Managing Patients With Gut Failure After Bariatric Surgery
Managing Patients With Gut Failure After Bariatric Surgery
This is a retrospective analysis of prospectively collected database for patients referred to 2 major academic centers with GF, defined by loss of nutritional autonomy that required or indicated the need for TPN. History of BS was used to identify the study population. Core data was pooled from a computerized database and chart review was conducted to obtain pertinent medical information. Most of the reconstructive and transplant operations were performed by the primary author. Institutional Review Board (IRB) approval was obtained from both centers with an honest broker for data management.
With development of GF after BS being the only inclusion criteria, the study population was categorized into 3 groups according to the newly introduced types of GF. Each type was defined by patient presentation, underlying etiology, and associated pathophysiology. Patients with Type-I GF experienced acute catastrophic gut loss due to vascular occlusion. Type-II GF patients suffered major long-lasting technical complications, including gastrointestinal fistulae, loss of gut continuity, and bowel obstruction after multiple corrective surgical interventions. In Type-III, failure to thrive with the development of GF was insidious with progressive course due to clinical dysfunctional syndromes, including motility disorders, restrictive intolerance in absence of mechanical pathology, and gut malabsorption.
The nomenclature of the different bariatric and transplant procedures has been fully defined. The term 'restorative surgery' defines collective procedures that are utilized to restore nutritional autonomy, including autologous reconstruction, intestinal lengthening, and transplantation. The term 'autologous reconstruction' is limited to the description of the different surgical techniques that are utilized to re-establish continuity of the native gastrointestinal tract. Intestinal lengthening is a rehabilitative rather than a reconstructive procedure to enhance gut adaptation.
TPN failure was defined by multiple line infections, central vein thrombosis, and hepatic injury. Nutritional autonomy indicates freedom from TPN and fluid replacement.
The study was conceived with the first BS patient who was referred to the University of Pittsburgh Medical Center (UPMC) for intestinal transplantation, 5 years after the 1990 inception of the program. With the evolving concept of gut rehabilitation, increased volume of referral was observed particularly in recent years (Fig. 1).
(Enlarge Image)
Figure 1.
Total number of patients referred to the UPMC and CCF with gut failure after bariatric surgery substratified according to type of gut failure (A) and type of the bariatric procedure (B). BPD, biliopancreatic diversion; CCF, Cleveland Clinic Foundation; DS, duodenal switch; UPMC, University of Pittsburgh Medical Center.
Over the past 20 years, 1500 adults with GF were referred for gut rehabilitation at UPMC (1995–2015) and Cleveland Clinic Foundation (CCF) (2010–2015). Of these, 142 (9%) had BS that preceded TPN dependency and made up the referral population. BS was performed at different regional, national, and international centers. Causes of GF were Type-I in 63 (44%), Type-II in 45 (32%), and Type-III in 34 (24%). Yearly referrals according to type of GF and BS are shown in Figure 1A and 1B, respectively. Types of primary BS, as documented in medical records, were malabsorptive in 6 (4%), restrictive in 27 (19%), and combined in 109 (77%) with surgical illustrations shown in Figure 1, Supplemental Digital Content, http://links.lww.com/SLA/A846. At referral, 14 patients had prior conversions with jejunoileal bypass (JIBP) to gastric banding (GB) in 4, GB to Roux-en-Y gastric bypass (RYGB) in 9, and sleeve gastrectomy (SG) to RYGB in 1.
Upon referral, the obtained medical records were reviewed. Patients with recent catastrophic events and complex abdominal pathology underwent a hospital-to-hospital transfer. Urgent abdominal re-exploration was required for Type-I GF patients with residual infarcted bowel, stapled visceral organs, and intra-abdominal infection. Outpatient candidates underwent thorough evaluation to assess nutritional status, residual gut, splanchnic circulation, and surgical candidacy. Of the 142 referrals, 131 (92%) were evaluated and comprised the study cohort; 73(56%) at UPMC and 58 (44%) at CCF. The remaining 11 patients were denied access by health care plans (n = 8) or died before transfer (n = 3).
