Prognostic relevance of persistent haematuria in patients with lupus nephritis.

Background: This study aimed to investigate the significance of persistent haematuria in lupus nephritis (LN). Methods: A total of 178 patients with LN were enrolled and regularly followed up from December 2016 to September 2022. Follow-up duration was from LN diagnosis to the patient's latest visit or terminated at the onset of endpoint events if they occurred. Their clinical and laboratory data as well as outcomes were monitored and analysed from the enrolment point until the final visit. Results: A total of 138 of 178 (77.5%) patients with LN presented with haematuria at the initial diagnosis. During follow-up, 34 patients (19.1%) had no haematuria, 72 (40.4%) patients initially diagnosed but later resolved and 66 (37.1%) exhibited persistent haematuria. Furthermore, patients with persistent haematuria showed higher Systemic Lupus Erythematosus Disease Activity Index scores (12.3 ± 2.9 versus 10.7 ± 3.6; P = .002), lower haemoglobin (92.7 ± 25.4 g/l versus 102.0 ± 21.1; P = .010), albumin (24.9 ± 6.8 g/l versus 27.5 ± 7.6; P = .024) and decreased C3 {median 0.4 g/l [interquartile range (IQR) 0.3-0.5]} versus 0.5 [0.3-0.6]; P = .009} and C4 [median 0.1 g/l (IQR 0.0-0.1) versus 0.1 (0-0.1); P = .012] and elevated blood urea nitrogen and serum creatinine [median 8.9 mmol/l (IQR 6.7-14.8) versus 5.7 (4.4-8.8) and median 90.0 μmol/l (IQR 64.5-132.3) versus 62.0 (48.0-80.0); all P < .001, respectively] compared with those without persistent haematuria (n = 106). In time-dependent Cox proportional hazards regression models, persistent haematuria exhibited a significant overall time-averaged effect on composite clinical endpoints, with an adjusted hazard ratio of 3.40 (95% CI 1.06-10.90; P = .039). Notably, this risk effect was not fixed but showed an increasing trend over time-at 1 months of follow-up, the risk of adverse outcomes in patients with persistent haematuria had increased to 10.37-fold. Conclusions: Almost 37.1% of patients exhibited persistent haematuria in LN. Furthermore, persistent haematuria independently predicted adverse outcomes, with significantly higher risks for all-cause mortality and progression to renal replacement therapy.