MURAKAMI, SOH MD; SAITOH, MASAHIRO MD; KUBO, TAKESHI MD; KOYAMA, TETSUO MD; KOBAYASHI, MASAKI MD
Objective: To identify women with severe preeclampsia or severe gestational proteinuria at high risk of having underlying renal disease.
Methods: Between 1980 and 1999, 86 Japanese women who had severe hypertension, severe proteinuria, or both during pregnancy had postpartum needle biopsies of their kidneys. Diagnoses before biopsies were severe preeclampsia in 74 women and severe gestational proteinuria in 12. We compared clinical characteristics, such as antepartum hematuria and postpartum proteinuria, and maternal and neonatal outcomes with regard to presence of renal disease.
Results: Nineteen of 86 women (22.1%, 95% confidence interval [CI] 13.9%, 32.3%) were diagnosed with underlying renal disease. Immunoglobulin A nephropathy was present in 12. Women with renal disease had a significantly earlier onset of proteinuria than those without (median 11 versus 32 weeks' gestation, P < .001). Eighteen of 19 women with renal disease had proteinuria, hypertension, or both before 30 weeks' gestation. Ten of 12 women with severe gestational proteinuria (83.3%, 95% CI 51.6%, 97.9%) had underlying renal disease. Eight of the 19 women had antepartum hematuria, and seven had elevated serum immunoglobulin A levels. In women with severe preeclampsia, onset before 30 weeks' gestation was the best predictor of underlying renal disease (odds ratio 34.1, 95% CI 3.8, 304.5). Women with renal disease had lower rates of severe hypertension (nine of 19 versus 59 of 67, P < .01) and small-for-gestational-age infants (four of 19 versus 34 of 67, P < .05) than those without renal disease.
Conclusion: Women who had gestational proteinuria or preeclampsia before 30 weeks' gestation were more likely to have had underlying renal disease.
Renal disease has been recognized as a risk factor for preeclampsia and pregnancy-induced hypertension.1,2 However, it is difficult to diagnose underlying renal disease without renal biopsy in women who develop hypertension, proteinuria, or both for the first time during pregnancy. Renal biopsy yields a definitive histologic diagnosis, which can then provide guidance for managing future pregnancies as well as remote prognosis.3,4 Only a few studies have analyzed the rate of underlying renal disease in this population using renal biopsy,3,5–7 possibly because physicians are generally reluctant to do biopsies because they are invasive procedures and can have serious complications.
Therefore, it is useful to identify women at high risk of having underlying renal disease and to clarify their prognoses. We studied the presence of underlying renal disease by postpartum renal biopsy in 86 women with severe preeclampsia or severe gestational proteinuria, and compared women with renal disease with those without for clinical characteristics and pregnancy outcomes.
Materials and Methods
Preeclampsia was diagnosed according to the criteria established by the National High Blood Pressure Education Program Working Group.8 Proteinuria was defined as excretion exceeding 300 mg in a 24-hour urine collection, measured by the pyrogallol red method,9 or in some outpatients, a concentration of 30 mg/dL (1+ on dipstick) or more in a random urine specimen that is usually correlative.8 Severe preeclampsia was defined as sustained hypertension exceeding 160 mmHg systolic or 110 mmHg diastolic pressure with proteinuria as described and sustained blood pressure of 140/90 to 160/110 mmHg with severe proteinuria exceeding 2.0 g in 24 hours. Gestational proteinuria means proteinuria detected for the first time during pregnancy without hypertension, and the severe range was set at excretion over 2.0 g in 24 hours.
Between 1980 and 1999, 97 Japanese women with severe preeclampsia or severe gestational proteinuria were admitted to our hospital. They were considered to have indications for percutaneous needle biopsy of the kidney. Exclusion criteria included patient refusal (n = 6), polycystic kidney (n = 1), and thrombocytopenia (n = 4). Among the 97 women, 86 (74 with severe preeclampsia and 12 with severe gestational protein-uria) were enrolled after giving written informed consent. Sixty-two women were nulliparous and 24 were multiparous. We began renal biopsies as a method of clinical management for women with severe preeclampsia or gestational proteinuria on the basis of previous reports.2–7 Ethics committee approval was not necessary for renal biopsies when our institutional ethics committee was established.
