Article Text

Download PDFPDF

Clinical predictors of fetal and maternal outcome in Chinese patients with systemic lupus erythematosus
  1. M Y Mok1,
  2. P Y Leung1,
  3. T H Lao2,
  4. Y Lo1,
  5. T M Chan1,
  6. W S Wong1,
  7. C S Lau1
  1. 1Department of Medicine, Queen Mary Hospital, Pokfulam Road, Hong Kong
  2. 2Department of Obstetrics and Gynaecology, Queen Mary Hospital, Pokfulam Road, Hong Kong
  1. Correspondence to:
    Dr M Y Mok
    4/F, Professorial Block Department of Medicine, Queen Mary Hospital Pokfulam Road, Hong Kong;

Statistics from

Maternal and fetal outcome of pregnancy in patients with systemic lupus erythematosus (SLE) depends very much on the composition of the patient cohort being analysed.


We performed a retrospective study to identify the clinical predictive factors for unfavourable pregnancy outcome in our Chinese patients with SLE from a homogeneous cohort being followed up at the university rheumatology clinic during the period 1982–96. Ninety one pregnancies from 66 patients were identified. Antiphospholipid antibodies (aPL), including lupus anticoagulant (LAC) and/or IgG anticardiolipin antibodies, were present in 18/66 (27%) patients. In 33/80 (41%) pregnancies resulting in live births, patients had a past history of lupus nephritis. Only one patient had impaired renal function, two had significant proteinuria ⩾1 g/day, and four had stable hypertension while receiving antihypertensive drugs at conception. Active lupus was present within 6 months before conception in 12/91 (13%) and at conception in 7/91 (8%) pregnancies.

The live birth rate and fetal loss rate was 80/91 (88%) and 11/91 (12%), respectively. Fetal loss was found to be associated with serum anti-Ro (p = 0.04), anti-La (p = 0.007), and significant proteinuria at conception (p = 0.03). aPL were not found to relate to a single episode of fetal loss (p = 0.06), but they were more prevalent in patients with recurrent miscarriages (odds ratio (OR) = 14.3, p = 0.008), of which LAC was the strongest predictive factor (OR = 23.3, p = 0.002). Poor fetal outcome was found in 27/80 (34%) live births. There were 13 premature births and 14 fetuses had intrauterine growth retardation (IUGR). Significant proteinuria during pregnancy (p = 0.006) and flare in gestation (p = 0.001) was found to be related to prematurity. Multivariate analysis identified flare during pregnancy, in particular in the second trimester, as the strongest risk factor (relative risk (RR) = 5.5, 95% confidence interval (CI) 1.4 to 21.9, p = 0.02). Active lupus both within 6 months before (p = 0.01) and at conception (p = 0.002) was found to predispose to IUGR. The latter was identified by multivariate analysis as the most predictive factor (RR = 22.1, 95% CI 2.5 to 215.1, p = 0.008). Pre-eclampsia was present in 8/80 (10%) pregnancies and no particular risk factor was found.

Thirty three pregnancies were recorded in the 27 patients with a past history of lupus nephritis. The live birth rate, fetal loss rate, and rate of poor fetal outcome (prematurity in four, IUGR in six) were not particularly different from those in patients with non-renal SLE and were 94%, 6%, and 30%, respectively. Predictive factors for poor fetal outcome in this subset of patients included significant proteinuria at conception (p = 0.004) and during pregnancy (p = 0.01) and active SLE within 6 months before conception (p = 0.007) and at conception (p = 0.03).


The fetal loss rate in our cohort lies near the lower end of the 11–46% range which was previously reported.1 This was partly because our patients had a lower prevalence of aPL (27%) than the 34–44% reported in other cohorts2 and partly related to the small proportion of patients with active SLE at conception. Better live birth rates have been reported in patients with inactive rather than active disease at conception.3–5 Our reported rate of poor fetal outcome and the associated risk factors were similar to those previously reported.1,6–9 The fetal loss rate in our patients with a past history of lupus nephritis was, however, lower than those reported (13–46%)3 because the renal disease of our patients was quiescent at conception; active nephritis at conception has been reported to give a higher risk of fetal loss.10 Furthermore, patients in our clinic with impaired renal function were generally discouraged from becoming pregnant.

In conclusion, adverse pregnancy outcome was more common in patients with active disease in gestation. Fetal outcome is favourable in patients with a past history of lupus nephritis if they have normal renal function, no significant proteinuria, and satisfactory blood pressure control.


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.