Transplantation

Islet transplantation (ITx) is rapidly becoming an attractive and successful approach for the treatment of selected patients with type 1 diabetes (¹). Recent reports have shown insulin independence for more than 2 years after transplantation using a glucocorticoid-free immunosuppressive regimen (^2,3). Injection of purified islets into the recipient’s liver through the portal vein has to date been found to be the most effective route for islet transplantation (⁴). Data from our clinical islet program has shown that multiple sequential islet transplants can be safely performed through the portal vein, provided that care is taken with islet purification and attention is given to portal venous pressure monitoring during islet infusion (⁵).

Despite increasing success with ITx using the intraportal route, perturbations in liver function have not been well characterized after this procedure. Early reports of major complications after injection of nonpurified islets into the portal vein, including portal vein thrombosis, bleeding esophageal varices, and massive hepatic infarction, raised concerns about the intraportal route initially (^6,7). The recent success and improved safety profile of ITx are attributable to improvements in islet isolation, low endotoxin enzyme blends, purification techniques, and the use of less diabetogenic immunosuppressive drug combinations (^8–10). The promise of ITx is now a reality, but the process still carries potential risk. In the current study, we focus on the factors that might lead to elevation of liver enzymes and the impact of these enzyme elevations on islet graft outcome.

PATIENTS AND METHODS

Patient Selection, Islet Preparation, and Transplantation

Changes in liver enzymes were examined in 84 consecutive allogeneic islet-alone intrahepatic transplant procedures performed in 42 recipients between March 1999 and November 2002. Patients were selected for transplantation on the basis of severe hypoglycemic unawareness or poorly controlled blood glucose despite an optimum insulin regimen. All patients received steroid-free, sirolimus-based immunosuppression together with low-dose tacrolimus (^1–3). The 42 recipients, 19 men and 23 women, with type 1 diabetes and absent C-peptide pre- and poststimulation, had a mean age of 42.7±1.5 years (at the time of the first transplant) and a mean duration of diabetes of 26±2 years. Of the 84 consecutive procedures, 42 were first transplants (mean islet mass, 417,271±21,069 islet equivalent [IE]), 33 were second transplants (367,592±18,615 IE), and 9 were third transplants (339,681±10,848 IE). Islets were isolated from cadaveric pancreata by Liberase (Roche Molecular Biochemicals, Indianapolis, IN) perfusion and gentle enzymatic digestion in a Ricordi chamber (¹¹). Islets were purified with the use of continuous gradients using xenoprotein-free media as previously described (¹). Access to the portal vein was achieved using a percutaneous transhepatic technique under ultrasound and fluoroscopic guidance (⁵). Portal venous pressure was recorded during the infusion. Patients were usually discharged the following day once ultrasound had confirmed absence of portal vein thrombosis and the hemoglobin value and liver enzymes were acceptable.

Study Design

Total bilirubin (normal, <20 μM), aspartate aminotransferase (AST) (normal, <40 U/L), alanine aminotransferase (ALT) (normal, <50 U/L), and alkaline phosphatase (ALP) (normal, 30–130 U/L) were assessed for the first 40 days post-ITx (three times weekly for the first 2 weeks and then twice weekly). Graft characteristics such as IE, IE per kilogram, packed cell volume, islet purity, total ischemia and culture time, and the effect of different Liberase lots on the liver function test (LFT) elevation were assessed after the first ITx. Parameters of graft function including decrease in daily insulin requirement, fasting C-peptide secretion, acute insulin response (AIR) to glucose, and AIR to arginine were assessed 1 month after first ITx. Of the 84 consecutive procedures, 14 were excluded from the overall analysis because of insufficient LFT data (which were not monitored frequently in the initial subjects).

Statistical Analysis

Data are expressed as mean±SE or as median and range where applicable. Analysis of variance was applied for statistical analysis. When equal variance tests failed, Kruskal-Wallis or Dunn’s methods were used. Proportions were compared using chi-square test with Yates continuity correction. A value of P <0.05 was deemed statistically significant. SigmaStat 2.0 (Jandel Scientific, San Rafael, CA) was used for statistical calculations.

