| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Section of Diabetes, Endocrinology, and Metabolism, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, Nebraska 68198-3020
Correspondence: Address all correspondence and requests for reprints to: Jennifer L. Larsen, M.D., Department of Internal Medicine, 983020 Nebraska Medical Center, Omaha, Nebraska 69198-3020. E-mail: jlarsen{at}unmc.edu
 |
Abstract |
|---|
)
|
|
I. Introduction |
|---|
) (2).
|
Kidney transplant markedly improves patient survival in the diabetic ESRD patient compared with dialysis, especially when performed early (4, 5, 6, 7). Therefore, the impact of adding a pancreas graft, as with simultaneous pancreas-kidney transplant, should compare patient survival to that of kidney transplant alone. Kidney graft failure, from whatever cause, also results in increased mortality as soon as the diabetic patient returns to dialysis (5). Thus, if kidney graft survival is lower after pancreas transplant, life expectancy will also be decreased, although not as immediate or as well tracked.
This review will discuss the three main types of pancreas transplantation (see Tables 1 and 2): 1) simultaneous pancreas-kidney transplant, in which the pancreas and kidney are transplanted from the same deceased donor; 2) pancreas-after-kidney transplant, in which a cadaveric, or deceased, donor pancreas transplant is performed after a previous, and different, living or deceased donor kidney transplant; and 3) pancreas transplant alone for the patient with type 1 diabetes who usually has severe, frequent hypoglycemia, but adequate kidney function. The indications for each procedure will be discussed, but contraindications to pancreas transplant are often the same for all procedures. Pancreas transplant alone and pancreas-after-kidney transplant candidates must have stable, adequate kidney function at the time of transplant, as both the transplant operation and immunosuppression can otherwise cause an immediate further decline of renal function. Absolute contraindications to transplantation of any type include active malignancy or infection, recently treated malignancy not meeting the minimum disease-free observation period as suggested by the Clinical Practice Guidelines of the American Society of Transplantation (8), psychiatric disease so severe or unstable that the stress of a large surgery would likely result in marked decompensation, and subjects unable or unwilling to take immunosuppressant medications regularly such that graft failure would be certain.
|
|
Living donor pancreas transplant is one type of pancreas transplant that will not be discussed in detail, as it represents only 0.5% of pancreas transplants performed (2). In this procedure, a hemipancreatectomy is performed on a living donor, often a relative of the recipient, and then implanted as a segmental graft in a recipient with diabetes. Islet cell mass in the recipient is less than with a whole-organ transplant. The University of Minnesota has the largest series and reports a 1- and 5-yr patient survival of 90%, and a 1-yr pancreas graft survival of 75% (10). The potential risk of diabetes in the donor, who also now has a smaller pancreas, continues to be a concern. In one series, six of 104 living donors from 1978–1997 required diabetes treatment or had an elevated hemoglobin A1C (A1C) after donation, a 5% diabetes rate. However, no details were given on whether additional screening for asymptomatic diabetes or impaired glucose tolerance was performed or how long the donors were followed to determine whether this rate represents the entire risk (10). In a random group of eight living pancreas donors who were 9–18 yr after their surgery, 50% had diabetes (mean age 50), with greatest risk in individuals with body mass index 
27.8 kg/m2 (11). More extensive testing using oral glucose tolerance tests and 24-h blood glucose and urine C peptide profiles was performed in 28 donors before and 1 yr after hemipancreatectomy (12). In this study, seven of 28 had evidence of abnormal glucose tolerance by oral glucose tolerance test, but none were found to have diabetes at the time of the study. Most concerning was that mean glucose was higher and urine C peptide was significantly lower 1 yr after hemipancreatectomy. These studies suggest a greater future risk of diabetes in donors after hemipancreatectomy, although the degree of risk still needs to be better defined. Additional centers are performing this operation, as well as simultaneous living-donor pancreas-kidney transplant, first described in 1996 (13). A recent report of six cases of simultaneous living-donor pancreas-kidney transplant from one center reported a 1-yr pancreas graft survival rate of 83% (14). In this study, all donors had normal glucose tolerance at 1 yr, but this is likely too early to determine the entire risk to the donor. Long-term benefits to the recipient of living-donor pancreas transplant of any kind, if established, must be balanced against both short- and long-term risks to donors and recipients before this procedure can be advocated.
