Chapter 13: Acute Renal Failure in Pregnancy

Acute renal failure (ARF) can present in multiple complicated medical conditions but is predominantly acquired in hospitalized patients. This is not a rare medical condition, with as many as 5% of all hospitalized patients having some degree of ARF. With respect to the obstetric patient, however, ARF has become an uncommon complication of pregnancy in developed countries. It is estimated that the current incidence of ARF complicating pregnancy approximates 1 per 10,000 pregnant women. In 3 successive periods of 10 years between 1958 and 1987, Stratta and colleagues have reported a continued decrease in ARF requiring emergency renal dialysis from a rate of 1 in 3000 in 1958 to 1 in 15,000 pregnancies in 1987. They documented 81 cases of ARF in pregnancy of which 11.6% experienced irreversible renal damage. The majority of these cases resulted from complications of either severe preeclampsia or eclampsia. Possible explanations for this dramatic trend in this subset of patients includes ready availability of prenatal care and the legalization of medical abortions as main factors responsible for this reduction in the incidence of ARF requiring dialytic support. However, in underdeveloped countries of the world, ARF remains a frequent complication of pregnancy and has an attendant maternal mortality surpassing 50%. In these countries, ARF has a bimodal distribution with peaks in the first and third trimester, presumably reflective of the continued practice of illegal abortions, the lack of access to quality prenatal care, and the occurrence of preeclampsia or eclampsia. Whatever the explanation, ARF in pregnancy can be the result of any of the disorders, which lead to severe renal dysfunction in nonpregnant patients or may result from disorders that are unique to the pregnant condition.

An understanding of the dramatic changes that occur normally in renal architecture, function, and blood flow is essential to the correct diagnosis and management of renal disease in the pregnant patient (Table 13-1).

Anatomic Changes

There is a marked increase in kidney size during pregnancy. This increase in size is primarily due to the increase in renal vascular volume and in the capacity of the collecting system. Hormonal influence is the most likely cause of the dilation of the urinary collecting system. In addition, increased production of prostaglandin E₂ (PGE₂) that inhibits urethral peristalsis, and mechanical obstruction by the enlarging uterus and distended iliac vessels (particularly on the right side) contribute to these changes that are evident as early as the first trimester and may continue for as long as 12 weeks postpartum.

Changes in Renal Blood Flow, Glomerular Filtration Rate, and Renal Tubular Function

Substantial increases in renal blood flow occur beginning early in the first trimester. This increase in renal blood flow is caused by both an increase in cardiac output and a decrease in renal vascular resistance. Renal vasodilatation is believed to be the most important mechanism for the dramatic rise in renal blood flow. Estimates of renal vascular resistance reveal a 50% decrease by the end of the first t-rimester. The underlying physiologic processes and cause of this pregnancy-related renal-vasodilation are not clearly understood. The large increases in the concentration of PGE₂ and prostacyclin (PGI₂) are believed to contribute to this effect, but they do not appear to be the only mechanism. Prolactin may be a hormonal mediator of renal vasodilatation in pregnancy. Estimation of renal plasma flow from p-aminohippuric acid clearance studies indicate an effective renal plasma flow of 809 mL/mm in the first trimester, 695 mL/mm in the last 10 weeks of pregnancy, and 482 mL/mm during the postpartum period. The most important consequence of this increase in renal blood flow during pregnancy is a dramatic rise in glomerular filtration rate (GFR). An increase in GFR of approximately 45% is seen as early as the end of the first trimester, and unlike renal plasma flow, this increase in GFR is maintained until term.

Despite this marked increase in GFR, the renal tubules are not only able to preserve normal sodium (Na⁺) balance but are able to achieve a cumulative Na⁺ retention of 500 to 900 mEq during the course of a normal pregnancy. Further, pregnant women maintain a normal sodium balance in settings when sodium intake is either increased or decreased. Pregnant women also maintain normal water balance and retain the ability to produce appropriately concentrated or dilute urine despite a significant alteration in the thirst and argenine vasopressin release thresholds during normal pregnancy.

Similarly potassium (K⁺) metabolism in pregnancy is unchanged. There is a physiologic requirement for the retention of approximately 350 mEq of potassium for the developing fetal-placental unit along with the significant expansion of maternal red cell volume. This increase in potassium retention occurs despite the dramatically elevated levels of aldosterone in the plasma of pregnant patients.

Pregnancy results in a respiratory alkalosis with a decrease of approximately 10 mm Hg in the arterial Pco₂. The slight respiratory alkalosis is compensated with an increased excretion of bicarbonate by the kidneys, resulting in a decrease in the plasma bicarbonate level to 18 to 20 mEq/L.

