Cluster-based prognostication in glioblastoma: Unveiling heterogeneity based on diffusion and perfusion similarities

Abstract

Background: While the association between diffusion and perfusion magnetic resonance imaging (MRI) and survival in glioblastoma is established, prognostic models for patients are lacking. This study employed clustering of functional imaging to identify distinct functional phenotypes in untreated glioblastomas, assessing their prognostic significance for overall survival.

Methods: A total of 289 patients with glioblastoma who underwent preoperative multimodal MR imaging were included. Mean values of apparent diffusion coefficient normalized relative cerebral blood volume and relative cerebral blood flow were calculated for different tumor compartments and the entire tumor. Distinct imaging patterns were identified using partition around medoids (PAM) clustering on the training dataset, and their ability to predict overall survival was assessed. Additionally, tree-based machine-learning models were trained to ascertain the significance of features pertaining to cluster membership.

Results: Using the training dataset (231/289) we identified 2 stable imaging phenotypes through PAM clustering with significantly different overall survival (OS). Validation in an independent test set revealed a high-risk group with a median OS of 10.2 months and a low-risk group with a median OS of 26.6 months (P = 0.012). Patients in the low-risk cluster had high diffusion and low perfusion values throughout, while the high-risk cluster displayed the reverse pattern. Including cluster membership in all multivariate Cox regression analyses improved performance (P ≤ 0.004 each).

Conclusions: Our research demonstrates that data-driven clustering can identify clinically relevant, distinct imaging phenotypes, highlighting the potential role of diffusion, and perfusion MRI in predicting survival rates of glioblastoma patients.

Keywords: diffusion; glioblastoma; machine learning; perfusion; prognostic biomarker.

Figure 1.

Graphical representation of the study structure. Only the training set was used to calculate cutoffs, for clustering and model training. The test dataset was used as a holdout dataset to provide an unbiased evaluation of the final model.

Figure 2.

Percentages of high imaging parameters within the 2 clusters in the different tumor compartments. Cluster 1 shows a significantly higher percentage of patients with high diffusion values in all compartments, whereas cluster 2 shows higher perfusion values in all compartments (chi-tests, P-value after Bonferroni correction < 0.001).

Figure 3.

Kaplan–Meier plots of OS in the training data set (A) and test data set (B), stratified to low or high-risk groups according to clustering. Significance was calculated using the log-rank test.