Comparison between REVEAL Lite 2.0 and COMPERA 2.0 for risk stratification in Pulmonary Arterial Hypertension

Background Risk stratification is the cornerstone of the management of pulmonary arterial hypertension (PAH). Current ESC/ERS guidelines recommend using the COMPERA 3-strata risk model at baseline, and the COMPERA 2.0 4-strata model at follow-up. However, the guidelines did not take into consideration other available risk scores such as REVEAL Lite 2. Research Question Is REVEAL Lite 2 better at discriminating risk than the COMPERA risk assessment models at baseline or follow up evaluations? Study Design and Methods We analyzed data from the consecutively enrolled PAH patients between the period of June 2011 to February 2022 in the PAH registry at our expert Pulmonary Hypertension Center. Patients were stratified according to REVEAL Lite 2, COMPERA 3-, and 4-strata risk scores at baseline and follow up to predict the composite outcome for lung transplantation or death. ROC curves in predicting the binary outcome at 3, 5 and 7 years were plotted. Areas under the curve (AUC) of the scores were compared by the chi-square test. The performance of the scores was determined by Harrel’s C-statistic. Results 296 patients were included for baseline and 196 for follow up evaluation. The overall transplant-free survival in our patient population at 1, 3, 5, and 7 years was 93.6%, 81.3%, 75.1%, and 68.8% respectively. At baseline, the C-statistic of REVEAL Lite 2 was 0.74 (0.69, 0.80), compared to 0.68 (0.63, 0.74) for the COMPERA 4-strata model, and 0.63 (0.58, 0.69) for the COMPERA 3-strata model. All C-statistic differences between REVEAL Lite 2 and the other models were statistically significant at baseline. Interpretation Our analysis showed that REVEAL Lite 2 was better at baseline at discriminating risk in our patient population. Future guidelines should consider including REVEAL Lite 2 in the management algorithm to help clinicians to make informed decisions. Further analysis in larger cohorts could help validate these findings.