Renal Disease in Pregnancy, an annotated bibliography

Here's an article on a placental site tumor resulting in secondary membranous nephropathy. More a curiosity than truly relevent as the diagnosis was made after delivery.

Also don't forget the review of renal biopsy in pregnancy I did a week ago. I reviewed this article.

Sebestyen et al. Successful management of pregnancy with nephrotic syndrome due to preexisting membranous glomerulonephritis: a case report. Fetal Diagn Ther (2008) vol. 24 (3) pp. 186-9

Sebestyen presents an interesting case. The patient first presented with proteinuria during her third trimester and undergoes a cesarian section at 36 weeks due to pre-eclampsia. At the time she had 3 grams of proteinuria on a 24-hour sample and a blood pressure of 170/100. Six weeks after delivery her blood pressure returned to normal but she remained proteinuric. She ultimatly had a kidney biopsy and was diagnosed with membranous glomerulonephritis. The patient was non-adherent with her treatment and variously treated with cyclosporin, cyclophosphamide and steroids. She aborted a pregnancy at 11 weeks due to cyclosporin exposure. [It is fascinating reading the literature and appreciating the widely varying recommendations and practices regarding the fetal toxicity of renal therapeutics]

The patient was lost to follow-up until she presented at 22 weeks gestation in a third pregnancy with nephrotic syndrome (5g in 24 hours), hypertension and decreasing renal function. The fetus was eutrophic (50% for femur length and biparietal diameter. Explanation).

The patient was started on daily methylprednisolone 16 mg and azathioprine 100 mg. Additionally she was treated with antihypertensives and diuretics. Subsequently the methylprednisolone dose was increased to 32 mg and then to 64 mg/day. Albumin was added to the diuretics. The patient did well until the 29th week when the albumin began to fall. They noted decreased fetal growth rate. In the 33rd week, fetal growth stopped and the patient went for cesarean section. A small for gestational age but otherwise healthy boy was delivered. The boy needed a few days of oxygen and was discharged home after 21 days.

After the patient stopped breast feeding (the article used the term ablactation, new one for me) she started cyclophosphamide and ACEi and went into remission within 3 months. Her subsequent history was marked by multiple relapses and ultimately end-stage renal disease requiring dialysis.

The authors felt this patient was remarkable for the use of a lower dose of steroids, 250 mg methylprednisolone pulses rather than 1g pulses. They felt the lower doses allowed them to avoid some of the side effects of glococorticoids:

• Intrauterine growth retardation (that's a stretch since the kid was small for gestational age)
• Suppression of fetal adrenal glands
• Periventricular leukomlacia

They attribute the success of the low dose steroids to the use of azathioprine, which is not absolutely contraindicated after the twelfth week.

I am skeptical of the claim that they were able to effectively control symptoms with the lower dose steroids and azathioprine. It looks to me like the patient has progressive worsening of the blood pressure, albumin, proteinuria and renal function. The decrease in the proteinuria may have more to do with the decrease in the GFR than the effectiveness of the therapy.

Malik et al. Repeated pregnancies in patients with primary membranous glomerulonephritis. Nephron (2002) vol. 91 (1) pp. 21-4

Malik presents a retrospective study on patients with MGN with multiple pregnancies with the hypothesis that repeated pregnancies worsen the prognosis of membranous nephropathy and increase the danger of pregnancy. Nine patients and 51 pregnancies were reviewed, 30 of the pregnancies were after the diagnosis of MGN and 21 were before.

All the patients came from a single tertiary care hospital and had their biopsies from May 1985 to December 1997. None of the patients had a renal biopsy during pregnancy.

The first table shows the gross outcomes for the pregnancies:

The data looks pretty promising. One women, patient 6, had terrible pregnancy outcomes:

1. Fetal loss at 37 weeks associated with pre-eclampsia in the first pregnancy
2. Spontneous abortion prior to 22 weeks with the second pregnancy
3. Live-birth complicated by prematurity, pre-eclampsia and cesarean sections. Infant with low birthweight.
4. Live-birth complicated by prematurity, pre-eclampsia and cesarean sections. Infant with low birthweight.

