Efficacy and safety of the dual SYK/JAK inhibitor cerdulatinib in patients with relapsed or refractory B‐cell malignancies: Results of a phase I study

Important clinical advances were made in recent years for the treatment of patients with relapsed/refractory B-cell malignancies with approvals of the B-cell antigen receptor (BCR)-targeted therapies ibrutinib and idelalisib,1 as well as venetoclax.2 Furthermore, inhibition of SYK (which is upstream of BTK and PI3K on the BCR signaling pathway) has demonstrated efficacy in early clinical trials.3 Although outcomes for patients have improved with the introduction of targeted therapies, there remains an unmet need for improved, durable, and complete responses. Hence, novel agents with differentiated mechanisms of action are currently under clinical development. As with other malignancies, tumor B cells interact with their microenvironment and are exposed to pro-inflammatory/oncogenic cytokines that signal in part via Janus kinase (JAK) family members.4 Dual suppression of BCR signaling, as well as supportive cytokine JAK/STAT signaling, may be an effective strategy to treat B-cell malignancies. Cerdulatinib was therefore developed as a first-in-class dual SYK/JAK kinase inhibitor. In preclinical studies, dual SYK/JAK inhibition demonstrated greater antitumor activity in B-cell lymphoma cell lines relative to inhibition of either target alone.5 Consistently, the extent to which cerdulatinib inhibited BCR and B cell IL-4 signaling following oral dosing significantly correlated with tumor response in patients with relapsed/refractory B-cell malignancies.6 Herein, we report the phase I dose-escalation study of cerdulatinib in patients with relapsed/refractory B-cell malignancies, detailing the pharmacokinetics, safety, and initial efficacy. Forty-three patients were dosed with cerdulatinib in the phase I, dose-escalation study between October 2014 and February 2016 at five participating study centers. Of the dosed patients, 8 had chronic lymphocytic leukemia (CLL)/small lymphocytic leukemia (SLL), 13 had follicular lymphoma (FL; including one transformed FL grade 3B), and 22 had aggressive B-cell non-Hodgkin lymphoma (NHL; diffuse large B-cell lymphoma [DLBCL], n = 16 and mantle cell lymphoma, n = 6). Of the 43 dosed patients, 36 received drug with sufficient duration for pharmacokinetic and efficacy assessments. Baseline patient demographics and characteristics are delineated in Supporting Information Table 1. Escalating QD and BID dose schedules were explored (Figure 1; Supporting Information Table 2). Cerdulatinib demonstrated rapid absorption with a median Tmax of 3 hours. Concentrations rose up to the Tmax of 3 hours and declined thereafter with a terminal T1/2 that averaged 13.4 hours across all patients on Day 1. With QD dosing, the Cmax increased with dose up to 40 mg but did not increase thereafter, up to the highest dose of 100 mg. The steady-state cerdulatinib pharmacokinetic parameters were dose-proportional from 15 to 30 mg, greater than dose-proportional from 30 to 40 mg, and less than dose-proportional at doses >40 mg. For QD doses from 40 to 100 mg, the average steady-state Ctrough was 0.69 μM, and steady-state Cmax was 1.48 μM. A physiologic-based pharmacokinetic model suggested pH-dependent drug solubility was likely responsible for the saturation in exposure with QD dosing. BID dosing was therefore explored as a means to increase exposure. This was initiated at 15 mg BID, then 20 mg BID, and finally 45 mg BID. The maximum administered dose was 45 mg BID, which achieved a steady-state Ctrough of 1.09 ± 0.50 μM and, interestingly, more limited maximum exposure (Cmax, 1.19 ± 0.71 μM) relative to QD dosing. For each dose level, percent inhibition of SYK and JAK signaling was estimated as previously described,6 and near complete inhibition of these pathways was achieved at steady-state Ctrough at tolerated exposures (Supporting Information Table S2). The overall median treatment duration was 7.9 weeks (range, 1-165; Supporting Information Table S3). Ten patients underwent dose reduction due to an AE, and 13 had their treatment temporarily interrupted due to an AE. The most common AEs associated with treatment interruption were fatigue (n = 5) and gastrointestinal events (n = 3). Although no clear relationship between dose and the incidence of common treatment-emergent AEs was evident, symptoms typically resolved by temporarily withholding cerdulatinib for up to 2 weeks and resuming therapy at the next lower dose. Twelve patients (most with a diagnosis of CLL or FL) remained on cerdulatinib for >24 weeks, and five patients (all with FL) remained on cerdulatinib for >60 weeks (Supporting Information Figure S1). Across cohorts, patients received the majority of their intended dose (relative dose intensity, mean 90% and median 99%). Overall, the most frequently observed grade 3+ AEs were anemia (16%), fatigue (14%), and diarrhea (9%; Supporting Information Table S4). There were no consistent observations of liver abnormalities or neutropenia. Three dose-limiting toxicities were observed during the study. One patient in the 