OBES SURG (2010) 20:386–392 DOI 10.1007/s11695-009-0013-6
CASE REPORT
Diagnosis and Treatment of Atypical Presentations of Hiatal Hernia Following Bariatric Surgery Brody A. Flanagin & Myrosia T. Mitchell & William A. Thistlethwaite & John C. Alverdy
Received: 22 April 2009 / Accepted: 7 October 2009 / Published online: 24 October 2009 # Springer Science + Business Media, LLC 2009
Abstract Bariatric surgery dramatically alters the normal stomach anatomy resulting in a significant incidence of hiatal hernia and gastroesophageal reflux disease. Although the majority of patients remain asymptomatic, many complain of severe heartburn refractory to medical management and additional highly atypical symptoms. Here, we describe the diagnosis and treatment regarding four cases of symptomatic hiatal hernia following bariatric surgery presenting with atypical symptoms in the University Hospital, USA. Four patients presented following laparoscopic Roux-en-Y gastric bypass or duodenal switch/pancreaticobiliary bypass (DS) with disabling and intractable midepigastric abdominal pain characterized as severe and radiating to the jaw, left shoulder, and midscapular area. The pain in all cases was described as paroxysmal and not necessarily associated with eating. All four patients also experienced nausea, vomiting, and failure to thrive at various intervals following laparoscopic bariatric surgery. Routine workup failed to produce any clear mechanical cause of these symptoms. However, complimentary use of multidetector CT and upper gastrointestinal contrast studies The authors of this paper disclose no grant support or other form of assistance in writing this manuscript. J. C. Alverdy (*) Department of Surgery, Section of General Surgery, The University of Chicago, 5841 S. Maryland Avenue MC 6090, Chicago, IL 60637, USA e-mail:
[email protected]
eventually revealed the diagnosis of hiatal hernia. Exploration identified the presence of a type I hiatal hernia in all four patients, with the stomach staple lines densely adherent to the diaphragm and parietal peritoneum. Operative intervention led to immediate and complete resolution of symptoms. The presence of a hiatal hernia following bariatric surgery can present with highly atypical symptoms that do not resolve without operative intervention. Recognition of this problem should lead to the consideration of surgery in cases where patients are dependent on artificial nutritional support and whose symptoms are poorly controlled with medication alone. Keywords RYGB . Roux-en-Y gastric bypass . GERD . Gastroesophageal reflux disease . DS . Duodenal switch . Pancreaticobiliary bypass Abbreviations RYGB Roux-en-Y gastric bypass GERD Gastroesophageal reflux disease DS Duodenal switch/pancreaticobiliary bypass CT Computerized tomography EGD Esophagogastroduodenoscopy UGI Upper gastrointestinal series CCK Cholecystokinin MRCP Magnetic resonance cholangiopancreatography 5-HIAA 5-Hydroxyindole acetic acid VIP Vasoactive intestinal peptide LAGB Laparoscopic adjustable gastric banding
B. A. Flanagin : W. A. Thistlethwaite Pritzker School of Medicine, The University of Chicago, Chicago, IL, USA
Introduction
M. T. Mitchell Department of Radiology, The University of Chicago, Chicago, IL, USA
Bariatric surgery results in a significant alteration in the proximal foregut anatomy and can result, in many cases, in
OBES SURG (2010) 20:386–392
the development of a hiatal hernia and severe gastroesophageal reflux disease (GERD). In the case of Roux-en-Y gastric bypass (RYGB), complete transection of the stomach allows the newly formed stomach pouch to recoil supradiaphragmatically as the opposing force of the nascent stomach and its attachment to the spleen and adjacent structures is lost. Yet, in general, the presence of a hiatal hernia rarely causes pain unless it is of the paraesophageal (type II) variety [1]. Also, in general, it is highly unusual for complications other than GERD to occur as a result of a type I sliding hiatal hernia, which represents the majority of hiatal hernias following bariatric surgery [2, 3]. In fact, most patients who have a hiatal hernia following RYGB or other bariatric procedures remain asymptomatic or complain of heartburn only. Finally, in many cases, hiatal hernias are present during bariatric surgery and either remain unrecognized or are purposely left untreated, assuming that symptoms will resolve with weight loss or acid diversion. The stomach is normally held in its proper anatomic position by several ligamentous attachments. The phrenoesophageal membrane, which attaches near the squamocolumnar junction, directly prevents