Islet transplantation, the subject of considerable ongoing discussion and investigation, has no long-term data (>5 yr) to compare with whole-organ pancreas transplant outcomes, so it will not be included in this review.
|
II. The Pancreas Transplant Candidate |
|---|
). Yet more women than men receive pancreas transplant alone, suggesting that women may be at higher risk for frequent hypoglycemia as a complication of diabetes, the most common indication for this procedure. The average age of pancreas transplant recipients has gradually increased at all centers for all transplant types. When the period of 1987–1992 was compared with 1999–2001, recipients over the age of 45 increased from 9% to 24% for simultaneous pancreas-kidney transplant, 9% to 29% for pancreas-after-kidney transplant, and 11% to 25% for pancreas transplant alone, although pancreas transplant-alone recipients are generally younger than other recipients (2). Pancreas transplant recipients with type 2 diabetes are generally older than those with type 1 (45 vs. 39 yr; P = 0.001) (17).
Most pancreas transplant recipients are Caucasian because Caucasians are also at highest risk for developing type 1 diabetes. Yet the number of non-Caucasian pancreas transplant recipients is increasing. For example, African-Americans represented 4% of recipients in 1987–1990 and 8% of recipients in 1996–2000 (17).
The mean duration of diabetes before transplant is 23–27 yr, depending on the category (Table 2). Almost by definition, all pancreas transplant candidates have had diabetic complications to be eligible for transplant: simultaneous pancreas-kidney or pancreas-after-kidney transplant candidates already have ESRD, and pancreas transplant alone candidates usually have frequent, severe hypoglycemic episodes that result from one or more complications. Macrovascular disease, whether or not symptomatic, is present in most, as markers of vascular disease such as carotid intima media thickness, and C reactive protein are increased in kidney and pancreas transplant candidates compared with age-matched controls or type 1 diabetes patients without nephropathy (18, 19, 20, 21, 22, 23). Whether diabetic complications stay the same, accelerate, or regress after pancreas transplant is one of the most important questions to be answered and will be discussed in Section VI.
|
III. Indications for and Types of Pancreas Transplantation |
|---|
and 2) (2). Although there is no consensus statement specifying all indications, the usual indication for this procedure at most centers is a type 1 diabetes patient with ESRD and adequate cardiac reserve who either has no option for a living kidney donor or desires to receive both organs simultaneously rather than waiting for a pancreas after the kidney transplant is completed. In simultaneous pancreas-kidney transplant, the pancreas and kidney are usually obtained from the same deceased donor; therefore, changes in kidney function can be used to determine whether rejection is occurring in either organ. Uncommonly, a cadaveric pancreas graft is transplanted at the same time as a living donor kidney to avoid two hospitalizations while reaping the benefits of living-donor kidney transplant (24). Even less commonly, simultaneous living-donor pancreas-kidney transplant has been performed as described above, but has the disadvantage of lower islet mass and greater risk to the living donor (13, 14).
|
), but simultaneous pancreas-kidney transplant has the best 1-yr and long-term pancreas graft survival rate of any pancreas transplant procedure. Whereas 1-yr pancreas graft survival rates were the same for simultaneous pancreas-kidney (83%), pancreas-after-kidney (82%), or pancreas transplant alone (80%) performed 1999–2001 (Fig. 3) (17), simultaneous pancreas-kidney transplant had a slightly greater 1-yr as well as long-term graft survival rate in those reported 1999–2002, 84% vs. 78%, for the other two categories (2). A number of centers report even higher (>90%) 1-yr pancreas graft survival rates after simultaneous pancreas-kidney transplant than the average for all UNOS data (25). Finally, simultaneous pancreas-kidney transplant has equal or better 1-yr patient and kidney graft survival rates compared with kidney transplant alone in diabetic recipients (Fig. 2) (26). Thus, simultaneous pancreas-kidney transplant does not have a negative impact on the success of the kidney graft. The retransplant rate is the number of individuals, out of the total, who have had a previous transplant of the same type; the retransplant rate after simultaneous pancreas-kidney transplant has remained quite stable at 1% (2).
|
B. Pancreas-after-kidney transplantation
This is the second most common pancreas transplant procedure (Fig. 1 and Table 2). The indication for this procedure is a patient with type 1 diabetes who has identified a living donor for kidney transplant and wants to plan a later pancreas-after-kidney transplant, or the type 1 diabetes patient who already has a kidney transplant that has stable graft function, desires the potential benefits of normoglycemia, and has the cardiac reserve to undergo the procedure. The initial kidney transplant can be obtained from either a deceased or living donor, but a living donor is preferred, when available, because it offers the best short- and long-term patient and graft survival for diabetic recipients (32). As a result, the number of living-donor kidney transplants has increased. Pancreas graft survival with pancreas-after-kidney transplant has also improved, which has further increased the interest in this procedure. The number of pancreas-after-kidney transplants has increased from 11% of all transplants in 1987–1998 to 18% in 1999–2000, whereas the number of simultaneous pancreas-kidney transplants performed has stayed the same, limited by the number of available deceased kidney donors (33).
