Simpson, C.

Abstract Rationale: The 7th World Symposium on Pulmonary Hypertension highlighted the importance of patient-reported outcomes in clinical assessment of pulmonary arterial hypertension (PAH). Patient-reported outcomes, such as health-related quality of life (HRQOL), have been associated with morbidity and mortality in PAH, but the added value of HRQOL measures to validated risk prediction models remains unknown. Methods: Adults with RHC-proven PAH from the Redefining Pulmonary Hypertension through Pulmonary Vascular Disease Phenomics (PVDOMICS) Study, a multi-center observational study, were included in this study. The Pulmonary Hypertension Association Registry (PHAR) was used as a validation cohort. HRQOL was measured using the emPHasis-10 score, where higher scores reflect worse HRQOL. Noninvasive risk assessment was measured by REVEAL Lite 2 risk score, categorized into 3-strata. Cox proportional hazard models were used to model survival evaluating the interaction between risk strata and HRQOL scores, adjusting for age, sex, and incident diagnosis. Harrell's C-statistics were calculated to compare discriminative performance. Results: From the PVDOMICS and PHAR cohorts, 265 and 2671 individuals with PAH and complete emPHasis-10 and REVEAL Lite 2 scores were included. Both cohorts had similar demographics, with an average age of 52±13 and 56±16 years, 75% and 72% female, 81% and 75% White, in the PVDOMICS and PHAR cohorts respectively. Over 87% of PVDOMICS were prevalent cases of PAH, while in the PHAR half were incident. Most individuals (62%) in PVDOMICS were REVEAL Lite 2 low risk, with 16% intermediate and 22% high risk; by contrast, in PHAR 33% were low, 29% intermediate, and 38% high risk. REVEAL Lite 2 scores demonstrated weak to moderate correlation with emPHasis-10 scores in both cohorts (PVDOMICS r=0.48 p<0.001; PHAR r=0.33 p<0.001). Notably, the full range of emPHasis-10 scores were observed in the low-risk individuals across both cohorts. In the PVDOMICS adjusted Cox models, higher emPHasis-10 scores predicted poorer survival (p=0.047) particularly among low-risk individuals (interaction versus high-risk group p=0.08). Similar adjusted models in PHAR demonstrated emPHasis-10 to be a confounder (p=0.001) without significant interaction by risk strata (low vs high-risk interaction p=0.94), but the greatest change in median survival time was observed in the low-risk group (Figure 1). Addition of emPHasis-10 to REVEAL Lite 2 scores improved discrimination in predicting survival (PVDOMICS: C-statistic 0.71 vs 0.73; PHAR: C-statistic 0.67 vs 0.69). Conclusions: Across two multicenter cohorts of incident and prevalent PAH, HRQOL as measured by the emPHasis-10 score improved discrimination for survival, particularly among individuals with lower risk scores.