Gastrointestinal Disease in Pregnancy

Esophagus

The esophagus normally maintains an alkaline pH, mainly by peristalsis of orally produced saliva and by the lower esophageal sphincter’s preventing reflux of gastric contents. Van Thiel and colleagues found decreased tone of the gastroesophageal sphincter during pregnancy and with the use of progesterone-containing oral contraceptives. ^, However, Fisher and associates found no change in lower esophageal pressure in early pregnancy, but the sphincter’s responses to hormonal and physiologic stimuli were reduced. These investigators observed in opossums decreased responses of the lower esophageal sphincter circular muscle to gastrin and acetylcholine. Whether arising from reduced sphincter tone or response to stimuli during pregnancy, the potential for gastrointestinal reflux appears to increase compared with that of nonpregnant individuals.

Stomach

The stomach receives, mixes, and propels food. It has mechanical and secretory properties. Peristaltic contractions are triggered by the gastric pacemaker located between the fundus and the corpus on the greater curvature with a normal frequency of three cycles per minute. Gastric emptying is difficult to measure in pregnancy studies, as it involves the use of radioisotopes in test meals. Measuring serial blood levels of acetaminophen, an agent that is poorly absorbed from the stomach but quickly absorbed from the small intestine, Macfie and coworkers found that gastric emptying was not delayed in pregnancy. Another study found no differences in pregnant individuals compared with nonpregnant individuals, with the exception of delayed emptying 2 hours after delivery. Using hydrogen breath testing, there were no delays in gastric emptying, but there were increases in orocecal transit time as pregnancy advanced. Thus, contrary to traditional thinking, these indirect observations suggest that gastric motility is not altered in pregnancy. Studies of gastric acid secretory function during pregnancy have produced conflicting results, showing that acid production is decreased, unchanged, or increased.

Small Intestine

Small intestinal transit time is prolonged during pregnancy and during the luteal phase in nonpregnant women. It may be caused by elevated progesterone levels during pregnancy and their relaxing effect on smooth muscle. The teleologic advantage is presumed to be increased nutrient absorption. Vitamin B ₁₂ absorption and transluminal transport of some amino acids in animals are increased during pregnancy. ^, The weight of the small intestine is increased during pregnancy in animals and is probably accounted for by the larger mucosal surface, supporting the theory of increased absorptive capacity.

Colon

Functional changes of the colon in human pregnancy are inadequately studied. In pregnant animals, colonic transit time is increased, and the smooth muscle is less responsive to stimulation. ^, Because progesterone is known to inhibit colonic smooth muscle activity, it appears plausible that colonic motor activity is diminished during human pregnancy.

Epidemiology

Between 50% and 80% of pregnant women experience nausea and vomiting during pregnancy. Annually, 4 million women in the United States and 350,000 in Canada are affected. Symptoms usually begin between 4 and 10 weeks’ gestation and resolve by week 20. A minority (3%) of patients experience significant nausea and vomiting only in the third trimester. One-half of pregnant women experience nausea in the morning, whereas nausea peaks in the evening in 7% of patients, and 36% experience symptoms constantly.

The term morning sickness is misleading. A prospective study suggested that nausea occurring only during the morning affected 2% of patients, whereas 80% of symptomatic patients had nausea or vomiting throughout the day. Demographic factors associated with vomiting include first pregnancy, younger age, fewer than 12 years of education, nonsmoking status, and obesity. Nausea and vomiting are more common in Western countries and Japan and are rare in Africa, Asia, and Alaska (among Native Americans), indicating that factors other than pregnancy contribute to the pathogenesis.

When compared to asymptomatic women, those with nausea and vomiting are more likely to develop complications including preeclampsia, but they otherwise have favorable birth outcomes.

