MP55-04 NO-RESPONSE AT MPMRI AFTER NEOADJUVANT PEMBROLIZUMAB IS A PROXY OF ADVERSE PATHOLOGICAL AND ONCOLOGICAL OUTCOMES IN PATIENTS TREATED WITH RADICAL CYSTECTOMY

Abstract

INTRODUCTION AND OBJECTIVE: The PURE01 study is testing neoadjuvant pembrolizumab in MIBC patients before radical cystectomy (RC). Despite final results are pending, preliminary data have shown promising response rate, with 42% of patients harboring pT0N0 at RC. Nevertheless, a non-negligible proportion (46%) of treated patients still exhibit persistent or progressive disease (pT2-4 or N+) at RC. The identification of no-response (NR) at the post-pembrolizumab mpMRI may represent a valid proxy for predicting adverse pathological and oncological outcomes METHODS: The PURE01 study is recruiting cT2-4N0 MIBC patients eligible for 3 cycles of neoadjuvant pembrolizumab at subsequent RC. We included only patients with available mpMRI before and after neoadjuvant immunotherapy (IO). NR at mpMRI was established by comparing pre and post IO mpMRI and was defined as the presence of any kind of suspicious lesion, based on morphologic, diffusion-weighted imaging, and contrast enhancement evaluations. Multivariable regression analyses tested the association between NR at mpMRI and presence of pT2-4, as well as lymph-node invasion (LNI) at RC, after adjusting for cT stage and variant histology. Kaplan-Meier and Cox regression analyses tested the effect of NR at mpMRI on clinical recurrence after RC, after accounting for neoadjuvant rescue chemotherapy [CHT], adjuvant CHT, and presence of carcinoma in situ RESULTS: Of all enrolled individuals, 82 met the inclusion criteria. No statistically significant differences emerged between response and NR at mpMRI in terms of age, gender, presence of variant histology or time from TURB to pembrolizumab (all p>0.2). NR at mpMRI after pembrolizumab was identified in 23 (28.1%) patients. NR at mpMRI reached independent prediction status for either pT2-4 disease (OR: 20.7, p=0.0001) and LNI (OR: 7.0, p=0.003) at RC. Finally, NR at mpMRI was associated with higher rate of recurrence at 1 year after treatment (21 vs. 3.4%, p=0.002), with an adjusted hazard ratio of 9.5 (95% CI: 1.7-52, p= 0.009) CONCLUSIONS: Results from out interim analysis are of utmost importance because they set a red flag for patient with NR at mpMRI after neoadjuvant IO. Herein, we found that the presence of any MRI sign of residual disease is associated with a remarkable risk of adverse pathological and unfavorable oncological outcomes. Alternative options in these patients, such as rescue CHT or adjuvant radiation therapy should be further explored in the upcoming trials Source of Funding: none