Gastrointestinal Diseases in Pregnancy

Changes to the Physical Exam During Pregnancy

Abdominal assessment during pregnancy is modified because the abdominal viscera are displaced by the expanding gravid uterus. For example, the location of maximal abdominal pain and tenderness from acute appendicitis migrates superiorly and laterally as the appendix is displaced by the growing gravid uterus (see “Appendicitis” below). A rigid abdomen with rebound tenderness remains a valid indicator of peritonitis during pregnancy, but abdominal wall laxity and the interposition of the gravid uterus between the appendix and the anterior abdominal wall in late pregnancy may mask the classic signs of peritonitis.

Changes to Laboratory Assessments During Pregnancy

Laboratory testing and results must also be considered through the lens of the pregnant population. The normal leukocytosis of pregnancy must be considered during a workup for infectious diarrhea. Similarly, the physiologic or dilutional anemia of pregnancy can confound the assessment of acute blood loss and hence anemia from an episode of hematemesis.

With pregnancy also comes a mild dilutional hypoalbuminemia and electrolyte changes, particularly mild hyponatremia (see Appendix 1 , “Normal Values in Pregnancy”). In addition, the alkaline phosphatase appears elevated during pregnancy due to a placental isoenzyme. The erythrocyte sedimentation rate (ESR) is physiologically elevated and is a less reliable monitor of inflammatory activity during pregnancy. In addition, the C-reactive protein (CRP) is elevated in pregnancy with a peak in the second trimester.

Stool studies are also commonly performed in the course of assessing GI diseases, including inflammatory bowel disease (IBD), although reference ranges for these are unchanged during pregnancy.

Endoscopy During Pregnancy

Endoscopy is often performed in the evaluation of abdominal symptoms in nonpregnant patients. The procedure is chosen based on the area that is thought to have the highest likelihood of pathology corresponding to the symptoms or to the patient's known history of prior disease or imaging findings.

To evaluate the upper GI tract—including the esophagus, stomach, and proximal duodenum—an upper endoscopy, also known as an esophagogastroduodenoscopy (EGD), is performed. To evaluate the colon and the terminal ileum, a colonoscopy is preferred. The small intestine beyond the reach of the EGD and proximal to the reach of the colonoscopy (which includes the distal duodenum, jejunum, and the majority of the ileum) can be evaluated via a video capsule or traversed with deep enteroscopy techniques such as a double-balloon enteroscopy as described further on. If patients require only rectal or left colon evaluation, they can undergo a flexible sigmoidoscopy.

Fetal monitoring should be considered at a gestational age when intervention for nonreassuring fetal status would be desired by the patient and obstetric provider (often at 24 weeks). Informed consent is particularly important during pregnancy. The patient should be informed of the benefits and apparent safety of endoscopy, but she should also be cautioned that the potential fetal risks are incompletely characterized. General measures to increase the risk/benefit ratio of endoscopy during pregnancy are listed in Box 53.1 . The available data on the safety of endoscopy during pregnancy are reviewed in the following text for each of the procedures.

Esophagogastroduodenoscopy

An EGD or upper endoscopy to evaluate the esophagus, stomach, and proximal duodenum does not require any preparation other than nothing to eat for a period of a few hours prior to the procedure. Anesthesia is routinely required for sedation, as upper endoscopy can cause discomfort that cannot usually be tolerated otherwise.

EGD is recommended during pregnancy for hemodynamically significant upper GI bleeding. It is rarely helpful in cases of nausea and vomiting or even in HG during pregnancy. EGD is reserved for atypical situations, such as severe and refractory nausea and vomiting accompanied by significant abdominal pain, hematemesis, or signs of gastroduodenal obstruction.

The benefits of EGD are generally thought to outweigh the risks in pregnant patients with strong indications. In a case-controlled study of 83 EGDs performed during pregnancy at a mean gestational age of 20 weeks, EGD did not cause any maternal complications and did not lead to labor; 70 of 74 patients (95%) delivered healthy infants. In this study, the four adverse pregnancy outcomes—three stillbirths or deaths from severe prematurity and one spontaneous abortion—occurred in high-risk pregnancies and were unrelated to EGD temporally or etiologically. No live-born infant had any congenital malformations. This study and two other studies of 60 pregnant patients undergoing EGD during the first trimester and of 30 pregnant patients undergoing EGD during the remainder of pregnancy suggest that EGD can be safe in pregnant patients with a strong indication for an EGD.

