Summary Basis of Decision for Zydelig

Clinical Pharmacology

Absolute bioavailability of idelalisib (the medicinal ingredient in Zydelig) is unknown but is expected to be high based on pharmacokinetic parameters and metabolite profiles. Limited data indicate that administration with food increases the exposure [area under the curve (AUC)] to idelalisib, however specific recommendations regarding food intake in relation to Zydelig administration are not necessary as the changes in idelalisib pharmacokinetics were modest and in the pivotal clinical studies Zydelig was administered without regard to food. Idelalisib is metabolized primarily by aldehyde oxidase and to a lesser extent by cytochrome P450 (CYP) 3A and glucuronidation by uridine 5'diphospho glucuronosyltransferase (UGT) 1A4. The primary metabolite, GS-563117 is an oxidative product that is inactive against phosphatidylinositol 3-kinase (PI3K), but is an irreversible inhibitor of CYP3A. Excretion of idelalisib and its metabolites is predominantly via the feces. The mean terminal half-life of idelalisib ranged from 6 to 12 hours.

Although there was no statistically significant effect of age on idelalisib pharmacokinetics, clinical safety findings (including serious adverse events, discontinuations and deaths due to adverse events) occurred more frequently among elderly patients. Idelalisib has not been studied in the pediatric population.

Preliminary data of a dedicated pharmacokinetic study indicate 1.7 fold increases in exposure in healthy volunteers with moderate and severe hepatic impairment; however patients with hepatic impairment were excluded from safety and efficacy studies of Zydelig due to its well-established hepatotoxic risk. Safe dosing recommendations can therefore not be made in patients with severe hepatic impairment. In healthy volunteers, idelalisib had no clinically significant effect on the QT interval and renal impairment had no clinically significant effects on idelalisib pharmacokinetics.

Idelalisib is a weak substrate of CYP3A, and the primary circulating metabolite, GS-563117, is a strong CYP3A inhibitor. In clinical drug-drug interaction studies, ketoconazole, a strong CYP3A inhibitor, increased idelalisib exposure [area under the curve from time zero extrapolated to the infinite time (AUC_inf)] by 79%, while rifampin, a strong CYP3A inducer, reduced idelalisib exposure by 75%. In a clinical drug-drug interaction study, exposure to a CYP3A substrate, midazolam, was increased by 440% with concomitant idelalisib, consistent with the identification of GS-563117 as an irreversible strong inhibitor of CYP3A. Therefore, concomitant use of Zydelig with narrow therapeutic index CYP3A substrates and with strong CYP3A inducers is not recommended. Warning statements are included in various sections of the Product Monograph including in Warnings and Precautions and Drug Interactions.

For further details, please refer to the Zydelig Product Monograph, approved by Health Canada and available through the Drug Product Database.

Clinical Efficacy

Follicular Lymphoma

The safety and efficacy of Zydelig was assessed in Study 101-09, a Phase II, single-arm, multicentre clinical study that enrolled 125 refractory and pre-treated indolent non-Hodgkin's lymphoma (iNHL) patients. Four disease subtypes under the umbrella classification of iNHL were eligible for enrollment: follicular lymphoma (FL) (72 patients); small lymphocytic lymphoma (28 patients); marginal zone lymphoma (15 patients); and lymphoplasmacytic lymphoma/Waldenstrom's macroglobulinaemia (10 patients). Only the FL subpopulation had a sufficient number of patients for regulatory evaluation.

During the study, patients received 150 mg of Zydelig taken orally twice daily until there was evidence of disease progression or unacceptable toxicity. Tumour response was assessed according to the revised International Working Group response criteria for malignant lymphoma. The primary endpoint was overall response rate (ORR) as assessed by an independent review committee (IRC). Secondary endpoints included duration of response (DOR), progression-free survival (PFS), and overall survival (OS).

In the FL sub-population, the median age was 62 years (range 33 to 84), 54% were male, and 90% were Caucasian. At enrollment, 92% of patients had a baseline Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1. The median time since diagnosis was 4.7 years and the median number of prior treatments was 4 (range 2 to 12). The most common prior combination regimens were:

1. rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) 49%;
2. bendamustine and rituximab (BR) 50%; and
3. rituximab, cyclophosphamide, vincristine, and prednisone (R-CVP) 28%.

At baseline, 33% of patients had extranodal involvement and 26% had bone marrow involvement. Twenty-one patients (23.2%) had disease classified as Grade 1, 39 (54.2%) had disease classified as Grade 2, and 12 (16.7%) had disease classified as Grade 3a.

The primary endpoint, ORR, was achieved in 39 patients (54.2%). A complete tumour response was observed in 6 patients (8.3%). The median DOR was not reached; however, the Kaplan-Meier curve suggests a median DOR of at least 6 months.

Overall, the efficacy analysis for monotherapy in the FL population demonstrates a tumour response that is substantial; however, the durability of the response could not be assessed with confidence due to the short median time of drug exposure. Overall, the efficacy observed in the pivotal study is premature and final results from this study will be needed for confirmation of clinical benefit.