All patients received antenatal care every 4 weeks in the first and second trimesters, every 2 weeks between 29 and 35 weeks' gestation, and weekly thereafter. Seven women who had hypertension or proteinuria detected on their first visits were included in the study because they had not been noted before.
Renal biopsies were done under ultrasound guidance 1 to 2 weeks after delivery. No women had therapy for renal diseases before biopsy because no marked deterioration of renal function was found. The specimen from each biopsy was sectioned and stained with hematoxylineosin and immunohistochemical methods. Each diagnosis was determined by consensus of two nephrologists who specialize in renal pathology.
After biopsy, women were assigned to one of two groups, those with underlying renal disease and those without. Clinical characteristics and maternal and neonatal outcomes were compared between groups. Fetal growth restriction (FGR) was determined on the basis of gestational age–birth weight standard for the Japanese population.10
We used presence of antepartum hematuria (hematuria), elevated serum immunoglobulin (Ig)A level (elevated IgA), and onset of preeclampsia before 30 weeks' gestation (early onset) as predictors for underlying renal disease in women with severe preeclampsia. In addition to calculations of positive predictive values, multiple logistic regression analysis was done for women with hematuria and early onset, using SAS-PC software (SAS Institute, Cary, NC). Because of a correlation between elevated IgA and early onset, the software failed to perform multiple logistic regression analysis with the three predictors described above. The 95% confidence intervals (CI) were calculated for the estimated proportions, such as the rate of underlying renal disease, using an approximation of the normal distribution. For non-normal distributions, such as onset of proteinuria, the Mann-Whitney U test was used. Means of data with an approximately normal distribution and proportion were compared using two-tailed Student t and Pearson χ2 tests, respectively, which are contained in the Statistical Package for the Social Sciences for PowerMac 6.1J (SPSS Inc., Chicago, IL).
One woman had transient macrohematuria after biopsy. No others had any complications. Nineteen of 86 women (22.1%, 95% CI 13.9%, 32.3%) had underlying renal disease. Histologic diagnoses of renal diseases were IgA nephropathy (n = 12), non-IgA nephropathy (n = 3), minimal change (n = 2), interstitial nephritis (n = 1), and benign nephrosclerosis (n = 1). Twelve of 19 women (63.2%) with renal disease and 50 of 67 (74.6%) without were nulliparous (statistically not significant). Nine of the 19 women were from the group of 74 with severe preeclampsia (12.2%, 95% CI 5.7%, 21.8%) and ten were from the group of 12 with severe gestational proteinuria (83.3%, 95% CI 51.6%, 97.9%). Among women with severe preeclampsia, four of 52 nulliparas (7.7%, 95% CI 2.1%, 18.5%) and five of 22 multiparas women (22.7%, 95% CI 7.8%, 45.4%) had underlying renal disease (not significant). In multiparas with repeated preeclampsia, six of seven women (85.7%, 95% CI 42.1%, 99.6%) had renal disease, and 11 of 17 (64.7%, 95% CI 38.3%, 85.8%) did not (not significant).
The median (range) gestational ages at onset of proteinuria in women with renal disease and those without were 11 (6–33) and 32 (6–38) weeks' gestation, respectively. Women with renal disease had significantly earlier onsets of proteinuria than those without (P < .001) (Figure 1). At onset of hypertension, the median (range) gestational ages of women with renal disease and without were 27 (6–34) and 32 (6–38) weeks' gestation, respectively (not significant).
Eighteen of 19 women with renal disease and 28 of 67 (41.8%) without had hypertension, proteinuria, or both before 30 weeks' gestation (Figure 2). In the remaining woman with renal disease, proteinuria and hypertension developed at week 30 of gestation.