100-mg QD cohort experienced persistent grade 3 nausea despite optimal management. In the 45-mg BID cohort, one patient experienced grade 3 fatigue and one experienced grade 3 elevated serum amylase and lipase, which was associated with abdominal pain, but a normal CT; events in the 45-mg BID cohort resolved with dose reduction. The 45-mg BID dose was deemed not tolerated and, in consultation with clinical investigators, pharmacokinetic data across all cohorts were used to develop a preliminary model, which showed the 35-mg BID dose resulted in predicted average exposures ~15% lower than that observed with 45 mg BID, with Cmin levels predicted to be 25% lower. The 35-mg BID starting dose, with subsequent permissible dose reductions deemed necessary by the investigator, was reviewed by regulatory authorities prior to the initiation of the phase IIa study. In addition to the dose-limiting toxicities described above, serious AEs considered possibly/probably related to treatment included two cases of Pneumocystis jiroveci pneumonia in patients with CLL (at 30 mg QD [grade 5] and 50 mg QD [grade 4]). Prophylaxis for Pneumocystis pneumonia was instituted following the second case, with no subsequent cases reported. Additionally, one grade 5 lung infection was observed, also in a CLL patient. Hypotension (45-mg QD cohort, n = 1 [grade 3]), pyrexia (65-mg QD cohort, n = 1 [grade 2]), febrile neutropenia (45-mg BID cohort, n = 1 [grade 4]), and pancytopenia (45-mg BID cohort, n = 1 [grade 3]) were also observed. The maximum percent change in tumor volume while on therapy is presented in Figure 1B (left). Two patients with FL who initiated the study at 45 mg BID achieved a complete response to cerdulatinib after dose reduction to 30 and 15 mg BID within the first 2 therapy cycles. Five patients achieved a partial response (CLL, n = 3; FL, n = 1; transformed FL grade 3B, n = 1), with response typically manifesting within the first two therapy cycles. The single evaluable patient with transformed FL grade 3B achieved a partial response and remained on study for 140 days. All three patients with CLL who achieved a response were on therapy for >200 days. Of the five CLL patients who did not respond to cerdulatinib, two withdrew from treatment due to Pneumocystis pneumonia, and two had rapidly progressed on prior ibrutinib therapy and came off study before tumor reassessment. Patients with aggressive NHL tended to be highly refractory, with 50% of patients having progressive disease as best response to their last therapy. These patients were also typically treated at suboptimal cerdulatinib exposures and often did not remain on study through the first scheduled on-treatment CT scan at the end of Cycle 2. The single DLBCL patient who achieved a > 50% reduction in target lesions presented with a new lesion and therefore came off study due to progressive disease. Overall, the greatest clinical response was observed in patients with FL and CLL/SLL. Figure 1B (right) shows representative PET (top; FL grade 3B) and CT (bottom) scans of FL patients prior to (baseline) and following two therapy cycles. In conclusion, this phase I study showed the dual SYK/JAK inhibitor cerdulatinib was generally well tolerated and demonstrated promising antitumor activity in heavily pretreated patients with relapsed/refractory B-cell malignancies. An ongoing phase IIa study is assessing the antitumor activity of cerdulatinib in patients with B-cell or T-cell NHL. This study was sponsored by Portola Pharmaceuticals, Inc. We thank the patients who participated in this study, as well as their families and the research staff at each site. Preparation of the manuscript was supported by Iwona Bucior, PhD, of Portola Pharmaceuticals, Inc. Data analysis was supported by Andreas Betz. Medical writing and editorial support were provided by Kimberly Brooks, PhD, CMPP, of SciFluent Communications, and were financially supported by Portola Pharmaceuticals, Inc. GPC, PBC, GM, JML, KD, YK, AST, MB, AP, AB, and JTC are employees and stockholders of Portola Pharmaceuticals, Inc. PAH received research support and consultant fees from Portola Pharmaceuticals, Inc. The other authors declare no potential conflicts of interest. Contributed to the design of the study: All authors equally. Collected data: PAH, IWF, NWJ, JAB, GPC, GM, JML, and MRP. Analyzed and interpreted the data: PAH, IWF, NWJ, JAB, GPC, PBC, GM, JML, KD, MB, AP, JTC, and MRP. Performed statistical analyses: PAH, GPC, GM, JML, KD, MB, and AP. Wrote the first draft: PAH, GPC, GM, and MB. Critically revised the manuscript: All authors equally. Approved the final manuscript: All authors equally. Table S1. Baseline Patient Characteristics Table S2. Summary of Pharmacokinetic and Pharmacodynamic Results Table S3. Patient Disposition Table S4. Summary of TEAEs Figure S1. Cerdulatinib duration of treatment. Duration of cerdulatinib treatment for patients with MCL, FL, DLBCL, and CLL/SLL are shown, as indicated. Weeks on cerdulatinib is indicated on the x-axis. *Indicates two FL patients who achieved complete remission of disease. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.