Disease activity in systemic lupus erythematosus patients with end-stage renal disease: systematic review of the literature

Despite several years of intense investigation, there continues to be controversy about the value of clinical, demographic and histologic features in prediction of outcomes of lupus nephritis. In addition, contemporary treatments have reduced the risk of progressive renal injury and thus may have altered the prognostic significance of some of these factors. We have therefore re-examined the predictive value of variables previously associated with an increased risk of renal insufficiency by studying 65 patients with severe lupus nephritis treated with intensive regimens of intravenous pulse cyclophosphamide or methylprednisolone. Five clinical features at study entry were each associated with an increased probability of doubling the serum creatinine: age greater than 30 years, Black race, hematocrit less than 26%, serum creatinine greater than 2.4 mg/dl, and C3 complement less than 76 mg/dl. By multivariate survival analysis, serum creatinine, hematocrit and race emerged as the strongest set of independent clinical predictors; the other clinical and demographic factors, including age and C3 complement did not contribute significantly to outcome predictions in the context of these three variables. Renal biopsy evaluation offered additional prognostic information and showed that patients with severe active and chronic histologic changes were at increased risk for developing renal insufficiency. The combination of cellular crescents and interstitial fibrosis was particularly ominous. Outcome predictions based on the strongest clinical model (serum creatinine, hematocrit and race) were significantly enhanced by the addition of renal pathology data. Consideration of these prognostic factors may contribute to decisions regarding the type and intensity of immunosuppressive therapy for patients with lupus nephritis.

To determine if differences in morbidity of systemic lupus erythematosus (SLE) as measured by (1) important renal disease, (2) number of hospitalizations, and (3) neurologic disease can be explained by race, socioeconomic status (SES), or measures of compliance. The interrelationship of black race, SES, and the physician's assessment of compliance as risk factors for morbidity was examined in a cohort of 198 patients with SLE (179 female, 115 black). SES was measured with Nam-Powers scores for education (years), income, and job status, and source of insurance; compliance was assessed by physician global assessment and percent of protocol visits kept. Morbidity outcomes were important renal disease (creatinine level 1.5 mg/dL or greater, renal failure, nephrotic syndrome), neurologic involvement, and number of hospitalizations. The Johns Hopkins Rheumatology Faculty Practice, in which both private and clinic patients are seen. Black patients had significantly lower SES on all measures (p less than 0.0001) and were also less compliant by physician global assessment (odds ratio [OR] = 0.39, p = 0.002). Univariate analyses showed that blacks had a higher frequency of important renal disease (OR = 2.07, 95% confidence interval [CI] 1.05 to 4.11) and hypertension (OR = 1.80, 95% CI 1.01 to 3.23). Important renal disease was associated with the physician global assessment of compliance (p = 0.009) and hypertension (p less than 0.001). Multiple regression models for important renal disease, including race, physician global assessment of compliance, hypertension, SES, age, and gender, identified significant associations with only physician global assessment of compliance (OR = 0.40, 95% CI 0.17 to 0.91) and hypertension (OR = 5.37, 95% CI 2.40 to 11.98); black race was not significant (OR = 1.60, 95% CI 0.68 to 3.76). The second morbidity measure, number of hospitalizations, was associated with renal disease, neurologic disease, mouth ulcers, duration of disease, and public insurance but not with black race, in the best log-linear model. Neither race, SES variables, nor physician assessment of compliance was significantly associated with neurologic disease, the third morbidity measure. These data fail to support an independent association of black race with morbidity in SLE; rather, they suggest that noncompliance (as measured by physician global assessment) and type of medical insurance are important factors in morbidity. Classical epidemiologic measures of SES (education, income, occupation) do not appear to be significant confounders of the relationship of race to morbidity in SLE.

Studies of proliferative lupus nephritis (PLN) suggest that African-Americans have a poorer prognosis than Whites. However, no study has simultaneously examined socio-economic status. We studied rates of progression of PLN among a tri-ethnic population with respect to socio-economic status and race/ethnicity. A retrospective cohort study was carried out using individual and census-based neighbourhood data. Consecutive patients in urban tertiary care centres with biopsy-proven PLN were studied. The main outcome was time to doubling of serum creatinine. Among 128 patients with PLN, the percentage of patients who did not double their serum creatinine at 5 years was 67.0% (+/-4.8%) and at 10 years was 58.9% (+/-5.7%). In bivariate analyses, residence in a poor neighbourhood was positively associated with progression (P = 0.03), as was African-American and Hispanic race/ethnicity (P = 0.01). Residence in a poor neighbourhood remained associated with progression of disease after adjustment for age, sex, creatinine, hypertension, cyclophosphamide treatment and race/ethnicity [relative risk (RR) 3.5, 95% confidence interval (CI) 1.2-11, P = 0.03]. After adjustment for poverty and insurance, the RR for African-American race/ethnicity was reduced from 3.5 to 2.7 and was not statistically associated with progression of disease in the full model (P = 0.10). A similar reduction in RR from 5.5 to 3.6 was seen for Hispanic race/ethnicity, but this retained statistical significance (P = 0.03). Poverty is an important risk factor for progression of PLN, independent of race/ethnicity. Hispanics have an elevated risk similar to or greater than African-Americans. Given these findings, some of the poorer prognosis of African-American patients with PLN may result from socio-economic rather than biological or genetic factors.