ESMO 2014: Pazopanib Improves Progression-Free Survival in Phase II Study in Patients with Advanced GIST

The benefit observed in patients resistant to treatment with imatinib and sunitinib

The results of PAZOGIST, a randomised phase II study of pazopanib plus best supportive care (BSC) vs BSC alone in patients with unresectable metastatic and/or locally-advanced gastrointestinal stromal tumours (GIST), who are resistant or experienced toxicity to previous treatments with standard doses of imatinib and sunitinib, show an improvement in progression-free survival (PFS) in favour of pazopanib arm. The study was presented by Prof. Jean-Yves Blay of the University Claude Bernard Lyon I, Centre Léon Bérard, Lyon, France during the Proffered Paper session in Sarcoma at ESMO Congress 2014 in Madrid, Spain.

GIST is the most common mesenchymal neoplasm of the gastrointestinal tract. In unresectable and metastatic or locally-advanced disease, imatinib followed by sunitinib, then regorafenib represent the standard treatments in first-, second- and third-line treatments, respectively.

Pazopanib is an active treatment in soft tissue sarcomas, but it has never been evaluated in a randomised setting in advanced GIST.

In this open-label, multicentre phase II study, eligible adult patients with adequate organ functions were randomly assigned 1:1 to receive pazopanib plus BSC or BSC alone. Randomisation was stratified by number of prior drugs (2 vs ≥ 3). Switch to pazopanib was allowed for patients from BSC arm with a progressive disease.

Investigator-assessed PFS

Investigator-assessed PFS. © Jean-Yves Blay.

The primary endpoint was PFS. It was planned to include 80 patients to detect an improvement in the 4-month PFS rate from 15% in the BSC arm alone to 45% in the pazopanib plus BSC with 5% two-sided α error and 80% power. Secondary objectives included overall survival (OS), objective response rate at 4 months, best response rate and tolerance.

It was required that patients have GIST diagnosis documented histologically, measurable disease according to the RECIST criteria, performance status (PS) ECOG ≤ 2, adequate organ functions, and absence of known contraindication to pazopanib administration.

For data analysis, it was needed to record 42 PFS events, and 80 randomised patients were planned. An interim analysis with a futility stopping rule was planned after 4-month follow-up of 27 randomised patients.

In previous interim analysis in September 2012, based on both efficacy and safety results, the Independent Data Monitoring Committee recommended that study enrollment should be pursued until the targeted sample size.

From April 2011 to December 2013, 81 patients were randomised: 40 to pazopanib plus BSC and 41 to BSC arm alone.

Median age was 65 years in the pazopanib plus BSC arm and 59 years in the BSC arm. The primary localisation in pazopanibplus BSC vs BSC arms were small intestine (35.9% vs 43.6%), stomach (30.8% vs 30.8%), colon/rectum (7.7% vs 5.1%), mesentery (2.6% vs 2.6%), oesophagus (0% vs 2.6%) and other (23.1% vs 15.4%). Regarding disease status at inclusion, there were 30.8% vs 53.8% of patients with locally-advanced disease in pazopanib plus BSC vs BSC arms, respectively.  

The intent-to-treat analysis based on investigator-assessed progression showed a significant improvement in PFS with 4-month PFS rate of 47.7% for pazopanib plus BSC vs 19.5% for BSC (HR: 0.56, stratified log-rank p = 0.02).

Following progression assessed by investigators, 36 of 41 patients allocated to BSC arm received pazopanib. Median duration from inclusion to switch was 2.2 months and median PFS from date of switch 3.6 months.

The best overall response rate assessed by central review showed partial response in 0% vs 2.4%, stable disease in 84.2% vs 70.7% and progressive disease in 15.8% vs 26.8% of patients in the pazopanib plus BSC and BSC arms.

At least one serious adverse event was experienced by 52.5% of patients receiving pazopanib plus BSC vs 14.6% of patients receiving BSC alone. In the pazopanib plus BSC arm, the more frequent serious adverse events were gastrointestinal disorders (17.5%), deterioration of global health status (15%) and pulmonary embolism (12.5%).


Prof. Blay concluded that pazopanib deserves further evaluation in this population of patients. When combined with BSC, it improves PFS in patients with advanced GIST resistant to imatinib and sunitinib. The 4 month PFS rate was 50% in the pazopanib arm. Toxicity was consistent with that reported with pazopanib in other indications. The OS  data will be available at beginning of 2015.

Dr Stefan Sleijfer, who discussed the study results, said that the introduction of imatinib in advanced GIST led to a response rate higher than 50%, median OS of 5 years (before imatinib it was 9 months) and >10% benefits for more than 10 years from first-line imatinib, results that made a  cancer a chronic disease. Despite the great success of imatinib, the vast majority of patients develop resistance. Subsequent treatment options based on randomised phase III studies are sunitinib in second-line with a response rate of 7% and median PFS of 6-7 months, and regorafenib in third-line with a response rate of 5% and median PFS of 4-5 months. Pazopanib is active in non-adipocytic soft tissue sarcoma and targets VEGFR1-3, PDGFRa, and c-KIT.

Dr Sleijfer further said that based on the results of the PAZOGIST phase II study, a further study of pazopanib is needed. Appropriate design of subsequent study would be a phase III trial of pazopanib vs regorafenib in GIST patients failing to imatinib and sunitinib.

He also said that imatinib-treated GIST has transformed from a homogeneous disease with one major tumour driver to a heterogeneous disease with multiple drivers differing in tyrosine kinase inhibition sensitivity. It is unlikely that one drug will induce prolonged stable disease in majority of patients after imatinib failure, as illustrated by short PFS and low response rate from KIT-targeting agents after failure to imatinib.

New treatment approaches would be combination treatments that are difficult due to toxicity, inhibition of KIT-signaling downstream from KIT, KIT degradation, and individualised treatment based on mutational profile of dominant clone.

For future treatment in GIST after imatinib failure it would be important to identify drug based on molecular characteristics of dominant clone, monitor molecular evolution of tumour cells during treatment, and adjust treatment if necessary. The requirements would be to know which KIT mutations causing imatinib resistance respond to other targeting TKIs (sunitinib, regorafenib, pazopanib, sorafenib, nilotinib, masitinib), tools to assess mutational profile dominant clone (ctDNA, molecular imaging), and randomised study in imatinib resistant GIST of traditional treatment (sunitinib followed by regorafenib) vs treatment based on mutational profile.

Dr Sleijfer concluded by stating that it is obvious to the sarcoma community that, in order to address these challenges, global collaboration is needed.


Abstract LBA45 - A randomized multicentre phase II study of pazopanib plus best supportive care (BSC) vs BSC alone in metastatic gastroIntestinal stromal tumors (GIST) resistant to imatinib and sunitinib

GlaxoSmithKline provided the study drug and research funding for this investigator-sponsored study.