herniation of the stomach through the diaphragmatic hiatus [4]. This membrane provides a recoil mechanism to “pull” the stomach back into its normal anatomic position following the “physiologic herniation” that occurs during eating. While many patients undergoing morbid obesity surgery experience GERD and often have a hiatal hernia at the time of surgery, GERD symptoms in the case of RYGB are not prominent because gastric acid production in the created proximal pouch is minimal, and the pouch is disconnected from the gastric body where the majority of parietal cells exist. In the case of duodenal switch/pancreaticobiliary bypass (DS), GERD symptoms often resolve because a majority of the acid-producing stomach is removed, and the patients rapidly lose weight, decreasing the intra-abdominal pressure that often contributes to GERD. Yet both the RYGB and DS alter foregut anatomy such that, if a patulous esophageal hiatus exists, the potential for subsequent hiatal hernia development is increased. Stomach transection in the case of the RYGB and creation of a tubularized stomach devoid of all lateral attachments in the case of the DS may significantly affect the function of the phrenoesophageal recoil mechanism and hence the function of the lower esophageal sphincter. There are no studies that have documented the prevalence of hiatal hernia following bariatric surgery nor are there any studies that document the degree to which physiologic reflux develops following RYGB or DS. Finally, whether or not nonspecific foregut symptoms of pain, vomiting, or inability to eat correlate with the presence of a hiatal hernia is unknown. Here, we describe four cases of patients following laparoscopic RYGB or DS who developed disabling and
387
intractable abdominal pain characterized as severe and radiating to the jaw, left shoulder, and midscapular area that was described as paroxysmal and not necessarily associated with eating. All four patients also experienced nausea, vomiting, and failure to thrive at various intervals following laparoscopic bariatric surgery. Routine workup failed to produce any clear mechanical cause of these symptoms. However, complimentary use of multidetector computerized tomography (CT) and upper gastrointestinal contrast studies eventually revealed the diagnosis of hiatal hernia. Exploration identified the presence of a type I hiatal hernia in all four patients, and subsequent operative intervention led to complete remittance of these unusual symptoms.
Case Reports Patient 1 A 49-year-old woman underwent laparoscopic DS in April of 2004 for treatment of morbid obesity. Her perioperative course was unremarkable, and she was discharge home on postoperative day 4. Within 1 month, she started experiencing sharp midepigastric pain with radiation to her chest between the shoulders and to the left neck which was exacerbated by eating or drinking. From May to September of 2004, she underwent CT scan, upper gastrointestinal series (UGI), and several esophagogastroduodenoscopy (EGDs) which were essentially normal. The patient had persistent pain, was unable to eat, and required total parenteral nutrition for nutritional support. She underwent laparoscopic exploration of the abdomen in an attempt to discover the source of her pain. A pouch of gastric fundus was tacked down and tethered to the spleen posteriorly; this was freed up to the angle of His. Intraoperative endoscopy at that time revealed no intraluminal abnormalities. Although a patulous diaphragmatic esophageal hiatus was observed, the fat pad of Belsey was not taken down, and the left and posterior mediastinum was not dissected and explored. The remainder of the exploration was negative. A feeding jejunostomy was placed at that time due to her inability to tolerate food or liquid by mouth. Her symptoms somewhat improved for approximately 2 months thereafter. The patient’s symptoms worsened, and her pain was now characterized as “phrenic pain” consisting of severe paroxysms of midscapular pain radiating to the left jaw and shoulder and occasionally down the left arm. The 24-h pH and esophageal manometry tests were negative for acid reflux, esophageal spasm, or an esophageal motility disorder. A CT in May of 2005 with axial and coronal reformations demonstrated that the staples along the greater curvature staple line were above the diaphragm, compatible with a small hiatal hernia, a finding confirmed by UGI exam (Fig. 1). The
388
OBES SURG (2010) 20:386–392
