Pazopanib (PZP) is a kinase inhibitor approved for the treatment of advanced soft tissue sarcoma (STS). There are no predictive biomarkers for PZP response.
Pre-treatment primary formalin-fixed paraffin-embedded tumours from PZP-treated STS patients (RMH-SARC, n = 38) with mixed histological subtypes were assessed for: (i) levels of PZP targets FGFR1 and PDGFR by immunohistochemistry (IHC), (ii) TP53 mutational status by Sanger sequencing and (iii) 730 cancer pathway-related gene expression. Consensus clustering was used to identify biological subgroups blinded to patient outcome. Multivariable Cox regression analyses of pre-defined biomarkers with patient outcome was determined for RMH-SARC and evaluated in an independent, PZP-naïve STS dataset (TCGA-SARC, n = 261).
Within RMH-SARC, seven cases were classified as low FGFR1 (F) and high PDGFRA (P) expression (F-Lo/P-Hi). Ten cases had TP53 mutation (TP53mut). These two patient subgroups had significantly worse progression free survival (PFS) and overall survival (OS) compared to their counterparts and were independent biomarkers for survival (Table 1). Among the 22 TP53wt cases which were not F-Lo/P-Hi, 3 molecular subgroups were identified and characterised with an optimal reduced list of 115 genes (Subgroups A, B, C). Integrating these multiple analytes into one decision tree classifier (defined as Pazopanib Activity and Response in SARComa - PARSARC) performed best at predicting post-PZP outcome. Subgroup A had a superior PFS and OS (median (m) PFS 13 months (ms); mOS 34ms) compared to the F-Lo/P-Hi (mPFS 1.4ms; mOS 1.7ms) and the TP53mut (mPFS 3.0ms; mOS 5.4ms) subgroups. In the independent PZP-naïve STS dataset (TCGA-SARC), these biomarkers were not associated with differential outcome, suggesting that the PARSARC classifier is a potential predictor for PZP benefit.Table: 485O
Univariate and multivariate analysis analysis of PFS and OS
|Progression Free Survival (PFS)||Overall Survival (OS)|
|F-Hi and/ or P-Lo||31||1||-||1||-||1||-||1||-|
|F-Lo and P-Hi||7||9.6|
We have developed the PARSARC classifier consisting of FGFR1 and PDGFRA IHC, TP53 sequencing and a 115 gene expression signature that appears to identify STS patients most likely to benefit from PZP therapy.
Clinical trial identification
Legal entity responsible for the study
The Institute of Cancer Research: Royal Cancer Hospital
National Institute for Health Research (NIHR) Liddy Shriver Sarcoma Initiative The Royal Marsden Cancer Charity
All authors have declared no conflicts of interest.