Article Text
Abstract
Introduction/Background To decrease immunosuppression and enhance T-cell activation in the tumour microenvironment, we conducted an open-label, investigator-initiated, multicohort, phase II trial (NCT03192059) of pembrolizumab with multimodal immunomodulation.
Methodology Chemotherapy-pretreated patients were recruited into two experimental cohorts (cervical carcinoma or endometrial carcinoma) and one exploratory cohort (uterine sarcoma). Patients received an immunomodulatory five-drug cocktail consisting of low-dose cyclophosphamide, aspirin, lansoprazole, vitamin D, and curcumin starting two weeks before radioimmunotherapy (figure 1). Pembrolizumab, 200 mg/dose, was administered on day 1 of each 21-day cycle from day 15 onwards; one of the tumour lesions was irradiated (8Gyx3) on days 15, 17, and 19. The primary endpoint was objective response rate (irORR) per immune-related response criteria (irRC) at week 26; a lower bound of its 90% confidence interval (CI) of >10% in either experimental cohort was considered successful.
Results Fifty patients were enrolled and treated across the cohorts (cervical, n=18; endometrial, n=25; sarcoma=7). Pathology review revealed that 3/7 sarcoma patients had carcinosarcoma. At week 26, the irORR was 11.1% (90%CI, 2.0 to 31.0) in cervical cancer, 12.0% (90%CI, 3.4 to 28.2) in endometrial cancer, and 14.3% (90% CI, 0.7 to 52.1) in uterine (carcino)sarcoma. The best overall response rate per RECIST v1.1 was 22.2% (90%CI, 8.0 to 43.9), 12.0% (90%CI, 3.4 to 28.2), and 28.6 (90%CI, 5.3 to 65.9). Median PFS was 13.4 weeks (11.3 to 26.1), 13.1 weeks (13.1 to 19.4), and 34.3 weeks (95%CI, 5.6 to 77.9) (figure 2A-C). Grade≥3 treatment-related adverse events were reported in 10 (55.6%), 9 (36.0%), and 4 (57.1%) patients. Overall, there was one (2.0%) possible treatment-related death. Health-related quality of life was generally stable over time. Multi-parameter immune monitoring characterised the patients and revealed changes throughout study treatment.
Combined waterfall and swimmer plot of the (A) cervical cohort, (B) endometrial cohort, and © (carcino) sarcoma cohort. Each row (i.e., response bar + characteristics + swimmer lane) corresponds to one patient. Waterfall plot showing best percentage change from baseline in the sum of diameters of the target lesions best overall response is indicated by color coding of bars and includes assessment of target, nontarget, and new lesions. The dotted lines at -30% abd +20% indicate thresholds for partial Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST v1.1). Swimmer plot (event chart) for tumor response (response category indicated by color coding), progressive disease per RECIST v1.1, safety, time on study, and death. The solid lines at week 3 and week 26 indicate the first pembrolizumab dose and timing of the primary endpoint, respectively
Conclusion PRIMMO did not show sufficient evidence of a positive risk-to-benefit ratio to recommend a confirmatory phase III trial.