Ideally, pancreas-after-kidney transplant should be performed when the kidney graft is stable. However, outcomes are not different between those receiving a pancreas graft early (<4 months), compared with later (>4 months), after kidney transplant (34). One-year patient survival rate with pancreas-after-kidney transplant is comparable to other pancreas transplant categories above (96%). In 2001, 1-yr pancreas graft survival with pancreas-after-kidney transplant was reported to be similar to simultaneous pancreas-kidney transplant, but long-term graft survival is still better with simultaneous pancreas-kidney (Fig. 3) (2, 17). One-year kidney graft survival is higher in pancreas-after-kidney transplants compared with simultaneous pancreas-kidney transplant, perhaps because of greater use of living-donor kidney grafts, but also the transplant procedure itself selects for recipients whose kidney function is stable and adequate after kidney transplant (17, 34).
C. Pancreas transplant alone
Pancreas transplant alone is the least common pancreas transplant procedure performed (5%; Table 1). Frequent, severe, hypoglycemic events are the most common indication for this procedure. The American Diabetes Association position statement suggests that indications for pancreas transplant (in the absence of kidney failure) are "frequent, acute and severe metabolic complications (hypoglycemia, hyperglycemia, and ketoacidosis) requiring medical attention" as well as "clinical and emotional problems with exogenous insulin therapy that are so severe as to be incapacitating; and consistent failure of insulin-based management to prevent acute complications," and centers performing this procedure generally evaluate possible candidates on a case-by-case basis (35).
Pancreas transplant-alone recipients are the youngest of all pancreas transplant recipients, which may explain why they also have the best 1-yr patient survival rate (Fig. 2; 99% in 1999–2001). One-year pancreas graft survival rate improved to 80% in 2001, similar to that reported after both pancreas-after-kidney transplants and simultaneous pancreas-kidney transplants, but is slightly lower at 78% in 2002 (Fig. 3) (2, 17). Optimally, creatinine clearance should be more than 70 ml/min to be considered for pancreas-only transplant, as a rapid decline in renal function can occur with impaired renal function, especially when cyclosporine-based immunotherapy is used (36, 37). Even with careful patient selection, kidney function may still deteriorate over time. At 1 yr, 2–8% of pancreas transplant-alone recipients underwent kidney transplant (17). Yet, by 10 yr after solitary pancreas transplant, the pathological changes of diabetes can reverse (38).
|
IV. Surgical Procedure Variations, Immunosuppression, and Immediate Complications |
|---|
)). With BD, urine amylase can be used as a marker of graft function. Biopsies of the pancreas graft are also easily obtained across the bladder wall through a cystoscope with this procedure. In early pancreas transplants, when rates of rejection were high, these were seen as advantages as they facilitated frequent monitoring for pancreas graft rejection. Even now, with pancreas transplant alone or pancreas-after-kidney transplant, where the kidney cannot be used to monitor pancreas rejection, BD may still be useful. However, this procedure also creates potential complications. Metabolic acidosis occurs in most, and extracellular volume depletion is common, occasionally severe enough to require hospitalization; both complications are due to the loss of sodium bicarbonate-rich pancreatic secretions into the urine (41). Oral sodium bicarbonate therapy is required in almost all and minimizes these complications in most. Additional problems that can complicate BD include bladder leak, reflux pancreatitis, particularly with neurogenic bladder, chemical cystitis/urethritis, frequent bladder infections, duodenitis in the connecting segment, bladder tumors, bladder calculi, urethral stricture, urethral erosion, epididymitis, prostatitis, and prostatic abscess (42, 43). Frequency of urological complications is high, 50–77%, but rarely results in graft or patient loss (42, 43, 45).
|
) (46, 47). Roux-en-Y was used predominantly in early ED procedures, but most centers no longer prefer a Roux-en-Y connection (2). There is less need for monitoring the pancreas graft, overall, because immunosuppression has improved and frequency of rejection episodes has decreased after pancreas transplant of all types. Thus, more new transplants are using ED at the outset, particularly with simultaneous pancreas-kidney transplant (77%), but also pancreas-after-kidney transplant (54%), and pancreas transplant alone (54%) as reported to UNOS from 1999–2002 (2). As long-term negative consequences of BD continue to emerge, many previous BD transplants have also been converted to ED, 9% by 1 yr and 15% by 3 yr for transplants performed 1996–2000 (2). Indications for enteric conversion surgery are frequent episodes of severe extracellular volume depletion, severe metabolic acidosis, urological complications, or problems with the duodenal segment. The hospitalization generally lasts 6–30 d (mean 12 d), with ED leak being the most common complication. The results of conversion surgery are good: 100% patient survival, 96% pancreas graft survival, and resolution of nearly all the indications for the surgery by 22-month follow-up in one series (48).