Fig. 1 Case 1. Multidetector CT axial image (a) shows enlargement of the area of distal esophagus and suggests that the greater curvature staples (arrow) may be above the diaphragm. The coronal CT reformation (b) clearly shows the staples (arrow) above the diaphragm (arrowheads), compatible with a hiatal hernia. Subsequent singlecontrast UGI exam (c) confirms the presence of a hiatal hernia. The proximal aspect of the gastric pouch (arrows) projects above the diaphragm (arrowheads). Single-contrast UGI exam performed after postoperative recurrence of symptoms (d) demonstrates the suture line of the gastric pouch (arrow) above the diaphragm (arrowheads), compatible with recurrent hiatal hernia. GE reflux was observed during the exam, which coincided with an episode of the patient’s pain
patient underwent laparoscopic repair in July of 2005, whereby the entire gastroesophageal (GE) junction was taken down. A hiatal hernia was not apparent at initial visualization of the GE junction. After complete takedown of the GE junction, the fat pad of Belsey, and the posterior attachments, the mediastinum was entered, and 5 cm of esophagus was mobilized. It could then be seen that a small portion of the stomach was clearly supradiaphragmatic and that the staple line of the greater curvature of the stomach was densely adhesed to the left diaphragmatic peritoneum and muscle. Intraoperative endoscopy confirmed that the GE junction had been repositioned well below the diaphragm without any tension. The crura were intact with thick viable muscle and, again after complete dissection, the hiatal opening was found to indeed be enlarged and patulous. Sutures were placed posteriorly and anteriorly to close the crura snuggly over a 60-Fr bougie. Sutures were then placed to affix the esophagus to the diaphragmatic muscle circumferentially with the GE junction lying 3 cm below the diaphragmatic hiatus. The following morning, the patient’s pain was completely eliminated as were all of her other symptoms. She tolerated a soft diet without any problem and was discharged on postoperative day 3.
In April of 2006, having had 11 months of complete symptom relief and eating regular food, the patient developed severe viral gastroenteritis with dry heaving, retching, and vomiting for 36 h. The viral syndrome completely resolved; however, the identical symptoms of phrenic (midscapular, left shoulder, left jaw) pain returned, and the patient could again not eat well and began to lose weight. UGI revealed recurrence of the hernia (Fig. 1). Despite a long period of observation and multiple attempts to relieve her symptoms nonoperatively, she failed to thrive and began losing 3–5 lb of weight per week. She again underwent operative repair of the hiatal hernia with Surgisis® mesh that resulted in relief of symptoms. The operation was performed laparoscopically, and the patient was discharged 24 h postoperatively. Patient 2 A 35-year-old woman presented in August of 2005 with onset of diffuse abdominal pain and progressive nausea and vomiting approximately 2 years after undergoing laparoscopic biliopancreatic diversion with duodenal switch for treatment of morbid obesity. Her postoperative course was
OBES SURG (2010) 20:386–392
uncomplicated, and she had lost approximately 150 lb up to the time of presentation. She had developed this severe right-sided abdominal pain with recurrent vomiting after a single episode of heavy lifting 8 months postop. UGI with small-bowel follow-through at that time revealed a competent and patent duodenoileostomy without evidence of leak or any other abnormalities. At the time of presentation, she was pale and in mild distress. Her pain was atypical in nature, localized in the midepigastrium with radiation towards the back between the shoulder blades. Multidetector CT scan of the abdomen also showed no evidence of bowel obstruction or anastomotic leak; however, the coronal reconstructions demonstrated a small hiatal hernia (Fig. 2). A UGI exam the next day again showed no evidence of anastomotic abnormalities, and a hiatal hernia was not appreciated. Colonoscopy and EGD the following day also revealed no diagnostic abnormality. The patient was taken to the operating room the next day for laparoscopic repair of her hiatal hernia and open repair of the umbilical hernia. Laparoscopy revealed a significant herniation of the posterior fundus of the stomach through the esophageal hiatus, which was reduced and secured with posterior followed by anterior crural repair over a 60-Fr bougie. Significant adhesions were seen between the staple line of the greater curvature of the stomach and the left diaphragmatic crus, which were taken down completely. Complete mobilization of 5 cm of distal esophagus was performed, and all adhesions around this area were divided. The patient symptoms immediately resolved, and she remained asymptomatic at her 2-year follow-up. Patient 3 A 28-year-old woman underwent a laparoscopic Roux-en-Y gastric bypass in February of 1999 for treatment of morbid