All patients underwent a thorough multidisciplinary evaluation. Assessment included clinical presentation, review of organ systems, gut anatomy, nutritional status, comorbidities, and behavioral health issues. In addition to the type of GF-associated clinical presentation, chronic abdominal pain was a common theme in all patients. Biochemical and hematologic testing included battery assessment of protein-calorie malnutrition, nutritional deficiencies, and thrombotic disorders. Radiologic and endoscopic evaluations were used to map the altered gut anatomy and residual intestinal length. Visceral angiograms were obtained in selected patients to delineate vascular anatomy and guide further therapy. Hepatic assessment, including liver biopsy was required for patients with severe morbid obesity, malabsorptive BS, chronic TPN therapy, hostile abdomen, and portomesenteric venous thrombosis. Cardiopulmonary assessment was guided by the existence of comorbidities and other surgical risk factors. In recent cases, full assessment of skeletal health was conducted.
All patients received full medical, nutritional, metabolic, and psychosocial support. TPN was optimized or initiated to restore nutritional deficiencies. Patients with renal, respiratory, and cardiopulmonary insufficiency were subjected to intense preoperative management.
The primary objectives of surgical management were to restore nutritional autonomy, reduce need for visceral allotransplantation, and avoid recurrence of severe obesity. Restitution of gut anatomy was offered with the premise to normalize gut physiology. Surgical strategies were guided by type of GF, residual gut anatomy, and surgical candidacy. A first-stage operation, including foregut reconstruction, was considered for Type-I GF patients with ultrashort gut syndrome and intra-abdominal infection. With Type-II GF, a single stage autologous reconstruction was planned for patients with gastrointestinal fistulae and disconnected gut with possible reversal of the bariatric procedure particularly in patients with short gut syndrome. In patients with isolated gastrointestinal pathology, the altered bariatric anatomy was preserved. Intestinal lengthening was performed in selected patients with short gut syndrome to enhance gut rehabilitation. With Type-III GF, surgery was considered after failure of comprehensive multidisciplinary medical management and limited to reversal of BS when technically feasible. With all types, portal hypertensive surgery was performed for a few patients with portomesenteric venous thrombosis. Surgical options and potential complications, including relapse of obesity, were discussed with all patients and an informed consent was obtained surgical options and potential complications, including relapse of obesity, were discussed with all patients and an informed consent was obtained.
Autologous reconstruction was an organ salvage procedure guided by the embryonic development of the alimentary canal (Figure 2, Supplemental Digital Content, http://links.lww.com/SLA/A846). Technical feasibility was dictated by the complexity of the abdominal pathology, the altered anatomy of the gastrointestinal tract, and integrity of the visceral vascular blood supply. Primary foregut reconstruction was the mainstay of a definitive surgical treatment or a first-step procedure in preparation for visceral transplantation. Interposition of a visceral conduit was used to avoid exclusion of the native duodenum from the alimentary flow and optimize the absorptive capacity of the residual gut.
(Enlarge Image)
Figure 2.
Causes of the 3 different types of gut failure in the 131 study patients; catastrophic gut loss (A, Type-I), technical complications with loss of gut continuity (B, Type-II), and dysfunctional syndromes with different malabsorptive, restrictive, and motility disorders (C, Type-III). With Type-II, some patients have more than one pathology.
In patients with short gut syndrome, a conservative surgical approach was adopted to salvage the survived bowel. The reconstructive techniques aimed at maximization of the absorptive functions of the residual gut. In addition, intestinal lengthening with serial transverse enteroplasty (STEP) was performed in patients with less than 100 cm of small bowel.
Visceral transplantation was indicated for patients with ultrashort gut syndrome and TPN failure. The number of transplanted organs was dictated by the extent of gut loss and coexistence of liver failure. Donors were deceased and allografts were ABO identical. Human leukocyte antigen (HLA) match was random with positive T/B cell lymphocytotoxic crossmatch in 9 (36%) patients. No attempts were made to immunomodulate the allografts and no recipients received donor bone marrow augmentation.
The standard donor and recipient operations are described elsewhere. However, a few modifications were made to allow safe re-establishment of gut continuity particularly in patients with previously disrupted foregut and ultrashort gut syndrome.
Immunosuppression was tacrolimus based with 19 (83%) recipients pretreated with a single dose of antilymphocyte depleting agent. Bortezomib was recently used in 3 patients with preformed donor-specific antibodies. Full details were recently described.
Nutritional care was a crucial element of the postoperative management. TPN was continued until complete adaptation of native gut and full recovery of transplanted organs. TPN weaning was gradual and guided by dietary measures, body weight, and other nutritional markers. The GLP2 analogue 'teduglutide' was recently offered for patients with short gut syndrome who failed TPN withdrawal. Antimicrobial therapy was required for patients with active infection. Anticoagulation therapy was adopted for patients with prothrombotic syndromes.
