Background: Clues to immunotherapy response

A study analyzing pretreatment tumors from 91 patients with pleural mesothelioma found an association between DNA methylation subtypes and response to immune checkpoint inhibitors, as well as survival. Responders were relatively more common in the LOW subtype, while nonresponders were relatively more common in the CIMP subtype. Tumor methylation patterns were proposed as a potential clue for predicting treatment outcomes, but whether choosing treatment according to subtype improves survival remains to be tested separately.

The researchers began with two challenges: pleural mesothelioma has a poor prognosis, and clinical outcomes with immune checkpoint inhibitors targeting CTLA-4 and PD-1 remain insufficient. According to the paper’s abstract, no biomarker was then available to predict the efficacy of immune checkpoint inhibitors in pleural mesothelioma. The study asked whether tumor information collected before treatment could distinguish subsequent differences in response and survival.

Key findings: Divergent outcomes in LOW and CIMP

NIBIT-EPI-MESO is a retrospective, multicenter study. The researchers used multiomics to analyze pretreatment tumors from 91 patients with pleural mesothelioma who received immune checkpoint inhibitors in prior clinical trials or routine care.

The analysis identified four subtypes in order of progressively higher overall DNA methylation: demethylated, low methylation (LOW), intermediate, and CpG island methylator phenotype (CIMP). This classification was associated with response to immune checkpoint inhibitors and survival.

Responders were relatively more common in the LOW subtype, which had the longest median overall survival and the highest 3-year overall survival rate. Another feature observed in this subtype was an immune microenvironment rich in T cells and B cells. By contrast, nonresponders were relatively more common in the CIMP subtype, which had the least favorable survival outcomes, including median overall survival, and an immune-poor microenvironment. However, the jointly observed immune-cell composition and treatment outcomes alone do not establish a causal relationship between them.

The supplied abstract does not report the actual survival times or rates for each subtype. The available information therefore cannot quantify how much longer survival was in the LOW subtype or the size of the differences between subtypes. Nor should the results be interpreted to mean that every patient in a particular subtype responded or that none did.

Meaning and outlook: From prediction tools to clinical use

The researchers also developed a probabilistic classification tool to predict outcomes of immune checkpoint inhibitor treatment from methylation data. This is an attempt to connect pretreatment tumor characteristics with patient-specific outcome predictions.

The supplied abstract does not establish the tool’s predictive accuracy, whether it was tested in an independent patient cohort, or whether it underwent prospective validation. The distribution of patients across treatment regimens, follow-up duration, sample sizes by subtype, and results of adjustment for confounding factors also cannot be confirmed. Associations observed among patients who received immune checkpoint inhibitors are insufficient to determine whether a particular subtype benefits more from these drugs than from other treatments.

These findings provide a basis for further validation of methylation subtypes. Before they can guide treatment choices, researchers must assess whether predictive performance holds in other patient cohorts and whether decisions informed by subtype benefit patients. The study design and main findings can be found in the supplied abstract indexed in Europe PMC.