389
obesity, for which the perioperative course was unremarkable. She presented in October of 2003 with 6 weeks of intractable midepigastric pain radiating to the neck, nausea, and vomiting. Her pain was not alleviated by nonabsorbable antacids, H2 antagonists, or proton pump inhibitors. The following day, she experienced 4 h of intense (“10/10”) epigastric pain requiring morphine analgesia. She was readmitted at this time for further workup. Ultrasound and hepatobiliary scan revealed no evidence of cholecystitis. A UGI with small-bowel follow-through was unremarkable. Her pain failed to resolve, and she was admitted again for further workup 2 weeks later in mid-November 2003. A chest CT was obtained and was read as normal. She was discharged home on admission day 6 in stable condition and tolerating a regular diet well. She was scheduled at that time for follow-up in the gastroenterology clinic for possible MRCP. Over the next 2 years, the patient underwent a series of diagnostic and therapeutic attempts at an outside medical center to treat her severe pain. She had a 24-h pH and esophageal manometry study to rule out spasm which was normal, and the following medicines were administered to control her pain: tricyclic antidepressants, selective serotonin reuptake inhibitors, benzodiazepines, anticholinergics, nitrates, calcium channel blockers, and several proton pump inhibitors, all at varying doses and schedules. She had no relief of her pain and continued to vomit intermittently. CT images showed fullness of the GE junction, but the coronal reconstructions demonstrated a hiatal hernia (Fig. 3). She underwent operative repair, during which approximately 50% of her gastric pouch was found to be above the diaphragm and the lateral staple line densely adherent to the left hemidiaphragmatic crural peritoneum. The hernia was carefully reduced and the esophagus mobilized and repositioned into the abdomen. The crura were then closed posteriorly followed by suturing of the esophagus to the crura laterally over a 50-Fr bougie in order to prevent recurrence of the hernia. On postoperative day 1, the patient reported complete relief of her pain, and she was able to eat regular food without problem. Patient 4
Fig. 2 Case 2. Multidetector CT scan coronal reformation demonstrates cranial displacement of the proximal gastric pouch (arrow) above the diaphragm, compatible with a hiatal hernia
A 60-year-old woman underwent a laparoscopic biliopancreatic diversion with duodenal switch in May 2005 without incident. She had lost 122 pounds over 10 months when she presented in March of 2006 with symptoms of intermittent nausea, pain radiating to left shoulder, and paroxysms of midscapular pain unprovoked by meals. The patient developed failure to thrive, was unable to eat, and was admitted to the hospital for malnutrition. Parenteral nutrition was initiated and maintained until she underwent laparoscopic surgery. A UGI at that time demonstrated a transient sliding hiatal hernia (Fig. 4). Multidetector CT was
390
OBES SURG (2010) 20:386–392
Fig. 3 Case 3. Multidetector CT scan 3D volume-rendered oblique coronal image (a) and oblique sagittal image (b) clearly show the proximal gastric pouch and gastric suture material (arrow) above the diaphragm (arrowheads), compatible with a hiatal hernia
unremarkable. The patient was also found to have gallstones; however, she had no symptoms typical of biliary colic. The patient underwent laparoscopic hiatal hernia repair. At the time of exploration, the presence of a hiatal hernia was not obvious; however, upon complete mobilization of the esophagus and lateral and posterior attachments, the greater curvature staple line could be seen to be located supradiaphragmatically and was densely attached to the parietal peritoneum laterally and posteriorly. These adhesions were taken down, the esophagus completely mobilized, and the intraoperative endoscopy confirmed the location of the GE junction to be well below the diaphragmatic hiatus. Hiatal hernia repair was performed using nonabsorbable sutures posteriorly and anteriorly over a 60-Fr bougie. A laparoscopic cholecystectomy was also performed. On postoperative day 1, the patient noted immediate improvement in her symptoms and was able to eat soft foods. Four months following her hiatal hernia repair, she is symptom free and eating a regular diet with normal nutritional indices.
