Time to Treatment Prediction in Chronic Lymphocytic Leukemia Based on New Transcriptional Patterns

Abstract

Chronic lymphocytic leukemia (CLL) is the most frequent lymphoproliferative syndrome in western countries. CLL evolution is frequently indolent, and treatment is mostly reserved for those patients with signs or symptoms of disease progression. In this work, we used RNA sequencing data from the International Cancer Genome Consortium CLL cohort to determine new gene expression patterns that correlate with clinical evolution.We determined that a 290-gene expression signature, in addition to immunoglobulin heavy chain variable region (IGHV) mutation status, stratifies patients into four groups with notably different time to first treatment. This finding was confirmed in an independent cohort. Similarly, we present a machine learning algorithm that predicts the need for treatment within the first 5 years following diagnosis using expression data from 2,198 genes. This predictor achieved 90% precision and 89% accuracy when classifying independent CLL cases. Our findings indicate that CLL progression risk largely correlates with particular transcriptomic patterns and paves the way for the identification of high-risk patients who might benefit from prompt therapy following diagnosis.

Keywords: IGHV; RNAseq; chronic lymphocytic leukemia; gene expression; machine learning; prognostic factors; time to treatment prediction.

Figure 1

Heatmap showing the rank-transformed distribution of expression values for the 290 genes in the study cohort. Red-labeled samples on the left bar petain to C1 and blue-labeled samples pertain to C2.

Figure 2

Kaplan-Meier survival plots. The upper plots show the association of C1 (red curve) and C2 (blue curve) with TTT in the study (left) and validation cohorts (right). Corresponding p-values are 1.7 × 10⁻⁶ and 1.3 × 10⁻⁴. The lower plots show the association with TTT stratified by IGHV mutation status in the study (left) and validation cohorts (right). The blue line indicates C2 samples with mutated IGHV, the purple line indicates C2 samples with unmutated IGHV, the red line indicates C1 samples with mutated IGHV, and the green line refers to C1 samples with unmutated IGHV.

Figure 3

Heatmap showing the rank-transformed distribution of expression values for the 290 genes in the validation cohort. Red-labeled samples on the left bar petain to C1 and blue-labeled samples pertain to C2.

Figure 4

ROC curve of the boosted-tree Ensembl model for 5 year treatment need prediction (upper left). Kolmogorov-Smirnov plot for the same model (upper right). Precision-Recall plot for the 5 year not-treated (lower left) and treated (lower right) patients according to the same model. White dots in each graph indicate the probability threshold (in this case 50%), which is the point reflecting the best classification accuracy of the patients.