HALP Score for Predicting Renal Relapse in Lupus Nephritis: A Nested Case-Control Study

To evaluate the HALP (hemoglobin-albumin-lymphocyte-platelet) score at lupus nephritis (LN) diagnosis in predicting renal relapse (RR).

Nested case-control study, including patients aged ≥18 years with diagnosis of LN between 2010 and 2022. Two patient sets were included: training and validation. Each set had 2 groups of patients: RR and non-RR. Data were obtained from clinical records. The optimal cutoff value of the HALP score at LN diagnosis was established to predict RR. Cox regression analysis was used to associate HALP score at diagnosis with RR.

We included 53 LN patients in the training set and 74 LN patients in the validation set. The optimal cutoff value for HALP score at diagnosis was 23.5, with an area under the curve of 0.896, sensitivity of 91.9%, and specificity of 97.3% in the validation set. The median age of patients in this set was 31.0 years, mostly female (93%). In the validation set, LN patients with HALP score at diagnosis ≤23.5 compared with higher HALP score subjects showed a significantly higher baseline SLEDAI-2K (18 [interquartile range {IQR}, 14-20] vs. 14 [IQR, 11-17], p  < 0.001), Systemic Lupus Collaborating Clinics/American College of Rheumatology Damage Index at the end of the follow-up (1 [IQR, 0-4] vs. 0 [IQR, 0-1], p  = 0.002), chronicity index in renal biopsy (2 [IQR, 1-4] vs. 1 [IQR, 1-2], p  = 0.030), and significantly reduced time to RR (4.2 vs. 12.9 years, p  < 0.001). A HALP score at diagnosis ≤23.5 was associated with RR (hazard ratio, 18.2; 95% confidence interval, 5.3-30.1; p  < 0.001).

A HALP score ≤23.5 at LN diagnosis was an independent risk factor for RR.