In contrast, a Swedish national registry cohort study included 3052 pregnancies exposed to endoscopies and showed that exposure to any endoscopy during pregnancy was associated with an increased risk of preterm birth or small for gestational age neonate but not of congenital malformation or stillbirth. The authors and others note, however, that although this is by far the largest number of endoscopies performed during pregnancy, the limitations of a national registry study include lack of knowledge of indication for the procedure, of the position of the patient at the time of endoscopy, the type of sedation used, bowel preparation used, duration of or therapeutics used during the procedure, and maternal comorbidities, all of which could play a role in the outcomes that were seen.

Colonoscopy

Colonoscopy is performed to evaluate the entire colon and terminal ileum. Colonoscopy should be considered, particularly during the second trimester, only for very strong indications, such as suspected colon cancer or a new diagnosis of IBD that would require a change in management . A polyethylene glycol (PEG) balanced electroyte solution is generally preferred over a sodium phosphate (Fleet phospho-soda) solution to cleanse the colon for colonoscopy because sodium phosphate is associated with renal failure or electrolyte abnormalities in dehydrated or otherwise at-risk patients. Colonoscopic tattooing of colonic lesions is performed using India ink or methylene blue in the nonpregnant population, but methylene blue should not be used during pregnancy because of potential fetotoxicity. Endoscopists should avoid abdominal compression during colonoscopy as well as prone or supine positioning in late pregnancy.

As described earlier with regard to EGD, the Swedish national registry cohort study included 1109 lower endoscopies and found that exposure to any endoscopy during pregnancy was associated with an increased risk of preterm birth or neonates that were small for gestational age. These limitations are as discussed in the preceding text. Prior to this study, the two largest studies comprised 20 and 8 pregnant patients. In the study of 20 pregnant patients, 16 colonoscopies were performed during the second trimester. Two minor maternal procedural complications occurred, both of which related to mild transient hypotension; however, these did not lead to clinical sequelae. Fetal outcomes were generally favorable: all infants were born healthy except for one spontaneous abortion and one infant born with congenital defects.

Flexible Sigmoidoscopy

If patients require only rectal or left colon evaluation, they can undergo a flexible sigmoidoscopy without anesthesia and may not need any preparation. Older case reports, a gastroenterologist survey, and the only case-controlled study to date have shown no significant endoscopic complications. An unsedated, unprepped flexible sigmoidoscopy is likely the safest endoscopic procedure in pregnancy, but it should be considered only for strong indications such as rectal bleeding suspicious for cancer, strictures, or suspicion for a new diagnosis of IBD that would result in changes in treatment.

Video Capsule Endoscopy

Video capsule endoscopy (VCE) employs a single-use pill-shaped camera that is swallowed or in some cases deployed via upper endoscopy to survey the small intestine. The most common serious complication of VCE in the general population is capsule retention, the risk for which may theoretically be increased during pregnancy due to the displacement and impingement of bowel loops. A deep enteroscopy or surgery is performed to retrieve the capsule when this occurs. These options are not possible or not ideal in the pregnant patient; thus VCE must be considered only for extremely strong indications.

Few data exist on the safety of VCE during pregnancy. A review of the literature revealed only two cases performed during pregnancy, both of which had favorable outcomes. VCE in pregnancy will require more data but may be considered if a very strong indication exists, thus balancing the risks of capsule retention.

Deep Enteroscopy

Deep enteroscopy is performed utilizing specialized scopes such as double- or single-balloon enteroscopes to evaluate the small intestinal mucosa. These procedures are generally much longer, require much higher doses of anesthesia, and call for more external mechanical force. To date no reports have been published on the use of deep enteroscopy during pregnancy.

Nausea and Vomiting of Pregnancy

Nausea and vomiting of pregnancy (NVP) occurs in more than 50% of pregnancies; in some regions of the world, the prevalence is greater than 90%. It most commonly begins in the early to middle part of the first trimester and has almost always had its onset by 9 weeks of gestation. The common term morning sickness is a misnomer, as symptoms are rarely isolated to the morning (1.8%). In fact, for most women they occur throughout the day (80%).

