High Incidence of Gastric Bezoars in Cystic Fibrosis Patients after Lung Transplantation : Transplantation

Journal Logo

Original Articles: Clinical Transplantation

High Incidence of Gastric Bezoars in Cystic Fibrosis Patients after Lung Transplantation

Dellon, Evan S.; Morgan, Douglas R.; Mohanty, Sanjib P.; Davis, Ken; Aris, Robert M.

Author Information
Transplantation 81(8):p 1141-1146, April 27, 2006. | DOI: 10.1097/01.tp.0000205813.54136.85
  • Free


Bezoars are concretions composed of vegetable matter (phytobezoars), hair (trichobezoars), medications (pharmacobezoars), or other ingested substances (1). Over time, these accumulate in the gastrointestinal tract, most commonly in the stomach, and can manifest with symptoms ranging from nausea or non-specific abdominal pain to frank gastric outlet obstruction (2, 3). The diagnosis may be suggested by radiologic study (4), and is confirmed by upper endoscopy.

The overall incidence of bezoars is felt to be low, and is extremely rare in healthy individuals (1). Certain groups of at-risk patients have been identified and include patients with altered upper-GI anatomy due to surgery as well as those with psychiatric illness or cognitive impairment (2, 3). Impaired gastric motility, as in diabetes, has not conclusively been shown to be a risk factor (5).

The literature reporting gastric bezoars after transplant is limited. No bezoars have been seen in studies of gastrointestinal (GI) related complications after lung transplant (6, 7). The majority of previous reports have been limited to single case descriptions of bezoar formation following a liver (8), pancreas (9), bone marrow (10), and combined heart/lung transplantation (11). One study of postlung transplant gastroparesis found three patients with retained food in the stomach, one of whom had cystic fibrosis (12). In our clinical experience, however, patients with cystic fibrosis (CF) have a propensity to form bezoars after double lung transplantation. Given that patients with CF have altered GI secretions due to mutation in the cystic fibrosis transmembrane conductance regulator gene (CFTR) (13, 14), they may represent a high risk group for bezoar formation after lung transplant. Other potential risk factors include pancreatic exocrine dysfunction (14), chronic illness, postoperative gastroparesis, and multiple medications. The aim of this study, then, is to evaluate the incidence of bezoars after lung transplantation, with a particular focus on patients with CF, and identify associated risk factors.


This is a retrospective analysis of patients who underwent lung transplantation from December, 1992 through July, 2005 at the University of North Carolina, Chapel Hill. The records of these patients were screened, and patients who developed gastrointestinal bezoars in the posttransplant setting were identified. Their medical records were reviewed to extract demographic variables (age, gender, race), illness variables (indication for lung transplantation, date of transplant, date of death if applicable, organisms causing colonization or infection in the peri-transplant period, presence of diabetes), and bezoar-related variables (symptoms, time from transplant to diagnosis, type of bezoar, treatment, recurrence, and presence of gastroparesis or performance of a nuclear scintigraphic gastric emptying study).

A bezoar was defined as an endoscopically confirmed concretion of retained food products or other materials found in the stomach during upper endoscopy. No restrictions were placed on the indication for upper endoscopy. A bezoar size threshold was not utilized for inclusion or exclusion. Patients with abnormal plain radiographs with suggestive findings such as gastric distention or retained gastric contents, but a negative endoscopy, were excluded. Similarly, patients with epigastric symptoms who were given empiric treatment, for example a promotility agent or digestive enzyme, for a presumptive bezoar were also excluded. Patients undergoing repeat lung transplantation were included in the study. Those undergoing combined heart/lung transplant were excluded to minimize variability in the patient population, as well as difference in operative techniques, that could affect the outcomes.

For data analysis, patients were divided into two groups, and bezoar cases were compared with non-bezoar subjects. Descriptive statistics were used for characterization of the population, and the two groups were compared with the Chi-squared test for categorical variables and Student’s t test to compare means for continuous variables. Odds ratios (OR) were calculated with 95% confidence intervals (CI). Statistical significance was set at the P<0.05 level. Data was analyzed with SAS version 8 (SAS Institute, Inc., Cary, NC, USA). The study protocol was approved by the Institutional Review Board of the University of North Carolina.


Over the thirteen year study period, 215 patients underwent lung transplantation and were included in our study. Cystic fibrosis was the leading indication for lung transplantation (n=145, 67%), followed by chronic obstructive pulmonary disease and idiopathic pulmonary fibrosis (Table 1).

