Growth modulation index and RECIST-based STBSG-EORTC criteria for the assessment of drug activity in advanced soft tissue sarcoma (ASTS) patients (pts).

Abstract

10014 Background: Despite promising phase II trial results, most of new drugs failed to improve the overall survival (OS) in ASTS pts. The choice of the endpoint in early trials remains crucial. We have evaluated the Growth Modulation Index (GMI) as potential endpoint. The GMI is defined as the Time To Progression with the second (or n+1) line of chemotherapy (TTP2) divided by the TTP with the first (or n) line (TTP1). Methods: We have carried out a retrospective multicenter study in pts receiving second-line chemotherapy after failure/intolerance to doxorubicin-based regimens. Data collected included best response(s), TTP1 & TTP2 and OS. Treatments have been classified as "active" treatment according to the EORTC- STBSG criteria (3-month Progression-free rate >40% or 6-month PFR>14%: including trabectedin, ifosfamide, gemcitabine+docetaxel for all subtypes and weekly paclitaxel for angiosarcoma) versus non-active drugs. Comparisons used chi-2 tests and Log-rank tests. We performed a logistic regression analysis for identifying factors associated with longer GMI. Results: The study population consisted in 106 men and 121 women, the median age was 57 years. 110 pts (48.4%) have received "active drugs". The median TTP1, TTP2 and GMI were 197 days, 134 days and 0.75, respectively. The median OS was 446 days. 70 pts experienced GMI>1.33 (30.6%). There was a strong relation between best objective response and GMI (p<0.0001). The treatment with "active drug" was not associated with an improvement of the OS: 490 days 407 versus days (p=0.524). The median OS of pts with GMI<1, GMI=[1.00-1.33] and GMI>1.33 were 324, 302 and 710 days, respectively (p<0.0001). None of the following factors were associated with GMI>1.33 in logistic regression analysis: age, gender, histological subtypes, grade, metastasis locations, interval between diagnosis & metastasis, association with ifosfamide or dacarbazine in 1st-line regimen or treatment with "active drug" in second-line. Conclusions: GMI seems to be an interesting endpoint providing additional information compared to classical criteria. GMI>1.33 is associated with significant improvement of the OS.