EXPERIMENTAL AND CLINICAL TRANSPLANTATION

Objectives: Pancreas transplant is an effective long-term treatment modality for complicated type 1 diabetes mellitus. However, allograft failure or severe concomitant rejection remain an obstacle to successful transplant outcome, occurring in approximately 21% of recipients within 1 year. Most histologic studies investigating the cause of pancreas transplant failure have concentrated on the presence and severity of acute and chronic cellular or vascular rejection. After vascular thrombosis, graft pancreatitis is the second most frequent complication after transplant.

Materials and Methods: We conducted a retro-spective analysis, collecting information from a contemporaneously maintained database of patients after pancreas transplant.

Results: We identified 44 patients with rejected allografts from a database of 196 pancreas transplant patients (44/196, 22%). In these identified rejected allografts, 27 patients (61%) had histopathology reports containing 1 or more terms associated with pancreatitis, with the most common histologic finding was being fat necrosis (21/27, 83%), followed by inflammatory or neutrophil infiltrate (13/27, 48%). Sixteen of these patients (60%) had two 2 or more terms histology terms descriptive of pancreatitis records. Ten of the 44 rejected allografts, 10 patients had histologic evidence of vascular or cellular rejection. There was no significant difference in the proportions showing evidence of rejection between groups with (2/27 patients [26%]) and without (3/17 patients [18%]) descriptions of pancreatitis in their medical records (P = .70). When time from transplant to pancrea-tectomy was analyzed, a larger proportion of pancreatectomies occurred late for patients with descriptions of pancreatitis in their medical records versus patients without (17/26 [65%] vs 4/16 [25%]; P = .05).

Conclusions: This case series demonstrates that 61% of rejected allografts over a span of 13 years at a single center had histologic features of graft pancreatitis, suggesting that pancreatitis may be a contributory mechanism to graft failure.

Key words : Pancreatitis, Pancreas, Transplant

Pancreas transplantation has evolved to become an effective long-term treatment modality for com-plicated type 1 diabetes mellitus. However, allograft failure or severe concomitant rejection remain an obstacle to successful transplant outcome, occurring in approximately 21% of recipients within 1 year. Histologic analyses of the cause of pancreas transplant failure have focused on the presence and severity of acute and chronic cellular or vascular rejection as potential reasons for allograft loss.^1-3 After vascular thrombosis, graft pancreatitis is the second most frequent complication after pancreas transplant, particularly early after transplant.⁴ Histologic features of pancreatitis have also been reported in biopsies of the rejected allograft.^5,6

Graft pancreatitis has been classified into 2 types: acute and chronic, with acute graft pancreatitis being further distinguished as physiologic, early, and late.7 Physiologic acute pancreatitis is related to ischemia-reperfusion injury and is usually clinically silent. It is a timely and prognostically self-limited process. Early acute pancreatitis occurs within 3 months after pancreas transplant in 35% of recipients and is associated with high rates of graft loss (78%-91%). Clinical signs are pain, systemic inflammatory response, and hematuria in the allograft with exocrine drainage via the bladder. Therapy is multimodal and can be medical, radiologic, or surgical. Late acute graft pancreatitis occurs more than 3 months after transplant in 14% to 25% of recipients and represents an uncommon cause of graft loss. Typical clinical signs are pain, abdominal tenderness, and fever. Typical laboratory signs are hyperamylasemia, hyperglycemia, and raised creatinine levels if there is associated renal injury. Therapy is usually conservative unless there is the rare complication of a transplant pancreas pseudocyst.8 Chronic graft pancreatitis is difficult to distinguish from chronic rejection and is associated with graft loss in 4% to 10% of recipients. The potential role of graft pancreatitis in the process of failure and rejection has never been clearly delineated.

We aimed to describe the frequency of histologic features of pancreatitis (as previously described) in a series of rejected pancreas allografts from a single center.

Materials and Methods

A retrospective analysis was conducted on a contemporaneously maintained database containing records of all pancreas transplants undertaken at the Manchester Royal Infirmary since the initiation of its pancreas transplant program. All patients who had undergone transplant pancreatectomy from January 1997 through May 2010 were identified, and the histology records (as performed by experienced histopathologists) of each rejected pancreas allograft were obtained.

A literature search was undertaken to identify histopathologic terms most commonly associated with acute and chronic pancreatitis.^4-9 A total of 9 terms considered specific for either acute or chronic pancreatitis were identified (see Table 1). Some histopathologic terms that are commonly associated with pancreatitis but were considered to be nonspecific were not included in the final list (including “abscess” and “hemorrhage”).