In women with severe preeclampsia, the positive predictive values of hematuria, elevated IgA (above 400 mg/dL), and early onset of any renal diseases were satisfied by two of 12 women (16.7%, 95% CI 2.1%, 48.4%), four of six (66.7%, 95% CI 22.3%, 95.7%), and eight of 35 (22.9%, 95% CI 10.4%, 40.1%), respectively. Table 1 shows that the odds ratio (OR) for underlying renal disease in women with early onset was higher than that in women with hematuria.
At 6 months postpartum, the frequency of protein-uria was higher in women with renal disease than in those without, although the difference was not significant. Among women who had preeclampsia or gestational proteinuria before 30 weeks' gestation, however, women with renal disease had proteinuria at 6 weeks postpartum more frequently than those without (Table 2).
Table 3 shows the pregnancy outcomes. Women with renal disease had a lower rate of severe hypertension and FGR.
Our rate of underlying renal disease was 22.1%, a figure compatible with some studies5,11 but lower than others.3,6,7 That discrepancy should not be because of characteristics of our patient population because the rate was lowered to 12.1% after the subjects were restricted to those with severe preeclampsia. Therefore, extremely high rates of underlying renal disease (65.0% in early-onset preeclampsia3 and 71.7%6) seem to appear in a selected population, perhaps relating to the referral pattern of the hospital. Those studies might be confounded by a small series7 or a low rate of enrollment.6 This study had stringent criteria for diagnosis of preeclampsia and gestational proteinuria, and a higher proportion of women had renal biopsies. Therefore, the frequency of 22.1% could represent the actual rate of underlying renal disease.
In women with severe preeclampsia, only a few nulliparas (7.7%) had underlying renal disease. That result supports the findings of a previous report.5 Although that study found that 76% of multiparas had underlying renal disease, the frequency in our study was only 22.7%. Although the incidence of underlying renal disease was higher in women with repeated preeclampsia, the rate was still 35.3% (six of 17 women). Therefore, those discrepancies might result from differences in the populations studied.
IgA nephropathy, the most common primary glomer-ulonephritis, showed the highest prevalence (63.2%) in our study, which confirmed a previous report.3 Although the frequency of renal disease in Japanese women of childbearing age has not been clearly described, IgA nephropathy was most prevalent among renal diseases diagnosed before pregnancy (30 of 72 pregnant women, 41.7%).12 The higher frequencies of IgA nephropathy in preeclamptic women might indicate a strong association between those diseases.13
In the present study, severe gestational proteinuria strongly predicted underlying renal disease. In women with severe preeclampsia, underlying renal disease was found only in those with onset of hypertension, protein-uria, or both by 30 weeks' gestation. It is novel to consider the relationship between gestational age at onset of preeclampsia and the rate of underlying renal disease. Preeclampsia that develops before 20 weeks' gestation is defined as superimposed preeclampsia. Our findings suggest that women who develop preeclampsia before 30 weeks' gestation are also likely to have underlying renal disease.
IgA nephropathy can be suspected by the presence of hematuria, elevated serum IgA level, or both. In our study, the sensitivities for antepartum hematuria and elevated serum IgA were low for all renal diseases. A nephrologist would generally recommend aggressive blood pressure control in pregnant women with complicating renal disease, to prevent an increase in glomerular pressure.14 It is of great value to predict underlying renal disease in preeclamptic women by noninvasive clinical methods because the biopsy rarely is indicated during pregnancy with a few exceptions.15 In this study, onset of preeclampsia before 30 weeks' gestation was the best predictor of underlying renal disease in women with severe preeclampsia. According to that criterion the rate of underlying renal disease among our patients was 39.1% (18 of 46 women). We propose that women who have preeclampsia before 30 weeks' gestation and exhibit proteinuria at 6 weeks postpartum have an indication for postpartum renal biopsy for managing future pregnancies and their remote prognosis; 45.5% (15 of 33 women) had underlying renal disease, although four of 19 women with it were not diagnosed.