|
B. Portal vs. systemic venous drainage
The other variation in surgical procedure involves the location of the venous effluent of the pancreas graft. The first successful pancreas transplant procedure used BD of the ED, which, because of the distance, required that the graft be connected to the systemic rather than the normal portal venous system. When placed in the systemic circulation, called systemic venous drainage (SVD), the insulin secreted into the pancreatic venous effluent is not extracted immediately by the liver, as it would be if it emptied into the portal circulation. Systemic concentrations of insulin, both fasting and postprandial, are elevated as a result (49, 50, 51). Subsequently, a procedure was developed where the graft was placed in the portal circulation and the pancreatic duct was drained into the small intestine. This combined portal venous drainage (PVD) with ED procedure resulted in much lower peripheral insulin concentrations than pancreas transplant recipients with SVD (52), comparable to nondiabetic kidney transplants receiving similar immunosuppression (50, 53).
The PVD with ED drainage procedure is necessarily more physiological, but there has been considerable discussion about whether it changes outcomes. In all transplants reported to UNOS, outcomes were similar after PVD and SVD (17, 54, 55), but when recipients of pancreas-after-kidney or simultaneous pancreas-kidney transplant performed in 1995–2000 were analyzed, recipients of PVD or ED had greater patient mortality (27). Because PVD recreates normal physiology more than SVD, many have thought it would be beneficial to lipid metabolism or insulin actions. Some surgeons have suggested that there are other benefits as well, but there is little evidence to support this, as outlined below.
Nonrandomized, retrospective studies suggested an immune benefit of PVD over SVD with improved pancreas graft survival (56). However, when a randomized prospective study compared the two procedures, pancreas graft survival was the same (57). The most recent UNOS data suggest that 1-yr graft survival was not different between PVD and SVD for any pancreas transplant category (2).
Total fasting lipid concentrations are not different between individuals receiving PVD and SVD (58, 59), but lipoprotein composition may be. Hughes et al. (60) evaluated nonrandomized groups and outlined multiple differences in lipid profiles, particularly lower apolipoprotein B content of very-low-density lipoprotein (VLDL) and intermediate-density lipoprotein, and decreased intermediate-density lipoprotein mass and free cholesterol in PVD recipients compared with SVD recipients. However, in this study, cyclosporine dose was generally lower in PVD recipients, and serum creatinine was higher in SVD recipients (statistically significant at one time point), yet no analysis was performed to determine whether cyclosporine (dose or concentration) or creatinine contributed to the differences in lipoprotein content, independent of the procedure performed. In another study, no difference in apolipoprotein B content of VLDL was observed between those receiving SVD and PVD, but no other lipoproteins were analyzed (61). The one consistent finding is that cholesterol ester transfer (CET) is increased after SVD compared with PVD, as observed with other hyperinsulinemic states, and higher than nondiabetic kidney transplant-alone recipients (53, 62). Whether increased CET contributes to atherogenesis in this setting or others has not been established.
Could the chronic hyperinsulinemia following SVD itself increase vascular risk? Hyperinsulinemia correlates with vascular risk in many clinical studies, but in most of these studies, hyperinsulinemia is a marker for the more general state of insulin resistance, which is associated with many factors that contribute to atherogenesis (63). We have shown that carotid intima media thickness, a marker of overall cardiovascular risk, improves after simultaneous pancreas-kidney transplant with SVD (44, 64), and others have shown that progression of atherosclerosis was slowed in simultaneous pancreas-kidney transplant recipients, despite hyperinsulinemia (65). Whether PVD would result in even greater benefits is unknown, as these kinds of studies have not been compared between PVD and SVD. Without established significant advantages of PVD over SVD, most pancreas transplant recipients still receive SVD because there is greater surgical experience with SVD over PVD, greater flexibility with SVD to perform either ED or BD, and possibly less patient mortality with SVD in one report (27).
C. Immunosuppression
Gruessner and Sutherland (17) have recently reviewed the current immunosuppression protocols being used for pancreas transplant as reported to UNOS. The most common regimen in all pancreas transplant categories in 2000 was tacrolimus/mycophenolate mofetil (MMF; nearly 80%) with cyclosporine/MMF a distant second (5%–20%). The combination of tacrolimus/MMF has largely replaced cyclosporine/MMF because of some evidence of lower rejection rates, and better blood pressure and lipids (66), and MMF has largely replaced azathioprine because of decreased rejection rates (67). With the availability of sirolimus (rapamycin) and the successful use of sirolimus-tacrolimus for islet transplantation, increasing numbers of patients are being treated with a variety of sirolimus combinations. Of these combinations, tacrolimus-sirolimus is used the most, followed by cyclosporine-sirolimus, tacrolimus-sirolimus-MMF, MMF-sirolimus, and sirolimus only. Many centers still use corticosteroids in their immunosuppression protocol, which may allow a reduced dose of calcineurin inhibitor, but others have tried to move to a "steroid-free" protocol with the assumption that this will decrease risk of weight gain, glucose intolerance, dyslipidemia, and bone loss. This has not proven to be true, as calcineurin inhibitors can also cause dyslipidemia, bone loss, and glucose intolerance, including the ability to induce islet cell apoptosis as will be discussed later (68, 69).