Pathogenesis

The etiology is unknown. Genetic influences have been suggested by the concordance of frequency in monozygotic twins and the fact that family members are more likely to be affected. ^{, ,}

Human chorionic gonadotropin (hCG) has been implicated in causing nausea and vomiting. The close temporal association of peak hCG levels and nausea and vomiting symptoms and the fact that conditions associated with elevated levels (e.g., female fetal sex, twins, hydatidiform moles) have higher rates of hyperemesis suggest a causal relationship. ^, Of 15 prospective comparative studies published since 1990, 11 have shown a significantly higher level of hCG in women with hyperemesis. Estrogens also can cause nausea and vomiting. Women who experience nausea with oral contraceptive use have a higher incidence of nausea and vomiting during pregnancy. There have been 17 studies of the relationship between nonthyroid hormones and nausea and vomiting, and only hCG and estrogen have shown a consistent association.

Prostaglandin E ₂ may contribute to this disorder, in that levels are elevated during symptomatic episodes. Helicobacter pylori has been implicated as an etiologic agent for hyperemesis gravidarum; although a recent review demonstrated an association, there was considerable heterogeneity among the pooled studies.

Physiologic causes of nausea and vomiting of pregnancy have historically been underemphasized compared with the psychosocial factors, although these issues do have an association. Some have proposed that nausea and vomiting of pregnancy is an evolutionary adaptation, protecting the woman and fetus from potentially harmful foods, and is therefore a healthy adaptive response.

Physiologic studies suggest that gastric dysrhythmias may have a role. Electrogastrographic recordings indicate that women with nausea and vomiting of pregnancy have alterations of the typical nonpregnant frequency and correlate temporally with the symptoms. Altered gastric motility and concomitant symptoms of nausea and vomiting in nonpregnant women appear to worsen in the luteal phase, and this effect has been shown to be improved with the use of oral contraceptives, suggesting a hormonal contribution.

Diagnosis

Evaluation of the patient is based mainly on history. A basic tool has been validated in scoring severity and quality of life ( Table 63.1 ). The nausea usually lasts all day. Pain is usually absent unless recurrent retching leads to abdominal and rib muscle strain. The physical examination is unremarkable unless the patient is severely dehydrated. Laboratory studies are usually not indicated or helpful. However, they are important in evaluating other potential causes that may mimic the condition, including hepatitis, pancreatitis, pyelonephritis, and uncontrolled diabetes.

Management

Therapy can be instituted before the development of symptoms. Two studies have suggested that multivitamins at the time of conception may reduce the incidence of nausea and vomiting. ^, Nonpharmacologic treatment may include ginger (500 to 1000 mg in divided doses daily). ^, A recent meta-analysis suggests that ginger was associated with improvement in those with mild symptoms. This spice accelerates gastric emptying and stimulates antral contractions. Six randomized controlled trials of vitamin B ₆ have suggested superior or equivalent efficacy in improving symptoms over placebo, with no reports of adverse events. Acupressure, acupuncture, and electrical stimulation have shown conflicting results. Acupressure is associated with symptom improvement, but the benefits of the latter two are unclear.

Pharmacologic therapy may be effective ( Table 63.2 ). Vitamin B ₆ in dosages of 10 to 25 mg taken three times daily is probably the best initial treatment and has a low potential of side effects. Doxylamine (10 mg) and vitamin B ₆ were used until the early 1980s with apparently good effect. The compound pyridoxine-doxylamine (Bendectin) was removed from the market in 1983, and there was a doubling in hospital admissions for nausea and vomiting. Systematic reviews have confirmed that pyridoxine-doxylamine is associated with improvement in patients with mild or moderate symptoms. For other pharmacologic treatments, including antihistamines and phenothiazines, there are no well-controlled trials. Promethazine, prochlorperazine, chlorpromazine, and trimethobenzamide have shown some clinical efficacy, but their safety in pregnancy has not been proven. Metoclopramide improves gastric emptying and corrects gastric dysrhythmias, and it has been safe and effective in small series. Ondansetron has become the most frequently prescribed antiemetic for the treatment of nausea and vomiting of pregnancy in the United States. Two recent publications have suggested that this medication has either no significant increase in adverse fetal outcomes or a low risk for congenital anomalies. Because of the conflicting data, the American College of Obstetricians and Gynecologists suggests that patients be counseled regarding this and use for <10 weeks individualized based on risks and benefits.