This study showed that women with renal disease had a lower rate of FGR and severe hypertension than those without. Hypertension at conception or early in pregnancy is the major independent risk factor for fetal prognosis.16 In fact, FGR was more severe in women with severe hypertension (32 of 68 women) than in those with mild or no hypertension (six of 18), although the difference was not significant. Proteinuria in the range of nephrotic syndrome with marked hypoalbuminemia in the first trimester is another important risk factor for FGR,17,18 but none of our patients had this risk factor. Therefore, the reason for the lower rate of FGR in women with renal disease is that about half of them were normotensive.
1. Paller MS. Renal diseases. In: Burrow GN, Ferris TF, eds. Medical complications during pregnancy. 4th ed. 1995:264–84.
2. Cunningham FG, Cox SM, Harstad TW, Mason RA, Pritcard JA. Chronic renal disease and pregnancy outcome. Am J Obstet Gynecol 1990;163:453–9.
3. Ihle BU, Long P, Oats J. Early onset preeclampsia: Recognition of underlying renal disease. BMJ 1987;294:79–81.
4. Gaber LW, Spargo BH, Lindheimer MD. Renal pathology in preeclampsia. Baillieres Clin Obstet Gynaecol 1994;8:443–68.
5. Fisher KA, Luger A, Spargo BH, Lindheimer MD. Hypertension in pregnancy: Clinical-pathological correlations and remote prognosis. Medicine 1981;60:267–76.
6. Beller FK, Dame WR, Witting CH. Renal disease diagnosed by renal biopsy. Prognostic Evaluation. Contr Nephrol 1981;25:61–70.
7. Weiner CP, Bonsib SM. Relationship between renal histology and plasma antithrombin III activity in women with early onset preeclampsia. Am J Perinatol 1990;7:139–43.
8. Joint National Committee on Detection, Evaluation and Treatment of High Blood Pressure. National high blood pressure education program working group report on high blood pressure in pregnancy. Am J Obstet Gynecol 1990;163:1689–712.
9. Watanabe N, Kamei S, Ohkubo A, Yamanaka M, Ohsawa S, Makino K, et al. Urinary protein as measured with a pyrogallol red-molybdate complex, manually and in a Hitachi 726 automated analyzer. Clin Chem 1986;32:1551–4.
10. Ogawa Y, Iwamura T, Kuriya N, Nishida H, Takeuchi H, Takada M, et al. Birth size standards by gestational age for Japanese neonates. Acta Neonate Jpn 1998;34:624–32.
11. Reiter L, Brown MA, Whitworth JA. Hypertension in pregnancy: The incidence of underlying renal disease and essential hypertension. Am J Kidney Dis 1994;24:883–7.
12. Abe S, Amagasaki Y, Konishi K, Kato E, Sakaguchi H, Iyori S. The influence of antecedent renal disease on pregnancy. Am J Obstet Gynecol 1985;153:508–14.
13. Kincaid-Smith PS, Whitworth JA, Fairley KF. Mesangial IgA nephropathy in pregnancy. Clin Exp Hypertens 1980;2:821–38.
14. Ferris TF. Hypertension and preeclampsia. In: Burrow GN, Ferris TF, eds. Medical complications during pregnancy. 4th ed. 1995:1–28.
15. Packham D, Fairley KF. Renal biopsy: Indications and complications in pregnancy. Br J Obstet Gynaecol 1987;94:935–9.
16. Jungers P, Chauveau D. Pregnancy in renal disease. Kidney Int 1997;52:871–85.
17. Studd JW, Blainey JD. Pregnancy and the nephrotic syndrome. BMJ 1969;1:276–80.
18. Barcelo P, Lopez-Lilo J, Cabero L, Del Rio G. Successful pregnancy in primary glomerular disease. Kidney Int 1986;30:914–9.