Characteristics of the study population

A total of 17 patients (7.9%) were diagnosed with gastric bezoars by upper endoscopy. Of these patients with bezoars, 16 (94.1%) had cystic fibrosis. This represents 11% of CF patients undergoing lung transplantation. If patients who did not survive longer than 30 days after surgery were excluded, then 12% of the CF population receiving lung transplantation had bezoars. The other patient with a bezoar had primary ciliary dyskinesia (patient 11, Table 2). No patient who underwent lung transplantation for other indications was diagnosed with a bezoar. The odds ratio for a patient with cystic fibrosis having a bezoar in the postlung transplant setting was calculated to be 8.0 (95% CI: 1.11–100, P=0.04) compared to patients who underwent lung transplantation for other indications. The only other variable associated with bezoar formation was mean age; there was no difference in survival between the groups (Table 1).

Characteristics of patients found to have bezoars

Characteristics of the individual patients diagnosed with bezoars are presented in Table 2. The mean time to bezoar diagnosis after transplant was 49 days (range: 10 to 289 days), though 12 of the 17 patients (71%) were diagnosed within 33 days. If patient 6, an outlier, is excluded, the mean time until diagnosis was 34 days (±23.4). The annual incidence was unchanged from 1992 to 2005. All of the patients with bezoars were white, and all received bilateral lung transplantation. None of the patients had a history of bezoar, gastric disease, or gastric surgery prior to lung transplant. Patients underwent diagnostic upper endoscopy for a variety of symptoms and signs suggestive of bezoar formation such as nausea, early satiety, dyspepsia, altered drug levels, or radiographs performed in the fasting state showing gastric distention or prominent gastric contents (Fig. 1). Phytobezoars were found in all patients, though several (patients 9, 10, 14, 15, and 17) also had a pharmacobezoar component. The majority of patients were treated with a combination of high volume gastric lavage at the time of upper endoscopy, dietary alteration to either a liquid or a low residue diet, and medical therapy (prokinetic agents and digestive enzymes); these measures were generally continued for 2–3 months after symptoms had resolved. Endoscopically confirmed bezoars recurred in 5 patients (range: 17 days to 7 years later) in spite of these interventions, and were successfully treated with repeat therapy.

A radiographic and endoscopic image of the bezoar diagnosed in patient 15. On postoperative day 9, the routine daily chest x-ray (panel A) revealed gastric distention with retained gastric contents, though this was a film performed after an overnight fast. Bilateral pulmonary infiltrates are present, and a central line and chest tubes are in place. The following day, the patient underwent upper endoscopy (panel B) which revealed a large bezoar comprised of primarily of food (phytobezoar).

In the peritransplant period, patients were infected with or colonized by a number of pathogens typically seen in cystic fibrosis such as Burkholderia cepacia, Pseudomonas aeruginosa, methicillin-resistant Staphylococcus aureas, Alcaligenes xylosoxidans, Aspergillus species, Candida species, and others (Table 2). Immunosuppressant regimens were standardized by protocol at the time of transplant, and in almost all cases included cyclosporine A, prednisone, and either azathioprine or mycophenolate mofetil.


In our lung transplant population, gastric bezoars were common in cystic fibrosis patients, and occurred early in the posttransplant setting. Compared with patients undergoing lung transplantation for other indications, we found CF patients to be at significant risk for developing postoperative bezoars. Younger age was also associated with bezoars, likely reflecting the fact that CF patients are also younger when they receive a transplant compared with patients undergoing the surgery for pulmonary fibrosis or emphysema.

This observation of a high incidence of bezoars in CF patients after lung transplant has not been previously reported in the literature, and has clinical implications. Whereas we identified an at-risk population with an incidence of greater than 10%, bezoars are traditionally quoted as occurring in far less than 1% of patients in retrospective endoscopic series (1, 2, 5). It is possible that our study underestimated the actual rate of bezoar formation in this population because we did not include cases that were suspected by abnormal radiologic studies or patients who received empiric medical therapy. Further, because posttransplant symptoms such as nausea, vomiting, and abdominal pain are non-specific and have a broad differential diagnosis, the key to making the diagnosis of bezoar is inclusion in the differential. This study will allow practitioners to identify patients who are at risk for bezoar formation.

