The Norwegian Pancreas Transplantation (PTx) Study

1. INTRODUCTION AND BACKGROUND The first pancreas transplantation (PTx) was performed in Minnesota in 1966 by Kelly and colleagues (1). In recent years the number of procedures has grown considerably worldwide, and is now a well established treatment option for patients with diabetes mellitus with and without concomitant diabetic End-Stage Renal Disease (ESRD) (2-4). The indication for PTx is advanced and/or badly controlled diabetes mellitus ("brittle" diabetes, severe hypoglycemic episodes, "unawareness", etc). Solitary pancreas transplantation (SPT; without concomitant kidney transplantation) is usually classified as PTx alone (PTA), PTx after kidney transplantation (PAK) or PTx after islet transplantation (PAI). Kidney transplantation of the diabetic uremic population increases survival compared to long-term dialysis (5, 6). Transplant options for patients with diabetic end-stage nephropathy include simultaneous pancreas-kidney (SPK), live donor kidney (LDK) and deceased donor kidney (DDK) transplantation. SPK transplantation relieves not only the patient's uremia, but also alleviates the hyperglycaemic state of diabetes. Large international patient registries show that patient survival rates after SPK have reached more than 95% at 1 year and 87% at 5 years post-transplant, respectively (2). Nevertheless, PTx as treatment for type 1 diabetes has not gained the same popularity as transplantation of other organs, partly because PTx have been associated with a high rate of surgical complications; particularly bleeding, thrombosis and exocrine leakage. Furthermore, there has been a lack of reliable, non-invasive rejection monitoring instruments, and the invasive, percutaneous pancreas biopsies have been associated with a high rate of complications. The difficulties encountered with PTx have to some extent been compensated by a very selective attitude towards the donors, but thereby making pancreas grafts a scarce resource. In contrast to other abdominal transplantations such as liver transplantation (LTx) and kidney transplantation (KTx), where repeated biopsies have been used for immunosurveillance, percutaneous biopsies of the pancreas-graft have traditionally been avoided due to a high rate of biopsy-related complications (exocrine leaks/fistulas and bleeding episodes). Thus, fear of acute rejections and lack of adequate rejection markers, have led to a rather intensive immunosuppressive load in PTx recipients. Solitary pancreas transplantation (SPT) has traditionally been subjected to even higher complication and rejection rates, with inferior graft and patient survival - thus favoring the combined SPK procedure. This has been attributed to an even worse rejection monitoring capability, without a "reporter" allograft kidney. No biochemical markers have proven to be effective in rejection surveillance. Pancreas graft thrombosis is a feared complication in the postoperative course, partly due to the oversized vessels used (coeliac trunk/superior mesenteric artery/portal vein) in conjunction with the low blood flow through an isolated pancreas graft. In the native setting, these vessels also serve the intestines and spleen. Therefore, PTx poses a delicate balance between thrombosis and bleeding complications. The Norwegian experience: PTx is performed at one single national centre in Oslo, and from 1983 to date 300 procedures have been performed (7-11). In recent years, the activity has increased, reaching 28 in 2012. Approximately 9 out 10 PTx's have been SPK's. In the first period from 1983 through 1987, a duct-occluded segmental pancreas was used for transplantation. From 1988, the whole pancreas graft was used, and the exocrine secretion was drained by anastomosing the duodenal segment to the urinary bladder. This technical solution was chosen partly because it offered some sort of rejection monitoring, by urine amylase counts and cystoscopic pancreas biopsies. However, many patients suffered from chemical cystitis and metabolic acidosis, due to loss of bicarbonate. In 1998 the urinary bladder anastomosis was abandoned, in favor of the more physiological enteric anastomosis, the duodenal segment being connected to the proximal jejunum. However, this solution offered even less options to monitor upcoming rejections, as percutaneous biopsies was mostly avoided due to the previously mentioned hazards. The investigators have recently examined (12) all PTx's performed at the investigators' hospital during 2006-2010 (n=61; 59 SPK, 2 PTA). The investigators' overall surgical complication rate has decreased from earlier years, but we still suffer a substantial rate of reoperations (about 30% of patients), mainly caused by exocrine leakage, bleeding and vein thrombosis. When comparing the populations with or without