Epidemiology

Hyperemesis gravidarum occurs in approximately 0.5% of live births. ^, It is the second most common cause of pregnancy-related hospitalization and it can result in weight loss, electrolyte imbalances, and nutritional deficiencies. ^, It is more common in young, nulliparous women who are non-Caucasian. Factors shown to increase the rate of hospitalization include hyperthyroidism, psychiatric disorders, diabetes, gastrointestinal disorders, and asthma. Similar to the situation for nausea and vomiting, multiple gestations and singleton female fetuses were associated with a higher risk. Hyperemesis gravidarum is more likely in gestational trophoblastic disease, hydrops fetalis, and fetal karyotypic abnormalities, including triploidy and trisomy 21.

Symptoms typically begin during the first trimester. About 10% of patients are affected throughout pregnancy. This disorder leads to elective pregnancy termination in approximately 2% of affected pregnancies. Severe rare complications of hyperemesis include Wernicke encephalopathy, pneumomediastinum, and esophageal rupture. ^, Infants born to women with poor weight gain resulting from hyperemesis frequently have low birth weights. ^,

Diagnosis

There is no single accepted definition. Symptoms of hyperemesis gravidarum include dry mouth, sialorrhea, hyperolfaction, and altered taste. Physical examination usually reveals dry mucous membranes, poor skin turgor, and hypotension. Electrolyte abnormalities and ketonuria are often present. Severely affected individuals may have elevated levels of hepatic transaminases and abnormalities in renal function.

Management

Intravenous fluid resuscitation is the initial treatment for hyperemesis gravidarum. Up to 2 L of lactated Ringer’s solution should be infused over 3 to 5 hours and subsequently adjusted to maintain urine output greater than 100 mL/h. Thiamine (100 mg) should be given intravenously before the infusion of dextrose to avoid Wernicke encephalopathy. Electrolyte levels, including magnesium and ionized calcium, should be monitored regularly. Hyponatremia should be corrected with infusion of sodium containing fluids but not too rapidly because of concerns of central pontine myelinolysis. Furthermore, potassium should be added until hypokalemia is corrected.

Pharmacologic treatment includes the various antiemetics previously discussed. Less well-studied but proposed options in refractory cases may include transdermal clonidine and oral mirtazapine; however, fetal effects are largely unknown. As H. pylori infection appears to be more prevalent in patients with hyperemesis, evaluation for this may be reasonable. However, treatment benefit has not been adequately studied. Methylprednisolone and hydrocortisone have been used in recalcitrant hyperemesis with conflicting results. A randomized trial of oral methylprednisolone (initial dosages of 16 mg three times daily followed by a 2-week taper) resulted in a reduction in readmission to the hospital when compared with oral promethazine. Intravenous hydrocortisone 300 mg daily for 3 days followed by a taper over 1 week reduced vomiting episodes and subsequent readmission when compared with a regimen of metoclopramide. Another study compared the use of intravenous methylprednisolone followed by an oral prednisone taper plus promethazine and metoclopramide with the use of only the latter drugs and found no difference in the rates of rehospitalization (35%). Another randomized trial using oral prednisolone that was converted to intravenous hydrocortisone for no response was inconclusive, with a trend toward reduced vomiting, but patients had an improved sense of well-being and weight gain.

The mechanism by which steroids affect nausea is not clear. The nuclei of the solitary tract and raphe and the area postrema have glucocorticoid receptors and are known to participate in the regulation of nausea and vomiting responses. Corticosteroid exposure in the first trimester may be weakly associated with an increased rate of facial clefting, potentially accounting for one or two cases per 1000 treated women. Steroids therefore should be used with caution and avoided if possible during the period of fetal organogenesis.