Chronic Lymphocytic Leukemia

The safety and efficacy of Zydelig was assessed in chronic lymphocytic leukemia (CLL) in Study 312-0116. This study was a Phase III, randomized, double-blind, placebo-controlled study conducted in 220 patients with relapsed CLL who required treatment but were not considered suitable for cytotoxic chemotherapy based on one of the following criteria: Cumulative Illness Rating Score (CIRS) >6; estimated creatinine clearance (CrCl) <60 mL/min; Grade ≥3 neutropenia or Grade ≥3 thrombocytopenia resulting from myelotoxic effects of prior therapy with cytotoxic agents. Patients were randomized 1:1 to receive 8 cycles of rituximab (first cycle at 375 mg/m², subsequent cycles at 500 mg/m²) in combination with either an oral placebo twice daily or with 150 mg of Zydelig taken twice daily until development of disease progression or unacceptable toxicity.

The median age was 71 (range 47, 92) with 78.2% of patients over 65 years of age. 65.5% of patients were male, 90.0% were Caucasian, 64.5% had a Rai stage of III or IV, and 55.9% had Binet Stage C. Patients had a median CIRS score of 8; 81 (36.8%) had cardiac, 114 (51.8%) had respiratory, 87 (39.5%) had renal, and 93 (42.3%) had endocrine/metabolic comorbidities. Two hundred and six patients (93.6%) had three or more organs with comorbidities and 82 (37.3%) had severe (score of 3 or higher in any system) comorbidities. The median number of prior therapies was 3.0 (range 1-12). Nearly all (95.9%) patients had received prior anti-CD20 monoclonal antibodies. The most common (>15%) prior regimens were: bendamustine + rituximab (98 patients, 44.5%), fludarabine + cyclophosphamide + rituximab (75 patients, 34.1%), single-agent rituximab (67 patients, 30.5%), fludarabine + rituximab (37 patients, 16.8%), and chlorambucil (36 patients, 16.4%). Most patients had adverse cytogenetic prognostic factors: 43.2% had a 17p deletion (deletion on the short arm of chromosome 17) and/or tumour protein p53 (TP53) gene mutation, and 83.6% had an unmutated immunoglobulin heavy chain variable (IGHV) gene.

The primary endpoint was PFS, defined as the interval from randomization to the earlier of the first documentation of definitive progressive disease or death from any cause. Definitive disease progression was based on standard criteria; however, lymphocytosis alone was not considered evidence of disease progression. Other key efficacy outcomes defined in the study protocol included ORR and overall survival. Overall survival results were limited to 26 reported events (12% of the total patient population) at the second interim analysis and are not included in the Product Monograph. The PFS and ORR results reported in the Product Monograph were based on evaluation by an IRC.

There was a clinically significant benefit in PFS for Zydelig + rituximab versus placebo reported following the first and second pre-specified interim analyses. Data reviewed following the second interim analysis reported a hazard ratio of 0.18 [95% confidence interval (CI): 0.10, 0.32, p<0.0001] in favour of the Zydelig + rituximab arm. The median PFS was not reached in the Zydelig + rituximab arm and was 5.5 months in the placebo + rituximab arm.

Overall response rates were significantly higher in the Zydelig + rituximab (74.5%) than in the placebo + rituximab arm (14.5%). Improvements in responses in the Zydelig + rituximab arm were driven by partial responses; complete responses were not achieved in either arm. Subgroup analyses in PFS and ORR also gave consistent results in patients for whom current treatment options have limited efficacy (17p deletion, IGHV unmutated).

The efficacy data reviewed demonstrates substantial evidence of clinical effectiveness for Zydelig in the treatment of patients with relapsed CLL. It is not possible, however, to rule out a contribution of rituximab to the observed effects, supporting the decision to indicate Zydelig in combination with this agent.

For more information, refer to the Zydelig Product Monograph, approved by Health Canada and available through the Drug Product Database.

Clinical Safety

Follicular Lymphoma

The safety evaluation of Zydelig for the treatment of patients with follicular lymphoma was based on safety data from 352 patients with hematological malignancies. Of these 352 patients, 200 were diagnosed with iNHL, of which 146 received the indicated oral dose of 150 mg daily. In the overall safety database, the median duration of exposure to Zydelig was relatively short at months. Seventy-eight patients received Zydelig for ≥12 months.

In the population of iNHL patients who received Zydelig at 150 mg, the safety profile generally mirrored the overall safety database. Almost all patients had at least one treatment emergent adverse event (AE) (82% had any Zydelig-related AE and 71% of patients had a grade 3 or higher AE). Approximately 50% of patients had a serious AE, including: pneumonia (15%), diarrhea (11%), and pyrexia (9%). Serious AEs were considered Zydelig-related in 30% of patients. Deaths related to AEs occurred in 8.2% of patients. Adverse events leading to dose reduction and discontinuations occurred in 23% and 25% of patients, respectively. Specifically, the AEs of transaminitis, diarrhea/colitis and pneumonitis were identified as AEs of concern.

