PURE-01 is a phase 2 study of Pembro before RC in MIBC: the primary endpoint (pathologic complete response [pT0]) was met in advance (Necchi et al. JCO 2018). While the study is still recruiting patients (pts), we analyzed the interim long-term outcomes in terms of relapse-free survival (RFS) and the role of multiparametric bladder magnetic resonance imaging (mpMRI) in predicting major pathologic response (pT<2) in RC.
Pts with clinical stage T2-3bN0M0 received 3 courses of 200mg Pembro, before RC. Pts were staged and re-assessed with mpMRI before RC. 1-y RFS (overall and by pathologic stage) after Pembro and RC was compared with that from clinical stage-matched cohort of San Raffaele Urological Institute (URI, n=1021) and of the RISC multicenter database (n=719, Bandini et al, Eur Urol Oncol 2018). mpMRI were performed in all pts through bladder catheterization. The apparent diffusion coefficient (ADC) changes post-Pembro were matched with pathologic response, qPCR and genomic data from RC samples (FoundationONE).
70 pts have completed treatment thus far: with 27 pT0 (38.6%) and 44 pT<2 (62.9%), activity was confirmed. 21 pts had ≥1y follow-up and were analyzed. Overall, 1-y RFS (95%CI) in PURE-01, URI and RISC were: 95.2 (86.6-100), 79.0 (76-82), and 69.7 (66.2-73.4). 1-y RFS after neoadjuvant chemotherapy and RC in URI and RISC cohorts were: 76.7 (61-97), and 76.5 (71-83), respectively. 1-y RFS for pT3-4 and/or pN+ pts were: 88.9 (71-100), 71.2 (67-76), and 57.1 (52-62), respectively. 39 pts had measurable disease for pre-post mpMRI changes. Pts with pT<2 (n=16) on RC showed a median “mean ADC” of 0.98 mm2/s vs 0.84 mm2/s of non-responders (NR, n=23, p=0.098), despite an increased T-cell infiltrate. Pts with cell-cycle gene alterations in RC (n=8) were all NR and showed lower median “mean ADC” values (p=0.074).
The early efficacy data further support the role of single-agent IO as neoadjuvant therapy in MIBC. Pembro may be beneficial at long-term also in pathologically high-risk pts, thus questioning the value of PD-L1 for pt selection, as well as the role of pathologic response as surrogate of survival outcome post-checkpoint inhibition. mpMRI and genomic data might help identifying those pts who may need re-TURB only instead of RC.