Over 80% of patients undergoing bariatric surgery are women, approximately half of whom are of reproductive age. The most common procedure in the United States is the Roux-en-Y gastric bypass. Small bowel obstruction is one of many recognized postoperative complications. For such a serious condition, this entity presents with remarkable subtlety and is easily misdiagnosed, particularly in pregnant women. The consequences of late recognition can be life-threatening to both mother and fetus. We aim to decrease preventable maternal and perinatal morbidity and mortality by revealing diagnostic and therapeutic missteps related to Roux-en-Y gastric bypass-associated small bowel obstruction.
Approximately 125,000 Americans undergo bariatric surgery annually, 80% of whom are women. Roux-en-Y gastric bypass (RYGB) accounts for over two-thirds of bariatric operations in the United States. Reproductive health benefits after surgery include improved fertility and fecundity. Pregnancy is associated with less frequent gestational diabetes, hypertensive disorders, and excessive fetal growth. Authors typically recommend deferring pregnancy for 12-18 months after surgery, the period of maximal weight loss, to reduce the theoretical risk of fetal growth restriction while allowing women to receive the procedure’s full therapeutic benefit. Therefore, most operative complications encountered by obstetrician-gynecologists develop remote from surgery.
The majority of late sequelae represent mild nutritional deficiencies that readily respond to replacement therapy. In contrast, small bowel obstruction (SBO) is a well-recognized life-threatening late complication of RYGB. Case reports and series describing SBO during pregnancy are usually published in the surgical literature. Moreover, a recent review and a practice bulletin published by the American College of Obstetricians and Gynecologists on bariatric surgery and pregnancy specifically caution obstetricians regarding this complication. Nonetheless, potentially avoidable maternal and perinatal mortality associated with RYGB-related SBO still occurs. The purpose of this paper is to shed light on pitfalls associated with misdiagnosis of RYGB-associated SBO complicating pregnancy and reduce preventable morbidity and mortality.
Discussion
Unfamiliarity with late surgical complication of RYBG-associated SBO
Understanding RYGB-related SBO and its sequelae requires working knowledge of the procedure and resulting anatomy. A 15-25 mL proximal gastric pouch is created, which enhances weight loss by restricting the amount of food consumed. The jejunum is transected and the distal segment brought in either a retrocolic or antecolic location to be anastamosed to the gastric pouch, forming the Roux limb, which induces weight loss by malabsorption. The proximal jejunum is anastamosed to the Roux limb, forming the biliopancreatic (afferent) limb and common channel ( Figure 1 ). The surgery creates a number of mesenteric defects, which predispose to internal hernia formation. Although some surgeons advocate closing these defects to reduce hernia risk, the sutures may subsequently pull through the fatty mesentery without therapeutic benefit. In one report, there was no difference in the rates of internal hernia formation with vs without mesenteric closure.
SBO after RYGB occurs in up to 9.7% of patients. This complication is more frequent after laparoscopic (1.6-9.7%) than open RYGB (0-3.1%). Retrocolic Roux placement is associated with a greater likelihood (4.5-7.0%) of SBO than antecolic location (0.42-2%). SBO follows a bimodal temporal and etiologic distribution. Early postoperative obstructions most commonly result from adhesions. Not surprisingly, adhesions are rarely described in RYGB-associated SBO complicating pregnancy. Additional causes of early obstructions include ventral wall or trocar incisional hernias, jejunostomy stenosis, and hematobezoars.
Internal hernias are the leading cause of later SBO, generally observed between 1 and 3 years after surgery and are the most frequently reported cause of RYGB-associated SBO in pregnancy. Internal hernias usually occur at 1 of 3 locations: (1) between the Roux limb mesentery and transverse mesentery (Petersen’s space), (2) defect in the transverse mesocolon, and (3) jejunojejunostomy mesenteric defect ( Figure 2 ). Antecolic Roux placement prevents internal herniation through the transverse mesocolon, likely explaining the lower SBO frequency observed with this approach.
Less common causes of RYGB-associated SBO described in pregnancy include volvulus and intussusception. Volvulus is of particular concern since it can cause a “closed loop” obstruction ( Figure 3 ) involving the biliopancreatic limb and gastric remnant. Left untreated, the overdistended gastric remnant can blow out with a high mortality rate. Volvulus was implicated in 2 of 3 reported maternal deaths after RYGB-associated SBO. Despite these generalizations, RYGB-associated SBO can occur at any time from any cause.