Outside of patient 6, there was only 1 pre-mature infant with a cesarian section in 26 pregnancies. No fetal deaths and no spontaneous abortions. 

Looking at the progression of renal disease that might have been induced by the pregnancy the data is likewise reassuring. The authors could find no increase in proteinuria or serum creatinine. Only one mother had hypertension and the worst serum Cr was 2.5 mg/dL in patient 5.

The authors conclude:

...the outcome of repeated pregnancies in patients with MGN is good with 90% live births. Repeated pregnancies do not influence the course of MGN.

This report looks great. It never mentions what type of therapy, if any, the patients received. The authors also mentioned that patients with more severe renal disease may have been routed to a different hospital that was a center of excellence in renal disease, suggesting possible selection bias to more mild cases.

References of interest

1. Abe S et al. The influence of antecedent renal disease on pregnancy Am J Obstet Gynecol 1985; 153: 508-514. 
2. Jungers Membranous glomerulobephritis and Pregnancy. Clin Nephrol 1988; 29: 106-107
3. Surian M et al. Glomerular disease in patients with primary and secondary glomerular diseases Nephron 1984; 36:101-105.

Barcelo et al. Successful pregnancy in primary glomerular disease. Kidney Internationsl (1986) vol. 30 (6) pp. 914-9

Barcelo examined 66 pregnancies in 48 women from 1972 to 1981. All of the women had primary glomerular disease. The distribution was:

• MPGN 16
• FSGS 13
• IgA 10
• MGN 7
• Focal nephritis 2

The pregancy outcome were generally very good, with MPGN having the worst outcomes. In the 9 pregnancies with membranous nephropathy, 7 were full term, 2 were premature. No fetal deaths.

When looking at proteinuria, the authors found a inverse linear correlation between proteinuria and fetal weight. 

The authors noted an increase in obstetric complications with three risk factors: renal failure, nephrotic syndrome and hypertension:

The incidence of obstetric complications varied with the presence or absence of risk factors. Women with low proteinuria (≤2.5 g/24 hours), without hypertension or renal failure had a low complication rate: 2.8% (one still birth). In the nephrotic syndrome, obstetric complications reached 33% of the pregnancies (pre-term), the same as with moderate renal failure (33% abortion, pre-term) and lower than women with hypertension (62% of the pregnancies, abortion, pre-term).

The authors employed a matched control group to try to assess any progression of renal disease induced by the pregnancy. Like Malik, they were unable to show any damage or worsening of renal prognosis due to the pregnancy. No significant differences between the pregnant and control groups.

The authors concluded that primary glomerular disease was associated with adequate fetal outcomes without damaging the mothers' prognosis. They speculate that the correlation of proteinuria to birthweight may be due to decreased placental perfusion from intravascular volume depletion from the loss of oncotic pressure.

The authors conclude with a twisted double negative meant to maximally confuse the reader:

The high fetal survival rate and the lack of repercussions of pregnancy on maternal nephropathy in the majority of women reported here indicate that pregnancy in patients with primary glomerulonephritis, glomerulosclerosis without hypertension, or significant renal function impairment should not be advised against. On the other hand, it should be kept in mind that coexistent hypertension worsens the prognosis, and the nephrotic syndrome increases the incidence of pre-term deliveries and low birth weight infants, although fetal viability is not markedly affected.

Song and Valentino. A pregnant patient with renal vein thrombosis successfully treated with low-dose thrombolytic therapy: a case report. Am J Obstet Gynecol (2005) vol. 192 (6) pp. 2073-5

From the department of "Some docs will try anything" comes this case report of a patient who presents with membranous nephropathy at 9 weeks gestation. She had 21 grams of protein and was started on cyclosporine.  [It is fascinating reading the literature and appreciating the widely varying recommendations and practices regarding the fetal toxicity of renal therapeutics] The patient had an albumin of 1.4. The physicians did a full thrombophilia work-up and found elevated levels of lipoprotein A and fibrinogen. Based on these results they started the patient on prophylactic Lovenox (1.5 mg/kg/d).

