Key Abbreviations
Amniotic fluid AF
Amniotic fluid index AFI
Amniotic fluid volume AFV
Antiphospholipid syndrome APS
Intrauterine fetal death IUFD
Intrauterine growth restriction IUGR
Maximum vertical pocket MVP
Neonatal intensive care unit NICU
Odds ratio OR
Perinatal mortality rate PMR
Premature rupture of the membranes PROM
Twin-to-twin transfusion syndrome TTTS
Overview
Abnormalities of amniotic fluid volume (AFV) raise the concern for an underlying fetal or maternal complication during pregnancy or fetal/neonatal compromise. The perinatal mortality rate (PMR) approaches 90% to 100% with severe oligohydramnios in the second trimester and can exceed 50% with significant polyhydramnios in midpregnancy. Although these two extreme conditions are rare, other less drastic examples are more common and can impact pregnancy outcome. Efforts to study abnormalities of amniotic fluid (AF) are complicated by the fact that despite over 30 years of research in various human and animal models, little is known about the processes involved in normal AFV regulation. Many of the disease states associated with the extremes of AFV are better understood than the physiologic processes that maintain the normal state.
This chapter explores what is known about the normal mechanisms that effect the production and removal of AF, including fetal urination, swallowing, lung liquid, and intramembranous absorption. The normal changes in AFV and composition across gestation are reviewed, as well as AFV abnormalities that include oligohydramnios and polyhydramnios, along with the possible underlying causes and treatment modalities.
Amniotic Fluid Volume
Attempts to measure true AFV are difficult because of obvious limitations. To measure the actual volume of AF, an inert dye must be injected into the amniotic cavity via amniocentesis, and samples of amniotic fluid must be obtained to determine a dilution curve. Although the dye injection technique is considered the gold standard for determining actual AFV and is compared with other methods of estimating AFV, such as ultrasound, it is impractical to utilize an invasive test to assess AFV in clinical practice.
Despite these limitations, Brace and Wolf identified all published measurements of AFV in 12 studies with 705 individual AFV measurements; for each week of gestation, wide variation was seen in AFV ( Fig. 35-1 ). The greatest variation occurred at 32 to 33 weeks of gestation, when the normal range was 400 to 2100 mL (5th to 95th percentile); this represents a wide normal range. One of the most interesting findings of the Brace and Wolf study is that from 22 through 39 weeks of gestation, the average volume of AF (black dots on Fig. 35-1 ) remained unchanged despite an increase in fetal weight from about 500 g to 3500 g, a sevenfold increase. Using the dye dilution technique, others found that the normal range was less and the peak AFV occurred at 40 weeks of gestation, instead of 30 to 38 weeks. These studies suggest the AFV is closely regulated throughout pregnancy.
Ultrasound Assessment of Amniotic Fluid Volume
Ultrasound has largely replaced clinical assessment of AFV based on the Leopold maneuver or fundal height measurements. However, AF disorders should be suspected when the uterus measures too large or too small for the gestational age. Polyhydramnios may be present if the maternal uterus is large for gestational age (LGA) or if the fetus cannot be easily palpated or is ballotable. The diagnosis of oligohydramnios is a consideration when the fundal height is small for gestational age (SGA) or the fetus is easily palpated.
Early ultrasound estimations of AFV were made by measuring the maximum vertical pocket (MVP) of AF. Chamberlain and colleagues and Mercer and colleagues found that perinatal morbidity and mortality rates were increased with an MVP of less than 1 cm and 0.5 cm, respectively. These lower values of the MVP identified at-risk fetuses, but the low sensitivity for identifying the majority of pregnancy complications associated with oligohydramnios was unacceptable and prompted other investigators to select higher cut-off values.