reoperation, higher donor age had a significant negative impact. No significant effect of donor age on graft survival was observed. There was a tendency towards better results in female recipients, both regarding surgical complications and graft survival. The rejection rate (altogether about 30%) was significantly higher in the graft loss group. From late 2011, several measures have been implemented to improve outcome and reduce the rate of surgical complications. In line with most Tx centres in Scandinavia, the investigators have switched the prophylactic anticoagulation treatment from the investigators' traditional Macrodex® regime to a Heparin®/Fragmin® regime. However, from June 2016 we give Macrodex intraoperatively and at postoperative 1, and acetylsalicylic acid is started between postoperative day 3 and 5 instead of at day 7. These measures were undertaken since our venous thrombosis rate still seem to exeed 20 %. Several technical changes have also been implemented during recent years; more atraumatic graft procurement, preserving the entire coeliac arterial axis including the gastroduodenal artery, obtaining a long portal vein without the need for elongation, as well as extended in situ dissection by means of LigaSure. Due to the conventional lack of rejection monitoring parameters, the investigators launched an investigatory surveillance program, with protocol biopsies of the duodenal segment via double balloon enteroscopy (13). The impact and value of this program has yet to be investigated. Previous reports have described separate rejection of the pancreas or kidney in the SPK setting, and the gold standard for proving rejection of the pancreas is undoubtedly a biopsy of the pancreas itself. This encouraged us to further develop techniques for better surveillance, such as Endoscopic transduodenal UltraSound-guided Pancreas Biopsies (EUSPB). Inferior outcome of PTA and lack of valid tools for immunosurveillance in the absence of a simultaneous kidney graft, have led some centers to evolve the duodenoduodenostomy (DD) for drainage of the exocrine pancreas, making the EUSPB possible. There are many theoretical advantages with the DD, especially regarding rejection surveillance, and we have recently adopted this technique. The endoscopic access afforded by the DD also makes it possible to stent the pancreatic duct in case of exocrine leakage. Though, in recent years we have experienced a very low incidence of complications with the conventional percutaneous ultrasound-guided pancreas biopsy. In our latest, retrospective study (Horneland et al., Am J Transpl; 15(1): 242-50, 2015), focusing on the duodenoduodenostomy, there were no complications among 18 percutaneous pancreas biopsies performed. Immunosuppressive therapy: Over time, the induction therapy and maintenance immunosuppressive protocols have changed. From 1983 to 2000, all recipients received triple immunosuppressive regimens with cyclosporine, azathioprine and prednisolone (CS). During the last part of the 1990's azathioprine was substituted by Mycophenolate mofetil (MMF), and cyclosporine was substituted by tacrolimus. After 2000, the immunosuppression has been intensified by induction therapy both for PTx (Antithymocyte globulin (ATG)) and for kidney transplants alone (basiliximab). Thus in recent years, PTx recipients have received a quadruple immunosuppressive regimen, that includes tacrolimus, MMF, high dose CS and ATG. The dosage of ATG has been directed by T-cell counts. Rejections/Donor-specific antibodies (DSA): During recent years, with the quadruple immunnosuppressive regimen, our biopsy-verified rejection rate has been about 30% (12). In this study we will follow the routine protocol for treatment of rejections; primarily more CS (5-8 doses of SoluMedrol), secondarily more ATG (2-5 doses; T-cell directed) A recent study (14) assessed the role of post-Tx HLA antibody monitoring in the surveillance of PTx recipients, and the impact of DSA. Four hundred thirty-three PTx's were performed at the Oxford Transplant Centre (317 SPK/116 Sol-PTx). It was demonstrated that 39.8% of patients developed de novo HLA antibodies, of which 36.9% were de novo DSA. The development of antibodies to donor HLA, but not to nondonor HLA, was significantly associated with poorer graft outcomes, with 1- and 3-year graft survival inferior in SPK recipients, and interestingly even more so in Sol-PTx recipients. In a multivariate analysis, development of de novo DSA emerged as a strong independent predictor of pancreas graft failure. These findings have urged us to investigate de novo DSA development in the present study. Graft monitoring with microdialysis catheters: As stated in the previous paragraphs the complication rates following pancreas transplantation are high. Except maybe for severe hemorrhage, all