The Hook-Profet “embryo protection” theory suggests that the food aversion and expulsion are physiologic and in fact intended to avoid embryotoxic ingestions. This theory is supported further by the time course of NVP and its peak during the first trimester and resolution once most steps of embryogenesis have been completed.

The pathophysiology of NVP is unknown but is believed to be multifactorial. Human chorionic gonadotropin (hCG) has been proposed as an etiologic factor, because the serum hCG level peaks when HG is most severe, and the serum hCG level is higher in patients with HG than it is in other pregnant patients. In addition, conditions associated with higher levels of hCG—such as multiple gestations or gestational trophoblastic disease—are likewise associated with higher levels of NVP.

Hyperestrogenemia has also been proposed as a potential etiology, since elevated levels of estrogen have been associated with increased rates of nausea. In addition, women who have previously been more sensitive to nausea induced by estrogen-containing oral contraceptive pills have been found to be at higher risk for NVP. Evidence that nutrition and vitamin deficiencies may play a role is found in the relative reduction in NVP among women taking a prenatal vitamin at the time of conception as well as in the effectiveness of pyridoxine (vitamin B6) in reducing symptoms. Other postulated etiologic factors include Helicobacter pylori , gastric dysrhythmias, and hyperthyroidism.

Although the origins of NVP may be evolutionarily protective, it is important to avoid underestimating the significant adverse impact that these symptoms can have on a pregnant woman's quality of life. The Pregnancy-Unique Quantification of Emesis (PUQE) score was developed to quantify the severity of NVP and has been shown to correlate well with quality-of-life measures and overall health care costs.

Patients with NVP have only mild to moderate symptoms, no weight loss, and lack evidence of dehydration, vitamin deficiencies, and other nutritional deficiencies. Both NVP and HG are diagnoses of exclusion arrived at only after excluding other conditions by appropriate evaluation. However, in the absence of an evolution to true HG, most cases of NVP are diagnosed and treated based on symptoms alone.

The decision regarding whether to initiate behavioral or medical treatment for NVP is based on shared decision making between the patient and provider. Behavioral modifications include eating small, frequent meals to avoid gastric distention, which may trigger nausea. Some recommend a bland diet emphasizing salty crackers, soups, starches, and chicken while avoiding spicy, fatty, or fibrous foods. Other nonpharmacologic therapies that have been suggested include ginger (which has been shown to have benefit for nausea but not emesis) and acupressure (for which the most robust studies have shown no benefit). The choices for pharmacologic management are the same as for HG and are discussed further on.

Hyperemesis Gravidarum

HG is a severe pathologic form of NVP characterized by a greater than 5% loss of prepregnancy weight and otherwise unexplained ketonuria . HG is much less common than NVP and occurs in only about 0.5% of pregnancies. Like NVP, HG typically begins early in pregnancy; severe vomiting that begins after the first trimester is unlikely from HG and additional workup should be considered in such cases, including serum liver function tests, lipase levels, and imaging studies.

Aside from severe nausea and vomiting, symptoms of HG may include xerostomia, sialorrhea or ptyalism, and dysgeusia. Physical findings reflective of hypovolemia include dry mucous membranes, poor skin turgor, and orthostatic hypotension or hypotension. Serum electrolyte abnormalities include hyponatremia, hypocalcemia, and hypokalemia. Patients may demonstrate prerenal azotemia, and chronic vomiting of gastric contents may cause hypochloremic metabolic alkalosis. An increased hematocrit reflects hemoconcentration from hypovolemia. Patients with severe vomiting may develop abnormal liver function tests, particularly elevations of the serum aminotransferases. Patients may exhibit mild biochemical hyperthyroidism that manifests as low thyroid-stimulating hormone (TSH) levels as a result of elevated serum hCG levels, and inadequate nutrition may lead to vitamin or micronutrient deficiencies.

