ESMO 2014: Activity of Second-Line Dovitinib in Advanced and/or Metastatic Endometrial Cancer
Phase II study in patients with fibroblast growth factor receptor 2-mutated or –non-mutated endometrial cancer
The single-agent dovitinib demonstrated clinically meaningful activity in second-line treatment of patients with fibroblast growth factor receptor 2 (FGFR2)-mutated or –non-mutated advanced and/or metastatic endometrial cancer and a trend toward greater median progression-free survival (PFS) and overall survival (OS) in the FGFR2-mutated group. The results of a phase II study were presented by Prof. Gottfried Konecny of the Gynecologic Oncology Unit, UCLA Westwood Oncology Hematology, Los Angeles, USA during the Proffered paper session in Gynaecological Cancers at ESMO Congress 2014 in Madrid, Spain.
Activating mutations in FGFR2, identified in 10%-15% of primary endometrial cancers, are associated with disease progression and poor outcome.
Dovitinib (TKI258) is a potent tyrosine kinase inhibitor that targets FGFR, vascular endothelial growth factor receptor (VEGFR) and platelet derived growth factor receptor (PDGFR) and other kinases. Dovitinib demonstrated dose-dependent growth inhibition of FGFR2-mutated and FGFR2-non-mutated xenografts.
Dovitinib in advanced and/or metastatic endometrial cancer
In this phase II study, researchers evaluated dovitinib as a second-line therapy in patients with FGFR2-mutated or –non-mutated disease. Women who progressed after first-line chemotherapy for advanced and/or metastatic endometrial cancer underwent molecular prescreening for FGFR2 status and then clinical screening if eligible for the study (ECOG performance status - PS ≤ 2). Then they were treated with oral dovitinib on a 5-days-on/2-days-off schedule.
The primary endpoint was percentage of patients who are progression-free by investigator assessment after 18 weeks. The trial used a 2-stage design for each group; stage 2 could proceed if ≥ 8 of the first 20 treated patients (40%) met the primary endpoint. Secondary endpoints included overall response rate, disease control rate, duration of response, PFS, OS, safety, tolerability, pharmacokinetics, and pharmacodynamics.
Key eligibility criterion was progressive disease after first-line antineoplastic treatment for advanced and/or metastatic endometrial cancer. Eligible histologies were endometrioid, serous, clear cell, mucinous, adenosquamous, and mixed types. Prior antineoplastic treatment should include at least one cytotoxic agent and prior hormonal therapy was not considered as a line of treatment.
Response was assessed by local investigator every 6 plus/minus 1 weeks according to RECIST 1.1 criteria. Adverse events were assessed according to CTCAE v4.03.
FGFR2 analysis was performed on archival tumour blocks or fresh fixed tumour biopsies. FGFR2-mutated status was identified by Sanger sequencing of the 5 main hotspot mutation sites reported for endometrial cancer.
Of 248 pre-screened patients, 27 had FGFR2-mutated tumours (11%). The study enrolled 53 patients, of which22 had FGFR2-mutated disease and 31 FGFR2-non-mutated.
Among patients with FGFR2-non-mutated tumours, 17 had ECOG PS 1 vs 7 patients with mutated tumours. In the same group there were also slightly more patients with serous and clear cell adenocarcinoma histology and poorly differentiated tumours. Median relative dose intensity was similar.
All patients discontinued the treatment mostly due to progressive disease (66%) or adverse events (26%). Most frequent adverse events leading to discontinuation were deep vein thrombosis, pulmonary embolism, and small intestinal obstruction.
The observed 18-week PFS rates were 32% in patients with FGFR2-mutated tumours and 29% in patients with FGFR2-non-mutated. The 18-week PFS rate in the first 20 patients was 35% and 25% in the FGFR2-mutated and FGFR2-non-mutated groups, respectively. However, the study did not proceed to stage 2 based on the predefined criteria.
The 18-week PFS rates from a Kaplan Meier analysis were 48% in FGFR2-mutated and 38% in FGFR2-non-mutated group.
The disease control rate (≥ stable disease) was 64% (59% stable disease, 5% partial response) in the FGFR-2 mutated group and 51% (35% stable disease, 16% partial response) in the FGFR2-non-mutated groups, respectively.
Median PFS (4.1 vs 2.7 months) and OS (20.2 vs 9.3 months) trended to be higher in the FGFR2-mutated group.
Adverse events suspected to be study drug related were similar between the groups. The most common grade 3/4 adverse events were hypertension (17%) and diarrhoea (9%). Of the 5 on-treatment deaths, 4 were due to endometrial cancer and 1 was due to cardiac arrest. The safety profile was similar to that observed in other dovitinib trials.
Prof. Konecny concluded that single-agent dovitinib demonstrated clinically meaningful activity in both groups. There was a trend toward greater median PFS and survival in the FGFR2-mutated group. The overall safety profile was similar to that observed in other dovitinib trials. However, the incidence of thrombosis appeared more common in this patient population.
Dr Michael Bookman, who discussed the study results, said that accrual was biased toward high-grade and advanced-stage tumours with increased risk of recurrence. 40-50% of patients have already received pelvic radiation, with impact on haematologic reserve. Recurrent disease within the pelvis is less responsive to many treatments, due to prior surgery and/or radiation. Historical controls are necessary to define the primary study hypothesis, but the reference population may not have the same molecular profile as the enrolled population.
According to Dr Bookman, the investigators have provided a well-designed and carefully-conducted phase II experience with dovitinib in endometrial cancer. Patients were allowed one prior treatment, plus/minus pelvic radiotherapy, and appear similar to patients enrolled on GOG 229 phase II trials, contributing to a broader experience. As a multi-targeted tyrosine kinase inhibitor, dovitinib also inhibits VEGFR2. Without randomised combinations or stratification for prior anti-VEGF therapy, it is uncertain which pathway(s) accounted for the observed outcomes.
FGFR2 is mutated in 10% of early-stage endometrioid (type I) tumours, and associated with risk of recurrence. The prognostic significance of mutations in recurrent or metastatic disease is uncertain. However, in this trial, patients with FGFR2-mutated tumours had an improved prognosis, compared to non-endometrioid (type II) tumours.
Type I endometrial cancer is associated with frequent (non-overlapping) mutations in KRAS, CTNNB1, FGFR2, and PIKC3A, with associated pathway activation. In recurrent disease, targeting VEGF (with bevacizumab) demonstrates activity that appears equal or superior to multi-targeted tyrosine kinase inhibitors, including anti-FGFR2. Some tyrosine kinase inhibitors may not be effective in the setting of common activating mutations within the kinase domain of FGFR2. Understanding, and optimising, the net contribution of each pathway awaits randomised trials that incorporate stratification based on prior treatment and analysis of biospecimens.
The study sponsor was Novartis Pharmaceuticals.