The choice and type of antibody induction therapy being used are even more variable (17). Some patients receive no antibody induction therapy (7%), but most (>50%) receive some form of anti-CD-25 therapy. The number receiving no antibody induction is decreasing over time, and combination-antibody treatment is increasing, with nearly 50% of pancreas transplant-alone recipients receiving both T cell-depleting and nondepleting antibody therapies. Outcomes were not significantly different between regimens, but there was a trend toward better 1-yr graft survival with combination therapy (90%) compared with single antibody (81–86%) or no therapy (77%) in pancreas transplant-alone recipients, although the numbers are small. The immune suppression agents are often selected to improve graft survival, but they also have differential effects on blood pressure, lipids, weight gain, and glucose metabolism as will be reviewed in Section VII below.
|
V. Effect of Pancreas Transplantation on Patient Survival |
|---|
Whole-organ simultaneous pancreas-kidney transplant with normal graft function consistently improves 7- to 10-yr patient survival compared with deceased donor kidney transplant, simultaneous pancreas-kidney transplant with loss of pancreas graft function, or dialysis in type 1 diabetes patients waiting for a transplant (27, 70, 71, 72, 73). Age can affect outcome, as recipients over age 40 have lower patient survival after simultaneous pancreas-kidney or pancreas-after-kidney transplant than those under age 40 (27, 74). UNOS data show no specific threshold for age-related effects on patient survival after simultaneous pancreas-kidney transplant (2). In fact, recipients over age 50 may receive no benefit of simultaneous pancreas-kidney transplant on patient survival over kidney transplant alone (75). No gender or ethnic differences in patient mortality have been reported, but duration of diabetes also increases risk (74). Presence of neuropathy also predicts greater mortality in pancreas transplant recipients, but abnormal cardiorespiratory reflexes had the greatest impact on risk of mortality (74, 76, 77).
Although better patient and kidney graft survival have been attributed to improved glucose control after simultaneous pancreas-kidney transplant compared with cadaveric kidney transplant, both recipient and donor graft differences may also contribute. The type 1 diabetes patient who receives a cadaveric kidney transplant is generally older, more likely to be African-American, and have a longer duration of dialysis (73, 78). The donor used for kidney transplant alone was also older than the donor used for simultaneous pancreas-kidney transplant. The harvested kidney used for kidney transplant alone also had a longer cold ischemia time, on average, than the kidney used for simultaneous pancreas-kidney transplant (73, 78). Simultaneous pancreas-kidney transplant was associated with a greater rejection episode rate (15% vs. 9%) than kidney transplant alone, but despite this, the simultaneous pancreas-kidney recipients were less likely to need dialysis the first week after transplant, and had better long-term kidney graft survival compared with cadaveric kidney transplant recipients (78).
Living-donor kidney transplants have better patient and kidney graft survival than deceased-donor kidney transplants, for both diabetic and nondiabetic recipients. In fact, living-donor kidney transplant offers the same 8- to 10-yr patient survival as simultaneous pancreas-kidney transplant (73, 75, 79). Simultaneous pancreas-kidney transplant results in a higher patient mortality than living-donor kidney transplant the first year. Although patient mortality beyond the first year is lower in simultaneous pancreas-kidney transplant, it is not enough to result in greater patient survival overall compared with living-donor kidney transplant (73).
The outstanding patient and kidney graft survival outcomes after living-donor kidney transplant, along with increasing waiting list times for deceased donor kidneys, have caused many centers to prefer living-donor kidney transplant, when available, with or without a later pancreas transplant. Yet the risk of pancreas-after-kidney transplant may not be negligible. Patient mortality was reported to be greater at 4 yr in pancreas-after-kidney transplant recipients than in a matched cohort who had a kidney transplant and were on the waiting list for a pancreas (27). Some of the factors identified that increased mortality in pancreas-after-kidney recipients were increased recipient age and use of either PVD or ED without-Roux-en Y. Outcomes were not different in small vs. large transplant centers; therefore, less experienced programs were not overrepresented (27).