Patients with more severe NVP may benefit from drug therapy. Bendectin—which contains vitamin B6 and doxylamine, an antihistamine—was once commonly administered for NVP but was withdrawn in 1983 due to unproven allegations of teratogenicity. A subsequent meta-analysis reported no increase in the incidence of birth defects after in utero exposure to Bendectin. A delayed-release form of doxylamine-pyridoxine (Diclegis in the United States or Diclectin in Canada) is currently approved to treat NVP, and it has been shown in multiple randomized, controlled trials to be more effective than placebo in treating NVP. However, it is important to note that no studies have been performed to compare its effectiveness with the combination of doxylamine and pyridoxine taken concurrently as separate pills, and the cost of this option is likely to be significantly less than the combined delayed-release form.

If the initial approach with doxylamine and pyridoxine is ineffective, consideration can be given to an alternate antihistamine or to adding a dopamine antagonist. Dopamine antagonists—such as promethazine, prochlorperazine, and metoclopramide— are thought to work both by increasing gastric motility and by centrally suppressing emetic signaling, and they have been used with success. Fetal data are reassuring, with no increased rates of spontaneous abortion or malformation. Maternal side effects are rare but can include dystonic reactions or tardive dyskinesia, which may be irreversible even after discontinuation of the medication.

Finally, selective serotonin antagonists, such as ondansetron , have been increasingly used for NVP and HG due to their demonstrated greater effectiveness than both pyridoxine/doxylamine and metoclopramide in the control of emesis. These medications work by antagonizing peripheral vagal afferents that mediate the vomiting reflex. Although they are highly effective, there are important maternal and fetal considerations in counseling on the use of ondansetron during pregnancy. Maternal side effects include both QT prolongation and the potential for serotonin syndrome, especially when these agents are combined with other serotonin antagonists. Fetal concerns include possible associations with cleft palate and an increased rate of cardiovascular defects, although these findings were noted only in small studies and the absolute risk increase for both outcomes is very small.

Some patients will require inpatient management, particularly if the HG is refractory to even sublingual medications and they are unable to tolerate oral intake . Therapy for HG is initially focused on aggressive intravenous rehydration and the restoration of electrolyte deficiencies. Vitamins should be replaced and thiamine should be administered before dextrose to prevent Wernicke encephalopathy . After intravenous repletion, the diet is gradually advanced as tolerated, initially to salty fluids and then to a bland diet. Nutritional consultation is recommended. Nasoenteral or nasogastric feedings are useful alternatives in patients who cannot tolerate oral feedings. Parenteral nutrition is sometimes necessary. Pregnancy-specific considerations are discussed later in this chapter.

NVP and HG are classically associated with favorable pregnancy outcomes, although profound weight loss, lack of weight gain, and multiple hospital admissions can rarely predispose to low birthweight. Recurrence rates are estimated to be high (up to 80% in some studies) and usually with similar severity.

Gastroesophageal Reflux Disease

Heartburn occurs in 30% to 50% of pregnancies and is more common later in pregnancy. GERD also occurs more frequently among older patients, those with a higher body mass index, and patients of Caucasian or Native American descent. The development of GERD is due to several factors, including the following:

• 1

  Hypotonic LES due to estrogen and progesterone

• 2

  Delayed GI transit

• 3

  Gastric compression and increased intraabdominal pressure by the uterus

GERD manifests in pregnancy, as in nonpregnant patients, by heartburn, regurgitation, water brash, dyspepsia, hypersalivation, or, rarely, pulmonary symptoms such as chronic cough. Symptoms may be exacerbated by ingesting certain foods, such as acidic drinks or chocolate, or by recumbency. Complications of poorly controlled, longstanding GERD include esophagitis, dysphagia from reflux-induced esophageal stricture ( Video 53.1 ), and Barrett esophagus or adenocarcinoma. During pregnancy, GERD tends to be mild and of short duration, coincident with the pregnancy; therefore these complications are rare ( Fig. 53.1 ).

The diagnosis of GERD in adults is typically made based on symptoms. In the absence of alarming factors (GI bleeding, unexplained weight loss, dysphagia, odynophagia, a first-degree relative with GI cancer), no further testing is indicated. If alarming factors are present or the patient is refractory to treatment, an EGD can be considered. Esophageal manometry and ambulatory pH testing are not usually indicated during pregnancy as these diagnostic tests are reserved for symptomatic patients with normal EGD findings.