Transplant recipients required more diligent postoperative management. Of crucial importance, is immunologic intestinal allograft monitoring with a standard protocol. Intestinal rejection was diagnosed from histopathologic examination of mucosal and full-thickness biopsies. Standard hepatic and pancreatic chemistry profiles were performed periodically to assess organ functions. Diagnosis and treatment of post-transplant lymphoproliferative disorders (PTLD), cytomegalovirus, and graft versus host disease are described elsewhere.
Data were collated into a single electronic file. Demographics of the 142 referrals were limited to types of GF and BS. The descriptive and observation data of the 131 members of the study cohort and 116 surgically treated subjects were pooled according to type of GF. Clinical features of transplant recipients were tabulated according to allograft type.
The body mass index (BMI) was calculated as body weight in kilograms divided by height in meters squared. Based upon the cardiovascular risk, patients were classified into: underweight (<18.5 kg/m), normal weight (≥18.5–24.9 kg/m), overweight (≥25.0–29.9 kg/m), and obese (≥30 kg/m). Obesity is further classified into 3 categories; class I (≥30.0–34.9 kg/m), class II (35.0–39.9 kg/m), and class III (≥40 kg/m).
All values were presented as mean ± standard deviation, median, frequency, and percentage. For continuous variables, differences between groups were assessed with ANOVA and the nonparametric Kruskal-Wallis rank-sum test. Noncontinuous variables and differences in proportions were assessed using the Person's χ test.
The product limit (Kaplan-Meier) was used to calculate survival. With intent to treat, all-cause cumulative risk of mortality of the 131 members of the study cohort was calculated from date of BS until death or last follow-up (FU). For the 116 members of the surgical cohort, the overall survival after restorative surgery, including transplantation, was calculated from date of reconstructive surgery and substratified according to type of GF. Recipient and graft survival was calculated from date of transplant. Group comparison was performed using log-rank test.
Multivariable analysis was performed using multinomial logistic models. Statistical analyses were performed using R software (Version 3.1.1 (2014-07-10), Vienna, Austria) and assume a 5% level of significance.
Methods
Study Design
This is a retrospective analysis of prospectively collected database for patients referred to 2 major academic centers with GF, defined by loss of nutritional autonomy that required or indicated the need for TPN. History of BS was used to identify the study population. Core data was pooled from a computerized database and chart review was conducted to obtain pertinent medical information. Most of the reconstructive and transplant operations were performed by the primary author. Institutional Review Board (IRB) approval was obtained from both centers with an honest broker for data management.
Definitions
With development of GF after BS being the only inclusion criteria, the study population was categorized into 3 groups according to the newly introduced types of GF. Each type was defined by patient presentation, underlying etiology, and associated pathophysiology. Patients with Type-I GF experienced acute catastrophic gut loss due to vascular occlusion. Type-II GF patients suffered major long-lasting technical complications, including gastrointestinal fistulae, loss of gut continuity, and bowel obstruction after multiple corrective surgical interventions. In Type-III, failure to thrive with the development of GF was insidious with progressive course due to clinical dysfunctional syndromes, including motility disorders, restrictive intolerance in absence of mechanical pathology, and gut malabsorption.
The nomenclature of the different bariatric and transplant procedures has been fully defined. The term 'restorative surgery' defines collective procedures that are utilized to restore nutritional autonomy, including autologous reconstruction, intestinal lengthening, and transplantation. The term 'autologous reconstruction' is limited to the description of the different surgical techniques that are utilized to re-establish continuity of the native gastrointestinal tract. Intestinal lengthening is a rehabilitative rather than a reconstructive procedure to enhance gut adaptation.
TPN failure was defined by multiple line infections, central vein thrombosis, and hepatic injury. Nutritional autonomy indicates freedom from TPN and fluid replacement.
Referral Material
The study was conceived with the first BS patient who was referred to the University of Pittsburgh Medical Center (UPMC) for intestinal transplantation, 5 years after the 1990 inception of the program. With the evolving concept of gut rehabilitation, increased volume of referral was observed particularly in recent years (Fig. 1).
(Enlarge Image)
Figure 1.
Total number of patients referred to the UPMC and CCF with gut failure after bariatric surgery substratified according to type of gut failure (A) and type of the bariatric procedure (B). BPD, biliopancreatic diversion; CCF, Cleveland Clinic Foundation; DS, duodenal switch; UPMC, University of Pittsburgh Medical Center.
