1023MO - Decoding immunotherapy resistance: A novel multiparametric magnetic resonance imaging approach to stratify cancer patient outcomes

Background

This prospective study evaluated the use of multiparametric MRI (mpMRI) in non-invasively assessing the tumor microenvironment and predicting response to immune checkpoint inhibitors (ICIs) in patients with advanced solid tumors, as biomarkers are needed to improve patient selection for ICIs.

Methods

146 lesions from 40 patients were analyzed at baseline using joint diffusion-T2-relaxation imaging (DRI) and dynamic contrast-enhanced (DCE) imaging to obtain 11 parameters describing tumor microstructure. In addition, 23 baseline biopsies were evaluated using a next-generation immunohistochemistry panel, including PDL1, CD163, CD3, CD8, FOXP3, CD31 and Ki67. The best overall response rate (ORR) was evaluated as per RECIST 1.1. Dichotomized outcome was defined as progression (PD) vs no progression (SD/PR) at best response. LASSO was used to select mpMRI variables, and univariate and multivariate logistic regression models were fitted to predict the outcome.

Results

The ORR was 10% (4 out of 40 patients). Higher density of CD8 cytotoxic cells was independently associated with lower progression rate (odds ratio (OR) 0.31, 95% confidence interval (CI) 0.08–0.77). In multivariate analysis, higher tissue kurtosis representing complex tissue microstructure (OR 5.13, 95%CI 1.33–30.5), low T2 of the tissue indicating lower water content (OR 0.14, 95%CI 0.03–0.51) and higher volume fraction of the extracellular space from DCE (OR 3.32, 95%CI 1.17–12.3) were associated with progression. The multivariate mpMRI model showed promising results with an AUC [95%CI] of 0.85 [0.73, 0.97], which was evaluated using calibration and bootstrapping (AUC of 0.80 [0.69, 0.92]). Statistically significant differences in PFS between High vs Low mpMRI signature score groups were observed (log rank test: p=0.005).

Conclusions

mpMRI provides valuable information about advanced solid tumor microstructure and may serve as a useful tool for patient selection for treatment with ICIs. Biomarker discovery in phase 1 trials is challenging, but mpMRI could identify a subset of progressing patients resistant to ICIs. Further studies are warranted to validate these findings.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Elena Garralda, Raquel Perez Lopez, Paolo Nuciforo, Rodrigo Toledo.

Funding

AstraZeneca.

Disclosure

All authors have declared no conflicts of interest.