Abstract
The differential diagnosis of proteinuria and hematuria in pregnancy is broad and includes active lupus nephritis. Identification of the correct diagnosis often has a profound therapeutic impact on not only the mother but also the fetus. To date, relatively few reports exist on the role of renal biopsy during pregnancy among women with systemic lupus erythematosus (SLE). We present a case series of 11 pregnant women with SLE who underwent a renal biopsy to evaluate a presumptive flare of lupus nephritis. The electronic medical record was retrospectively analyzed for pre-biopsy serum creatinine, proteinuria, hematuria, antinuclear antibodies (ANA), and antibodies to double-stranded DNA (anti-dsDNA); histologic findings on renal biopsy; and the clinical course of each mother and fetus. From 2001 to 2012, 11 pregnant women with SLE flares during pregnancy underwent a renal biopsy at an academic tertiary medical center. At the time of biopsy, median gestational age was 16 weeks (range 9 to 27), median serum creatinine was 0.6 mg/dl (interquartile range 0.5 to 0.9), six (55%) had hematuria, and all had proteinuria >500 mg/24 hours. Proliferative lupus nephritis was found in 10 (91%) of 11 biopsies (five with ISN/RPS Class III; five with ISN/RPS Class IV). All but one individual underwent a change in management guided by information gleaned from renal biopsy. No apparent biopsy-related complications occurred to mother or fetus. Three women elected to terminate their pregnancy; although many factors were involved, the findings on renal biopsy informed the decision-making process. Among the remaining cases, there were three pre-term deliveries, one fetus with complete heart block, one in utero demise, and one maternal death. Renal biopsy is helpful at informing the management of patients with lupus nephritis during pregnancy.
Introduction
Systemic lupus erythematosus (SLE) most commonly affects women of childbearing age. 1 When disease activity is high during pregnancy, SLE is associated with worse clinical outcomes.1–4 For the fetus, these adverse events include neonatal lupus, intrauterine growth retardation, preterm birth, and even fetal death. For the mother, these include a flare in disease activity, preeclampsia, and a heightened risk of death.1–3,5 In addition, women with pre-existing kidney disease (serum creatinine ≥1.4 mg/dl), whether lupus related or not, are at increased risk of accelerated renal disease progression. 6
Among the many manifestations of SLE, lupus nephritis can be particularly challenging to manage during pregnancy, as the diagnosis of a flare is often difficult to make based on clinical judgment alone. A healthy pregnancy itself may present with symptoms that could be confused with SLE. Moreover, pregnancy-related kidney disease, such as pre-eclampsia or hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome, and other glomerulonephritides may mimic lupus nephritis. Thus, while the presence of hematuria, proteinuria, cellular casts, elevated serum creatinine, increased anti-double-stranded DNA (anti-dsDNA) titers, and/or low complement levels on routine laboratory work is suggestive of a lupus nephritis flare, a definitive diagnosis cannot be made unless a renal biopsy is performed. 7 The information gained from biopsy, particularly the underlying histopathology as well as activity and chronicity indices, may be valuable in guiding treatment.7,8
Notwithstanding, a renal biopsy is not without risk. Potential complications include pain, infection, bleeding, injury to surrounding organs, and death. 9 In pregnancy, one must consider how these risks might affect not only the mother, but also the fetus. Several studies have explored the association between lupus nephritis and pregnancy outcomes.2,3,10,11 Over the past 20 years, however, only one study has specifically addressed the role of renal biopsy in lupus nephritis during pregnancy. 12 In this case series by Krane et al., three women underwent a renal biopsy while pregnant. All were found to have diffuse proliferative lupus nephritis and subsequent management was guided by information gained from renal biopsy. 12
We describe 11 women with SLE who underwent a renal biopsy during pregnancy for further evaluation of possible active renal involvement. We also outline our approach to renal biopsy in the pregnant patient with SLE.
Materials and methods
Using a pathology database at The Johns Hopkins Hospital, a tertiary medical center in Baltimore, Maryland, all biopsies were screened to identify patients with lupus who underwent a kidney biopsy during a pregnancy. From May 2001 to March 2012, a total of 11 pregnant women with SLE underwent a renal biopsy at our institution. All 11 women were 18 years or older and pregnant at the time of biopsy, and each satisfied the revised American College of Rheumatology (ACR) lupus criteria.13,14 The decision to pursue a renal biopsy was made by the attending nephrologist and rheumatologist, based on their combined clinical judgment, and in consultation with the treating obstetrician.