At 26 weeks, the patient developed back and flank pain. Ultrasound showed right renal vein thrombosis with extension into the inferior vena cava. The patient was then switched to unfractionated heparin 10 units/kg/hr and a t-PA infusion.

The thrombolytics were working and daily ultrasounds showed progressive reductions in clot volume. On the seventh day, however, the fetal heart rate slowed and the patient went for emergency cesarean section. They stopped the t-PA and heparin in the OR. The baby was born at 720 g, and did well enough to be extubated after 18 days and go home neurologically intact 83 days after birth.

The mother's course was complicated by 800 mL of blood loss (amazingly little if you think about it), anemia and hypertension but, likewise did well.

I'm trying to draw lessons from this case and the only ones that jump out at me are that heavy proteinuria can cause renal thrombosis in membranous, just like as in non-pregnant patients, and that prophylactic LMW heparin failed, at least in this case. 

Katzir et al. Pregnancy in membranous glomerulonephritis--course, treatment and outcome. Clin Nephrol (2004) vol. 61 (1) pp. 59-62

Katzir presents a case of a 23 y.o. primigravid with previously diagnosed membranous nephropathy stage III/IV, who was admitted at 14 weeks gestation. The patient had initially been diagnosed with membranous at age of 14. She went into complete remission following use of the Ponticelli regimen (true Ponticelli, with alternating chlorambucil and steroids). She had been in remission for more than a year prior to getting pregnant.

Seven weeks into the pregnancy she developed hypertension (160/90) with 1.5 grams of protein on a 24-hour sample. She responded to methyldopa and a low-salt diet (135/70).

At 14 weeks, the patient was readmitted with symptomatic hypertension (160/100). At that time she had 10.6 grams of protein on a 24-hour sample, an albumin of 2.1, and a cholesterol of 324 mg/dL. Renal function remained normal. After informed consent they started IV methylprednisolone 1g a day for 3 days followed by 50 mg a day of prednisone for 4 weeks. This resulted in improved blood pressure, loss of facial edema and a decrease in proteinuria to 6g/day and an increase in albumin to 2.5 g/dL. 

The patient had a second 3 day pulse of 1 gram of methylprednisolone and then oral prednisolone 60 mg every other day. After the second month of therapy, 26th week gestation, the albumin rose to 2.9 g/dL and the proteinuria was down to 4.8 g/24-hours, and the blood pressure was down to 135/80.

At 34 weeks gestation was readmitted with hypertension and uterine cramps. She was induced and delivered by cesarian section. The infant did well and was healthy at 1 year follow-up.

This looks like an effective treatment of membranous nephropathy in pregnancy, in my eyes more compelling than the Sebestyen case report of lower dose steroids and azathioprine. Steroids alone are generally considered ineffective in membranous, hence the use of Ponteceli regimen and tacrolimus as the standard of care. I would describe the steroid pulses as a means to prolong the pregnancy to get the fetus to viability and after delivery allow the mother to seek definitive care.

Packham et al. Membranous glomerulonephritis and pregnancy. Clin Nephrol (1987) vol. 28 (2) pp. 56-64

Packham reviews 33 pregnancies in 24 women with membranous nephropathy.

• 17 women initially presented during pregnancy
• 7 women had the diagnosis ante-partum

Of the women who presented during pregnancy 11 of them, received biopsies at the end of the first trimester.

