single-agent parsaclisib. Comparable efficacy was observed in patients diagnosed with nodal, extranodal, or splenic MZL. Treatment with parsaclisib was generally well tolerated without unexpected toxicities. Updated data will be presented.