These “physiologic” anatomic and functional changes seen in pregnant patients’ renal systems have practical consequences. For example, dilated collecting systems make the diagnosis of an obstructive uropathy challenging. The increased GFR and tubular functions result in changes of the normal laboratory values for the commonly used serum tests of renal function. Examples include normal blood urea nitrogen (BUN) values in pregnancy average 8.7 ± 1.5 mg/dL and serum creatinine levels average 0.46 ± 0.13 mg/dL. Glucosuria is also common finding during normal pregnancy.

Acute renal failure (ARF) is a syndrome characterized by the rapid (hours to weeks) decline in renal function, resulting in the retention of nitrogenous waste products such as BUN and creatinine along with the inability to maintain normal fluid and electrolyte balances. Nonpregnant patients with ARF are often asymptomatic, but, when seen in pregnancy, ARF is rarely encountered in the absence of significant clinical findings or events complicating the gestation. ARF may complicate a host of diseases that, for purposes of diagnoses and management, are conveniently divided into 3 categories (Table 13-2).

• Diseases characterized by renal hypoperfusion in which the integrity of renal parenchymal tissue is preserved (prerenal azotemia, prerenal ARF). This is the most common form of ARF and has the best prognosis.

• Diseases involving renal parenchymal tissue (intrarenal azotemia, or intrinsic renal ARF).

• Diseases associated with an acute obstruction of the urinary tract (postrenal azotemia, postrenal ARF).

Most acute intrinsic renal azotemia is caused by ischemia or nephrotoxins, and is classically associated with acute tubular necrosis (ATN). Thus, in clinical practice, the term ATN is commonly used to denote ischemic or nephrotoxic ARF.

In prerenal ARF, impaired renal perfusion is the problem and this may be secondary to true intravascular volume depletion, decreased effective circulating volume to the kidneys secondary to impaired cardiac output, or due to agents that alter renal perfusion. Prerenal ARF may be rarely seen in the first trimester of pregnancy as a complication of severe hyperemesis gravidarum. In the second and third trimesters, severe blood loss as a complication of uterine hemorrhage in the antepartum, intrapartum, or postpartum periods is an important and not an uncommon cause of hypovolemia and subsequent ARF. It is important to remember that maternal bleeding may be concealed behind the placenta in some patients with serious placental abruption, and this condition may also be accompanied by varying degrees of a consumptive coagulopathy that can further complicate the degree of renal dysfunction.

Pregnancy is associated with a higher incidence of both bladder infections and pyelonephritis. This increased incidence of both upper and lower urinary tract infections, estimated to complicate approximately 2% of all pregnancies, is believed to result from both hormonal and mechanical changes that result in stasis within the urinary collecting system. Unlike their nonpregnant counterparts, pregnant women with pyelonephritis can manifest a substantial decrease in creatinine clearance and rarely may experience some degree of transient ARF. Patients with pyelonephritis in pregnancy, complicated with ARF, often have evidence of chronic renal parenchymal infection and recovery after appropriate antimicrobial therapy may be incomplete.

Severe preeclampsia and eclampsia account for the majority of ARF unique to pregnancy. Renal failure is unusual even with severe disease, unless there has also been significant blood loss with hemodynamic instablility or severe disseminated intravascular coagulopathy. Usually, the clinical picture of ARF seen in such patients is that of ATN, and typically resolves spontaneously within the first 2 weeks postpartum. Whereas in the short term, some of these patients with preeclampsia/eclampsia-induced ATN may require dialytic support, in general, even these more seriously affected patients experience complete recovery and have an excellent long-term prognosis. In those patients with ATN secondary to preeclampsia in whom the disease persists and long-term dialysis is necessary, frequently, it appears that pregnancy and/or preeclampsia have unmasked a chronic renal disorder or there has been some degree of renal cortical necrosis. When ARF develops in patients with severe preeclampsia or eclampsia, consideration should be given to effecting delivery as safely and as expeditiously as possible because maternal condition will frequently dramatically improve postpartum.

Preeclamptic/eclamptic patients, patients who develop placental abruption with or without coagulopathy, pregnancies in which there is a prolonged intrauterine fetal demise complicated with DIC, or women experiencing an amniotic fluid embolism are at an increased risk of developing renal cortical necrosis. Renal cortical necrosis, a pathologic process that destroys the renal cortex partially or completely while sparing the medulla, is heralded by the abrupt onset of oliguria or anuria that may be accompanied by flank pain, gross hematuria, and hypotension. This triad of anuria, gross hematuria, and flank pain is unusual in the other causes of renal failure in pregnancy. Renal cortical necrosis is not unique to pregnancy. However, pregnancy-related cases account for up to 70% of all cases. This entity is diagnosed based on the clinical presentation and can usually be established with ultrasonography or CT scanning of the kidneys. The characteristic findings are hypoechoic or hypodense areas in the renal cortex. There is no specific therapy that has been shown to be effective for patients with renal cortical necrosis with many women requiring chronic hemodialysis. Approximately 30% to 40% of patients experience at least partial recovery and have markedly compromised creatinine clearances.

