791PD - Multivariate analysis of cytokines and angiogenic factors (CAFs) and established prognostic parameters in metastatic renal cell carcinoma (mRCC) pat...

Date 01 October 2012
Event ESMO Congress 2012
Session Genitourinary tumors, non-prostate (renal cancer)
Topics Anti-Cancer Agents & Biologic Therapy
Renal Cell Cancer
Presenter Amado Zurita-Saavedra
Authors A. Zurita-Saavedra1, Y. Liu2, Y. Lin2, H.T. Tran3, L. Pandite4, J.V. Heymach5
  • 1Dept. Of Genitourinary Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston/US
  • 2Oncology Research And Development, GlaxoSmithKline, Philadelphia/US
  • 3Cancer Medicine, UT MD Anderson Cancer Center, 77030 - Houston/US
  • 4Clinical Lead, GlaxoSmithKline, Research Triangle Park/US
  • 5Thoracic/head And Neck Medical Oncology, MD Anderson Cancer Center, Houston/US




In mRCC, prognosis is still solely determined based on clinical criteria. Although multiple candidate biomarkers exist, none has yet been incorporated into practice. We identified CAFs associated with PFS in mRCC pts with ECOG PS £1 treated in a randomized, placebo-controlled, phase III clinical trial of pazopanib (Sternberg, JCO 2010, Tran, ASCO 2010 #4522). In this study, we evaluated the prognostic significance of these CAFs relative to established clinical parameters.


Seven candidate CAFs (IL-6, IL-8, VEGF, HGF, TIMP1, osteopontin [OPN], E-Selectin) in plasma obtained pretreatment, and 5 clinical variables indicative of poor prognosis (time from diagnosis to treatment <1 year, hemoglobin upper [U]LN [Heng, JCO 2009]) were correlated with PFS (n = 343; n = 225 pazopanib/n = 118 placebo) using uni- and multivariable Cox proportional hazard models. Correlations with ECOG PS, Motzer's (MSKCC) and Heng's risk criteria were also assessed. CAFs were measured using SearchLight (Aushon).


Median PFS for all pts was 29.9 weeks (95% CI 24.6–32.1), 39.6 weeks (95% CI 32.1-48.1) for those treated with pazopanib and 13.4 weeks (95% CI 12.1–19.1) for placebo. IL-6, IL-8, and OPN were prognostic [Table]. Considering treatment group, 2 clinical variables (hemoglobin ULN [HR 1.95, p = 0.0151]) were predictors of short PFS in multivariate analysis. Adjusting for those 2, only OPN and IL-6 in the placebo arm (OPN high 7.9 v low 24 weeks, p = 0.034; IL-6 high 7.9 v low 24.1 weeks, p = 0.03), and OPN alone in the pazopanib arm (high 31.3 v low 56.4 weeks, p = 0.0007), remained prognostic. Although the 3 clinical classifications were associated with PFS, none was as strong a prognostic marker as the CAFs [Table].


In mRCC pts with good PS, circulating OPN and IL-6 are prognostic for PFS independent of established clinical criteria.

CAF PFS (wks) P Clinical PFS (wks) P
IL-6 Low 24·0 <0·001 ECOG 0 18·4 0·144
High 9·9 1 13·0
OPN Low 24·0 <0·0001 MSKCC Good 24·0 0·011
High 8·4 Interm/Poor 12·1
IL-8 Low 23·9 0·002 Heng Good 24·3 0·139
High 8·4 Interm/Poor 12·6


Y. Liu: Stock ownership: Yes; GSK Membership on an advisory board or board of directors: No Corporate-sponsored research: No Other substantive relationships: No

. L.N. Pandite: Stock ownership: Yes; GSK Membership on an advisory board or board of directors: No Corporate-sponsored research: No Other substantive relationships: No.

J.V. Heymach: Stock ownership: No Membership on an advisory board or board of directors: Yes; GSK Corporate-sponsored research: Yes; GSK for biomarker analysis Other substantive relationships: No.

All other authors have declared no conflicts of interest.