Pancreas transplant alone was also reported to cause greater patient mortality than that observed in a matched cohort waiting for this procedure (27). There are limitations of this comparison because some individuals who are placed on the list may later, particularly in the case of pancreas transplant alone, change their minds and decide they feel well enough to forego transplantation. Mortality may not be the only variable worth considering in this particular group. Some patients have such severe, frequent hypoglycemia that they can no longer hold employment, drive, or leave their home unaccompanied because of the risk of unconsciousness or requiring third-party assistance to treat their hypoglycemic events. The social impact of these hypoglycemic episodes, including effects on short-term memory and other cognitive functions, may warrant the potential increased risk of mortality, especially when the reported overall mortality rate, even if higher than without transplant, is still quite low (1–2% at 1 yr) (2, 17).
In summary, patient survival after simultaneous pancreas-kidney transplant is consistently better than that observed after cadaveric-donor kidney transplant, with the possible exception of recipients over age 50. Although this advantage may, in part, be due to improved glucose after pancreas-kidney transplant compared with kidney transplant alone, differences between the recipients who undergo these procedures, and between the donor grafts used for these two procedures, likely also contribute to the difference in survival described between these two procedures. Mortality after simultaneous pancreas-kidney transplant is equal to living-donor kidney transplant alone after 10 yr, and both pancreas-after-kidney and pancreas transplant alone may increase 4-yr mortality compared with remaining on the waiting list for those procedures. In these cases, specific quality of life (QOL) concerns and impact of pancreas transplant on specific diabetic complications need to be weighed against potential early increase in mortality before these procedures are considered.
|
VI. Consequences of Pancreas Transplantation on the Management of Diabetes |
|---|
2. Insulin, C peptide, proinsulin, and glucagon secretion.
Many studies have evaluated insulin secretion and acute insulin responses after pancreas transplant. It is important to note that absolute insulin concentrations may not be comparable from study to study because there is no common standard for insulin assays, the cross-reactivity of the antibodies used is often not stated, and differences in renal function can alter C peptide clearance, in particular. Antiinsulin antibodies can persist after transplant and can increase total insulin concentrations as well (82). Only studies in whole-organ pancreas transplant recipients will be reviewed.
Fasting insulin concentrations are two to three times greater than normal after pancreas transplant performed with SVD, but decrease over the first 12–24 months with decreasing immunosuppressant doses. Glucose- and arginine-stimulated insulin concentrations are also increased, but these peak, stimulated concentrations do not decrease considerably over time (49, 50, 51, 61, 80, 81, 83). Two groups most clearly demonstrated that the hyperinsulinemia with SVD results from a delay in first-pass hepatic extraction (50, 80). In contrast, only mild elevations in insulin concentration are observed with PVD, similar to nondiabetic kidney transplant patients treated with corticosteroids (52, 53).
Could prolonged hyperinsulinemia due to SVD cause insulin resistance? Using a rat model of pancreas transplant that does not require immunosuppression, euglycemic-hyperinsulinemic clamp studies were performed, after streptozotocin-induced diabetes, to determine differences between PVD and SVD. Glucose utilization and hepatic glucose production were the same between the two procedures; therefore, hyperinsulinemia does not appear to cause insulin resistance in this setting (84).
Insulin is normally secreted in patterns of both low-frequency ultradian and high-frequency oscillations. Denervation of the pancreas graft did not affect the presence of low-frequency ultradian oscillations of insulin secretion, which were also stable over time (6 months and 2 yr), and similar to kidney transplant recipients (85). However, another group suggested that although frequency was unchanged, pulsation amplitude was increased compared with controls (86). This group also described no changes in frequency or amplitude of high-frequency pulsations, whereas another group reported both a greater frequency and amplitude of high-frequency pulsations after pancreas transplant compared with kidney transplant recipients (86, 87).
The ß-cells of a healthy pancreas graft display both normal first- and second-phase insulin secretion in response to iv glucose (50, 80, 88, 89, 90). Blunting of first-phase insulin secretion suggests impending graft failure as it is only observed after whole-organ pancreas transplantation with damage to the pancreas graft or with increased secretory demand, as with obesity, worsened insulin resistance, or high concentrations of immunosuppressant therapy (91, 92).
Although fasting C peptide concentrations are generally elevated, they are similar to those in nondiabetic kidney transplant recipients regardless of the venous drainage technique (80, 81, 93). C peptide concentrations after a mixed meal challenge have been reported to be greater than (93), less than (80), and the same as (81) those of kidney transplant recipients, so they may depend on the relative insulin resistance and/or renal insufficiency of both the pancreas and the kidney transplant controls being studied.
Increased proinsulin secretion relative to insulin or C peptide can be an early marker of ß-cell injury or failure. Several (83, 94, 95, 96), although not all (80, 90), investigators reported that proinsulin was increased after pancreas transplant. However, fasting insulin and proinsulin concentrations and proinsulin/insulin ratio fall over the first 24 months after transplant, with no evidence of deterioration of glucose control (83). Proinsulin to insulin and proinsulin to C peptide ratios are also similar to those reported in kidney transplant recipients (94, 95). Thus, the elevation in proinsulin after pancreas transplant most likely reflects mild insulin resistance and decreased renal clearance rather than deteriorating pancreatic graft function.
