1493P - Results of HRQOL data from palette: a randomized double blind phase III trial of pazopanib versus placebo in metastatic soft tissue sarcoma (STS) pa...

Date 29 September 2012
Event ESMO Congress 2012
Session Poster presentation I
Topics Supportive Care
Soft Tissue Sarcomas
Presenter Corneel Coens
Authors C. Coens1, W.T.A. van der Graaf2, J. Blay3, S.P. Chawla4, D. Kim5, R. Sanfilippo6, S. Manson7, M. Ouali8, S. Marréaud9, A. Bottomley1
  • 1Quality Of Life Department, EORTC HQ, 1200 - Brussels/BE
  • 2Medical Oncology /452, Radboud University Medical Centre MijmegenUniversity Medical Center St. Radboud, NL-6500 HB - Nijmegen/NL
  • 3University Claude Bernard Lyon I, Centre L, 69373 - Lyon cedex /FR
  • 4Med. Oncology, Sarcoma Oncology Center, 90403 - Los Angeles/US
  • 5Department Of Internal Medicine, Seoul National University Hospital, 110-744 - Seoul/KR
  • 6Cancer Medicine Department, Istituto Nazionale dei Tumori, 20133 - Milano/IT
  • 7R&d, GlaxoSmithKline, Uxbridge/UK
  • 8Statistical Department, EORTC HQ, 1200 - Brussels/BE
  • 9Medical Department, EORTC HQ, 1200 - Brussels/BE

Abstract

Background

HRQoL was an important secondary endpoint in this global double-blind, randomised phase III trial of pazopanib 800 mg QD versus placebo as second or later line treatment for advanced STS patients (N = 369) where progression free survival was significantly improved in the pazopanib arm (median: 4.6 versus 1.6 months; hazard ratio = 0.31; P < 0.001), but no statistically significant difference was observed in overall survival. Toxicity of pazopanib consisted mainly of fatigue, diarrhea, nausea weight loss and hypertension.

Methods

HRQoL was assessed using EORTC QLQ-C30 at baseline, week 4, 8 and 12 in patients who were progression free. The primary HRQoL endpont was the QLQ-C30 global health scale, analyzed using linear mixed modeling supplemented with sensitivity analyses using alternative scales, methods and imputation of missing data.

Results

Compliance of HRQoL was good in this trial, ranging from 94% at baseline to 81% at week 12. Placebo patients completed fewer questionnaires, largely due to a higher progression rate (median number of completed questionnaires- pazopanib: 3 vs placebo: 2). Differences in the global health scale between the two treatment arms assessed at weeks 4, 8 and 12 were not significant and did not exceed the pre-determined minimally important difference of 10 points (P = 0.291; average difference = 2.6 points). The EQ-5D data and sensitivity analyses mainly confirmed this finding. Among the other scales, the pazopanib arm reported significantly worse symptom scores for diarrhea (P < 0.001), loss of appetite (P < 0.001), nausea/vomiting (P < 0.001) and fatigue (P = 0.012). On the pazopanib arm, 74.8% of patients reported at least one minimally important worsening in any of those four symptom scales, compared to 51.2% on the placebo arm. In general, HRQoL scores tended to decline over time in both arms, indicative of the underlying disease. Exploratory factor analysis revealed that global health was associated mainly with functioning scales (Physical, Role and Social) and less with symptom scales except for fatigue.

Conclusion

The toxicity profile of pazopanib is reflected in the patients' self-reported symptoms but did not translate into worse overall global health while on-treatment, as patients maintained continued good functioning.

Disclosure

J. Blay: research support and honoraria from GSK, Novartis, Roche, MSD, PharmaMar.

S.P. Chawla: consultant GSK.

D. Kim: consultant GSK, MEK Advisory Board.

R. Sanfilippo: Honoraria from GSK.

S. Manson: employee of GSK, holds GSK shares.

A. Bottomley: Funding for EORTC research projects from Roche, Genentech, BMS.

All other authors have declared no conflicts of interest.