RESULTS

In all transplant procedures, LFT were within normal ranges before transplantation. LFT normalized in 90% of the recipients within 4 weeks posttransplant. The remaining 10% normalized within 2 months post-ITx. The mean total bilirubin remained within the normal range pre- and posttransplant (8±0.7 μM vs. 12±0.7 μM, P <0.001). In the 14 procedures with insufficient LFT data, all had limited monitoring posttransplant, but none had any values greater than 2.5 times above the ULN. In all procedures, the Gram stain of the infusate was negative and the endotoxin level was 0.014±0.002 endotoxin units/kg (i.e., less than the critical threshold of <5 endotoxin units/kg, based on recipient weight). Doppler examinations of the intra- and extrahepatic portal system were completed at times of transaminasemia, and no evidence of portal thrombosis was found.

Changes in AST

AST peaked at 7±0.5 days (range, 1–19 days) at a mean value of 162±23 U/L (range, 23–1,392 U/L), significantly greater than the baseline value (P <0.001). In further analysis, we found that 27% of the procedures peaked within the first 72 hr after islet infusion and the remaining peaked at later time points. In 54% of the transplants, the AST increased by greater than 2.5 times above the ULN, and in this group AST peaked at 245±37 U/L (range, 104–1,392 U/L). A 5-fold increase in AST was observed in 27% of the procedures and peaked at 352±69 U/L (range, 216–1,392 U/L). In 6% of the procedures, AST remained within the normal range. The AST values returned to normal after 19±1 days (range, 3–50 days). In the ITx procedures with peak AST greater than times above the ULN, 83% were after the first transplant, 11% were after the second transplant, and 6% were after the third transplant. Peak elevation of AST greater than five times above the ULN was seen after the first ITx in 15 of 35 procedures but was seen after the second ITx in only 2 of 24 procedures (χ²=6.68, P <0.01).

Changes in ALT

ALT showed a pattern similar to AST and peaked at 8±0.5 days (range, 2–17 days) with a value of 178±17 U/L (range, 17–667 U/L), significantly greater than the baseline value (P <0.001). In 56% of the transplants, the ALT increased by more than 2.5 times above the ULN. A 5-fold increase in ALT was observed in 24% of the procedures. In 10% of the transplant procedures, AST remained normal. The ALT values returned to normal after 23±2 days (range, 5–68 days). In the ITx procedures with elevation of peak ALT greater than five times above the ULN, 79% occurred after the first transplant and 21% occurred after the second transplant.

Changes in ALP

ALP peaked at 8±0.5 days (range, 2–19 days) at a value of 137±10 U/L (range, 60–319 U/L), which was significantly greater than the baseline value (P <0.001). In 12% of the transplants, the ALP increased by more than two times above the ULN. In the majority (69%) of the transplant procedures, ALP remained within the normal range. The ALP values returned to baseline after 20±2 days (range, 5–68 days).

The Relationship between Elevated LFT and the Graft Characteristics after First ITx

Patients were classified into three groups after their first islet transplant depending on the degree of elevation of AST, as follows: group 1, AST less than or equal to 2.5 times above the ULN (≤100 U/L) including the procedures with no elevation in AST; group 2, AST greater than 2.6 to less than five times above the ULN (101–199 U/L); and group 3, AST greater than or equal to five times above the ULN (≥200 U/L). These groups were examined in relation to the total IE, IE per kilogram, packed cell volume, purity, total ischemia time, and culture time. The results obtained showed no significant relationship between graft characteristics and the different groups with elevated liver enzymes (Table 1). In our ITx series, early recipients received freshly prepared islets with short culture time (1–2 hr). In the more recent ITx procedures, islets were cultured for longer periods (48–60 hr). The relationship between culture time and the peak AST shows no significant correlation (r²=0.06) (Fig. 1).

The Relationship between the Elevated LFT and Metabolic Parameters after First ITx

Fasting C-peptide, decrease in daily insulin requirements, 90-minute glucose after standard meal test (Ensure test), AIR to glucose, and AIR to arginine were performed 1 month after first ITx and compared to the AST groups defined above. No significant relationship was found between any of these metabolic parameters and the acute increase in AST (Table 2).

The Relationship between Different Liberase Lots and the Peak AST

In the 84 consecutive islets transplant procedures, 21 Liberase lots were used. We analyzed eight lots used in 68% of the ITx procedures. The results showed that the elevation of LFT in the transplant patients varied from one Liberase enzyme lot to another. Our observation suggests a possible association between enzyme lots and elevated LFT (Fig. 2).