Other clinical entities, more commonly associated with pregnancy and ARF with varying degrees of proteinuria, are seen in patients with HELLP (hemolysis, elevated liver transaminases, low platelets) syndrome, acute fatty liver of pregnancy, postpartum renal failure due to adult hemolytic uremic syndrome (HUS), and thrombotic thrombocytopenic purpura (TTP). Some investigators group these disease processes and preeclampsia under the same rubric of microangiopathic hemolytic processes of pregnancy, because they share a common histologic feature of anemia with evidence of red cell destruction on peripheral smears. The clinical manifestation and the nomenclature of the disease reflect the primary target organ. However, it is believed that there may be a unifying pathophysiologic process of profound vasoconstriction of arterioles resulting from an, as yet, unidentified “toxin” or mechanism likely involving the vascular endothelium. Consequently, in acute fatty liver of pregnancy, although this rare complication of pregnancy is associated with ARF in up to 60% of cases, the target organ is primarily the liver, with accompanying alterations of normal laboratory values such as transaminases, bilirubin, glucose metabolism, and various clotting factors (eg, antithrombin III, fibrinogen). Similarly, in postpartum renal failure resulting from HUS, the kidneys appear to be the target organs with resultant alterations in normal renal functions assays and, in comparison to renal dysfunction associated with HELLP syndrome or acute fatty liver of pregnancy, protracted ARF may result. Though HUS is generally a postpartum disease and HELLP syndrome is a form of severe preeclampsia usually confined to the late second or third trimesters, TTP almost always occurs antepartum and, although it may occur in the third trimester, many cases appear before 24 weeks of gestation. TTP is characterized by the pentad of microangiopathic hemolytic anemia, thrombocytopenia, renal insufficiency, fever, and neurologic abnormalities. The severity of renal dysfunction is usually mild particularly in comparison with the more striking neurologic involvement, fever, and thrombocytopenia.

Regardless of the underlying cause of ARF, the diagnosis usually hinges on serial analysis of BUN, creatinine, urinalysis, and urinary sediment analysis. It must be emphasized, however, that measurements of BUN and creatinine are relatively insensitive indices of glomerular function. For instance, the GFR may fall by 50% before the serum creatinine values rise. Conversely, a relatively large increment in the serum creatinine value reflects relatively small decrement in GFR in patients with preexisting chronic renal disease (Tables 13-3 and 13-4).

The clinical approach to the diagnosis of ARF involves a detailed history and physical; urinalysis; flow charts of serial blood pressures, daily weights, urine output, and intake, routine laboratory assessments; and special diagnostic procedures (see Table 13-2).

The clinical course of ATN can be divided into 3 phases, irrespective of the underlying disease process: the initiation phase, the maintenance phase, and the recovery phase. In the initiation phase, patients have been exposed to the “insult,” either ischemia or toxins, and while the picture is evolving, renal parenchymal injury is not yet established. The initiation phase is followed by the maintenance phase in which the renal parenchymal injury is established and GFR stabilizes at a value of 5 to 10 mL/min. Urine output is usually lowest during this period. This phase typically lasts 1 to 2 weeks, but may be prolonged for 1 to 11 months depending on the etiology before recovery, if any, is seen. Uremic complications usually arise during this phase. The recovery phase is heralded by a gradual restoration of the urine output and a decline in the serum BUN and creatinine. The recovery phase is frequently accompanied by a marked diuresis that will result in profound electrolyte disturbances unless closely monitored and corrected.

The initial therapy for acute renal failure (ARF) should be focused on correction of the specific etiology of the renal dysfunction, reestablishing normal fluid and electrolyte abnormalities, being cognizant of potential complications, and maintaining appropriate nutritional and medical requirements (Tables 13-5,13-6, 13-7). Prerenal azotemia is rapidly reversible upon restoration of renal perfusion. The source of the loss of fluid determines the composition of the replacement fluid. Hypovolemia caused by hemorrhage is ideally corrected with packed red blood cells. Isotonic saline is usually an appropriate replacement for plasma losses. Urinary or gastrointestinal fluids are generally hypotonic. Accordingly, initial replacement can be achieved with hypotonic solutions (eg, 0.45% saline) and subsequent replacement should be based on laboratory values. Serum potassium and acid base status should be monitored in all patients. In certain instances, use of insulin and glucose replacement may be necessary to lower the plasma potassium levels. Alternative therapies include sodium polystyrene sulfonate (Kayexalate) orally or as an enema or consideration of dialysis.

