Predicting cardiac adverse events in patients receiving immune checkpoint inhibitors: a machine learning approach

Abstract

Background: Treatment with immune checkpoint inhibitors (ICIs) has been associated with an increased rate of cardiac events. There are limited data on the risk factors that predict cardiac events in patients treated with ICIs. Therefore, we created a machine learning (ML) model to predict cardiac events in this at-risk population.

Methods: We leveraged the CancerLinQ database curated by the American Society of Clinical Oncology and applied an XGBoosted decision tree to predict cardiac events in patients taking programmed death receptor-1 (PD-1) or programmed death ligand-1 (PD-L1) therapy. All curated data from patients with non-small cell lung cancer, melanoma, and renal cell carcinoma, and who were prescribed PD-1/PD-L1 therapy between 2013 and 2019, were used for training, feature interpretation, and model performance evaluation. A total of 356 potential risk factors were included in the model, including elements of patient medical history, social history, vital signs, common laboratory tests, oncological history, medication history and PD-1/PD-L1-specific factors like PD-L1 tumor expression.

Results: Our study population consisted of 4960 patients treated with PD-1/PD-L1 therapy, of whom 418 had a cardiac event. The following were key predictors of cardiac events: increased age, corticosteroids, laboratory abnormalities and medications suggestive of a history of heart disease, the extremes of weight, a lower baseline or on-treatment percentage of lymphocytes, and a higher percentage of neutrophils. The final model predicted cardiac events with an area under the curve-receiver operating characteristic of 0.65 (95% CI 0.58 to 0.75). Using our model, we divided patients into low-risk and high-risk subgroups. At 100 days, the cumulative incidence of cardiac events was 3.3% in the low-risk group and 6.1% in the high-risk group (p<0.001).

Conclusions: ML can be used to predict cardiac events in patients taking PD-1/PD-L1 therapy. Cardiac risk was driven by immunological factors (eg, percentage of lymphocytes), oncological factors (eg, low weight), and a cardiac history.

Keywords: immunotherapy; lung neoplasms; programmed cell death 1 receptor.

Figure 1

SHAP summary plot for interpreting the impact of features on our model. Each row shows the impact of a single feature on the model’s predictions. Within each row, each dot represents a patient. Red means patients had a high feature value; blue means patients had a low value; gray means patients had a missing value. The position of the dot along the x-axis indicates whether that feature increased or decreased a patient’s predicted risk. When all the red dots are on the right, a high feature value was associated with increased risk. When all the blue dots are on the right, a low feature value increased risk. Statistical significance is indicated as follows: *p<0.05, **p<0.01, ***p<0.002. BMI, body mass index; Cr, Creatinine; DBP, Diastolic Blood Pressure; SBP, Systolic Blood Pressure; Hb, hemoglobin; SHAP, Shaply additive explanations.

Figure 2

Plot of the cumulative dynamic AUC-ROC of our model on PD-1/PD-L1 patients. Model’s ability to predict cardiac adverse events within 20, 40, 60, 80, 100, 120, and 140 days of index is shown. The model’s performance varied between 63% and 72% as the time window for predictions changed. AUC-ROC, area under the curve–receiver operating characteristic; PD-1, programmed death receptor-1; PD-L1, programmed death ligand-1.

Figure 3

Cumulative incidence of cardiac events in low-risk and high-risk PD-1/PD-L1 patients from the test set. Groups were stratified by our model’s median predicted HR. The cumulative incidence function was calculated using the method described by Aalen and Johansen, taking into account the competing risk of death. High-risk patients had a significantly higher incidence of cardiac events. PD-1, programmed death receptor-1; PD-L1, programmed death ligand-1.