The Pattern of Liver Enzymes after Sequential Islet Transplantation

We noticed higher elevations in the AST after the first ITx, compared with the second and third ITx. These elevations normalized within the first month post-ITx in 90% of the procedures (Fig. 3a). The mean peak AST after the first, second, and third transplants was 202±39, 120±21, and 102±23, respectively; whereas the median peak AST after the first, second, and third transplants was 166, 95, and 85, respectively. The difference in the median peak AST after sequential ITx was not statistically significant (P =0.06) (Fig. 3b). The pattern of transaminasemia did not correlate with any acute change in portal pressure after islet infusion (portal pressure increase of 3.7±1.0 mm Hg for procedures with a greater than 5-fold increase in AST vs. 3.7±1.9 mm Hg for procedures with a less than 2.5-fold increase in AST, P =not significant).

DISCUSSION

The field of ITx has shown dramatic improvements in the past few years, after the introduction of the Edmonton Protocol and with more recent protocol modifications. This protocol is based on avoidance of corticosteroids and on the use of sirolimus, low-dose tacrolimus, and an inductive anti–interleukin-2 receptor antibody to protect against rejection and recurrent autoimmunity (¹²). The 1-year insulin independence rate after islet transplantation improved from 8% to 85% in patients treated with the Edmonton Protocol (^1–3).

The results of the current study indicate that there is a significant elevation in liver enzymes occurring during the first week after islet infusion. These elevations normalize spontaneously during the first 4 weeks post-ITx.

A marked trend toward higher elevation in liver enzymes was noted after the first ITx compared with subsequent transplants, which is an unexpected finding. The greater elevation in liver enzymes after the first ITx may be explained by the fact that with the second and third transplants, the patients are already established on immunosuppressive drugs, which might protect against hepatocyte injury and avoid immunologic reactions around the newly infused islets. We found no significant relationship between elevated LFT and graft characteristics or graft function.

Although the exact cause for elevated LFT remains unclear, it seems that many factors may be involved. One explanation might relate to hypoxic injury to the presinusoidal hepatocytes after islet embolization. Another explanation could be that islets injected into the liver through the portal circulation might trigger an injurious inflammatory reaction. This reaction could lead to complement activation, platelet deposition, and proinflammatory cytokine release, resulting in islet apoptosis, cell death, and possible hepatocyte injury (^13–15). Endotoxin contamination of reagents used in islet isolation and purification may also contribute to islet injury and early graft loss (^16,17). However, in our transplant series, by using Liberase enzyme in islet preparation, the endotoxin content has been low.

When studying the different Liberase lots, there was a wide range in LFT elevation, suggesting differences in the enzyme activity and inter- and intralot variability. The limited number of isolations for each enzyme lot precludes a more detailed multivariate analysis at present. It was noted that Liberase lot G had a slightly higher enzyme activity (Wunsch, neutral protease ratio of 35). It is unclear whether this was associated with the pattern of LFT increase with this particular lot.

We found that the mean AST peaked on the seventh day posttransplant. This pattern suggests that LFT elevation is not an acute reaction but is rather more gradual. The response is self-limiting and resolves spontaneously during the first 4 weeks after ITx. It is not associated with thrombosis of any major branches of the portal vein based on Doppler examination. However, we cannot exclude a possible association with microthrombi in the terminal branches of the portal vein.

Toxicity to hepatocytes from the immunosuppressive drugs used for ITx could be an alternative indirect cause for LFT elevation. However, the pattern of LFT elevation is unlikely to be because of drug toxicity, as these agents are continued beyond the timeframe during which LFT return to normal.

In two previous described cases with partial portal vein thrombosis from our center (right branch and segmental branch), there was no initial elevation in LFT (²). Therefore, elevations in LFT cannot be used reliably to screen for portal vein thrombosis.

CONCLUSION

Islet transplantation is a relatively safe procedure that results in remarkable improvements of the patient’s glycemic control. The elevation in liver enzymes after ITx is an expected finding without apparent negative clinical consequence to the patient or to the functional outcome of the islet graft. More extensive multivariate analysis will be helpful once a larger cohort of islet patients have been followed long term.

Acknowledgments.

We thank the staff of the human islet isolation laboratory for providing us with the graft characteristics data and for their expertise in islet isolation; the staff of the organ procurement program in Alberta (Human Organ Procurement and Exchange) and across Canada for identifying cadaveric donors, Kathleen LaBranche for help with data collection, and the nursing coordinator staff of the clinical program for excellent patient care. We are also grateful to Sandra Blitz, Ph.D., for help with statistical analysis and Gregory Korbutt, Ph.D., who provided data on graft purity.

REFERENCES