Abstract 1533P
Background
The multi-tyrosine kinase inhibitor pazopanib is approved for the treatment of advanced soft-tissue sarcoma and has also shown activity in other sarcoma subtypes. However, biomarkers to prospectively identify patients who will benefit from the drug are lacking.
Methods
We studied patients with soft-tissue or bone sarcomas treated with pazopanib and whose tumors were investigated by whole-genome/exome and transcriptome sequencing as part of the MASTER program, a prospective observational study of the German Cancer Consortium. Molecular and clinical characteristics were correlated with progression-free survival (PFS) and overall survival (OS). In addition, the results of the discovery cohort were validated in an independent group of pazopanib-treated patients from MASTER and in pazopanib-untreated sarcoma patients from MASTER and The Cancer Genome Atlas (TCGA).
Results
A total of 109 patients representing 20 sarcoma subtypes were analyzed. There was no association between clinical parameters or DNA-based genetic alterations and PFS. Transcriptome sequencing in 62 patients who underwent molecular analysis before pazopanib treatment showed that expression of KDR (p=0.016, hazard ratio [HR]=0.31), NTRK3 (p=0.005, HR=0.44), and IGF1R (p=0.015, HR=1.98) correlated independently with PFS. A combination score based on the features KDR-high, NTRK3-high, and IGF1R-low allowed separation of three patient groups with significantly different PFS of 8.6, 3.0, and 1.5 months for the good (n=10), intermediate (n=42), and poor (n=10) responder groups, respectively (p<0.0001). Exploratory analysis of OS also revealed a significant difference between the three groups (p=0.001). Application of the score to an independent validation cohort of pazopanib-treated sarcoma patients (n=43) confirmed its significant impact on PFS (p=0.02), whereas no such correlation was found in two independent cohorts of non-pazopanib-treated patients (MASTER, n=77; TCGA-SARC, n=256).
Conclusions
We have developed a gene expression-based predictor of pazopanib treatment outcome in sarcomas. The ability to prospectively identify patients with the best or least benefit from the drug will be of value for treatment decisions in advanced sarcoma patients.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
German Cancer Research Center.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.