The clinical presentation of RYGB-associated SBO without bowel necrosis is surprisingly subtle, considering the consequences of delayed diagnosis. Patients typically complain of left upper quadrant or epigastric pain, which is often crampy and progressive in severity and may radiate posteriorly. Characterization of the pain is remarkably consistent whether the patient is pregnant or not, and if pregnant, regardless of gestational age. Nausea with or without emesis is common. Vital signs are usually normal. Physical examination typically reveals normal bowel sounds and no distension, although there may be tenderness over the area of pain. Laboratory studies including white blood cell count and differential, serum electrolytes, and liver enzymes are usually within normal limits, or expected ranges for pregnancy. Tachycardia, abdominal distension, elevated white blood cell count, left shift, elevated liver enzymes, amylase, or lipase levels should raise concerns for bowel necrosis or perforation. Thus, the patient’s surgical history and high clinical index of suspicion are required for early detection of RYGB-associated SBO.
We believe that greater awareness among pregnant women and their obstetricians may improve the timely diagnosis and treatment of this surgical complication, and hopefully reduce the frequency of adverse maternal and perinatal outcomes. Ideally, women of reproductive age with a prior RYGB will have a preconception visit with an obstetrician. At that visit, we recommend that patients be informed of the possibility and presenting symptoms of RYGB-associated SBO in pregnancy. Once pregnant, this conversation should occur again at an early gestational age. We request patients present in person for early evaluation of abdominal complaints that raise concern for a possible SBO. The patient is actively encouraged to advocate for herself, relating her surgical history and specific concern for possible SBO. This is particularly important if her complaints are minimized by a provider unfamiliar with their significance. This approach has directly led to timely diagnosis and treatment in at least 2 pregnancies.
We suggest that obstetricians clearly document the RYGB history in a prominent place in the prenatal record and familiarize themselves with the clinical implications for pregnancy. Ensure that any pregnant patient with RYGB and abdominal complaints, no matter how trivial, is evaluated expeditiously in person. It is our belief that the obstetrician should have a very low threshold to promptly obtain bariatric surgical consultation. Some authors contend that evaluation by a bariatric surgeon is mandatory. We suggest that when possible, women with a RYGB receive prenatal care and plan to deliver in a facility with a bariatric surgeon available for consultation. Adherence to these principles was pivotal in the timely diagnosis and treatment of 4 pregnant women with RYGB-associated SBO.
Misdiagnosis of RYGB-associated SBO for an obstetric disorder
The generally mild signs and symptoms accompanying early RYGB-associated SBO may be mistaken for common and benign pregnancy-related complaints by 1 of 3 pathways. The SBO may be the only pathology, but is mistaken for an unrelated condition of pregnancy. Second, the SBO can provoke a secondary obstetric complication with signs and symptoms similar to or masking those of the obstruction. Finally, an obstetric disorder may occur coincidentally with a SBO, with clinical features overlapping or overshadowing those of the obstruction.
One report described a woman who was treated with antacids for epigastric pain attributed to pregnancy-induced gastroesophageal reflux. Progressive pain accompanied by nausea and vomiting precipitated more extensive evaluation. The correct diagnosis of a strangulated internal hernia was made at the time of cesarean delivery with concurrent exploratory laparotomy. Speculatively, this misdiagnosis could occur at any time during pregnancy. One can easily envision a first trimester SBO presentation being attributed to nausea and vomiting of pregnancy. Likewise, a SBO occurring in the latter half of pregnancy could speculatively be confused with severe preeclampsia, placental abruption, or uterine rupture.
Three reports describe preterm contractions coincident to women presenting for evaluation of third trimester abdominal pain. Two women additionally noted nausea and vomiting. Patients at 30 and 32 weeks underwent exploratory laparotomy and repair of internal hernias; both received postpartum tocolysis with resolution of contractions. In these cases, managing clinicians neither attributed the presenting complaints to the incidentally observed contractions, nor ignored the concurrent and more serious diagnosis of SBO. The remaining patient progressed in preterm labor to deliver vaginally at 36 weeks. When her presumably labor-related symptoms persisted and progressed postpartum to include abdominal distension and loss of bowel function, SBO was suspected. At surgery, a volvulus involving the jejunojejunostomy with bowel necrosis was repaired.