Transaminitis occurred in over 40% of patients and elevations of alanine aminotransferase (ALT) or aspartate aminotransferase (AST) Grade 3 or higher occurred in 12-18% of patients. Generally, transaminitis was asymptomatic and successfully treated with dose interruption. Nearly all events resolved to Grade 1 or lower. There were no fatalities.

Diarrhea was the most common AE observed in all patients treated with Zydelig monotherapy (47%) and Grade 3/4 diarrhea/colitis was reported in 14% of patients in the monotherapy population. Time of onset for the more severe cases was later in treatment at a median onset of 5.3 months. Treatment consisted of interruption of Zydelig and various anti-diarrheal and anti-inflammatory drugs, including budesonide, mesalamine and systemic steroids, as well as supportive care. Resolution occurred in the majority of patients (85%) at a median of one month. There were no fatalities secondary to diarrhea/colitis.

Pneumonitis occurred in 2.5% of patients, unpredictably throughout treatment, and there was one fatality in the iNHL monotherapy population. Pneumonitis can be a fatal disease and its accurate diagnosis is sometimes difficult in the presence of other respiratory adverse events, such as pneumonia, that can occur commonly in this population.

Overall, in the monotherapy safety database, there is a moderate level of toxicity observed in the iNHL patient population which required a significant degree of drug modification (dose interruption, dose reduction and drug discontinuation). Although the level of toxicity does not preclude a positive benefit/risk assessment, prescribers and patients need to be fully informed of the risks associated with Zydelig use. Specifically, these risks include, but are not limited to, transaminitis, diarrhea/colitis and pneumonitis.

Chronic Lymphocytic Leukemia

The safety profile of Zydelig for the treatment of patients with relapsed chronic lymphocytic leukemia (CLL) was primarily based on the Phase III study previously described in the Clinical Efficacy section. In patients who were treated with Zydelig + rituximab, the most common AEs were pyrexia (35%, 38 patients), neutropenia (27%, 30 patients), and nausea (25%, 28 patients), whereas infusion-related reactions, fatigue, and cough were most common in the placebo + rituximab arm. The AEs most commonly reported as drug-related in Zydelig treated patients were diarrhea (10%, 11 patients), neutropenia (11%, 12 patients), and fatigue (9.1%, 10 patients). Grade 3 AEs assessed as related to Zydelig were: neutropenia (7.3%, 8 patients), diarrhea (3.6%, 4 patients), increase in alanine aminotransferase (ALT), pneumonitis (each 2.7%, 3 patients), increase in transaminases, and colitis (each 1.8%, 2 patients).

Deaths related to AEs occurred in 3 (2.7%) of Zydelig-treated patients. Two additional Zydelig-treated patients died during the study due to disease progression-related causes. Thirteen placebo patients died, primarily due to disease progression-related causes. Adverse events leading to dose reduction and discontinuations occurred in 15% and 10% of Zydelig-treated patients, respectively. Specifically, the AEs of transaminitis, diarrhea/colitis, neutropenia, rash and pneumonitis were identified as AEs of concern, although there were no deaths associated with these events.

Transaminitis occurred in 35% of patients and elevations of ALT or AST that were Grade 3 or higher occurred in 8% of patients. Generally, transaminitis was asymptomatic and successfully treated with dose interruption.

Diarrhea was common (21%) and grade 3/4 diarrhea/colitis was reported in approximately 5% of patients

Neutropenia was common in both treatment arms, but was more frequent in the Zydelig-treated patients (37% vs. 27%). Despite this augmentation in risk, Zydelig did not appear to be broadly immunosuppressive, and Zydelig-treated patients generally experienced stabilization or improvement of cytopenias observed at baseline.

Rash was reported in twenty Zydelig-treated patients, and Grade 3 rash was observed in 4 patients (4%), and was responsible for three dose reductions and one discontinuation.

Grade 3 pneumonitis occurred in 3.6% of Zydelig-treated patients and 0.9% of placebo-treated patients, with significant latency.

Conclusion

Rates for the selected AEs described above were lower for CLL patients treated with Zydelig compared to iNHL patients. Dose reductions, interruptions and discontinuations were also lower in CLL patients. These discrepancies were ascribed by the sponsor to a higher level of pre-treatment and associated morbidity in the iNHL patients, as well as implementation of more careful AE management practices in the later CLL pivotal study. Accordingly, the Product Monograph was revised to incorporate the most up-to-date grade-specific adverse event management practices from the CLL pivotal study protocol. Furthermore, the sponsor is currently developing a protocol for a dose optimization study in the iNHL population.

For more information, refer to the Zydelig Product Monograph, approved by Health Canada and available through the Drug Product Database.