Glucagon secretion immediately after transplant reflects both graft and native pancreas function. Fasting glucagon concentrations the first month after pancreas transplant can be three to four times normal, but decrease over the first 24 months after transplant, to slightly increased concentrations, similar to those of kidney transplant recipients who are given similar immunosuppression (81, 83, 97, 98, 99). Oral glucose suppresses glucagon secretion in pancreas transplant recipients more than controls, but is similar to that of nondiabetic kidney transplant recipients (81, 99, 100). Most importantly, glucagon secretion in response to hypoglycemia recovers after transplant, returning to normal in some studies (101), or less than normal, but improved in others, compared with type 1 diabetic controls (90, 97).
Several tests can serve as markers of those at risk for graft failure from whatever cause. In one study, an oral glucose tolerance test, performed on average 1.7 yr after simultaneous pancreas-kidney transplant that showed impaired glucose tolerance (using World Health Organization criteria), predicted risk of graft failure within 10 yr (102). Cytomegalovirus (CMV) infection also predicted greater risk of graft failure (P < 0.05). Mean 24-h glucose greater than 127 mg/dl at 1 yr was better than any other measure evaluated after transplant to predict those at high risk (93%) for pancreas graft failure within 4 yr (103).
In summary, glucose normalizes immediately after pancreas transplant, at the expense of hyperinsulinemia if SVD is used. Insulin secretion demonstrates oscillations despite denervation, as well as normal first- and second-phase secretion responses, unless there is a decrease in graft function or increased insulin resistance. C peptide concentrations are often slightly elevated, both basally and after mixed meal stimulus, but similar to those of nondiabetic kidney transplant recipients. Although fasting proinsulin is increased, it does not necessarily represent failing graft function, and glucagon response to hypoglycemia improves over time.
B. Hypoglycemia and counterregulatory hormone response
Many patients with longstanding diabetes have a history of severe hypoglycemic episodes before transplant. These episodes result, initially, from a decreased to absent glucagon response to hypoglycemia followed by a diminished epinephrine response to hypoglycemia over time. Glucose recovery in response to insulin-induced hypoglycemia is markedly improved after pancreas transplant compared with nontransplanted diabetic controls (97, 104). By 3 months after pancreas transplant, glucagon secretion and hepatic glucose production in response to hypoglycemia also return to normal (101, 104). Although epinephrine and GH responses to hypoglycemia improve after pancreas transplant, these do not return to normal (104, 105). Most importantly, hypoglycemia symptom score also returns to normal after pancreas transplant (105).
Episodic hypoglycemia was reported in a minority of recipients in the early months after pancreas transplant (82, 106, 107, 108, 109). In one study, seven of 12 individuals with repeated, documented hypoglycemic episodes had a hyperglycemic response to Sustacal followed by hypoglycemia in two, positive antiinsulin antibodies, and low free to total insulin concentrations, suggesting the possible role of antiinsulin antibodies in their hypoglycemia (82). In another study, only one of 10 reported to have symptomatic hypoglycemia episodes was found to have antiinsulin antibodies, but also had documented hypoglycemia after a mixed meal, 4 or more years after transplant (109). Glucose concentration after a 24-h fast was lower in the symptomatic group compared with those without symptoms of hypoglycemia, but the cause of the hypoglycemic symptoms was not established in another series (108). A recent case of hypoglycemia after pancreas transplant was reported to have nesidiodysplasia, another potential cause of hypoglycemia (110). Gastroparesis, the acute effects of corticosteroids on insulin secretion in the early posttransplant period, and improving counterregulation with improved recognition of hypoglycemic symptoms have also been proposed as contributing factors to the hypoglycemic symptoms reported in some patients after pancreas transplant. Symptoms generally diminish over time in most.
In summary, glucagon and symptom response to hypoglycemia return to normal or near normal over time, and epinephrine and GH responses improve but are not normal after pancreas transplant. Hypoglycemic symptoms and documented events are uncommon and tend to diminish over time but may be due to a variety of factors. It should be cautioned that glucagon secretion in response to hypoglycemia does not improve with either allo- or autotransplantation of islets into the liver, in human or animal studies, as described with pancreas transplant, and may be related to their location in the liver (111, 112). Thus, islet transplantation, commonly performed for treatment of hypoglycemia, may not be as effective a treatment for this indication.