There are several strengths of this study. First, by employing a stringent and objective requirement for the diagnosis of bezoar, endoscopic confirmation, we minimize the possibility of misclassification of cases. Second, the detailed lung transplant database available to us facilitated inclusion of all patients in the analysis. Third, in the context of the previously published literature, this study population is comparatively large.

In addition to these strengths, there are limitations to this study, most of which are inherent in this type of retrospective study design. For example, the relationship between gastroparesis and bezoars was not able to be fully explored as only a minority of patients underwent formal gastric emptying nuclear scintigraphy either before or after their bezoar was found. Additionally, since we relied on data from a single center over a long period of time, the effects of change in practice patterns and possible institutional or practitioner biases are operative. It may not be possible to generalize our findings to other centers or to the posttransplant cystic fibrosis population as a whole. Finally, there may have been unrecognized bezoars in patients who remained asymptomatic. If these patients were to be over-represented in the non-CF population, then our findings might have been altered.

What are the biologic and pathogenic mechanisms that may explain our findings? Since the vast majority of bezoar formation was in cystic fibrosis patients, it is reasonable to ask if there is something unique in that population. CFTR is actively transcribed in the epithelial cells of the small intestine, but there does not appear to be substantial gastric expression (13, 15). In the pretransplant setting, CF patients can have meconium ileus in infancy or distal ileal obstruction syndrome (DIOS) in adulthood related to abnormal secretions from mutations in the CFTR gene product (13, 14). However, to date there have been no published reports of bezoar formation, either in the small bowel or in the stomach, in cystic fibrosis patients at any point along their disease course. This implies that there may be additional factors specific to lung transplantation. Our data would support this, as most of our patients developed bezoars in a short period of time after their surgery.

There are several aspects of the peri-operative period that could predispose CF patients to bezoar formation. Surgical technique might play a role. Branches of the vagus nerve may be compromised during lung transplantation, and upper GI dysmotility has been reported as a result of this in combined heart/lung recipients (16). Operative techniques, however, have changed over time to attempt to improve outcomes, and omental wraps for the airway anastamoses, which can potentially injure the vagus nerve, are no longer commonly performed (17, 18). In CF patients, however, it is possible that transient vagus injury or stunning could occur as a consequence of the extensive surgical dissection sometimes necessary due to chronic lung infection and scarring which may not be seen in patients transplanted for other indications. Were this the sole mechanism, though, it would be expected that rates of bezoar formation would be more frequent in bilateral lung recipients and that some bezoars would be seen in non-CF transplant patients. Our data do not bear this out.

Postoperative gastroparesis has been noted in patients after lung and heart/lung transplantation (12, 19, 20), but again, were this the only mechanism, all patients regardless of indication would develop bezoars. In our patient population, four patients with bezoars had gastric emptying studies and were diagnosed with gastroparesis postoperatively. It is difficult to draw conclusions from this limited subset in a retrospective study. No specific association between cystic fibrosis and gastroparesis has been reported in the literature in the preoperative setting. It is plausible that chronic hyperglycemia or diabetes in a CF patient may predispose to delayed gastric emptying due to diabetic autonomic neuropathy (21). In the CF population, it is possible that chronic subclinical gastroparesis and transient vagus nerve compromise are risk factors for postoperative gastric dysmotility. This may act in concert with abnormal endogenous secretions (sinus, gastrointestinal) to create a milieu favorable to bezoar formation.

While it is appealing to theorize that immunosuppressants may play a role, there are no reports of pharmacobezoar formation due to cyclosporine, prednisone, or azathioprine. Cyclosporine, however, has been associated with slower gastric emptying than either tacrolimus or azathioprine (22–24). Additionally, because tacrolimus shares structural similarities with macrolides, it may increase gastric motility by acting as a motilin agonist; cyclosporine does not share this property (22). Multiple other medications are routinely used in the posttransplant setting including pancreatic enzymes and prophylactic antimicrobials. While it is possible that these medications, particularly the enteric-coated pancreatic enzyme formulations, may become involved in a preexisting bezoar, it is hard to postulate how they individually may contribute to bezoar pathogenesis.

Finally, one additional factor is unique to the CF population: CF airways are chronically colonized with multiple pathogenic organisms. It is possible, though unstudied, that ingestion of these bacteria from respiratory and sinus secretions could promote bezoar formation. Our data, however, did not detect any specific bacteria in the bezoar patients that are not typically seen in other cystic fibrosis patients. It is important to note that at our center, colonization with Burkholderia cepacia has not been a contraindication to lung transplantation in CF patients. Further, this is not likely the sole mechanism, as CF patients who have not undergone lung transplant do not develop bezoars.