complications have in common that they are difficult to detect. Accordingly, there is an emerging need for better monitoring of pancreas transplants. We consider that further improvement of surgical techniques and immunosuppressive protocols rely upon better monitoring tools. Microdialysis is a technique, which enables close to 'real time' monitoring of tissues and organs of interest. Depending on the membrane's pore size, metabolic substances (lactate, pyruvate, glucose and glycerol) and/or mediators of inflammation (cytokines, chemokines and complement factors) are sampled in a feasible way (15). So far, the method's ability to detect brain ischemia is the best validated (16). In the United States clinical application is so far restricted to neurointensive care units, as only the brain catheter (CMA 70, CMA Microdialysis AB, Stockholm, Sweden) is approved by the Food and Drug Administration for clinical use. However, there are more than 2000 clinical reports on microdialysis catheters, and in Europe the catheters are Conformité Européenne marked for a wider range of indications. We have done extensive clinical observation trials using microdialysis catheters in liver transplanted patients and we have inserted more than 200 catheters in hepatic tissue without experiencing any major complications (17-21).Graft thrombosis has been detected almost in 'real time' as elevated intrahepatic lactate and lactate to pyruvate ratio. Rejection has been detected several days before the rise in conventional blood markers (bilirubin and transaminases) by elevated lactate and with unchanged lactate to pyruvate ratio. Acute cellular rejections were detected with more than 80 % sensitivity and specificity. Ischemic complications like hepatic artery thromboses have been detected with 100 % sensitivity and specificity. We also revealed potentially specific biomarkers for ischemia (complement factor 5a) and rejection (CXCL-10) (19) . We have now implemented microdialysis as routine standard of care in pediatric liver transplants. We are also investigating the potential role of microdialysis in monitoring patients who have undergone Whipple's operation for pancreatic or duodenal cancer. Preliminary results show that leakages in the pancreaticojejunostomy can be detected at a very early time point, by increased concentrations of glycerol in samples collected from catheters positioned close to the enteroanastomosis.
2. OBJECTIVES OF THE STUDY Several studies have shown acceptable results after PTx by substituting ATG with basiliximab (14-18), which is considered to convey a considerably lower number of adverse events. However, the investigators' experiences with ATG in PTx (introduced in 2004) are good, and the investigators presumably gentle way of administrating the drug - directed by T-cell counts - is in fact unique (12). The potential advantages of reducing the overall corticosteroid (CS) load is obvious, as CS is a well-known pro-diabetic agent and causes severe long term adverse effects (14). On this background, we have very recently reduced our CS dosing (in the routine protocol) to a level corresponding to our Kidney Tx protocol (valid since 2009). Thus, we intend to prospectively investigate and compare a single cohort of our present PTx immunosuppressive protocol with previous (historical) cohorts. The rationale for the study is that; i) a high immunosuppressive load, and in particular CS, may be partly responsible for the high rate of PTx associated complications/reoperations; ii) a high immunosuppressive load is related to infectious complications; iii) improved PTx rejection surveillance by DD and EUSBP allows a lower-graded immunosuppressive protocol; iv) evaluating the surgical and medical measures made to our PTx programme during the past two years For detailed description of objectives/aims: See "Outcome measures" below.
3. STUDY DESIGN This is a prospective, single cohort observational study, aimed at using a historical control group as comparison. It will be conducted at our single, national centre for organ transplantation in Oslo. All pancreas recipients > 18 years of age, who fulfill the inclusion criteria, will prior to transplantation be asked for inclusion.
4. DURATION OF STUDY All consecutive PTx recipients during 3-5 years are planned to be enrolled, with a intended number of 60-80 patients. The study will continue until all patients have completed a minimum of 60 months of follow-up or have discontinued participation in the study.
5. NUMBER OF PATIENTS 60-80 patients will be enrolled in the study and all will receive the standard quadruple immunosuppressive regime with reduced CS dosing compared with previous cohorts (according to newly changed routine protocol).
6. SELECTION OF PATIENTS See "Eligibility Criteria" below.
7. DOSAGE AND ADMINISTRATION

