Liver disease, including biliary tract disease, affects approximately 9-37% of patients with cystic fibrosis (CF) and is the second most important cause of death after respiratory failure.1 Focal biliary cirrhosis and resultant portal hypertension are found in 0.8-1.7% of all patients with CF, with an increased incidence in patients 16-20 years old,2 although some studies have reported an incidence as high as 8%.3 Historically, the risk of pulmonary decompensation in patients with CF made liver transplantation risky. However, with improvements in CF treatment and survival, liver transplantation has become an acceptable option for patients with end-stage liver disease and portal hypertension secondary to CF. Multiple studies have established that CF patients have satisfactory survival after liver transplantation and have improvement in their pulmonary function compared to their pretransplant status.4–8
Pancreatic dysfunction is also widespread in patients with CF, with 85% of CF patients developing pancreatic exocrine insufficiency and up to 34% developing diabetes requiring insulin.9, 10 In addition, the immunosuppressive regimens needed for successful liver transplantation are also associated with the development of diabetes in 15% of recipients.11 Patients with CF and pancreatic insufficiency may be at an increased risk of developing diabetes after transplant, and those who already have diabetes may have worsening of their glycemic control.
There have been 4 case reports of successful combined liver-pancreas transplant for children with CF,12, 13 including 1 report using an en bloc technique.14 In this paper, we describe a series of 3 patients with CF who underwent en bloc liver-pancreas transplant and compare them to patients with CF who underwent liver transplant alone.
Since 1990, 214 pediatric liver transplants have been done at our institution. Nine consecutive patients transplanted for end-stage liver disease secondary to CF were reviewed. Patients' mean age was 13 (range, 4 to 19) years. Six patients received a transplant of the liver only (LO), while 3 patients underwent combined en bloc liver-pancreas (LP) transplantation. All 9 patients had pretransplant pancreatic exocrine insufficiency requiring enzyme supplementation. Five patients (55%) had preoperative insulin-dependent diabetes, including all 3 patients who underwent liver- pancreas transplantation.
All liver transplants were performed orthotopically using well-described standard techniques. Liver-pancreas transplants were performed en bloc. The LP graft was sewn in to the inferior vena cava in standard piggyback fashion. The portal vein of the recipient was anastamosed to the superior mesenteric vein below the neck of the pancreas graft (Fig. 1). An iliac artery Y graft from the donor was anastamosed to the celiac axis and superior mesenteric artery of the en bloc graft, which was then connected to a previously placed infrarenal aortic jump graft. The pancreas graft was draped directly over the native pancreas. After reperfusion of the graft, biliary continuity was reestablished with a Roux-en-Y duodenojejunostomy (Fig. 2).
Immunosuppression was a calcineurin inhibitor (cyclosporine before 1997 and tacrolimus after 1997) and a steroid taper with maintenance low-dose prednisone. Mycophenolate mofetil was added in select patients. Rejection was treated with a standard steroid pulse and taper, and steroid-resistant rejection was treated with OKT3.
For both LO and LP patients, 1-year patient survival was 100%. There was 1 death 14 months after transplant (LO) from Pneumocystis carinii pneumonia. Five-year survival was 83% for the LO patients and 100% for the LP transplant patients. There were 2 late deaths among the LO patients, 6 and 12 years after transplant. The first death was from noncompliance with insulin and end-stage diabetes, and the second was from end-stage pulmonary CF. There was a single graft loss in a LO patient 1 year after transplant from chronic rejection. This patient was successfully retransplanted.
Postoperative complications included 2 bile duct complications (1 leak and 1 stricture) among the LO patients; both required conversion to a Roux-en-Y hepaticojejunostomy. There were no episodes of vascular thrombosis or primary nonfunction. Postoperative infectious complications did not differ between the 2 groups. Infections developed in 7/9 patients (77%, 3 LP and 4 LO). Pneumonia developed in 5 patients (55%) (3 fungal and 2 bacterial). Of interest, 4 (44%) patients in the LO group developed cytomegalovirus infections. One patient developed sinusitis requiring operative drainage. Four (44%) patients developed acute rejection (2 LO, 2 LP); 1 case resolved with steroid pulse only, and 3 required additional treatment with OKT3. Chronic rejection developed in 2 LO patients; 1 required retransplantation. Both of these patients were teenagers at the time of transplantation and had issues with medication compliance. One patient had improvement after switching from FK-506 to sirolimus and compliance. The other patient switched from cyclosporine to FK-506 but still progressed to graft failure requiring retransplantation.
