Pregnancy in women with pre-existing lupus nephritis: predictors of fetal and maternal outcome.

BACKGROUND Only few data are available on pregnancy in patients with lupus nephritis (LN) diagnosed before conception. The aim of this study was to identify the risk factors for complicated pregnancy in women with pre-existing LN. METHODS In a multicentre study, we collected data on 113 pregnancies occurring in 81 women with pre-existing biopsy-proven LN. Primary outcomes were fetal loss including perinatal death and renal flares during and 12 months after pregnancy. Univariate and logistic regression analyses were used to identify predictors of outcomes. RESULTS Renal biopsy performed 7.2 +/- 4.9 years before pregnancy showed the following WHO classes: 6 patients in II, 8 in III, 48 in IV and 19 in V. At conception, most patients were in complete (49%) or partial (27%) remission. There were nine spontaneous abortions, one stillbirth and five neonatal deaths. Thirty-one deliveries were preterm. Birth weight was <2500 g in 34 newborns. During pregnancy or after delivery, there were 34 renal flares, most of which (20) were reversible. Three patients had a progressive decline of glomerular filtration rate (one on dialysis). At logistic regression analysis, the pregnancy outcome was predicted by hypocomplementaemia at conception (RR 19.02; 90% CI 4.58-78.96) and aspirin during pregnancy (RR 0.11; 90% CI 0.03-0.38). Renal flare was predicted by renal status (partial remission RR 3.0; 90% CI 1.23-7.34, nonremission RR 9.0; 90% CI 3.59-22.57). CONCLUSIONS Pregnancy can be successful in most women with pre-existing LN, even for those with a severe renal involvement at onset. Renal flares during and after pregnancy are not uncommon and can be predicted by renal status assessed before pregnancy. Normocomplementaemia and low-dose aspirin therapy during pregnancy are independent predictors of a favourable fetal outcome.