Over the past 20 years, 1500 adults with GF were referred for gut rehabilitation at UPMC (1995–2015) and Cleveland Clinic Foundation (CCF) (2010–2015). Of these, 142 (9%) had BS that preceded TPN dependency and made up the referral population. BS was performed at different regional, national, and international centers. Causes of GF were Type-I in 63 (44%), Type-II in 45 (32%), and Type-III in 34 (24%). Yearly referrals according to type of GF and BS are shown in Figure 1A and 1B, respectively. Types of primary BS, as documented in medical records, were malabsorptive in 6 (4%), restrictive in 27 (19%), and combined in 109 (77%) with surgical illustrations shown in Figure 1, Supplemental Digital Content, http://links.lww.com/SLA/A846. At referral, 14 patients had prior conversions with jejunoileal bypass (JIBP) to gastric banding (GB) in 4, GB to Roux-en-Y gastric bypass (RYGB) in 9, and sleeve gastrectomy (SG) to RYGB in 1.
Study Population
Upon referral, the obtained medical records were reviewed. Patients with recent catastrophic events and complex abdominal pathology underwent a hospital-to-hospital transfer. Urgent abdominal re-exploration was required for Type-I GF patients with residual infarcted bowel, stapled visceral organs, and intra-abdominal infection. Outpatient candidates underwent thorough evaluation to assess nutritional status, residual gut, splanchnic circulation, and surgical candidacy. Of the 142 referrals, 131 (92%) were evaluated and comprised the study cohort; 73(56%) at UPMC and 58 (44%) at CCF. The remaining 11 patients were denied access by health care plans (n = 8) or died before transfer (n = 3).
Patient Evaluation
All patients underwent a thorough multidisciplinary evaluation. Assessment included clinical presentation, review of organ systems, gut anatomy, nutritional status, comorbidities, and behavioral health issues. In addition to the type of GF-associated clinical presentation, chronic abdominal pain was a common theme in all patients. Biochemical and hematologic testing included battery assessment of protein-calorie malnutrition, nutritional deficiencies, and thrombotic disorders. Radiologic and endoscopic evaluations were used to map the altered gut anatomy and residual intestinal length. Visceral angiograms were obtained in selected patients to delineate vascular anatomy and guide further therapy. Hepatic assessment, including liver biopsy was required for patients with severe morbid obesity, malabsorptive BS, chronic TPN therapy, hostile abdomen, and portomesenteric venous thrombosis. Cardiopulmonary assessment was guided by the existence of comorbidities and other surgical risk factors. In recent cases, full assessment of skeletal health was conducted.
Management Strategy
All patients received full medical, nutritional, metabolic, and psychosocial support. TPN was optimized or initiated to restore nutritional deficiencies. Patients with renal, respiratory, and cardiopulmonary insufficiency were subjected to intense preoperative management.
The primary objectives of surgical management were to restore nutritional autonomy, reduce need for visceral allotransplantation, and avoid recurrence of severe obesity. Restitution of gut anatomy was offered with the premise to normalize gut physiology. Surgical strategies were guided by type of GF, residual gut anatomy, and surgical candidacy. A first-stage operation, including foregut reconstruction, was considered for Type-I GF patients with ultrashort gut syndrome and intra-abdominal infection. With Type-II GF, a single stage autologous reconstruction was planned for patients with gastrointestinal fistulae and disconnected gut with possible reversal of the bariatric procedure particularly in patients with short gut syndrome. In patients with isolated gastrointestinal pathology, the altered bariatric anatomy was preserved. Intestinal lengthening was performed in selected patients with short gut syndrome to enhance gut rehabilitation. With Type-III GF, surgery was considered after failure of comprehensive multidisciplinary medical management and limited to reversal of BS when technically feasible. With all types, portal hypertensive surgery was performed for a few patients with portomesenteric venous thrombosis. Surgical options and potential complications, including relapse of obesity, were discussed with all patients and an informed consent was obtained surgical options and potential complications, including relapse of obesity, were discussed with all patients and an informed consent was obtained.