Subsequently, Phelan and others proposed a four quadrant assessment of AF referred to as the amniotic fluid index (AFI). After 20 weeks, the uterus is divided into four equal quadrants, as shown in Figure 35-2 . The deepest pocket of AF is measured in each quadrant, making sure that the ultrasound transducer is perpendicular to the floor and that fetal body parts and umbilical cord do not interfere with the vertical measurement ( Fig. 35-3 ). The sum of the MVP in each quadrant equals the AFI. Moore and Cayle performed a cross-sectional study of 791 normal pregnancies; the 5th and 95th percentile of the AFI varied for each gestational age. At the 95 th percentile, the AFI at 35 to 36 weeks’ gestation was 24.9 cm, and it was 19.4 cm at 41 weeks’ gestation. The variation in the AFI at the 5th percentile was less than that of the 95th percentile, but it still varied by as much as 2.5 cm. Finally, the investigators reported the interobserver and intraobserver variation to be 3.1% and 6.7%, respectively, which is acceptable for this commonly performed procedure. Comparing the ultrasound estimation of the AFV by the AFI ( Fig. 35-4 ) with the actual measured volume (see Fig. 35-1 ) demonstrates very similar appearing curves.
Studies that have compared estimates of AFV by ultrasound (MVP and AFI) with actual measurements taken by the dye-dilution technique demonstrate that MVP and AFI measurements are poor predictors of actual AFV. Dildy and colleagues found that the AFI overestimated the actual volume in 88% of cases at lower volumes, and it underestimated the actual volume in 54% of cases at higher volumes. However, this difference should not alter clinical practice. Magann and colleagues reported a sensitivity of 10% (specificity 96%) for an AFI measurement of less than 5 cm (oligohydramnios), and the sensitivity was 5% (specificity 98%) for MVP up to 2 cm. For cases of suspected polyhydramnios, an AFI greater than 20 cm had a sensitivity of 29% (specificity 97%), as did an MVP of greater than 8 cm (specificity 94%). The MVP method had fewer false-positive tests compared with the AFI. Based on these findings, the authors concluded that the MVP is superior to the AFI. In contrast, Moore found the AFI superior to the MVP for identifying cases of oligohydramnios but found the two methods similar at predicting polyhydramnios. In a recent review, Moise found that the MVP was superior to the AFI in diagnosing oligohydramnios using an MVP less than 2 cm. Although the MVP appears to be the preferred method to diagnose oligohydramnios near term, the vast majority of research on ultrasound measurement of AFV utilizes the AFI.
Differences in ultrasound technique, specifically the pressure of the transducer on the maternal abdomen, can affect the accuracy of the ultrasound measurement of AF. Low pressure can result in a higher AFI, compared with moderate pressure, whereas high pressure on the maternal abdomen can result in a decrease in the AFI measurement. Despite overwhelming evidence that current ultrasound methods are poor predictors of abnormal AFV, clinical practice continues to include the use of weekly or twice-weekly ultrasound estimates of AFV to assess fetal status.
For many years, investigators have tried with mixed success to demonstrate the utility and applicability of ultrasound estimation of AFV in relation to perinatal outcome. Early work by Chamberlain and colleagues found that when the MVP was less than 1 cm, a marked increase was seen in perinatal morbidity and mortality that persisted even after correcting for birth defects.
Amniotic Fluid Formation
Fetal Urine
The main source of AF is fetal urination. In the human, the fetal kidneys begin to make urine before the end of the first trimester, and production of urine increases until term. Many different animal models have been used to study fetal urine production. The fetal sheep provides an excellent model for comparative human study owing to its similar fetal weight at term, its sufficient size to allow catheter placement, and the fact that the sheep fetus has a low risk of premature labor after catheter placement. In the fetal sheep, urine production has been reported to be approximately 200 to 1200 mL/day in the last third of pregnancy. Efforts to measure human fetal urine production have been accomplished by ultrasound measurements of the change in fetal bladder volume over time. Wladimiroff and Campbell initially measured three dimensions of the fetal bladder every 15 minutes and reported a human fetal urine production rate of 230 mL/day at 36 weeks of gestation, which increased to 655 mL/day at term. Others found similar volumes using the same technique. Interestingly, using the same technique but measuring the change in volume every 2 to 5 minutes, Rabinowitz and colleagues found fetal urine production to be much greater than previously predicted (1224 mL/day). This was confirmed by three-dimensional (3-D) ultrasound and computer modeling. Fetal urine-production rates from several studies are shown in Figure 35-5 . Human fetal urine-production rate appears to be approximately 1000 to 1200 mL/day at term, which suggests that the entire AFV is replaced more frequently than every 24 hours.