Total parenteral nutrition has been used for hyperemesis gravidarum with reportedly improved outcomes. Significant complications of lipid emulsions have been reported, primarily with the use of cottonseed oil, and they have included placental infarction, ketonemia, and increased uterine activity. Subsequent case reports using soybean-based or soybean and safflower oil–based emulsions have not shown the same adverse events. There is an increased risk for complications directly related to therapy, including infusion catheter sepsis (25%) and venous thrombosis (3%). ^, Centrally placed venous catheters have higher morbidity rates (50%) compared with peripherally inserted lines (9%). Peripherally inserted central catheters have an overall complication rate of 50% and a severe complication rate of 22%. ^, The use of total parenteral nutrition should be considered a last resort.

Adverse outcomes for women with hyperemesis and low maternal weight gain compared with those for patients without hyperemesis include higher rates of small-for-gestational-age fetuses, low birth weight, prematurity, and 5-minute Apgar scores less than 7. Among women who experienced hyperemesis gravidarum in their first gestation, 15% to 19% will be affected in the second pregnancy, compared with 0.7% who had unaffected first pregnancies. ^,

Epidemiology

Heartburn is the most common gastrointestinal complaint in Western populations, with up to 20% of individuals having symptoms on a weekly basis. During pregnancy, up to 80% of women experience heartburn at some point, and 25% are affected daily. ^, The prevalence and severity progressively increase during pregnancy and resolve after delivery. ^, The prevalence of heartburn symptoms is 22% for patients in the first trimester, 39% for those in the second trimester, and 72% for patients in the third trimester. Factors associated with increased risk for heartburn symptoms were increasing gestational age, prepregnancy heartburn, and parity.

Pathogenesis

The exact cause of gastroesophageal reflux disease (GERD) during pregnancy is unknown. The resting lower esophageal sphincter (LES) pressure is decreased during pregnancy, and progesterone mediates LES relaxation in the setting of elevated estrogen levels. Other factors may include increased intraabdominal pressure, decreased gastric emptying (causing gastric fluid reflux), and ineffective esophageal motility to clear the esophageal acid.

Diagnosis

The symptoms of gastroesophageal reflux in pregnancy are the same as in the general adult population. They can include substernal burning, epigastric discomfort, dyspepsia, mild dysphagia, and regurgitation. Patients may have extraesophageal manifestations, such as hoarseness, chronic cough, chronic laryngitis, and asthma. Complications of GERD, such as gastrointestinal bleeding and esophageal strictures, are rare during pregnancy.

The diagnosis is usually made by clinical symptoms. Upper gastrointestinal endoscopy is rarely needed, except in exceptional cases of symptoms refractory to medical management. The endoscopic appearance of GERD ranges from no mucosal breaks to severe ulceration of the distal esophagus ( Fig. 63.1 ).

Management

Lifestyle and dietary changes should be initiated for mild symptoms. Patients should avoid eating within a few hours of bedtime, elevate the head of the bed 6 inches, sleep on the left shoulder, and avoid alcohol, cigarette smoking, caffeine, chocolate, and peppermint.

Medication is commonly used as first-line therapy for GERD in the general adult population, but in pregnancy, some drugs may theoretically be associated with teratogenic effects. Most medications for GERD are not routinely or rigorously tested in pregnant women, and their risks and benefits should be discussed with the patient. Because of concerns of teratogenicity, efforts should be made to avoid all unnecessary medications during the first 10 weeks of gestation. Table 63.3 shows the US Food and Drug Administration (FDA) classification of drug therapy for GERD and peptic ulcer disease during pregnancy. Most drugs are FDA category B.

Antacids are considered safe and can be expected to relieve symptoms in 30% to 50% of pregnant women. Sucralfate is a mucosal protectant and felt to be safe because it is locally acting and minimally absorbed. Metoclopramide can also be used to help gastric emptying and increase LES pressure, although the duration of use should be short because of concerns about tardive dyskinesia.