7.1 Immunosuppression

The single cohort study group will receive our routine immunosuppressive regimen based on ATG, tacrolimus, mycophenolate mofetil and corticosteroids as follows:

7.1.1 ATG (Thymoglobulin): Initiated at day 0 (the first dose preoperatively) at a dose of 2,5 mg/kg i.v. Later dosing is directed by T-cell counts once daily. The T-cells are kept suppressed for 10 days post-Tx, and new doses of 1,0-2,5 mg/kg i.v. is given whenever the T-cell count rises above 0,050 x109. Altogether, 2-4 doses of ATG is usually needed.

7.1.2 Tacrolimus:

Initiated at day 0 (the first dose preoperatively) at a dose of 0,06 mg/kg x 2 p.o., later adjusted to achieve steady state whole-blood trough levels as follows:

Month 1-3 8-12 ng/ml Month 3-6 4-8 ng/ml

7.1.3 Mycophenolate mofetil (MMF): MMF will be given 1000 mg twice daily. It can be reduced to 750mg twice daily in case of adverse events and further down to 500mg in case of persisting adverse events.

7.1.4 Corticosteroids: Day 0 : Methylprednisolone 250 mg i.v. (peroperatively) Day 1-14: Prednisolone 20 mg x 1 p.o. Day 15-28: Prednisolone 15 mg x 1 p.o. Day 29-60: Prednisolone 10 mg x 1 p.o. Day 61- 180: Prednisolone 7,5 mg x 1 p.o. Day 181 - : Prednisolone 5 mg x 1 p.o.

7.2 Concomitant Treatments

7.2.1 Required treatment

1. Prophylaxis against the development of Pneumocystis carinii, with trimethoprim-sulfa is required for all patients during the first 6 months of treatment. ii) Prophylaxis against Cytomegalovirus (CMV) with valganciclovir for 3 months, if the donor is CMV + and the recipient is CMV ÷. - By all other CMV constellations, preemptive valganciclovir treatment is given, based on weekly CMV-PCR analyses (cut off: CMV-PCR count > 0). iii) Antibiotic prophylaxis with meropenem (2 doses) and vancomycin (1 dose) at day 0. (iv) Proton pump inhibitor (pantoprazole/esomeprazole) is given for at least 2 months post-Tx.

7.2.2 Prohibited treatment

1. Investigational study drugs ii) NSAID's should be avoided iii) Terfenadine, cisapride, astemizole, pimozide, cimetidine and ketoconazole are not allowed.
2. ANALYSES OF RESULTS

8.1 Power calculations

By conventional presumptions (Power 1-β=80%, α=0,05) based on binomial distribution, 3-400 patients (and 3-400 historical controls) would be needed to detect a 33% relative change in rejection/complication rates.

However, regarding practicability:

(i) It is totally unrealistic - for any Tx-center in the world - to include 3-400 PTx patients, during any reasonable time frame. In Oslo, the investigators are by far the highest volume center in Scandinavia. The investigators' 28 PTx's performed in 2012 represent 5,6 p.m.p. (per million population), which actually is far higher than any other country in the world. Even if the investigators cooperated with all the other PTx centers in Scandinavia (Uppsala/Göteborg/Helsinki), the potential would not exceed 50 patients per year.

(ii) The maximally realistic number of PTx patients to be included in Oslo during a reasonable time frame (2-3 years) will be 60-80.

(iii) Thus, the investigators' intentions with regard to statistical Power have to be more modest - for a single cohort observational study. And a doubled rejection rate can be detected at 80% Power with only 42 patients.

(iv) The prospects/visions of this study consists of a lot more than detecting significant changes in rejection/complication rates. The simultaneous biopsy strategy (D- + P- + K-biopsies) is unique. And the 'molecular biology' analyses of these simultaneous biopsies and blood samples have the potential to provide new insights. Furthermore, "new" potential rejection markers (C-peptide; CRP/Amylase/Lipase combined parameter) will be explored.

8.2 Statistical methods in data analysis

• The analysis of categorical parameters will consist of:

1. Comparisons of groups using the Fisher exact test. ii) Confidence intervals of 95% of the percentage of incidence of these events.

• The loss of grafts and deaths will be analyzed by the Kaplan-Meier method for estimating the time to events.
• Continuous (non-categorical) variables will be analyzed by student t-tests and chi-square tests.