Histologic records of patients who had undergone transplant pancreatectomy were reviewed for the inclusion of each of these terms. Each identified term found was assigned a value of 1, and a total “pancreatitis score” was calculated for each rejected allograft. Repeated terms in the same histology record were only counted on 1 occasion per case. Histologic records were also analyzed for statements indicating the clear presence of chronic or acute rejection and vascular thrombosis.

We also obtained individual patient demographic information and data on surgical procedures for all patients who had undergone pancreas transplant. For patients who had undergone more than 1 pancreas transplant removal, each rejected allograft was treated as 1 event.
Statistical analyses were conducted with StatsDirect Statistical Software (StatsDirect, Altrincham, UK).

Of 196 pancreas transplants conducted during the study period, 44 patients who had pancreas transplant removal (22%) were identified from the database. Of these, 27 patients (61%) had histo-pathology records containing 1 or more terms associated with pancreatitis and thus were deemed to show evidence of pancreatitis (Table 1). Two patients had 2 failed pancreas transplants during the study period.

Patient demographic information for both the patients without rejected allografts and those with rejected allografts are shown in Table 2. Available details on surgical care and procedures are shown in Table 3. No statistically significant differences were shown between any of the groups regarding demographic and available surgical data. All patients were given anticoagulants as part of perioperative care.

In the 27 of 44 rejected allografts (61%) with evidence of pancreatitis, the most common histologic finding was fat necrosis (21/27, 83%), followed by inflammatory or neutrophil infiltrate (13/27, 48%) (Figure 1). In 16 of 27 rejected allografts with evidence of pancreatitis (60%), 2 or more terms descriptive of pancreatitis were shown in the histology records based on the validated list of terms (Figure 2).

In total, 10 of the 44 rejected allografts had clear histologic evidence of vascular or cellular rejection. No significant differences were shown between patients who had (7/27 patients; 26%) and those who did not (3/17; 18%) have documentation of pan-creatitis regarding evidence of rejection (P = .7).

When the time from transplant to pancreatectomy was analyzed, a larger proportion of pancreatectomies occurred late (≥ 7 days posttransplant) in patients with pancreatitis versus those without (17/26 [65%] vs 4/16 [25%]; P = .05) (Figure 3). Data were not available for 2 patients.

In this single center study, 44 patients (22%) had graft loss over the study period of 13 years. This is a relatively high proportion and includes several graft loss during the learning curve of establishing a new transplant program. Of these rejected pancreas allografts, 27 (61%) had histopathology records containing 1 or more terms associated with pancreatitis, which we deemed as evidence of pancreatitis. In total, 14% of rejected allografts over the 13-year study period had histologic features of graft pancreatitis, suggesting that pancreatitis may be a contributory mechanism to graft failure. There is inevitable difficulty in definitively identifying those pancreata that showed evidence of primary pancreatitis as opposed to the other inflammatory causes of pancreas transplant failure (eg, sepsis or inflammation secondary to rejection). In addition, the relative paucity of studies defining the histologic classification of pancreatitis must also be considered, with the majority of classification systems being clinical in nature.^9-15 Transplant clinicians and pathologists remain reliant on classifications of pancreatitis related to the nontransplant population, further hindering the understanding of patients within the transplant population.

The exact nature and timing of the features of pancreatitis related to graft failure clearly need further definition. However, the effects of ischemia-reperfusion injury may be important in driving the proinflammatory process. Further work may be required to delineate whether any preventative strategies for ischemia-reperfusion injury may ameliorate the effects of graft pancreatitis.

Despite these limitations, it appears that docu-mentation of histologic features of pancreatitis, as either a primary cause or sequelae of other pathologies, is a common finding in rejected pancreas allografts. The pancreatectomies in the patients with documented pancreatitis occurred significantly later than in patients who had no documentation, suggesting that different pathologic mechanisms are responsible. Early graft failure often reflects the effects of acute thrombosis, major hemorrhage, nonperfusion of the organ before transplant or at organ retrieval, or (rarely) hyperacute rejection. This is likely to be a different subgroup from those who lose their pancreas over weeks and months and more than likely driven by distinct molecular mechanisms.5 Given the relatively low incidence of clear rejection in the allografts showing signs of pancreatitis, we hypothesize that graft pancreatitis is a more common and important complication than previously understood. There is a need for further definition and understanding of the complex pathophysiology in transplant pancreatitis in this subgroup of patients.

Volume: 13 Issue: 5 October 2015

FULL TEXT