The feasibility of conducting randomised controlled trials in patients with rare cancer, specifically desmoid tumours, was demonstrated in two studies presented at the 2018 American Society of Clinical Oncology (ASCO) Annual Meeting, held 1 to 5 June in Chicago, USA. The studies also showed that sorafenib and pazopanib provided clinical benefit in these patients, for whom there is currently no standard of care, and also extended progression-free survival (PFS).
Mrinal M. Grounder of Memorial Sloan Kettering Hospital in New Your, USA and colleagues conducted the phase III Alliance A091105 trial (NCT02066181), a randomised, double blind, trial of sorafenib versus placebo in patients with desmoid tumours, while Maude Toulmonde presented findings on behalf of the French Sarcoma Group from the randomised phase II DESMOPAZ (NCT01876082) trial of oral pazopanib or i.v. methotrexate/vinblastine in patients with progressive desmoid tumours.
Desmoid tumours occur rarely and generally are non-fatal but can cause significant morbidity, including cause pain, loss of mobility, bowel obstruction, and visceral organ compromise. They most often affect young patients and arise in the connective tissue at any bodily location.
Sorafenib extends PFS over placebo
In the ALLIANCE A091105 trial, 87 patients with unresectable progressive or symptomatic desmoid tumours were randomised 2:1 to receive oral sorafenib at 400 mg or placebo. The investigators also stratified these cohorts according to pain level and disease site. Patients in the placebo cohort were allowed to crossover to sorafenib upon demonstrating progressive disease (PD) according to RECIST v1.1. The study was powered to detect a hazard ratio (HR) of 0.4 favouring sorafenib, assuming 15 month median PFS for sorafenib and 6 month median PFS for placebo with a minimum 75 patients and 90% power (1-sided α = 0.025, 1 interim analysis), which would require 52 PFS events.
The primary endpoint was PFS. Other endpoints included response rates (RR), toxicity, overall survival (OS), patient-reported outcome (PRO), and the identification of tumour and imaging biomarkers.
In all, 50 patients were treated with sorafenib and 37 received placebo. The median patient age was 37 (range 18 to 72) years, 69% were female, 62% were chemotherapy-naive, and 9% of patients had received prior radiotherapy. Tumour location and pain level were balanced between arms.
Analysis of the first 75 evaluable patients was performed at a median follow-up of 26 months. The partial RR with sorafenib was 33%, with durable responses lasting 2 to 28 months versus a RR of 21% (duration 4 to 17 months) with placebo (p = 0.3). The one-year PFS rates were 87% with sorafenib compared to 43% with placebo. Median PFS was not reached (NR) with sorafenib versus 9.4 months (95% confidence interval [CI] 5.7, NR) for placebo, hazard ratio [HR] 0.14 (95% CI 0.06, 0.33) (p < 0.0001). The Data Safety Monitoring Board recommended unblinding the trial and patients on placebo were offered sorafenib. Further analyses of PRO, MRI T2 signal, and tumour biopsy are ongoing.
Thirty-eight percent of sorafenib patients and 30% of placebo patients remain on treatment, whereas 18% of sorafenib patients discontinued treatment due to toxicity; no patients in the placebo arm discontinued. Toxicities grades 3/4 occurred in 33% of patients on sorafenib versus 14% of those on placebo. The most frequently reported grades 3/4 toxicities included rash, hypertension, fatigue, and pain.
More patients receiving pazopanib showed tumour shrinkage than those on methotrexate/vinblastine
The French Sarcoma group designed DESMOPAZ as a multicentre non-comparative randomised phase II trial based on a two-stage optimal Simon’s design to assess the safety and efficacy of pazopanib in adult patients with advanced desmoid tumour. Patients were required to have documented PD according to RECIST v1.1 based upon two images performed within a 6-month interval and to provide samples of tumour tissue at baseline; an on-treatment tumour biopsy at cycle 2 was optional. Forty-eight patients were randomly assigned to receive pazopanib at 800 mg/day orally continuously, and 24 were assigned to receive methotrexate at 30 mg/m² plus vinblastine at 5 mg/m² i.v. once a week for 6 months and every 15 days for the next 6 months. Both treatments were administered until PD, whereupon cross-over was allowed, or unacceptable toxicity occurred or for a maximum of one year.
The primary endpoint was 6-month non-PD rate according to RECIST v1.1. Pazopanib could be considered to have activity if a minimum 31 of 43 patients demonstrated 6-month non-PD. As of December 2017, 72 patients had been enrolled in the 12 centres of the French sarcoma group, including 46 female and 26 male patients. The patients’ median age was 40 years (range 18 to 79) and they had undergone a median of one (ranges 0 to 3) previous lines of treatment. Of the 48 patients in the pazopanib group, 46 patients had assessable data, as did 20 of the 26 patients receiving methotrexate/vinblastine.
Central pathological and radiological review revealed that a total of 38 (82.6%) assessable patients in the pazopanib arm had tumour shrinkage; 17 (37%) patients achieved partial response (PR), and 21 (45.7%) patients showed stable disease (SD). With methotrexate/vinblastine, 11 (55%) assessable patients showed tumour shrinkage; of these 5 (25%) achieved PD, and 6 (30%) had SD. With pazopanib and methotrexate/vinblastine, the 6-month non-PD rates were 86% (95% CI 72.1, 94.7) and 50% (95%CI 27.2, 72.8), respectively.
The investigators of the Alliance A091105 trial concluded that these findings demonstrated that phase III studies in rare cancers are feasible and may be rapidly conducted in a National Clinical Trials Network group setting.
They noted that the occurrence of spontaneous responses on placebo was consistent with observation trials in patients with desmoid tumours. Nevertheless, sorafenib significantly improved PFS over placebo, and the study exceeded the primary PFS endpoint. According to the study authors, sorafenib may represent a new first- or subsequent-line standard of care in select patients with desmoid tumour.
According to the French Sarcoma Group, the primary endpoint of the DESMOPAZ study was met, with pazopanib demonstrating meaningful clinical activity in patients with progressive desmoid tumour.
The ALLIANCE A091105 trial was supported by U10CA180821, U10CA180882, U24CA196171, R01FD005105 grants.
The DESMPAZ study was sponsored by the Institut Bergonie.
Gounder MM, Mahoney MR, Van Tine BA, et al. Phase III, randomized, double blind, placebo-controlled trial of sorafenib in desmoid tumors (Alliance A091105). J Clin Oncol 36, 2018 (suppl; abstr 11500).
Toulmonde M, Ray-Coquard IL, Pulido M, et al. DESMOPAZ pazopanib (PZ) versus IV methotrexate/vinblastine (MV) in adult patients with progressive desmoid tumors (DT) a randomized phase II study from the FrenchSarcoma Group. J Clin Oncol 36, 2018 (suppl; abstr 11501).