Lung Liquid
Fetal lung liquid also plays an important role in AF formation. For years, it was presumed that actual movement of AF into the fetal lungs occurred; however, recent data offer no support for this concept. In fact, throughout gestation, the fetal lungs produce fluid that exits the trachea and is either swallowed or leaves the mouth and enters the amniotic compartment. In fetal sheep experiments, the lungs have been reported to produce volumes of up to 400 mL/day, with 50% being swallowed and 50% exiting via the mouth. Although we do not have direct measurements in humans, the presence of surfactant in the AF near term provides evidence for the outward flow of lung liquid. During normal fetal life, fetal breathing movements provide a “to-and-fro” movement of AF into and out of the trachea, upper lungs, and mouth with a net outward movement of fetal lung liquid into the AF.
Amniotic Fluid Removal
Fetal Swallowing
In the human, fetal swallowing begins early in gestation and contributes to the removal of AF. In the fetal sheep, swallowing has mostly been measured in the latter half of pregnancy and appears to increase with increasing gestational age. Sherman and colleagues reported that the ovine fetus swallows volumes of 100 to 300 mL/kg/day in episodes that last 2 minutes. In the term ovine fetus, that volume represents a daily swallowing rate of 350 to 1000 mL/day for a 3.5-kg fetus. This is obviously more than the adult sheep, which drinks 40 to 60 mL/kg daily.
Many different techniques have been used to determine swallowing rates in the animal model, including repetitive sampling of injected dye and actual flow probe measurements. For obvious reasons, actual measurement of human fetal swallowing is much more difficult. Human fetal swallowing was studied in the distant past by injecting radioactive chromium–labeled erythrocytes and Hypaque (Amersham Health, Princeton, NJ) into the amniotic compartment, and swallowing rates of 72 to 262 mL/kg/day were found in studies in the 1960s. Abramovich injected colloidal gold into the human amniotic compartment and found that fetal swallowing increased with advancing gestational age. He also found similar swallowing rates to those previously reported. Obviously, similar studies could not be performed today, but this information is helpful in our understanding of human fetal swallowing. Fetal swallowing does not remove the entire volume of fluid that enters the amniotic compartment from fetal urine production and lung liquid; therefore other mechanisms of AF removal such as intramembranous absorption must occur.
Intramembranous Absorption
One major stumbling block to the understanding of AFV regulation was the discrepancy in volume between fetal urine and lung-liquid production and its removal by swallowing. If the measurements and estimates of AF production and removal were accurate, at least 500 to 750 mL/day of excess fluid would enter the amniotic compartment, which should have resulted in acute polyhydramnios. This does not occur under normal conditions (see Fig. 35-1 ); therefore a second route for AF removal has been suggested, namely the intramembranous pathway. This process describes the movement of water and solutes between the amniotic compartment and the fetal blood, which circulates through the fetal surface of the placenta. The large osmotic gradient ( Fig. 35-6 ) between AF and fetal blood provides a substantial driving force for the movement of AF into the fetal blood. Intramembranous absorption has been described in detail in the fetal sheep and has also been demonstrated in the rhesus monkey fetus. Several anecdotal studies suggest that intramembranous absorption also occurs in humans. Heller and Renaud and colleagues each injected labeled amino acids into the amniotic compartments of women, who were shortly thereafter delivered by cesarean section. Both groups found high levels of the amino acids concentrated in the placenta within 45 minutes of injection. They concluded that the amino acids had to have been absorbed by some route other than swallowing in order to explain the rapid absorption into the fetal circulation within the placenta. Intramembranous absorption could easily explain this movement. This route of absorption is now being actively investigated, and researchers have noted that 200 to 500 mL/day leaves the amniotic compartment under normal physiologic conditions. In addition, it has been reported that absorption through the intramembranous pathway can increase almost tenfold under experimental conditions in sheep. Figure 35-7 summarizes all currently identified avenues for fluid entry and exit from the amniotic compartment and measured or estimated volumes. The flow of fluid into and out of the amniotic cavity appears to be in a state of balance. Recent work on the mechanisms associated with intramembranous absorption suggests that four intramembranous transport mechanisms act in concert and include (1) a unidirectional bulk transport of AF and solutes out of the AF into the fetal circulation, (2) passive bidirectional diffusion of solutes, (3) passive bidirectional water movement, and (4) unidirectional transport of lactate into the AF. In spite of these new findings, overall AFV regulation still needs further investigation.