In the 33 pregnancies the outcomes were not as rosy as other studies have shown:

• 8 pregnancies resulted in fetal death (24%)
• 6 occurred in the first trimester
• 3 spontaneous abortions (1 at 14 weeks, 2 in the first 12 weeks)
• 4 therapeutic abortions
• 1 still birth at 22 weeks
• 14 pregnancies resulted in premature infant (32-36 weeks)
• 60% of pregnancies that went beyond the first trimester had premature delivery
• 2 babies were born before 32 weeks
• Two congenital abnormalities
• hydrocephalus (one of the therapeutic abortions)
• hare lip

The mother's outcomes were relatively benign

• three patients had decrease in renal function
• 1 lost half of her kidney function, the loss persisted after delivery but remained stable on azathioprine and warfarin. The patient ultimatly stopped those drugs and progressed to ESRD
• 1 mother had a twin pregnancy and was diagnosed with mild membranous in the first trimester. At 34-weeks she developed ARF, severe hypertension. She was induced and after delivery went for a repeat kidney biopsy which demonstrated crescentic GN. The patient was treated with cyxlophosphamie and had a good outcome with stable renal function.
• 1 mother had worsening renal function at 14 weeks. Her biopsy revealed fibrinoid crescents. She later spontaneously aborted the fetus. Renal function later normalized.
• Hypertension occurred in 15 (46%) of the pregnancies.
• usually in the third trimester
• resolved after delivery in all but 3 women.

The authors felt that proteinuria was the most important risk factor:

In this, the largest study ever done on membranous nephropathy the authors found, outcomes worse in both the fetus and mothers. This may be attributable to the length of the study. The authors commented that neonatal treatment and anti-hypertensive therapies have gotten much better and this may explain the fact that no child or fetus died after  1975 (the patients were collected from 1964-1986).

Lindheimer and Katz. Gestation in women with kidney disease: prognosis and management. Baillieres Clin Obstet Gynaecol (1994) vol. 8 (2) pp. 387-404.

Lindheimer and Katz. Two of the biggest characters in my fellowship at University of Chicago. Whenever I was with them they would trade stories that left them both in stitches and mostly bewildered me. Adrian Katz was my mentor in clinic for my first year of fellowship and his last year of clinical medicine. It was an interesting pairing.

This is a book chapter masquerading as a journal article. It's huge. The chapter leads off with three conclusions the authors derived from reviewing 9 articles covering over 1,000 pregnancies:

1. women with renal disease but ony mild decreases in renal function and normal blood pressure at conception will do well and the pregnancy does not adversly affect their renal prognosis. They specifically exclude lupus, MPGN, scleroderma and pariarteritis from this conclusion.
2. If the patient has a creatinine ≥1.5 and >3 mg/dl or hypertension at conception about a third of patients will lose further renal function during pregnancy and have further loss of function post-partum.
3. Women with a creatinine ≥3 mg/dl are frequently infertile and when they do conceive the liklihood of a successful outcome is low with high maternal morbidity.

The remainder of the chapter/article goes on to provide data to support the above conclusions. They lead off reviewing Katz's own article from 1980 covering 121 pregnancies with intact renal function (Cr ≤ 1.4) and normal blood pressure (ony 20% had hypertension). The series excluded lupus nephritis. The outcomes are summarized below and have excellent outcomes:

The graph on the left is pregnancy outcome, on the right is the long-term follow-up for the mothers.

Ptoteinuria was common and was often nehrotic range. Katz found that the prpteinuria was usually well tolerated and did not influence the pregnancy outcome.

• 121 pregnancies
• 5 still births
• 6 neonatal deaths
• 91% success rate

Impressive, especially considering that most of these deliveries were in the 60's and 70's. He quotes a study by Abe, with more contemporary neonatal care that noted no perinatal deaths in mothers with IgA despite 20% preterm deliveries. 76% of the infants were of adequate size for gestational age, so renal disease does not necessarily result in small babies.

The 5 patients that ended up oin ESRD did so 2-8 years after the pregnancy and their diagnosis was a rogues gallery of renal badness: amyloidosis, FSGS, ADPKD, and 2 cases of crescentic GN.

