Differential response rates in early-phase cancer clinical trials (EPCCT)

Abstract

Background: The primary objective of EPCCT (phase I and non-randomised phase II trials) is to determine the safety and tolerability of new therapeutic agents. Response rates (RR) in these trials have typically been reported at around 10-15%. Increasingly RR and survival outcomes are now investigated in EPCCT as primary or secondary objectives. Methods: Retrospective data analysis was performed on patients (pts) enrolled onto an EPCCT between January 2018 and December 2019 at The Christie NHS Foundation Trust, UK. Data on demographics, prior systemic treatment, sites of disease, performance status, comorbidities, types of therapy, RR, progression free survival (PFS), and overall survival (OS) were collected. Statistical analyses were performed with univariable and multivariable models. Objective response rate (ORR) was defined as the proportion of pts with complete response (CR) and partial response (PR). Duration of response (DOR) was from initial response to progressive disease (PD). Disease control rate (DCR) was defined as CR+PR+ stable disease (SD). Results: A total of 247 pts were treated across 46 EPCCTs. Median age 61 years; 57% female. Sixty-six percent of pts had ≥2 lines of treatment and the majority were ECOG PS 0/1 (98%). Eighty-one percent of pts had ≥2 sites of metastatic disease, and 13 major tumour types were included. Monotherapy trials (159 pts) were predominantly targeted therapies (TT; 60%), or immunotherapies (IO; 20%). Combination therapy trials (88 pts) were TT-based (68%) or IO-based (32%). Data for RR analyses was available for 231 pts. ORR across all trials was 15% (CR 2%) and DCR was 63%. The median DOR was 8.3 months (mos) (95% CI: 7.0 – 9.7) with 28% of pts responding for >6 mos and 7% for >12 mos. ORR in pooled IO treated pts was 27%, DCR was 65% with sustained response >6 mos seen in 37% of these pts. ORR in pooled TT treated pts was 9.4%, DCR was 60% and sustained response > 6 mos seen in 25% of pts. ORR for IO v TT treated pts was significantly different, p=0.007 (pearson chi square), but no significant difference was seen for DCR. Median PFS for all patients was 5.0 mos (95% CI: 4.1 – 6.0) and OS was 10.4 mos (95% CI: 8.4 – 13.0). OS for those with a PR is not reached (HR for PR v PD, 0.006 (95% CI: 0.002 – 0.18). Pts with SD appear to have significantly better OS compared to those with PD (14.6 v 4.2 mos, HR 0.2 (95% CI: 0.1 – 0.3). Multivariable Cox proportional hazards analysis for OS was significant for male gender (HR 1.9, p=0.002), presence of liver metastasis (HR 2.0, p=0.001), low Hb (HR 0.8, p=0.03) and log (LDH) (HR 1.9, p<0.001). Conclusions: Two-thirds of pts enrolled on EPCCTs benefitted in terms of DCR with significant OS improvement in those with PR and SD. Higher ORR were seen in pts receiving IO-based treatments however DCR was similar in IO and TT pts. Gender, presence of liver metastases, Hb count and LDH level contributed significantly to survival differences.