At our center, there has not been a protocol for routine bezoar surveillance or primary prevention in the postlung transplant population. Patients are referred for endoscopic evaluation and treatment when symptoms or imaging suggest the possible formation of a bezoar, and when other conditions in the differential diagnosis are excluded. There are a variety of approaches to treat gastric bezoars. These include endoscopic techniques to remove particulate matter (snares, nets), large volume water or saline irrigation, carbonated beverage-irrigation, and enzymatic digestion (25–31). While these strategies are reasonable, they require an invasive procedure (upper endoscopy) during which complications may occur (32). Aspiration can be a devastating complication after lung transplantation, and the risk for this may be increased with large volume lavage. One of the patients in this study (patient 15 in Table 2) underwent large volume lavage on two occasions for treatment of recurrent bezoar and subsequently developed pulmonary infiltrates that were attributed to aspiration as bronchoscopy and lung biopsies were negative for rejection.

After endoscopic treatment, patients at our center are now prescribed a combination regimen to prevent bezoar formation. It is an empiric “cocktail” of metoclopramide (10 milligrams four times daily), sodium bicarbonate (648 milligrams twice daily) and pancrelipase enzymes. We use a powder preparation of enzymes containing 16,800 U lipase, 70,000 U protease, and 70,000 U amylase in 0.7 gm (Viokase®, Axcan Scandipharm) administered as 3 grams four times daily. These enzymes are prescribed in addition to their normal digestive enzyme regimen. In addition, patients are also provided with an aggressive bowel regimen to minimize any effect of constipation. The rationale for this strategy is that the combination of digestive enzymes, a promotility agent, and a neutralizing agent will act in concert to prevent a nidus from forming. Enzymes break up concretions and prokinetics clear them from the stomach. The use of bicarbonate counteracts the low-pH milieu that can promote bezoar formation (25), and it may also help to make the mucus secretions in CF less viscous. There is no current standard for how long these medications are used, but they are generally discontinued several months after symptoms have resolved. No routine imaging or endoscopy is done for follow-up testing. This approach raises the question of whether these are the best medications, and whether similar medications should be used in all CF patients after lung transplantation for primary prevention of bezoar formation.

In conclusion, this paper reports a high incidence of bezoar formation early after lung transplantation in the cystic fibrosis population. Further investigation is needed to understand the multifactorial pathogenesis in this select population. It is possible that factors acting in combination include acute and chronic gastric dysmotility, immunosuppressants, chronic bacterial colonization, and abnormal respiratory and GI secretions. Strategies for primary prevention of bezoars in cystic fibrosis patients undergoing lung transplantation may be beneficial. Additionally, as these observations have not previously been reported, confirmation in a multi-center investigation would be welcome.


The authors would like to express their gratitude to Drs. Peadar G. Noone and Thomas M. Egan for contributions to the current study.