When a pregnant patient with a RYGB and abdominal complaints fails to respond as expected to appropriate treatment of the presumptive cause, she should be expeditiously reassessed, preferably by a bariatric surgeon. If SBO is not suspected or diagnosed until the bowel becomes ischemic or strangulated, resulting maternal compromise may be associated with fetal compromise, precipitating urgent cesarean delivery. If a woman with an RYGB and abdominal complaints undergoes cesarean, we recommend the bowel be completely inspected at that time by a bariatric surgeon or other surgeon experienced with complications of RYGB. These recommendations may prevent further diagnostic and therapeutic delays, including avoiding unnecessary reoperations.
Misdiagnosis of RYGB-associated SBO for a medical or other surgical disorder
SBO in pregnancy after RYGB can be confused with medical and surgical disorders for the same reasons that it is mistaken for obstetric complications. Efthimiou and colleagues reported a woman at 24 weeks’ gestation who underwent RYGB 9 years earlier, followed by 2 uncomplicated pregnancies. After 7 days of inpatient treatment for colicky abdominal pain, nausea, and vomiting diagnosed as gastroenteritis, persistent symptoms and clinical deterioration precipitated care transfer to her bariatric surgeon. Emergent surgery identified and repaired the suspected internal hernia. Unfortunately, refracting preterm labor after surgery resulted in delivery and neonatal death.
A patient with a history of RYGB and urolithiasis presented at 33 weeks with left flank pain and hematuria. Urology consultation was obtained, pain was treated with morphine, and an intravenous pyelogram showed no stones. As her pain progressed over the next 24 hours, bariatric surgical consultation was obtained. Emergent surgery confirmed and treated an internal hernia. The patient had an otherwise unremarkable postoperative and pregnancy course.
Obesity, female sex, pregnancy, and rapid weight loss after bariatric surgery are well-known risk factors for developing gallbladder disease. Some surgeons will perform a cholecystectomy at the time of RYGB to avoid subsequent gallstone formation. Other surgeons will only perform cholecystectomy in the presence of symptomatic gallstones. There is no consensus regarding these 2 approaches. Thus, cholelithiasis may be commonly observed in obstetric patients with a previous RYGB. Although RYGB-associated SBO may present similarly to biliary colic, and gallstones may be detected during evaluation of the symptomatic patient, we recommend maintaining SBO as the top differential diagnosis. Ideally, a bariatric surgeon will be involved in the initial evaluation of the patient. If cholecystectomy is performed, the bariatric surgeon should examine the bowel to rule out SBO as either the source of the abdominal complaint or a concurrent complication. The following 2 cases illustrate contrasting outcomes for this scenario.
Gazzalle et al described a 33-week pregnant patient with epigastric pain, nausea, and vomiting in the context of a RYGB 2 years earlier. Ultrasound examination of the gallbladder showed slight distension, significant wall thickening, cholelithiasis, and fluid in the right iliac fossa and hypochondrium. Despite continued symptoms the day after a laparoscopic cholecystectomy, she was discharged home. Within hours of seeking care at a different hospital, she progressed in preterm labor to deliver vaginally. Immediately postpartum, exploratory laparotomy revealed and allowed repair of a jejunal volvulus with 150 cm of necrotic bowel.
A 35-week pregnant woman with a RYGB presented with crampy upper abdominal pain, and guarding and distension on physical examination. Ultrasound showed free abdominal fluid and gallstones with normal bile ducts. Concurrent cesarean delivery and exploratory laparotomy were performed. An internal hernia and volvulus were reduced, with all bowel remaining viable.
Mechanical ampullary obstruction by gallstones is one of the most common causes of acute pancreatitis. Despite multiple risk factors, this disease is rare during pregnancy. If it is diagnosed in a pregnant woman with RYGB, we would first consider SBO. Elevated amylase or lipase levels suggest small bowel necrosis or perforation. A closed loop obstruction of the biliopancreatic limb can cause a secondary pancreatitis. Two case reports describe women at 30-31 weeks’ gestation with Roux-en-Y anatomy, epigastric pain, nausea, vomiting, and elevated serum amylase levels leading to the diagnosis of acute pancreatitis. After inpatient medical management of pancreatitis for 2 and 3 days, respectively, deteriorating clinical conditions precipitated transfers to other hospitals. Emergency laparotomy in the first patient followed intrauterine fetal death, revealing an internal hernia of the entire small bowel, of which 61 cm were necrotic and resected. The patient died 3 hours postoperatively. The other women also underwent emergency surgery showing a gangrenous and perforated volvulus of a closed loop from duodenum to ileum, edematous pancreas, and fat necrosis of the lesser omentum. Neonatal death occurred after delivery via concurrent cesarean. The patient died 4 days later.