The risks and benefits of a renal biopsy were discussed with each patient and written informed consent was obtained. Each biopsy was performed, with the patient in prone position, by an attending nephrologist via a percutaneous approach using an 18-gauge semi-automated needle and real-time ultrasound guidance. For biopsies performed after December 2001 (nine patients), a needle guide device for the ultrasound probe was also used. At our institution, it has been standard practice for patients to have a blood pressure ≤160/100 mmHg, platelet count ≥50,000, partial thromboplastin time <42.1 seconds, and international normalized ratio <1.3 before going forth with a renal biopsy. If necessary, pharmacologic and/or transfusion therapy is provided to achieve these thresholds. Two to three cores were obtained per biopsy and all patients were monitored overnight in the hospital for any procedure-related complications. This included bed rest for the first six hours post-biopsy, frequent vital signs, including pre- and post-biopsy fetal heart rate monitoring for patients at 23 weeks’ gestation or greater, and hemoglobin measurements at approximately 8 and 16 hours post-procedure. Follow-up imaging was ordered only if clinically indicated. 15
Information on the clinical course of each woman and the outcome of her fetus were extracted from the mother’s electronic medical record. The presence of proteinuria was based on either a 24-hour urine collection and/or a random spot urine protein-to-creatinine ratio. 16 Biopsies were read by attending pathologists at our institution using the International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification system. 17 We defined renal biopsy complication as bleeding requiring packed red blood cell (RBC) transfusion or therapeutic intervention performed by radiology or surgery; hemodynamic instability necessitating vasopressor support or transfer to the intensive care unit; inadvertent injury to surrounding organs or the fetus; or death to the mother or fetus. This study was approved by the Johns Hopkins University School of Medicine Institutional Review Board.
Results
Baseline characteristics of 11 pregnant women with systemic lupus erythematosus at the time of renal biopsy
C: Caucasian; AA: African American; A: Asian; AC: African Caribbean; Cr: creatinine; ANA: antinuclear antibody; anti-dsDNA: anti-double-stranded DNA; aPL-Ab: antiphospholipid antibody; RBC: red blood cells; hpf: high-power field; N: no; Y: yes; n/a: not available.
Had clinical features suggestive of lupus nephritis (proteinuria, low complement, and positive anti-dsDNA for Patient 7; and proteinuria for Patient 11) prior to pregnancy onset, but not biopsy-proven until renal biopsy performed during pregnancy.
Findings on renal biopsy and subsequent changes in treatment among 11 pregnant patients with systemic lupus erythematosus
Presence of crescents.
Presence of fibrinoid necrosis.
ISN: International Society of Nephrology; RPS: Renal Pathology Society; mod: moderate; n/a: not applicable; AZA: azathioprine; CsA: cyclosporine; CYC: cyclophosphamide; HCQ: hydroxychloroquine; MMF: mycophenolate mofetil; P: prednisone.
Clinical outcomes of both mother and fetus following histologic evidence of active lupus nephritis during pregnancy
Neonate transferred to neonatal intensive care unit (NICU) following delivery.
Cr: serum creatinine (mg/dl); ARDS: acute respiratory distress syndrome; n/a: not available.
Patient 10 also had an active ISN/RPS Class IV lupus nephritis on renal biopsy but was further along in her pregnancy (19 weeks’ gestation) at the time of tissue diagnosis. She elected to proceed with prednisone and azathioprine. However, at 21 weeks, she was noted to have worsening proteinuria (17 grams on 24-hour urine collection). Further evaluation revealed a large thrombus involving her gonadal vein, inferior vena cava, and right common iliac vein. The renal veins were not affected. Given these interval adverse clinical developments, she chose to terminate her pregnancy and was subsequently treated with intravenous heparin and cyclophosphamide. She was later found to have a pulmonary embolus as well as lower extremity cellulitis, both of which resolved with medical management.
Patient 6 had markedly aggressive disease with crescents present in 16 of 29 glomeruli on renal biopsy. Termination of pregnancy was considered as it was felt that she would benefit most from intravenous cyclophosphamide, which was deemed unsafe for the fetus at this stage in her pregnancy (23 weeks’ gestation). However, the patient wanted to continue her pregnancy through delivery, and was thus treated with high-dose intravenous steroids (methylprednisolone 1000 mg daily for three days) and azathioprine. She also received aspirin 81 mg once daily and subcutaneous enoxaparin 40 mg twice daily for thromboprophylaxis given her history of antiphospholipid antibodies and active nephrotic-range proteinuria. Unfortunately, at 25 weeks’ gestation, she suddenly developed hypotension, shortness of breath, and subsequent cardiac arrest. Despite aggressive resuscitation efforts, the patient and her fetus did not survive. Autopsy was not performed; the major diagnostic considerations were pulmonary embolism and sepsis to account for the demise.
