Objectives: Pulmonary arterial hypertension is a life-threatening complication of systemic lupus erythematosus. However, there is no algorithm to identify those at high risk. We aimed to develop a prediction model for pulmonary arterial hypertension in lupus patients that provides individualized risk estimates.
Methods: A multicenter, longitudinal cohort study was undertaken from January 2003 to January 2020. The study collected data on 3,624 consecutively evaluated patients diagnosed with lupus. The diagnosis of pulmonary arterial hypertension was confirmed by right heart catheterization. Cox proportional hazards regression and least absolute shrinkage and selection operator were used to fit the model. Model discrimination, calibration, and decision curve analysis were assessed for validation.
Results: Ninety-two lupus patients developed pulmonary arterial hypertension (2.54%) at a median follow-up of 4.84 years (interquartile range, 2.42-8.84). The final prediction model included five clinical variables (acute/subacute cutaneous lupus, arthritis, renal disorder, thrombocytopenia, and interstitial lung disease) and three autoantibodies (anti-RNP, anti-Ro/SSA and anti-La/SSB). A 10-year pulmonary arterial hypertension probability-predictive nomogram was established. The model was internally validated by C statistic (0.78), the Brier score (0.03), and a satisfactory calibration curve. According to the net benefit and predicted probability thresholds, we recommend annual screening in high-risk (> 4.62 %) lupus patients.
Conclusion: We developed a risk stratification model using routine clinical assessments. This new tool may effectively predict the future risk of pulmonary arterial hypertension in patients with systemic lupus erythematosus.
Keywords: pulmonary arterial hypertension; risk prediction model; systemic erythematosus lupus.