Abstract
BACKGROUND/PURPOSE: Approximately half of patients with systemic lupus erythematosus (SLE) develop lupus nephritis (LN), a major cause of morbidity and early mortality in that disease. Prolonged renal inflammation is associated with irreversible kidney damage which confers a 30% risk of end stage renal disease (ESRD), making early, aggressive treatment mandatory. Failure to achieve therapeutic response or recurrence of renal flare often prompts repeat biopsy. However, the role of repeat biopsy in determining long-term renal prognosis remains controversial. For this reason repeat biopsies are usually not utilized unless clinical evidence of refractory or recurrent disease is already present, despite known mismatches between clinical and biopsy findings. The current study quantifies the degree to which histopathologic worsening between first and second biopsies and duration between them predicts ESRD and death. METHODS: Medical records of 141 LN patients with more than one biopsy were obtained from a single large urban medical center. Cases were attained using billing codes for diagnosis and procedures from 1/1999-1/2015. Biopsy worsening was defined as unfavorable histopathologic classification transitions and/or increased chronicity; if neither were present, the patient was defined as non-worsening. We used Cox proportional hazard models to study the relationship between ESRD and survival adjusting for covariates which included age at first biopsy, gender, race, initial biopsy class, and initial induction therapy. RESULTS: Of 630 patients screened, 141 had more than one biopsy. Advancing chronicity was detected in 48 (34.0%) and a renal class switch to worse grade of pathology was found in 54 (38.3%). At least one of these adverse second biopsy features was reported in 79 (56.0%) patients. Five years following initial biopsy, 28 (35.4%) of those with worsening histopathology on second biopsy developed ESRD, compared to 6 (9.7%) of non-worsening patients and 10 (12.7%) of patients with worsening histopathology had died compared to 2 (3.2%) of non-worsening patients. Biopsy worsening was associated with a significantly greater 15-year risk of ESRD (Hazard Ratio 4.2, p=0.0001) and death (Hazard Ratio 4.3, p=0.022), adjusting for age, gender, race, biopsy class, and treatment. Time between first and second biopsies was <1year in 32 patients, 1-5years in 81, and >5years in 28. Over a 15-year period, those with <1year between first and second biopsies (presumably enriched for patients with early clinical signs of progression) had a significantly greater risk of ESRD (Hazard Ratio 13.7, p<0.0001) and death (Hazard Ratio 16.9, p=0.0022) after adjusting for age, gender, race, biopsy class, and treatment. CONCLUSION: A repeat renal biopsy demonstrating worsening pathology increases the risk of ESRD and death more than four-fold compared to non-worsening patients. Given known potential mismatch between biopsy and clinical data, repeat biopsies may add important information and justify changes in treatment not considered on clinical grounds. Earlier detection of poor prognostic signs in those without early clinical deterioration might improve outcomes in enough patients to reconsider cost effectiveness of routine repeat biopsy.