Finally, Patient 11 was treated with methylprednisolone, azathioprine, and hydroxychloroquine after renal biopsy revealed proliferative glomerulonephritis with mild to moderate activity. Her fetus had multiple anomalies and was severely compromised with evidence of poor cardiac function five days prior to the biopsy. The fetal heart rate was initially normal following the biopsy; however, the fetus died the following day from heart failure. Following the diagnosis of fetal demise the patient underwent induction of labor with subsequent extreme preterm delivery of a stillborn infant. Both mycophenolate mofetil (in place of azathioprine) and lisinopril were then added to her medication regimen. At the time of discharge, her serum creatinine remained stable at 1.0 while her proteinuria decreased to 4959 mg/g Cr (from 17,000 mg/g Cr). The five remaining women delivered healthy children.
Table 3 also depicts serial values of maternal serum creatinine, including both pre-biopsy and at one year post-delivery. Patient 10 had the greatest increase in serum creatinine, from 1.1 mg/dl to 2.3 mg/dl. Most of the other women experienced smaller rises in serum creatinine. It is difficult to determine whether these observed increases, during this interval from pregnancy to the postpartum period, were due to lupus nephritis, the physiologic lowering of creatinine during pregnancy, or a combination thereof.
None of the 11 women with histologic evidence of active lupus nephritis had a reported complication as a result of performance of a renal biopsy during pregnancy. Mean hemoglobin level prior to biopsy was 9.9 ± 0.8 g/dl, with the lowest hemoglobin being 8.2 g/dl. Mean post-biopsy hemoglobin was similar at 9.6 ± 0.5 g/dl. Three women were on aspirin therapy, with doses held seven days before and restarted seven days following biopsy. Only one had a delay in procedure because of aspirin therapy. Our standard protocol is to hold subcutaneous heparin for deep vein thrombosis prophylaxis the night before and the morning of renal biopsy, restarting 48 hours post-procedure. Finally, pre-biopsy platelet counts ranged from 125,000 to 470,000 cells/mm3.
Discussion
Lupus nephritis is often challenging to manage because of the inability to predict underlying histopathology based on clinical features alone. 7 Such challenging diagnostic and therapeutic considerations are no less compelling during pregnancy. Moreover, one must consider not only the impact of lupus nephritis on pregnancy and health of the fetus, but also how pregnancy can affect lupus nephritis and the progression of kidney disease.1,2,18 This case series highlights the importance of renal biopsy in helping guide the management of lupus nephritis in pregnancy.
In the case of other preexisting primary renal diseases that may co-exist with pregnancy, such as membranous nephropathy or focal segmental glomerulosclerosis, significant increases in proteinuria can generally be attributed to that particular disease once preeclampsia has been ruled out. In a patient with SLE, however, an increase in proteinuria could be a manifestation of class II (mesangial), III (focal proliferative), IV (diffuse proliferative), or V (membranous) disease, or a combination thereof.17,19 Among those with antiphospholipid antibodies, the possibility of thrombotic microangiopathy must also be considered.1,20 Moreover, other systemic causes of proteinuria including thrombotic thrombocytopenic purpura, HELLP syndrome, and acute fatty liver of pregnancy constitute additional diagnostic possibilities. Thus, the differential diagnosis of proteinuria in the pregnant patient with SLE is broad and sometimes cannot be determined without performing a renal biopsy. Furthermore, distinction between these different kidney disorders is critical, as the treatment of lupus nephritis usually entails initiation or escalation of immunosuppressive therapy whereas the other diagnostic considerations require alternate treatment modalities.1,21
The normal physiology of pregnancy further complicates the clinical assessment of renal function, as changes in renal hemodynamics that occur with pregnancy generally lead to an increase in glomerular filtration rate by 40% to 65%. 22 Given that creatinine production does not change, serum creatinine levels are therefore expected to decrease in pregnancy. Moreover, an increase in proteinuria can also be seen as a result of this increase in glomerular filtration rate, with up to 300 mg per day considered normal in pregnancy.22,23 These physiologic changes among pregnant women can make the diagnosis of a lupus nephritis flare that much more challenging.
In this instructive case series of 11 pregnant women with SLE, we demonstrated how information garnered from renal biopsy was used to direct treatment for each individual. The implications of this clinical experience are significant, inasmuch as medications used to treat lupus nephritis in pregnancy (i.e. corticosteroids, azathioprine, and/or calcineurin inhibitors) are themselves not without risk.1,18,24 Recently, Ritchie et al. reported that the two leading causes of maternal death among pregnant women with SLE were infection and disease activity, thus highlighting how immunosuppressive therapy should be used judiciously during pregnancy. 25 All but one patient in this informative case series underwent a change in management as a result of findings on renal biopsy.