1. Andrus CH, Ponski JL. Bezoars: classification, pathophysiology, and treatment. Am J Gastroenterol 1988; 83: 476.
2. Zamir D, Goldblum C, Linova L, et al. Phytobezoars and trichobezoars: a 10-year experience. J Clin Gastroenterol 2004; 38: 873.
3. Erzurumlu K, Malazgirt Z, Bektas A, et al. Gastrointestinal bezoars: a retrospective analysis of 34 cases. World J Gastroenterol 2005; 11: 1813.
4. Ripolles T, Garcia-Aguayo J, Martinez MJ, Gil P. Gastrointestinal bezoars: sonographic and CT characteristics. AJR Am J Roentgenol 2001; 177: 65.
5. Ahn YH, Maturu P, Steinheber FU, Goldman JM. Associate of diabetes mellitus with gastric bezoar formation. Arch Intern Med 1987; 147: 527.
6. Lubetkin EI, Lipson DA, Palevsky, et al. GI complications after orthotopic lung transplantation. Am J Gastroenterol 1994; 91: 2382.
7. Gilljam M, Chaparro C, Tullis E, et al. GI complications after lung transplantation in patients with cystic fibrosis. Chest 2003; 123: 37.
8. Schwarzenberg SJ, Freese DK, Payne WD, et al. Gastric bezoar after hepatic transplantation. J Pediatr Gastroenterol Nutr 1989; 9: 119.
9. Boggi U, Vistoli F, Morelli L, et al. Small-bowel obstruction due to bezoar following pancreas transplantation with portal-enteric drainage: a case report. Transplant Proc 2004; 36: 575.
10. Ben-Yehuda A, Ben-Yehuda-Salz D, Or R, et al. Intestinal obstruction caused by bezoar: a rare complication after bone marrow transplantation. Bone Marrow Transplant 1989; 4: 137.
11. Reid KR, McKenzie FN, Menkis AH, et al. Importance of chronic aspiration in recipients of heart-lung transplants. Lancet 1990; 336: 206.
12. Austin JH, Gougoutas CA, Schulman LL. Short air bubble in the gastric fundus during fasting: radiographic sign of gastroparesis after lung transplantation. J Thorac Imaging 2000; 15: 65.
13. Turcios NL. Cystic fibrosis: an overview. J Clin Gastroenterol 2005; 39: 307.
14. Eggermont E. Gastrointestinal manifestations in cystic fibrosis. Eur J Gastroenterol Hepatol 1996; 8: 731.
15. Crawford I, Maloney PC, Zeitlin PA, et al. Immunocytochemical localization of the cystic fibrosis gene product CFTR. Proc Natl Acad Sci USA 1991; 88: 9262.
16. Au J, Hawkins T, Venables C, et al. Upper gastrointestinal dysmotility in heart-lung transplant recipients. Ann Thorac Surg 1993; 55: 94.
17. Bonser RS, Fragomeni LS, Kriett JM, et al. Technique of clinical double-lung transplantation. J Heart Transplant 1988; 7: 298.
18. Khaghani A, Tadjkarimi S, al-Kattan K, et al. Wrapping the anastomosis with omentum or an internal mammary artery pedicle does not improve bronchial healing after single lung transplantation: results of a randomized clinical trial. J Heart Lung Transplant 1994; 13: 767.
19. Berkowitz N, Schulman LL, McGregor C, Markowitz D. Gastroparesis after lung transplantation: potential role in postoperative respiratory complications. Chest 1995; 108: 1602.
20. Sodhi SS, Guo JP, Maurer AH, et al. Gastroparesis after combined heart and lung transplantation. J Clin Gastroenterol 2002; 34: 34.
21. Camilleri M. Advances in diabetic gastroparesis. Rev Gastroenterol Disord 2002; 2: 47.
22. Maes BD, Vanwalleghem J, Kuypers D, et al. Differences in gastric motor activity in renal transplant recipients treated with FK-506 versus cyclosporine. Transplantation 1999; 68: 1482.
23. Verleden GM, Besse T, Maes B. Successgul conversion from cyclosporine to tacrolimus for gastric motor dysfunction in a lung transplant recipient. Transplantation 2002; 73: 1974.
24. Eagle DA, Gian V, Lauwers GY, et al. Gastroparesis following bone marrow transplantation. Bone Marrow Transplant 2001; 28: 59.
25. Rider JA, Foresti-Lorente RF, Garrido J, et al. Gastric bezoars: treatment and prevention. Am J Gastroenterol 1984; 79: 357.
26. Sechopoulos P, Robotis JF, Rokkas T. Gastric bezoar treated endoscopically with a carbonated beverage: case report. Gastrointest Endosc 2004; 60: 662.
27. Kato H, Nakamura M, Orito E, et al. The first report of successful nasogastric Coca-Cola lavage treatment for bitter persimmon phytobezoars in Japan. Am J Gastroenterol 2003; 98: 1662.
28. Silva FG, Goncalves C, Vasconcelos H, Cotrim I. Endoscopic and enzymatic treatment of gastric bezoar with acetylcysteine. Endoscopy 2002; 34: 845.
29. Ladas SD, Triantafyllou K, Tzathas C, et al. Gastric phytobezoars may be treated by nasogastric Coca-Cola lavage. Eur J Gastroenterol Hepatol 2002; 14: 801.
30. Bonilla F, Mirete J, Cuesta A, et al. Treatment of gastric phytobezoars with cellulase. Rev Esp Enferm Dig 1999; 91: 809.
31. Wang YG, Seitz U, Li ZL, et al. Endoscopic management of huge bezoars. Endoscopy 1998; 30: 371.
32. Eisen GM, Baron TH, Dominitz JA, et al. Complications of upper GI endoscopy. Gastrointest Endosc 2002; 55: 784.

Bezoar; Lung transplantation; Cystic fibrosis; Incidence

Copyright © 2006 Wolters Kluwer Health, Inc. All rights reserved.