Then this table. Wow that is awesome:

This table summarizes the 6 studies published after Katz's 1980 article and generally support Katz's initial conclusion that as long as the renal function is pretty good and the blood pressure is okay going into pregnancy the outcomes are acceptable.

The article looks at the data on IgA nephropathy, where there is some controversy. A number of authors (Kincaid-Smith,  Becker GJ, Packham 1, 2, 3, 4) have reported worsening of proteinuria, blood pressure and renal function during pregnancy. This was associated with fetal loss in 27% and on biopsy, woman who had a history of pregnancy tended to have more severe lesions. Other studies have shown beter outcomes with IgA and a lack of adverse outcomes compared to their nulliparous peers:

So Katz and Lindheimer conclude that for IgA:

Thus the observations of most investigators is that women with IgA nephropathy who are normotensive and have preserved renal function should anticipate few problems and that there is no convincing evidence linking gestation to progression of their disease.

The authors then describe some conflicting data on FSGS but generally reinforce their initial conclusions that patients with intact renal function be counseled that pregnancy is generally safe and a good outcome should be expected.

Lupus is a condition where, if the disease has been active in the 6 months prior to pregnancy, the prognosis is more guarded. They also mention a progressive post-partum renal failure due to thrombotic lesions in patients with anti-phospholipid antibody syndrome.

The authors feel that woman with scleroderma or periarteritis should be counseled against pregnancy and go as far as suggesting therapeutic abortions if the condition is present in the first trimester. Successful pregnancies with scleroderma have been managed with ACEi, a drug considered to be teratogenic.

The chapter then discusses women with more severe renal disease going into pregnancy. Bear and Kincaid-Smith (unable to find the ref) both reported permanent and severe losses of renal function associated with pregnancy in roughly haf the pregnancies. More contemporary studies have largely supported the worse outcomes with typically a quarter of patients suffering acceleration of renal disease. 85% of pregnancies were successful.

The next section is on nephrotic syndrome. The authors oint out that albumin physiologically falls 0.5-1.0 g/dL in pregnancy, so patient with proteinuria may slip into nephrotic syndrome as the serum albumin falls further. This is called cyclic nephrotic syndrome of pregnancy. Though there is some disent in the literature, the consensus is that nephrotic syndrome alone does not result in poor outcomes.

Kand L then examine the renal biopsy. They recounted the 1965 paper "Bleeding after renal biopsy in Pregnancy" which described poor outcomes after 77 renal biopsies in pregnancy:

• 16.7% gross hematuria
• 4.4% perirenal hematomas
• one maternal death

Kand L discuss their own reviews of the literature which found major biopsy complications occur during transperitoneal renal biopsies done during c-sections, biopsies with uncorrected coagulation abnormalities and biopsies with uncontrolled hypertension. They thern reported the excellent results at Chicago Lying-in Hospital with 3.5% hematuria following 400 biopsies.

In 1987 Packham and Fairley reported on 111 biopsies with a hematuria rate of only 0.9%.

They conclude that biopsey is safe in pregnancy.

They recommended biopsy in the following situation:

• sudden deterioration of renal function before 32 weeks
• symptomatic nephrotic syndrome before 32 weeks

They woud defere biopsy when there is

• proteinuria alone, with normal blood pressure and normal renal function.
• hematuria alone, with normal blood pressure and normal renal function.

They argue that active sediment (red and or white cell casts) is a relative indication for biopsy as a diagnosis of scleroderma would prompt termination of the pregnancy.

Management

• basic metabolic profile
• BUN
• albumin
• fasting lipid profile
• 24-hour urine for:
• creatinine
• protein
• uric acid
• AST/ALT, LDH
• PT, PTT, INR
• CBC d/p

The patient should be followed every 2 weeks until week 32 and weekly after that. Renal parameters should be checked at least every month.

Summary: