Ruxolitinib + capecitabine in advanced/metastatic pancreatic cancer after disease progression/intolerance to first-line therapy: JANUS 1 and 2 randomized phase III studies

Study design

JANUS 1 and JANUS 2 were randomized, double-blind, placebo-controlled, multinational studies that evaluated ruxolitinib or placebo in combination with capecitabine in patients with advanced/metastatic pancreatic cancer after disease progression or intolerance to first-line therapy (JANUS 1 and JANUS 2 ClinicalTrials.gov registration numbers: and , EudraCT registration numbers: 2014–000293-20 and 2014–000294-39, respectively). Randomization was performed centrally via an interactive response technology system and was stratified by mGPS (1 vs 2) and Eastern Cooperative Oncology Group performance status (ECOG PS; 0 or 1 vs 2). All investigators, investigational site staff, patients, and study personnel were blinded to treatment allocation.

Each study was conducted in accordance with the study protocol, Declaration of Helsinki, Good Clinical Practice guidelines, and applicable regulatory requirements. Study protocols and amendments were approved by the relevant institutional review board or independent ethics committee. All patients provided written, informed consent before study participation.

Patients

Adult patients with radiographically evaluable advanced pancreatic adenocarcinoma that was inoperable/metastatic, according to Response Evaluation Criteria In Solid Tumors (RECIST), version 1.1, were enrolled. All patients must have received one prior chemotherapy regimen for advanced/metastatic disease. Patients who previously received neoadjuvant/adjuvant chemotherapy with fluoropyrimidine-based chemotherapy were eligible. Fluoropyrimidine-containing therapies were allowed as the first-line regimen, provided the patient had discontinued treatment for reasons other than disease progression. Patients were required to have a baseline mGPS of 1 (CRP >10 mg/L and albumin ≥35 g/L) or 2 (CRP >10 mg/L and albumin <35 g/L) on a 3-point (0–2) scale, with higher scores indicating worsening prognosis; and an ECOG PS between 0 and 2, on a 5-point scale, with 0 indicating no symptoms and higher scores indicating a worsening ability to conduct activities of daily living.

Patients were excluded if they had received more than one prior regimen for advanced/metastatic disease or single-agent capecitabine in the first-line setting; had brain/central nervous system metastases; a history of seizures; any chronic/active infection requiring systemic treatment; a history of, or active, viral infection (hepatitis B, hepatitis C, or human immunodeficiency virus); any other malignancy ≤2 years (except protocol-specified skin and genitourinary carcinomas); prior treatment with a JAK inhibitor, recent or ongoing radiation therapy, or concurrent anticancer therapy; recent (≤3 months) or ongoing partial or complete bowel obstruction, unless related to the disease under study and surgically corrected; were unwilling to undergo blood transfusion; or had any clinically significant disease or finding that, in the investigator’s opinion, precluded study participation.

Treatment

In both studies, patients were randomly assigned, in a 1:1 ratio, to receive ruxolitinib 15 mg twice daily or matching placebo, orally self-administered as three 5-mg tablets in two daily doses approximately 12 h apart for 21 days. In addition, capecitabine 2000 mg/m² per day was orally self-administered in approximately two equal doses, for the first 14 days of each cycle in both treatment groups (Fig. 1). Treatment was initiated within 3 days of randomization and consisted of continuous 21-day cycles until intolerance or disease progression. Ruxolitinib could be escalated to a maximum dose of 25 mg twice daily at the investigator’s discretion if the patient had stable neutrophil and platelet laboratory parameters (ie, ≤ grade 1) for two complete cycles. Protocol-specified dose interruptions and adjustments also were permitted for safety reasons at the discretion of the investigator. Following withdrawal from the study, patients returned 30 to 35 days later for a safety visit. Survival assessment was planned every 6 weeks until completion of final primary analysis, after which assessments were to occur every 12 weeks.

Assessments

The primary endpoint in each study was OS, defined as the time from randomization to death from any cause. Secondary endpoints included progression-free survival (PFS), defined as the time from randomization to disease progression (per RECIST 1.1) or death from any cause if earlier; overall response rate (ORR), defined as the proportion of patients who had a partial response or complete response per RECIST 1.1; and safety and tolerability. Radiographic tumor assessments were performed every 6 weeks until progressive disease, and treatment response was assessed according to RECIST 1.1 by blinded investigator radiologic review. Subgroup analyses for OS were also conducted to investigate consistency of treatment effects. Safety and tolerability were assessed by adverse events (graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events [CTCAE], version 4.03, with the exception of death [CTCAE severity grade 5], which was collected as an outcome), laboratory results (hematologic and chemical), physical examination, vital signs, and electrocardiography.

The exploratory endpoints of JANUS 1 and 2 were to evaluate and compare the treatment groups with respect to: (i) body weight, (ii) health-related quality of life (HRQoL), and (iii) pharmacodynamics and tumor markers. HRQoL was assessed using the 45-item Functional Assessment of Cancer Therapy – Hepatobiliary questionnaire (FACT-Hep), version 4, which assessed five areas of quality of life (QoL): physical well-being, social/family well-being, emotional well-being, functional well-being, and additional concerns [18]. Responses were based on a 5-point scale, with a higher score indicating a higher QoL (0 = not at all to 4 = very much). In addition, for both studies, a planned exploratory endpoint was to evaluate the pharmacokinetics of ruxolitinib, capecitabine, and the active metabolite of capecitabine, fluorouracil, when administered alone and in combination. Exploratory analyses also were planned for clinically relevant tumor markers and markers of inflammation, which included but was not limited to CRP, CA19–9, and CA125. Other markers such as vascular cell adhesion molecule-1 and adiponectin that were measured will be reported separately.

Statistical considerations

It was calculated that a total of 262 deaths in JANUS 1 and 227 deaths in JANUS 2 would provide 90% and 85% power, respectively, to detect a 33% reduction in the risk of death (hazard ratio [HR] = 0.67) in the ruxolitinib group (vs placebo) with a two-sided type 1 error of 5%. For each study, an event-driven interim analysis for futility and efficacy was planned when approximately 170 (JANUS 1) or 148 (JANUS 2) total deaths had occurred. In both JANUS 1 and JANUS 2, an interim analysis was planned to be conducted by an external unblinded statistician prior to a review by an independent data monitoring committee, which then made recommendations to the sponsor as to the continuation or early termination of the studies; the planned one-sided p-value threshold for futility was ≥0.7465 for JANUS 1 and ≥0.7672 for JANUS 2. In JANUS 2, the planned interim futility and efficacy analyses were not conducted.

The efficacy analysis was assessed in the intent-to-treat population, which comprised all patients randomized to the study. The safety analysis was based on the safety population, which comprised all patients who received at least one dose of the assigned study treatment.

For both studies, the Kaplan-Meier method was used to estimate OS and PFS. For the OS analyses, patients alive at the time of analysis were censored at the time of last contact. The stratified log-rank test was used to analyze the OS and PFS differences between treatment groups. HRs and associated 95% confidence intervals (CIs) were estimated based on the stratified Cox regression model using Efron’s method of accounting for ties. ORR was compared between treatment groups using the Miettinen and Nurminen method for stratified data. The analyses of OS, PFS, and ORR were stratified by mGPS status (1 vs 2) and ECOG PS (0 or 1 vs 2). For subgroup analyses of OS, the HR and 95% CIs were provided for each subgroup based on the Cox proportional hazards model.

Exploratory analyses were conducted using a two-sided 5% significance level. Changes and percentages in body weight were summarized by visit and treatment group. The proportion of patients achieving a minimally important difference (ie, a 7-point increase from baseline) in FACT-Hep total score was analyzed using a logistic regression model with mGPS and ECOG performance status as covariates. Descriptive statistics, change, and percentage change from baseline in tumor marker levels were measured by treatment group at each scheduled visit. The pharmacokinetics of ruxolitinib, capecitabine, and fluorouracil (the active metabolite of capecitabine) were planned to be evaluated.

Unless otherwise noted, all analyses were conducted with SAS Version 9.0 or later (SAS Institute, Inc., Cary, NC, USA).

Efficacy

Exploratory analyses

No clinically relevant difference was seen between treatment groups in the largest confirmed percentage change in baseline body weight (mean in kilograms [standard deviation]) in both studies (JANUS 1: ruxolitinib group, −1.4 [5.45]; placebo group, −2.3 [5.60]; p = 0.4308; JANUS 2: ruxolitinib group, −0.2 [6.09]; placebo group, −2.4 [5.00]; p = 0.9218). The proportions of patients achieving a minimally important difference (ie, a 7-point increase from baseline) in FACT-Hep total score were similar between treatment groups and both studies (JANUS 1: ruxolitinib group, 28.6%; placebo group, 32.5%; odds ratio, 1.22; 95% CI, 0.735–2.031; p = 0.48; JANUS 2: ruxolitinib group, 25.6%; placebo group, 39.5%; odds ratio, 1.737; 95% CI, 0.587–5.331; p = 0.38).

The levels of the pharmacodynamic biomarker, CRP, were similar between groups at baseline (mean CRP [mg/L]: JANUS 1: ruxolitinib group, 39.5; placebo group, 38.5; JANUS 2: ruxolitinib group, 36.0; placebo group, 32.1). Cycle 2 Day 1 median CRP reductions from baseline were larger for the ruxolitinib group vs the placebo group in both studies (JANUS 1: −57.4% vs −33.3%; JANUS 2: −55.2% vs −33.3%). The number of patients with reductions in tumor markers CA19–9 and CA125 from baseline at Cycle 2 Day 1, the last visit where patients were reliably measured, were similar between groups in JANUS 1 and numerically higher in the placebo group in JANUS 2 (CA19–9: JANUS 1: ruxolitinib group, 53/114 [46.5%], placebo group, 53/110 [48.2%]; JANUS 2: ruxolitinib group, 12/26 [46.2%], placebo group, 17/30 [56.7%]) (CA125: JANUS 1: ruxolitinib group, 29/114 [25.4%]; placebo group, 27/110 [24.5%]; JANUS 2: ruxolitinib group, 7/26 [26.9%], placebo group, 12/30 [40.0%]). As the studies were terminated, pharmacokinetic analyses were not performed.

Safety and tolerability

Safety population included 307 patients in JANUS 1 (ruxolitinib plus capecitabine, 153; placebo plus capecitabine, 154) and 85 patients in JANUS 2 (ruxolitinib plus capecitabine, 42; placebo plus capecitabine, 43). Median treatment durations were 42 days (range, 1–400) for patients receiving ruxolitinib plus capecitabine and 42 days (1–392) for patients receiving placebo plus capecitabine in JANUS 1; the corresponding treatment durations were 40 days (1–251) and 43 days (1–247), respectively, in JANUS 2. Median dose intensity was 97.5% (range, 10.9–1000.0) for patients receiving ruxolitinib plus capecitabine (one patient was dosed for 3 days only and was treated as received all of the study treatment, which led to the high dose intensity) and 97.4% (12.6–425.0) for patients receiving placebo plus capecitabine in JANUS 1; the corresponding median dose intensity was 97.0% (54.9– 125.8) and 97.7% (48.0–109.0), respectively, in JANUS 2. Over 90% of patients had at least one treatment-emergent adverse event (TEAE) in each of the treatment groups in both studies; over 70% had at least one grade 3/4 TEAE (Table 3). The most common nonhematologic and hematologic TEAEs are summarized in Tables 4 and 5, respectively.

Over 60% of patients in the ruxolitinib plus capecitabine groups and over 45% of patients in the placebo plus capecitabine groups of both studies had serious TEAEs (Table 3). Serious TEAEs occurring in ≥6% of patients in the ruxolitinib plus capecitabine group or the placebo plus capecitabine group, respectively, in either study were abdominal pain (JANUS 1: 4% vs 12%; JANUS 2: 10% vs 5%), dyspnea (JANUS 1: 3% vs 3%; JANUS 2: 10% vs 2%), nausea (JANUS 1: 3% vs 3%; JANUS 2: 7% vs 9%), dehydration (JANUS 1: 5% vs 3%; JANUS 2: 7% vs 2%), and vomiting (JANUS 1: 5% vs 3%; JANUS 2: 2% vs 12%). A total of 34 and 10 fatal TEAEs were reported in JANUS 1 and JANUS 2, respectively.