Oligohydramnios
The incidence of oligohydramnios varies depending on which definition is used; reported rates vary between 1% and 3%. The incidence of oligohydramnios is much higher (19% to 20%) among women undergoing antepartum testing for an underlying maternal or fetal indication. Three studies reported actual measurements of AFV for oligohydramnios from less than 200 mL to 500 mL. With the advent of ultrasound estimation of AFV, multiple thresholds have been reported. In clinical practice, an MVP less than 1 to 2 cm or an AFI less than 5 cm are commonly used as criteria for the diagnosis of oligohydramnios.
Chamberlain and colleagues reported a fiftyfold increase in PMR for pregnancies with an MVP of less than 1 cm. This report was instrumental in raising concern about the risk of stillbirth and neonatal mortality in the presence of oligohydramnios. A second, less often reported finding of that study was that 40% of the cases with oligohydramnios also had other confounding factors such as intrauterine growth restriction (IUGR), maternal hypertensive disorders, and congenital malformations. Other investigators have reported that oligohydramnios in the prolonged pregnancy has an increased risk of meconium staining of the AF, fetal distress in labor, and low 1-minute Apgar scores.
The PMR approaches 100% when the AFV is greatly decreased early in pregnancy, especially in midpregnancy. The cause of the decrease or absence of AF largely determines the perinatal outcome ( Box 35-1 ). With renal agenesis, virtually 100% of newborns die because of pulmonary hypoplasia. AF is required for fetal lung development during certain periods of early and mid gestation. If premature rupture of the membranes (PROM) results in a loss of all AF, neonatal survival will vary based on the gestational age when the membranes ruptured and whether intraamniotic infection was the cause of the membrane rupture. Oligohydramnios can also occur with maternal conditions such as hypertensive disorders or the antiphospholipid syndrome (APS). In these cases, if the fetus is large enough to survive outside of the uterus, there may be little impact on perinatal outcome other than the consequences of prematurity.
Fetal Conditions
- •
Renal agenesis
- •
Obstructed uropathy
- •
Spontaneous rupture of the membranes
- •
Premature rupture of the membranes
- •
Abnormal placentation
- •
Prolonged pregnancy
- •
Severe intrauterine growth restriction
Maternal Conditions
- •
Dehydration-hypovolemia
- •
Hypertensive disorders
- •
Uteroplacental insufficiency
- •
Antiphospholipid syndrome
A common incidental finding is the existence of a low AFI in an otherwise normal pregnancy. Because the diagnosis of oligohydramnios has been associated with poor perinatal outcomes, many women who are at or near term are sent to labor and delivery to be considered for induction solely because of a low AFI. Frequently, their cervical examination is unfavorable for induction, and an induction is attempted in spite of this; often this ends in a cesarean delivery for failed induction. Although the evidence for induction in the prolonged pregnancy is solid (see Chapter 36 ), the term or preterm patient with isolated oligohydramnios may not need immediate delivery. Having said this, borderline oligohydramnios has been found to be a predictor of SGA newborns with an increase in admissions to the neonatal intensive care unit (NICU) but no other significant morbidities. Lagrew and colleagues reported that 41% of women with oligohydramnios, as determined by the AFI, had a normal AFI 3 to 4 days later. They also found that a normal AFI measurement was valid for 1 week, which suggests that the test need not be repeated more often except in certain high-risk situations. Magann and colleagues examined 1001 high-risk women who underwent antepartum testing. They found that those with an AFI of less than 5 cm (19% of cases) had similar outcomes to those whose AFI was in the normal range, and these researchers concluded that an AFI of less than 5 cm was not an indication for delivery. Rainford and colleagues examined 232 women at more than 37 weeks of gestation who had an AFI of less than 5 cm (19%). They found outcomes to be no worse when compared with those whose AFI was in the normal range. In fact, the risk of meconium staining of the AF was found to be increased (35% vs. 16%) in the normal group. Finally, Casey and colleagues examined 6423 women at more than 34 weeks’ gestation with an AFI of less than 5 cm and found an increase in intrauterine fetal death (IUFD), admissions to the NICU, neonatal death, low birthweight (LBW), and meconium aspiration syndrome (MAS) compared with women with an AFI of greater than 5 cm. If the birth defects and IUGR were removed, no difference was seen in admissions to the NICU, neonatal death, or respiratory distress syndrome (RDS). This suggests that IUGR and birth defects contributed to the increased morbidity and mortality, not the oligohydramnios. All cases with oligohydramnios should be evaluated for evidence of IUGR and should be followed with antepartum testing.