Surgical Reconstruction
Autologous reconstruction was an organ salvage procedure guided by the embryonic development of the alimentary canal (Figure 2, Supplemental Digital Content, http://links.lww.com/SLA/A846). Technical feasibility was dictated by the complexity of the abdominal pathology, the altered anatomy of the gastrointestinal tract, and integrity of the visceral vascular blood supply. Primary foregut reconstruction was the mainstay of a definitive surgical treatment or a first-step procedure in preparation for visceral transplantation. Interposition of a visceral conduit was used to avoid exclusion of the native duodenum from the alimentary flow and optimize the absorptive capacity of the residual gut.
(Enlarge Image)
Figure 2.
Causes of the 3 different types of gut failure in the 131 study patients; catastrophic gut loss (A, Type-I), technical complications with loss of gut continuity (B, Type-II), and dysfunctional syndromes with different malabsorptive, restrictive, and motility disorders (C, Type-III). With Type-II, some patients have more than one pathology.
In patients with short gut syndrome, a conservative surgical approach was adopted to salvage the survived bowel. The reconstructive techniques aimed at maximization of the absorptive functions of the residual gut. In addition, intestinal lengthening with serial transverse enteroplasty (STEP) was performed in patients with less than 100 cm of small bowel.
Visceral Transplantation
Visceral transplantation was indicated for patients with ultrashort gut syndrome and TPN failure. The number of transplanted organs was dictated by the extent of gut loss and coexistence of liver failure. Donors were deceased and allografts were ABO identical. Human leukocyte antigen (HLA) match was random with positive T/B cell lymphocytotoxic crossmatch in 9 (36%) patients. No attempts were made to immunomodulate the allografts and no recipients received donor bone marrow augmentation.
The standard donor and recipient operations are described elsewhere. However, a few modifications were made to allow safe re-establishment of gut continuity particularly in patients with previously disrupted foregut and ultrashort gut syndrome.
Immunosuppression was tacrolimus based with 19 (83%) recipients pretreated with a single dose of antilymphocyte depleting agent. Bortezomib was recently used in 3 patients with preformed donor-specific antibodies. Full details were recently described.
Postoperative Care
Nutritional care was a crucial element of the postoperative management. TPN was continued until complete adaptation of native gut and full recovery of transplanted organs. TPN weaning was gradual and guided by dietary measures, body weight, and other nutritional markers. The GLP2 analogue 'teduglutide' was recently offered for patients with short gut syndrome who failed TPN withdrawal. Antimicrobial therapy was required for patients with active infection. Anticoagulation therapy was adopted for patients with prothrombotic syndromes.
Transplant recipients required more diligent postoperative management. Of crucial importance, is immunologic intestinal allograft monitoring with a standard protocol. Intestinal rejection was diagnosed from histopathologic examination of mucosal and full-thickness biopsies. Standard hepatic and pancreatic chemistry profiles were performed periodically to assess organ functions. Diagnosis and treatment of post-transplant lymphoproliferative disorders (PTLD), cytomegalovirus, and graft versus host disease are described elsewhere.
Data Management and Statistical Analysis
Data were collated into a single electronic file. Demographics of the 142 referrals were limited to types of GF and BS. The descriptive and observation data of the 131 members of the study cohort and 116 surgically treated subjects were pooled according to type of GF. Clinical features of transplant recipients were tabulated according to allograft type.
The body mass index (BMI) was calculated as body weight in kilograms divided by height in meters squared. Based upon the cardiovascular risk, patients were classified into: underweight (<18.5 kg/m), normal weight (≥18.5–24.9 kg/m), overweight (≥25.0–29.9 kg/m), and obese (≥30 kg/m). Obesity is further classified into 3 categories; class I (≥30.0–34.9 kg/m), class II (35.0–39.9 kg/m), and class III (≥40 kg/m).
All values were presented as mean ± standard deviation, median, frequency, and percentage. For continuous variables, differences between groups were assessed with ANOVA and the nonparametric Kruskal-Wallis rank-sum test. Noncontinuous variables and differences in proportions were assessed using the Person's χ test.
The product limit (Kaplan-Meier) was used to calculate survival. With intent to treat, all-cause cumulative risk of mortality of the 131 members of the study cohort was calculated from date of BS until death or last follow-up (FU). For the 116 members of the surgical cohort, the overall survival after restorative surgery, including transplantation, was calculated from date of reconstructive surgery and substratified according to type of GF. Recipient and graft survival was calculated from date of transplant. Group comparison was performed using log-rank test.
Multivariable analysis was performed using multinomial logistic models. Statistical analyses were performed using R software (Version 3.1.1 (2014-07-10), Vienna, Austria) and assume a 5% level of significance.
