Abstract 592P
Background
Antitumor activity is an important goal of Ph1 trials of INDs. We reviewed our series to find associations of RECIST response to baseline clinical variables and to define a predictive score.
Methods
We analyzed our pt treated in Ph1 trials, except ClinPharm, from Nov08 to Dec16. 28 baseline clinical (age, sex, ECOG, BMI, hypothyroid, HBP, DM, clots, #prior lines, liver mets, #sites) and analytical (Hb, PLT, ANC, lymphs, alb, AlkPhos, ALT, AST, WBC, TBil, Creat, GGT, LDH, ANC/lymph, PLT/lymph, Ca, Mg) values were collected; also objective response (OR), toxicity and therapy type. The variables associated to OR in the univariate analysis (UA, p<0.05, chi-square test) were included in the step-wise logistic regression multivariate analysis (MA). The ones found to keep statistically significant were included in a predictive score of antitumor activity. Kaplan-Meier survival curves were compared with log-rank test. p values are 2-sided.
Results
773 consecutive pt treated in 85 ph1 trials in START Madrid-CIOCC were included. Mean age was 58.7 y (range: 18-87); 48,1% were male. ECOG 0, 55.8%; 1, 42.2%. 29% pt had GI tumours, 20% lung, 14% breast, and 12% GYN. 131 pt (17.3%) received IO, 303 (39.9%) targeted and 325 (42.8%) chemo drugs, and no statistically significant differences in OR rate or mOS were seen. 103 of 730 evaluable pt achieved an OR (14.1%), and 271 pt (48.8%) stable disease; their mOS was 10.8 m (95%CI: 9.8-10.8). The UA of the 28 variables showed that BMI>25 kg/m2 (p=.018), <3prior lines (p=.000), normal AlkPhos (p=.000), AST (p=.033), GGT (.001), and LDH (.025), and ANC/lymph <5 (p=.047) were predictive of OR. Of these, BMI>25kg/m2 (p=0,021), <3 prior lines (p=0.002) and normal AlkPhos(p=0.012) were found to be independent factors associated to OR in the MA. The presence of 0 to 3 altered factors was associated to worsening OS (X2, 18.1; p=.000). A score of 0-1 vs. 2-3 unfavorable factors was predictive of OR (17% vs. 7.6%; X2, 11.9, p=.003) and OS (11.6 m vs. 8.6 m; log rank, 7.8, p=0.005).
Conclusions
A high number of pt might derive clinical benefit (OR+SD, 51.2%) from ph1 trials. A practical score based on BMI, #prior lines and AlkPhos (0-1 factors) was predictive of antitumor activity (OR, OS). If prospectively validated, it will help in pt selection.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.