Attempted medical management of RYGB-related SBO
Despite the appropriateness of medical management of SBO in certain circumstances, RYGB-associated SBO generally requires surgical exploration for diagnosis and treatment. Even if the SBO responds to conservative measures of bowel rest, nasogastric decompression, and intravenous fluids, common causes of SBO such as adhesions, mesenteric defects, and stenoses remain, leaving the patient at risk for recurrent and potentially catastrophic SBO. Importantly, a nasogastric tube cannot decompress the isolated gastric remnant and biliopancreatic limb. In the face of a closed loop obstruction, small bowel ischemia may occur within as few as 6 hours, followed by necrosis and perforation or gastric remnant blowout, with lethal consequences.
Loar and colleagues reported a pregnant woman at 25-6/7 weeks with a prior laparoscopic RYGB and possible SBO. A computed tomography (CT) scan showed distension of the entire small bowel and colon with air fluid levels. She was treated with bowel rest and nasogastric decompression. After several days she experienced black emesis and melanotic diarrhea. Flexible sigmoidoscopy showed no obstruction and symptoms improved. At 26-5/7 weeks, the patient underwent emergent cesarean through a Pfannenstiel incision for premature ruptured membranes and umbilical cord prolapse. The authors noted that “healthy loops of small bowel were visualized with no evidence of intra-abdominal infection,” however, the authors did not describe the degree of exploration and examination of the bowel in greater detail. On the third postoperative day, marked clinical deterioration occurred with multiorgan failure and septic shock. Exploratory laparotomy identified necrosis of the entire small bowel and right colon, midgut volvulus about multiple adhesions, and distal small bowel perforation. The patient died after life support was discontinued. The infant survived.
When surgery is performed for possible RYGB-associated SBO, the entire pelvis and abdomen is explored for additional surgical pathology. The entire small bowel must be inspected. Adhesions should be lysed, mesenteric defects closed, stenoses corrected, and obstructions relieved. Many bariatric surgeons approach this procedure laparoscopically, even during pregnancy. Guidelines regarding nonobstetric surgery in pregnancy and laparoscopic surgery in pregnancy are available elsewhere and are of particular importance to the obstetrician caring for affected women.
Delayed or missed diagnosis of RYGB-related obstruction related to medical imaging
Imaging in women with RYGB and abdominal complaints is only recommended if the diagnosis of SBO is uncertain, otherwise the time taken will only delay needed surgery. As noted by Higa et al, the “work-up should include noninvasive studies as indicated, but is not complete unless an exploration is performed.” Moreover, a false-negative study may conceivably defer potentially life-saving surgery.
There are a number of ways to potentially minimize imaging related missteps. First, obtain bariatric surgical consultation before ordering imaging. If imaging is ordered, ensure that the interpreting radiologist is clearly informed of the patient’s RYGB history, symptoms, and suspected diagnosis. Otherwise, subtle findings suggesting obstruction may be missed. Specifically, we suggest noting this clinical information in the written request for imaging.
Choose the imaging modality with the greatest SBO detection rates in bariatric patients. Plain abdominal radiographs are most often normal. CT with contrast is the most sensitive imaging modality, although 1 early report noted a 20% false-negative rate. Subsequent studies showed 92-100% detection rates when images were interpreted by bariatric surgeons, experienced radiologists, or both ( Figure 4 ). Of the 13 reported pregnant women with surgically proven RYGB-associated SBO who underwent preoperative CT imaging, signs of obstruction were prospectively noted in 12 (92%) patients. There should be no hesitancy to obtain a medically indicated CT in a pregnant women because of concerns over fetal radiation exposure. Magnetic resonance imaging is not well-studied in this situation, although there is one case report of a successful diagnosis in pregnancy. Finally, we caution against allowing incidental positive findings to divert attention from more concerning inciting pathology, such as SBO.