Follow-up ranged from 11 months to 11 years, with a mean of 4 years. All 3 liver-pancreas transplant patients are off insulin and do not require pancreatic digestive enzyme supplementation. These patients have normal posttransplant fasting blood sugars. Of the patients who received liver transplants alone, 2 were insulin dependent prior to surgery and remain insulin dependent after transplant. Of the 4 nondiabetic patients who received LO, 3 developed insulin-dependent diabetes mellitus within 3 months after transplant. Patients receiving LO were more likely to be diabetic and require exogenous pancreatic enzymes after transplant than those receiving LP (83% vs. 0%, P = 0.04, and 100% vs. 0%, P = 0.01, respectively).
Traditionally, pancreas transplantation has been reserved for patients with end organ damage from type I diabetes mellitus. However, pancreas transplants have been utilized as treatment for other diseases, including upper-abdominal malignancies15 and chronic pancreatitis.16, 17 Combined liver-pancreas transplants are also rare. One series describes a group of 5 patients with various causes of end-stage liver disease and type II diabetes mellitus who successfully underwent simultaneous but separate liver and pancreas transplants.18 Another case report describes 2 pediatric patients with end-stage liver disease and type I diabetes mellitus who benefited from en bloc liver-pancreas transplantation.19
In CF, there are 4 case reports of either successful multivisceral or liver-pancreas transplants. The first multivisceral transplant described in a CF patient was pancreas-liver-kidney transplant in 1994, for a 21-year-old male with pancreatic insufficiency requiring pancreatic enzymes and insulin, as well as cirrhosis and renal failure.12 This patient had normal pancreatic endocrine and exocrine function 24 months after transplant. Subsequently, 2 case reports describe 3 patients (13, 22, and 25 years old) with CF, end-stage liver disease and diabetes mellitus who underwent either simultaneous separate liver-pancreas13 or en bloc liver-pancreas transplants.14 Similarly, all 3 patients had improvement of their liver function and resolution of portal hypertension, and they no longer need either pancreatic endocrine or exocrine supplementation. There is also a case report of a successful combined liver and small intestine transplant in a child CF and short-gut syndrome after multiple bowel resections for meconium ileus.20 Postoperatively the child has not had any episodes of rejection and requires pancreatic enzyme supplementation but has not developed diabetes mellitus.
The risk of CF-related diabetes has increased dramatically in the last 2 decades as survival of CF patients has improved. More than 25% of patients over the age of 20 develop diabetes.21 For a patient with CF, developing diabetes leads to pulmonary decline22 and increases mortality 6-fold.23 A prospective trial of children with CF followed 2 groups of children with and without diabetes mellitus over 5 years. The group that developed diabetes mellitus had a progressive decline in their forced expiratory volume at one second and forced vital capacity over the 5 years and had twice as many deaths as the control group.24 Patients with CF have an increased risk of developing diabetes after transplantation due to the preexisting pancreatic damage and the diabetogenic effects of steroids and calcineurin inhibitors. In a study of 77 patients undergoing lung transplantation, the prevalence of diabetes increased from 28.6% before transplant to 49.7% after transplant.25 In the current series, the prevalence of diabetes in the LO patients increased from 33% before transplant to 83% posttransplant and decreased from 100% to 0% in the LP transplants.
In this series, the combined liver-pancreas transplant patients clearly benefited from the pancreas transplant, as they were significantly less likely to require pancreatic enzymes or insulin after transplant compared to the LO group, without an increase in complications or a decrease in survival. This benefit not only prevents the long-term complications of diabetes mellitus and improves quality of life associated with diabetes, but it also may prevent pulmonary decline and decrease mortality.
The benefit of the en bloc technique is a simpler operation with fewer anastamoses, along with the possible benefits of the physiologic insulin release into the portal circulation. Possible advantages of the en bloc technique include portal venous drainage, which will increase hepatotrophic insulin flow to the liver and decrease the risk of venous torsion,18 and an immunologic benefit of the liver acting as a sink for circulating antigen and inducing tolerance to pancreatic graft.26, 27 There are no case reports that describe a downside to combined liver-pancreas transplants, but this fact may represent a reporting bias. Clearly, repeat liver transplantation in the patients receiving en bloc grafts may be more difficult, as the entire graft would have to be removed. We have not encountered this situation to date in our patient population.
In conclusion, patient and graft survival is equivalent among patients with end-stage liver disease secondary to CF who receive either LO or LP transplants. Patients who receive LP transplants are more likely to remain off insulin and exogenous pancreatic enzymes in the transplant setting. Patients with CF who do not develop diabetes mellitus live longer and have better pulmonary function than patients who develop diabetes mellitus. Patients with CF and end-stage liver disease requiring liver transplantation may benefit from combined liver-pancreas transplantation, particularly if they have preexisting pancreatic endocrine and exocrine insufficiency.