E2126. Value of Pre-treatment Imaging and Early Restaging in Assessment of Impact of Immunotherapy (IT) in Advanced Stage Malignancy
  1. Milos Janicek; Boston Medical Center
Identifying the group of patients with advanced stage malignancy benefiting from IT is very challenging in setting of multiple prior treatments, interventions as well as toxicity. In our study, we explored COMBINED value of pre-treatment imaging and two early restaging scans as a predictor of length of stay on treatment as a surrogate measure of IT positive impact.

Materials and Methods:
Cohort of 50 patients with advanced solid tumors, majority of aerodigestive system, treated on standard protocol with Pembro at BMC in 2015-2018. Last pre-IT scan, baseline, and two early restaging scans were analyzed utilizing modified imRECIST score derived from S of 4 discrete lesions. Slope of progression curve was calculated as % score change per time interval. Progression (P) of disease defined as >10% change between scans, regression as >10% decrement of score, remaining deemed Stable (S). The length of stay on treatment was defined by clinical evaluation of IT effect and toxicity.

Patients (pts) starting IT with progressive disease (dz) showed only trend of longer stay on IT compared to pts with stable or regressing dz, median = 218 vs 157 days, P=0.189, T-test. Pts with Progression (P) on two restaging scans showed significantly shorter stay on IT compared to Pseudo-progression, mean = 125 vs 249 days, P=0.014, T-test. The most favorable group started therapy as stable dz and showed regression or stable (S) dz on 2 restaging scans, mean = 362 days, P=0.0001, T- test.

Robust imRECIST scoring of early restaging scans can predict length of effective treatment. Pseudo-progression was seen in 20% of our pts on Pembro with similar outcome to most favorable group. Each drug needs to be evaluated individually to set benchmark criteria. Our novel approach including scoring of pre-IT scans suggests only trend of more prolonged IT effect in setting of progressive disease before initiation of the treatment. The most valuable predictor of TXI is the 2nd restaging scan. Our study offers a model of validation of PRE-Tx scan IN COMBINATION with early restaging scans predicting favorable impact of IT. Identifying group of pts who may not benefit from IT could avoid unnecessary toxicity and allow timely selection of alternative therapies.