Evaluation and Treatment of Oligohydramnios
When the diagnosis of oligohydramnios is made in the second trimester, it is vitally important to obtain a complete history and physical exam and to perform a targeted ultrasound to help identify a cause (see Box 35-1 ). The patient should be questioned for any history consistent with rupture of the membranes, leakage of clear or bloody fluid, or wetness of her underwear. If there is a question of possible rupture of the membranes (ROM), a sterile speculum examination should be performed in an attempt to obtain fluid that can be examined for evidence of rupture. Specific tests include evidence of ferning on microscopy examination, a neutral pH on nitrazine paper, and pooling of fluid in the posterior vagina. Commercial tests are also available that screen for ROM: these include AmniSure (Qiagen, Hilden, Germany) and ROM Plus (Clinical Innovations, Murray, Utah), which check for certain proteins from the AF in the vagina. These tests reportedly have a higher sensitivity and specificity than either the fern or nitrazine tests. Next, a targeted ultrasound should be performed to assess the amount of AF present; evaluate fetal anatomy, including the kidneys and bladder; and determine appropriate interval growth. If the fetus is normally grown with kidneys and bladder visualized, more often than not, the amniotic membrane has ruptured prematurely (PROM). If kidneys and bladder cannot be seen, the diagnosis is most likely renal agenesis. Renal agenesis is uniformly fatal, whereas PROM can have a reasonable prognosis if it occurs after fetal viability and if infection is not present.
Although severe oligohydramnios has an increased PMR later in the third trimester, it is still not as high as earlier in pregnancy. Other studies have reported similar increases in perinatal mortality associated with oligohydramnios, but most have not corrected for other underlying medical conditions. Because of the increase in perinatal morbidity and mortality associated with oligohydramnios in prolonged pregnancy, delivery is recommended (see Chapter 36 ). As discussed earlier, the patient who presents with isolated oligohydramnios in the third trimester may be a candidate for expectant management.
Several investigators have attempted to treat oligohydramnios with the oral administration of water in the hope of “hydrating” the fetus through the mother. Animal studies have demonstrated that a close relationship exists between the hydration status of the mother and the fetus. Attempts to dehydrate the mother have resulted in dehydration of the fetus, and in some cases vice versa. In human pregnancies, Goodlin and colleagues found that the maternal intravascular volume was low in cases of idiopathic oligohydramnios, and that by increasing the maternal intravascular volume, the oligohydramnios resolved. In a randomized study, the oral administration of water increased the AFI in women with oligohydramnios. The treatment group that drank 2 L of water within 4 hours of a repeat AFI measurement had a significantly greater increase in AFI on repeat testing (6.3 cm) than the control group (5.1 cm). A follow-up study of women with a normal AFI demonstrated that the amount of water ingested could increase or decrease their AFI. As illustrated in Table 35-1 , the AFI significantly increased in the oral hydration group compared with the control group. The control group was given what was thought to be a “normal” volume of water to drink, but the AFI actually decreased and urine osmolality increased. This suggests that the mothers were actually dehydrated during the control portion of the study. Both groups demonstrated that the AFI can be influenced by increasing or decreasing water intake orally.
