作者
Niven Mehra,Karim Fizazi,A. Douglas Laird,Philippe Barthélémy,R. Delva,Tanya B. Dorff,Marco Maruzzo,A. Stirling,J-P. Machiels,Herlinde Dumez,Vincent Renard,Lee A. Albacker,Julia F. Hopkins,Hsiang‐Cheng Chen,Marielena Mata,Nicola Di Santo,Cynthia G. Healy,Inge M. van Oort,Consuelo Buttigliero,Johann S. de Bono
摘要
Phase 2 TALAPRO-1 enrolled patients (pts) with RECIST1.1 measurable soft tissue disease, progressive CRPC, and DDRalt likely to sensitize to PARP inhibitors (PARPi) (ie, ATM, ATR, BRCA1, BRCA2, CHEK2, FANCA, MLH1, MRE11A, NBN, PALB2, RAD51C). Pts received ≥1 taxane-based chemotherapy and progressed on ≥1 novel hormonal therapy. Primary endpoint is objective response rate (ORR; central review). Based on a prior interim analysis (IA) of 43 evaluable pts enrolled by 02/12/19, TALA showed encouraging antitumor activity and was well tolerated. This interim biomarker analysis was to assess germline/somatic origin of tumor DDRalt (tDDRalt) and associations with efficacy. Tumor tissue was tested using FoundationOne®. Saliva samples were sequenced using Ambry CustomNext-Cancer panel, which included 9/11 genes in the tDDR panel. Known/likely pathogenic tDDR variants (tDDRalt) used to support molecular eligibility for study were categorized as of germline (also present in saliva), somatic (not present in saliva), or unknown origin. Categorization was limited to short variants. This IA contained investigator-entered data ("as-is", cutoff 12/12/19). Of 37 evaluable pts, with results from both tumor and investigational germline analysis, 20 (54%) had ≥1 tDDRalt of germline origin, 16 (43%) exhibited only tDDRalt of somatic origin, and 1 (3%) was nonassessable for origin (FANCA not in germline panel). The most common tDDRalt in the 37 pts were BRCA2 (11 germline; 7 somatic) and ATM (3 germline; 6 somatic). Observed ORR was numerically higher for pts with tDDRalt of germline origin (7/20 [35%]) than with only somatic DDRalt (1/16 [6.3%]), although this difference was not significant (odds ratio [95% CI] 8.1 [0.8–387.7]; P = 0.053, 2-sided Fisher's exact test). Based on this IA, prostate cancers bearing germline DDRalt may be more sensitive to TALA than those bearing only somatic DDRalt. Potential reasons include possible gene-specific imbalances in origin of alterations and/or sensitivity to PARPi, and lower tumor subclonality of germline DDRalt.