Strategies for renal biopsy in the pregnant patient with systemic lupus erythematosus
The use of omega-3 fatty acids can be associated with an increased risk for bleeding. 29
RBC: red blood cells; hpf: high-power field; anti-dsDNA: anti-double-stranded DNA.
It is also our clinical experience that renal biopsy should be avoided after 28 weeks’ gestation, as the potential risks to the fetus are likely too great at this stage of pregnancy. We have previously demonstrated that the risk of major complication (defined as bleeding necessitating packed RBC transfusion, intervention by radiology or surgery, hemodynamic instability requiring the use of vasopressors and/or transfer to a higher level of care, or death) in a cohort of 219 patients with lupus was low, at 2.7%, which is comparable to that of patients without lupus who undergo kidney biopsy. 15 In pregnancy, there is the added concern of potential harm to the fetus. If severe bleeding were to occur as a result of the biopsy, not only would blood flow to the placenta be compromised, but shielding the fetus from radiation exposure during angiography could also prove challenging. Fortunately, previously published reports have shown that the overall complication rate of performing a renal biopsy in the pregnant woman is low if performed prior to 28 weeks’ gestation. Packham and Fairley reported a complication rate of 4.5% among 111 renal biopsies in 104 pregnant women, four of whom had lupus nephritis. 30 Although their definition of complication was broader (clinical perirenal hematoma, macroscopic hematuria, loin pain, and/or insufficient tissue sample), the one woman who developed a perirenal hematoma ultimately required seven units of blood transfusion, went into premature labor, and delivered an infant who died 48 hours later. She did not, however, have lupus nephritis. It should be noted that in this case series, only 10 biopsies were performed under ultrasound guidance, which is now the standard of care. 30 In another case series of 15 patients of whom eight had lupus nephritis, no complications were noted other than one case of gross hematuria. 31 Finally, Day et al. reported that among 20 women (seven with lupus nephritis) undergoing a renal biopsy during pregnancy, only one had a complication and this was post-biopsy hematuria that spontaneously resolved. 32 These studies provide support that renal biopsy is safe in pregnancy, particularly when performed at 28 weeks’ gestation or earlier. More recently, Piccoli et al. published a systematic narrative review on the risks and benefits of renal biopsy in pregnancy that included the above three studies, among others. While they reported a slightly higher risk of complications at 7% with possible peak at 25 weeks’ gestation, there was a high degree of heterogeneity among the studies included in their review. 33 Regardless, the decision to pursue a renal biopsy in the pregnant woman with SLE should not be taken lightly. Thus, in patients with thrombocytopenia (<100,000 platelets) refractory to therapy and/or coagulopathy requiring anticoagulation, empiric therapy should be considered instead, as the risk of biopsy is likely too high in these patients to undertake the procedure with sufficient safety.
The following limitations are acknowledged. First, the sample size in absolute terms is small, with only 11 patients. Despite its small size, however, to our knowledge this is the largest single case series to date of patients with lupus undergoing a renal biopsy during pregnancy. In addition, each case illustrates how renal biopsy helped to determine not only the severity and class of lupus nephritis but also guided treatment. Second, the data were collected retrospectively. Third, we do not have data on pregnant women with SLE who were either not offered, or who were offered but declined, a renal biopsy with which to compare outcomes to these 11 patients. Fourth, the cohort was assembled at a tertiary medical center with specialists both in SLE and lupus nephritis. As such, the patient population seen at our institution might be different, and possibly constitute the more severe end of the disease spectrum, compared to other medical institutions. Despite these limitations, our study provides valuable insight on the role of renal biopsy in managing lupus nephritis during pregnancy.
In conclusion, the management of lupus nephritis in pregnancy is complicated because one must consider the consequences to not only the mother but also the fetus. While most studies recommend that a patient with SLE be in remission for at least six months prior to conception, it remains possible to develop a flare of lupus nephritis during pregnancy.1,24 When this occurs, renal biopsy can be an invaluable tool in helping determine the underlying class and severity of lupus nephritis, and more importantly, guide management. In our case series, these histologic findings, derived from the pregnant patient with lupus, enabled a greater understanding of the active disease processes that affected the mother and allowed for a more informed therapeutic approach designed to optimize clinical outcomes. For these reasons, we firmly believe that renal biopsy should be a part of the evaluation process in pregnant patients with SLE who present with findings suggestive of active lupus nephritis.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement
The authors have no conflicts of interest to declare.