Fig. 4 Case 4. UGI exam shows gastric mucosal folds (thick arrow) above the diaphragm (arrowheads), compatible with hiatal hernia. The gastric staple line (thin arrows) is still below the diaphragm, as is the gastric fundal remnant. The hernia was very transient and not manifest during most of the UGI exam. CT scan did not identify this hernia
Discussion Among the many disorders that accompany morbid obesity, the presence of GERD is highly prevalent and often a major disabling problem that is completely treated by weight reduction surgery. In fact, RYGB has been proposed by some as the procedure of choice for patients with medically refractory GERD who are morbidly obese [5, 6]. Although the presence of GERD has been traditionally considered to be a relative contraindication for laparoscopic adjustable gastric banding (LAGB), Dolan et al. have recently proposed that occult hiatal hernias are present in many of these patients and that hiatal hernia repair concomitant with LAGB is both safe and effective for morbidly obese patients with GERD [7]. While many morbidly obese patients are found to have GERD accompanied by a patulous esophageal hiatus during routine laparoscopic RYGB, it would require complete takedown of the GE junction and surrounding fat with subsequent intraoperative endoscopy to identify the position of this junction in each patient to fully determine the true prevalence of hiatal hernia in the morbidly obese patient undergoing bariatric surgery. Although GERD is diagnosed by 24-h pH testing and esophageal manometry, the presence of a hiatal hernia is frequently a difficult diagnosis and one that is often inaccurately diagnosed by EGD or UGI exam. Because the presence of an asymptomatic hiatal hernia is so common, foregut symptoms other than GERD are quite often dissociated from the hiatal hernia, and other causes of these symptoms are sought. This is especially true for a type I sliding hiatal hernia where, absent symptoms of GERD, pain, and vomiting are rare features. Yet in the case where a staple line is present at the angle of His, such as occurs following a sleeve gastrectomy during DS or a vertically stapled gastric pouch in a RYGB, takedown of the parietal peritoneum at the left crus could allow for adhesion formation between the staple line of the transected stomach and the left crus. The tubularized stomach of a DS, void of any lateral attachments, and the small detached pouch of a RYGB are more likely to recoil supradiaphragmatically and, if adhesed to the parietal
OBES SURG (2010) 20:386–392
peritoneum surrounding the left diaphragmatic crus, could produce the “phrenic” type of pain that was typical of our four patients. Although poorly described in the literature, the left shoulder, left jaw, and midscapular pain in these patients could possibly be due to esophageal spasm instead of traction on the peritoneum of the left crus. While we found no evidence of esophageal spasm in these patients, it is often a difficult diagnosis that remains a possibility in these cases. The striking immediate resolution of the very similar symptom complex in these patients following radical and definitive repair is noteworthy. Two of the patients were treated with a variety of agents including nitrates and calcium channel blockers without any symptomatic relief. We believe that the midepigastric and chest pain radiating to the left shoulder and jaw reported by these patients is referred, secondary to irritation of the phrenic nerve [8–10]. Given that patients undergoing bariatric surgery are more prone to herniation of the gastric pouch through the diaphragmatic hiatus, it is likely this “phrenic pain” is provoked via rubbing of the gastric staple line on the left diaphragmatic parietal pleura. Stimulation of the left somatic afferent fibers of the phrenic nerve would trigger symptoms of pain in a C3, C4, and C5 dermatomal distribution, thereby being perceived in the left shoulder and jaw region (Fig. 