Histamine ₂ (H ₂ ) receptor antagonists appear to be safe and effective in pregnancy. ^, Proton pump inhibitors are more effective than H ₂ blockers for healing esophagitis and treating GERD, but they are less well studied in pregnancy than H ₂ blockers. Therefore usage should be reserved for patients not responding to more conservative measures. Omeprazole is categorized as a class C drug by the FDA because of embryotoxicity and fetotoxicity in animals and because of a few case reports of fetal birth defects in humans. However, a meta-analysis of the risk for fetal malformations during the first trimester found no increase with the use of proton pump inhibitors in general and with omeprazole specifically by pregnant women. A large study from Europe comparing several proton pump inhibitors, including omeprazole, also found no increase in major fetal anomalies compared with patients not exposed to proton pump inhibitors. A Danish cohort study specifically assessing proton pump inhibitor usage during pregnancy did not suggest an increased risk for malformations. However, when a group exposed 4 weeks before pregnancy was included, there was a small but statistical increase in risk (odds ratio [OR] = 1.39; 95% confidence interval [CI], 1.10 to 1.76). Subsequent reviews of the accumulated data suggest that proton pump inhibitor therapy predisposes the fetus to minimal risk. ^, There is some concern regarding of an association between gastric acid suppression therapy in pregnancy at development of childhood allergy and asthma, but their role requires further confirmation.

A recommended algorithm for treating GERD in pregnancy is shown in Fig. 63.2 . Esophagogastroduodenoscopy (EGD) is usually avoided unless the patient is failing empiric therapy or if there is hematemesis associated with hypotension or requiring blood products, need for endoscopic therapy, or for suspected gastric outlet syndrome secondary to PUD. If there is no improvement with lifestyle modifications, use of antacids or sucralfate should be considered. Persistent symptoms, especially after the first trimester, can be treated with H ₂ blockers. If symptoms persist despite H ₂ blockers, consideration can be given to using proton pump inhibitors or undergoing upper endoscopy to assess the cause of GERD.

Epidemiology

The incidence of peptic ulcer disease (PUD) during pregnancy is estimated from case reports and retrospective studies. Epidemiologic studies suggest a decreased incidence, and in those affected, there appears to be an improvement during pregnancy. Clark found that in 313 pregnancies occurring after a diagnosis of PUD, 45% of patients had no symptoms, 43% improved, and 12% experienced no improvement. Hypotheses for this improvement include physiologic changes, including increased plasma histaminase or estrogen levels leading to reduced gastric acid, increased gastric mucosal protection induced by progesterone, immunologic tolerance allowing H. pylori colonization without injury, and elevated epidermal growth factor levels stimulating mucosal growth. Maternal alterations such as avoidance of alcohol, smoking, and nonsteroidal antiinflammatory drugs (NSAIDs); increased rest; and improved nutrition have also been postulated.

Diagnosis

Symptoms of PUD are similar to those in the nonpregnant population and include epigastric pain, anorexia, postprandial nausea, and vomiting. Duodenal ulcer pain can occur several hours after a meal or at nighttime and usually improves with eating. Gastric ulcers may not follow the classic pattern of symptoms ( Fig. 63.3 ), and older patients who use antiulcer medication and NSAIDs may be asymptomatic. GERD symptoms differ in that there may be a sense of substernal chest burning and fluid reflux, and these symptoms are often worse after meals. Results of the physical examination are nonspecific, but they are helpful in ruling out complicated ulcer disease or other causes of abdominal pain, including pancreatitis, cholecystitis, and appendicitis.

In the nonpregnant population, EGD is preferable mainly because of its ability to allow accurate diagnosis of ulcers and to provide histologic specimens that are tested for H. pylori infection. EGD (discussed later) is relatively safe in pregnancy.