Fluid management is particularly difficult in patients with ARF resulting from underlying severe preeclampsia as they are not necessarily volume depleted and are at increased risk for developing pulmonary edema. The response to oliguria in these patients is often to give fluids to protect the kidney from developing ARF. But in doing so the risk of pulmonary and cerebral edema may actually be greater since ATF is relatively rare with severe preeclampsia except when abruption, hemorrhage, sepsis or DIC is superimposed. With delivery these patients will almost always diurese after a day or two. Pulmonary artery catheterization to measure pulmonary capillary wedge pressure (PCWP) has been advocated as an alternative to tailoring fluid management in these patients, as central venous pressure is often inaccurate in this setting. But in recent years this type of monitoring has largely fallen out of favor as the risk often outweighs the benefits in this setting. Similarly, measurements of urinary indices may not accurately reflect the volume status of the patient and allow the clinician to differentiate between prerenal azotemia and intrinsic renal disease.

The single most important factor in the appropriate management of intrinsic renal azotemia is the optimization of both cardiovascular function and intravascular volume. The clinician must have an appreciation of the potential complications seen in patients with intrinsic renal azotemia and adopt an aggressive preventative management protocol in order to attempt prevention of significant worsening of ARF. Pregnancy represents a unique challenge in such patients as there are both maternal and fetal concerns. Regardless of the underlying etiology of ARF, there are general guidelines and principles governing its management in pregnancy. As discussed, most cases of ARF complicating pregnancy develop in the postpartum period, thus obviating concerns about fetal well-being. However, in cases of severe preeclampsia/eclampsia or any of the microangiopathic thrombotic processes complicated by ARF, expeditious delivery of the fetus should be carefully considered with appropiate and judicious use of blood product replacement. Oliguria in the pregnant patient is frequently mismanaged. Fluid replacement must be carefully monitored. Indeed, worsening of the medical condition resulting from efforts to reestablish urine production is a common complication of ARF in pregnancy. As with aggressive but inappropriate fluid replacement, the use of potent diuretics, such as furosemide, may increase urine output transiently without correcting the underlying pathology and, in certain instances, such therapy may exacerbate the disease and worsen the clinical condition. One example of this is the oliguric phase of severe preeclampsia that is managed most appropriately with support and minimal fluid replacement rather than aggressively giving fluid and/or diuretics with the intent to simply establish a greater amount of urine production.

The available data support an early and aggressive use of hemodialysis in ARF which is not responding to fluid replacement and diuretics. This is also applicable to the pregnant patient although the use of acute hemodialysis during pregnancy is quite rare. However, following delivery, it may indeed be necessary to use this important clinical modality, particularly for patients experiencing protracted renal failure such as may be seen with HUS or acute cortical necrosis complicating pregnancy. Dialysis has been shown to decrease maternal mortality and accelerate recovery of renal function. Typically, initiation of dialysis should be considered when the BUN reaches 50 to 70 mg/dL or when the serum creatinine is greater than 6 to 7 mg/dL. These are lower levels than for the nonpregnant patient (Table 13-8).

For women requiring dialysis for renal failure, the reported pregnancy outcomes have been generally poor. In 1980, the European Dialysis and Transplant Association (EDTA) reported a 23% successful pregnancy outcome in women conceiving after starting dialysis. Subsequently, there has been reported a possible improvement with peritoneal dialysis in pregnant women requiring dialysis when compared with the EDTA report. The question of continuing a pregnancy in patients either already dialysis-dependent or who have moderate renal insufficiency at conception but experience a rapid deterioration of renal function during pregnancy leading to chronic dialysis is critically important in counseling women with end-stage renal disease. Pregnancy termination with the hopes of improving renal function does not appear to be a good option. In a recent series of 82 pregnancies in women with chronic renal insufficiency, Hayslett reported that only 15% of women had a return to baseline renal function at 6 months postpartum. More recently, in a review of 2300 dialysis units in the United States, Okundaye and colleagues report that over a 4-year period, 2% of the female dialysis patients of childbearing age became pregnant. Of the hemodialysis women, 2.4% conceived whereas 1.1% of the peritoneal dialysis women conceived. Of 184 pregnancies occurring in women who conceived after starting dialysis, the infant survival rate was 40.2%. Comparatively, the infant survival rate for the 57 pregnancies in women starting dialysis after conception was 73.6%. There was no difference in outcome between women receiving hemodialysis versus peritoneal dialysis and these investigators do not recommend switching from one modality to another during pregnancy. However, they do report a possible association between improved outcome and increased time and frequency of dialysis. The major complication remains premature birth, resulting in both significant mortality as well as morbidity in the surviving infants.

The role of renal biopsy in pregnancies complicated with ARF is not without controversy. Although such a procedure is valuable in many nonpregnant patients and plays a critical role in establishing the specific renal pathology, renal biopsy has potentially significant complications that may be of more concern in the pregnant patient. Consequently, renal biopsy should rarely, if ever, be performed in a patient with ARF while pregnant.