5). The precise mechanism underlying the nausea, inability to eat, and requirement for artificial nutrition in these patients is perplexing. Despite the absence of mechanical obstruction, all of the patients developed failure to thrive and required enteral or parenteral nutritional support. In the postbariatric patient, it is important for the treating physician not to dismiss symptoms that impair adequate nutritional intake and to institute immediate nutritional support when indicated. Given the immediate relief of pain, Fig. 5 Illustration of possible mechanism by which a hiatal hernia causes phrenic pain. Irritation of the diaphragmatic parietal peritoneum by the gastric staple line following RYGB is shown here to be the cause of the referred pain in the left shoulder, jaw, and midscapular region via visceral afferent fibers innervating the diaphragm from C3 to C5
391
nausea, and inability to eat in these patients following reparative surgery, a neurophysiologic basis for the nausea as a result of the hiatal hernia is likely. The radiographic diagnosis of hiatal hernia in morbidly obese gastric bypass patients is technically challenging. Fluoroscopic diagnosis of a hiatal hernia by UGI exam relies on demonstrating that a portion of the stomach lies above the diaphragm. This is best made with the patient in a left posterior oblique position, with recumbent positioning and Valsalva maneuvers useful for identifying transient sliding hiatal hernias. When patient weight exceeds 300 lb, they are typically too large to fit within standard fluoroscopy units and, if they do, they often can be imaged only in a frontal projection. The mucosal fold patterns and gastric suture lines are often indistinct due to limited X-ray beam penetration of the abdomen, and the postsurgical stomach is difficult to fully distend during UGI exam. These factors make it difficult to accurately localize the position of the gastroesophageal junction relative to the diaphragm. In addition, these patients often have limited tolerance for supine positioning, making it more difficult to identify sliding hiatal hernias. The advantage of UGI exam, and also of small-bowel series, over CT scan is the ability to assess bowel function in addition to anatomy. Fluoroscopy can diagnose esophageal dysmotility, gastroparesis, small-bowel dysmotility, and postsurgical complications such as Roux segment inversion that are in the differential diagnosis for nausea and vomiting in the post-gastric-bypass patent. These entities will have normal anatomic findings on all static imaging studies including CT, and many will be occult on EGD and other invasive studies. CT scan diagnosis of hiatal hernia in morbidly obese gastric bypass patients has its advantages and its own technical challenges. Patients tend to be more tolerant of the
392
supine positioning for CT than they are for UGI exams, possibly because they perceive the time period as more finite and/or because they perceive the CT as a more important test. The scan is done during breath holding, which can help provoke a sliding hiatal hernia. CT table weight limits are typically greater than fluoroscopy table weight limits, allowing imaging of larger patients; however, the large size of these patients results in significant image degradation that, with nonhelical and with standard helical CT scanners, requires thicker scan slices which result in suboptimal coronal reformations. The technical advances offered by multidetector CT scanners with 16 channels or more allow high-resolution imaging of these patients that better delineates esophageal and gastric anatomy and yield multiplanar image reconstructions that maintain the high image resolution. As with UGI exam, the diagnosis of hiatal hernia by CT depends on demonstrating that a portion of stomach lies above the diaphragm. While the gastroesophageal transition is less obvious than with UGI exams because CT does not demonstrate as much mucosal detail, the