Treatment for GERD begins with lifestyle modifications, including the avoidance of caffeine, alcohol, smoking, and the use of nonsteroidal antiinflammatory drugs (NSAIDs), all of which are generally avoided in pregnancy in any case. Patients with GERD should elevate the heads of their beds and avoid wearing tight clothing. They should avoid recumbency after eating and discontinue oral intake 3 hours before bedtime.

Antacids—including calcium, magnesium, and aluminum-containing compounds —provide adequate control in up to half of women with GERD and are generally considered safe during pregnancy; alginates and agents containing sodium bicarbonate should be avoided.

Sucralfate is an aluminum salt that suppresses pepsin activity and is minimally absorbed in the GI tract. It has been studied in a randomized controlled trial and was found to be superior to lifestyle modifications alone in reducing symptoms of heartburn and regurgitation (90% improvement vs. 43% with lifestyle modification alone), with no adverse effects on neonatal outcomes.

Histamine ₂ receptor antagonists (H ₂ RAs) remain among the most commonly prescribed medications for pregnant women with symptoms of heartburn. Among the H ₂ RAs, the only medication that has been studied specifically in pregnancy is ranitidine. Twice-daily ranitidine has been shown to lead to a significant reduction in heartburn symptoms and antacid use compared with both baseline and placebo. Human data for other H ₂ Ras, including famotidine, have been reassuring, although animal data at supratherapeutic doses for cimetidine and nizatidine were of concern.

Proton pump inhibitors (PPIs) have been increasingly used for moderate disease because of accumulating evidence of relative fetal safety. The largest study to investigate the relationship of PPIs and fetal safety showed no significant increase in the rate of birth defects among the women exposed to PPIs versus those who had not, and this result likewise did not differ by type of PPI exposure. Although the US Food and Drug Administration (FDA) has historically rated omeprazole as FDA category C during pregnancy, a study in 2001 that involved 863 infants exposed to omeprazole during the first trimester revealed rates of stillbirth and congenital malformations comparable to those of unexposed controls. Thus the choice of this agent can be considered on par with other PPI medications.

Finally, it is worth noting that gastric acid secretions are necessary for the absorption of iron, and a large case-control study showed an association between the use of acid-suppressing medications (such as H ₂ RAs and PPIs) and iron deficiency anemia. This is particularly of concern for women who are already taking oral iron therapy for a diagnosis of iron deficiency anemia. Such women should be instructed to take the two medications at different times of day.

Peptic Ulcer Disease

Symptomatic peptic ulcer disease (PUD) is uncommon during pregnancy, and antecedent PUD often improves during pregnancy. However, the true incidence of PUD is unknown since the symptoms have significant overlap with GERD, and PUD can be definitively diagnosed only via endoscopy, which is less commonly performed during pregnancy.

Symptoms in pregnant women are similar to those in the nonpregnant population and include epigastric pain, which classically worsens after meals, anorexia, bloating, and belching. The most common cause of a duodenal ulcer is H. pylori infection. NSAIDs, which include aspirin, can cause both gastric or duodenal ulcers, although use of these is discouraged in pregnancy; therefore they are less likely to be the cause ( Fig. 53.2 ).

Definitive diagnosis of a gastric or peptic ulcer can be made by only performing an EGD, and endoscopic evaluation also permits testing for H. pylori ( Video 53.2 ). If H. pylori is diagnosed in nonpregnant patients, eradication therapy is recommended, for which the first-line regimen is quadruple therapy with a PPI, clarithromycin, amoxicillin, and metronidazole. Although PPIs, amoxicillin, and metronidazole have individually been shown to be safe for use in pregnancy, eradication therapy for H. pylori is typically deferred during pregnancy due to concerns for the safety of clarithromycin in pregnancy. If H. pylori is not identified and there is no concurrent history of NSAID use, consideration should be given to other rarer causes of PUD such as infectious causes, illicit substance use (cocaine, amphetamines), inflammatory or infiltrating diseases (sarcoidosis, Crohn disease [CD]), or acid hypersecretory states (Zollinger-Ellison syndrome).

Endoscopic therapy and/or surgery are rarely indicated for PUD during pregnancy unless for therapeutic purposes to control hemodynamically significant upper GI bleeding or other concerning symptoms. Both can be performed safely (see “Endoscopy,” earlier). However, data on the management of complicated PUD with these methods during pregnancy is limited to case reports.