suture lines that delineate the gastric pouch are clearly visible. Axial CT images may show soft tissue fullness or displacement of the suture line above the diaphragm; however, a small or early hernia can be difficult to identify on axial images alone. The coronal reformations and volume-rendered images that can be obtained from a multidetector CT scan allow excellent delineation of the relationship between the diaphragm and the gastroesophageal anatomy. The relationship of the suture line to the diaphragmatic crura is clearly visible, allowing definitive diagnosis of a hiatal hernia. A sliding hiatal hernia that manifests transiently could still be occult by CT. For this reason, we consider multidetector CT and UGI exam to be complementary studies. Although conjectural, our findings indicate that a hiatal hernia may cause vomiting, intractable atypical pain, and failure to thrive in some patients following bariatric surgery. When no other mechanical or functional basis for such symptoms can be identified, definitive repair of the hiatal hernia should be considered. Complete takedown of the esophageal hiatus is necessary, and we prefer to accomplish this laparoscopically. It is important to completely encircle the entire esophagus and establish the location of the squamocolumnar junction endoscopically. At least 3 cm of the esophagus should lie below the crura as established
OBES SURG (2010) 20:386–392
endoscopically. Mesh prosthesis may be necessary in cases where the crural muscle is not suitable for suture. Although GERD may be present preoperatively, there is no way to establish an effective antireflux procedure following a sleeve gastrectomy or RYGB. While GERD may have improved in our patients by simply repositioning the lower esophageal sphincter into the abdomen, the main purpose of the surgery was to improve their pain, nausea, and vomiting. In summary, the presence of a hiatal hernia following an RYGB or DS associated with nausea, vomiting, and phrenictype pain in patients who cannot thrive without nutritional support may be improved by surgical repair of the hiatal hernia. The mechanism by which a hiatal hernia causes failure to thrive in patients remains to be clarified. Careful evaluation of patients’ history and symptoms should guide the decision to operate.
References 1. Peridikis G, Hinder RA. Paraesophageal hiatal hernia. In: Nyhus LM, Condon RE, editors. Hernia. 4th ed. Philadelphia: Lippincott; 1995. p. 543–53. 2. Ott DJ, Gelfand DW, Chen YM, et al. Predictive relationship of hiatal hernia to reflux esophagitis. Gastrointest Radiol. 1985;10: 317–20. 3. Wright RA, Hurwitz AL. Relationship of hiatal hernia in endoscopically proved reflux esophagitis. Dig Dis Sci. 1979;24:311–3. 4. Rabine JC, Nostrant TT. Miscellaneous diseases of the stomach. In: Yamada T, Alpers DH, Laine L, Owyang C, Powell DW, editors. Textbook of gastroenterology, vol. 1. 4th ed. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 1455–65. 5. Frezza EE, Ikramuddin S, Gourash W, et al. Symptomatic improvement in gastroesophageal reflux disease (GERD) following laparoscopic Roux-en-Y gastric bypass. Surg Endosc. 2002;16:1027–31. 6. Perry Y, Courcoulas AP, Fernando HC, et al. Laparoscopic Roux-enY gastric bypass for recalcitrant gastroesophageal reflux disease in morbidly obese patients. JSLS. 2004;8:19–23. 7. Dolan K, Finch R, Fielding G. Laparoscopic gastric banding and crural repair in the obese patient with a hiatal hernia. Obes Surg. 2003;13:772–5. 8. Riedel HH, Semm K. The post-laparoscopic pain syndrome (author's syndrome). Geburtshilfe Frauenheilkd. 1980;40:635–43. 9. Scawn ND, Pennefather SH, Soorae A, et al. Ipsilateral shoulder pain after thoracotomy with epidural analgesia: the influence of phrenic nerve infiltration with lidocaine. Anesth Analg. 2001;93: 260–4. 10. Petchkrua W, Harris SA. Shoulder pain as an unusual presentation of pneumonia in a stroke patient: a case report. Arch Phys Med Rehabil. 2000;81:827–9.