C. Diabetic nephropathy
Risk of microvascular complications of diabetes is linked to glucose concentration (113). Thus, normalizing glucose after successful pancreas transplant might be expected to stabilize or reverse microvascular complications. Recurrent diabetic nephropathy is observed as early as 2 yr after kidney transplant in a diabetic recipient or upon failure of the pancreas graft after simultaneous pancreas-kidney transplant (114, 115, 116). Diabetic nephropathy has never been reported in a kidney graft when the graft is accompanied by a functioning pancreas graft. In fact, histological evidence of diabetic nephropathy in native kidneys can resolve between 5 and 10 yr after successful pancreas transplant in type 1 diabetes recipients as documented by prospective renal biopsies (38, 117). In summary, diabetic nephropathy can be prevented by a functioning pancreas graft, and pathological changes of diabetes can reverse over time after more than 5 yr of normal pancreas function.
D. Diabetic retinopathy
Most pancreas transplant candidates have had laser surgery for retinopathy before transplant (118, 119). This damage cannot be reversed. In fact, some patients experience short-term worsening of retinopathy immediately after transplant much like that described after sudden "tight" glucose control (120, 121, 122, 123). Some small and early studies did not show significant improvements in retinal complications after simultaneous pancreas-kidney transplant compared with pancreas transplant recipients who had lost graft function, type 1 diabetes controls, or kidney transplant alone (118, 124, 125, 127, 128, 130). Longer term studies that followed simultaneous pancreas-kidney transplant for 3 or more years demonstrated more consistent improvements including less progression of established neuropathy, fewer new vitreous hemorrhages, improved visual acuity, and less need for further laser surgeries compared with kidney transplant alone (118, 119, 120, 129, 130). In one study, 89% of those with unstable retinopathy were stable at follow-up with mean time since transplant of 5 yr, and all those with stable diabetic retinopathy were still stable (131).
One exception was a study of 20 simultaneous pancreas-kidney transplant recipients compared with 12 kidney transplants or simultaneous pancreas-kidney transplant recipients who had lost pancreas graft function. A number of the eyes, 48% of both groups, were excluded from analysis at the outset because of end-stage eye disease. Mean A1C of the pancreas transplant group was 6.2–6.5%, higher than most successful pancreas transplant patients, which usually have an A1C of 5%; therefore, many recipients were not always normoglycemic. Mean time of follow-up was not stated but was implied to be more than 3 yr. Blood pressure control and rate of smoking were also not stated. Neither the pancreas- or kidney-transplant group experienced many improvements in eye pathology or events, and there was no difference between the groups over time (132). Thus, patients with advanced retinopathy may or may not benefit from pancreas transplant, and normoglycemia is likely required for a benefit to be observed. The only report of changes in diabetic retinopathy after solitary pancreas transplant suggested early improvements observable within 6 months although there was no control group and only a small number of patients were studied (133).
However, cataracts can worsen after pancreas transplant and may be the most common long-term eye disease identified after transplant (128). From 40–55% of pancreas transplant recipients have been reported to require cataract surgery within 5 or more years, although many cataracts began before transplant (131, 134, 135).
In summary, diabetic retinopathy may worsen initially after pancreas transplantation with sudden improvement in glucose concentration; therefore, evaluation and treatment of preexisting retinopathy is important when pancreas transplant surgery is being considered. After 3 or more years of pancreas graft function, less retinal surgery is required after simultaneous pancreas-kidney transplant compared with kidney transplant alone in patients who do not already have end-stage eye disease. Lifelong eye surveillance examinations are required in all pancreas transplant recipients as laser surgery may still be required, particularly early after transplant surgery. Also, screening eye exams are needed to evaluate cataracts that can form or progress, particularly in any patient treated with corticosteroids.
E. Diabetic neuropathies
Both diabetes and renal failure can cause sensory neuropathy; therefore, peripheral sensory and motor neuropathy is present in the vast majority of individuals with diabetic ESRD. Symptoms of neuropathy were found in 86% and abnormal neurological exam in 94% of simultaneous pancreas-kidney transplant candidates in one large study (136). Peripheral sensory neuropathy improves after both simultaneous pancreas-kidney transplant and kidney transplant alone, but recipients of simultaneous pancreas-kidney transplant have even greater improvements compared with kidney transplant alone by 4–8 yr after transplant (70, 137, 138, 139, 140, 141). In fact, continued improvement in sensory and motor neuropathies can be observed as late as 10 yr after transplant (140). Equal improvements in sensory and motor neuropathy can occur after all three pancreas transplant procedures, suggesting that glucose concentration is the most important variable by which to observe these improvements (140). However, if the pancreas fails, nerve conduction velocity can worsen again to pretransplant levels within 2 yr (141). Those treated with nifedipine or angiotensin-converting enzyme (ACE) inhibitors were observed to have greater improvement, and those with prolonged uremia before transplant, obesity, or impaired renal graft function after transplant had less improvement overall (142). Weakness, due to a variety of factors including the effects of immunosuppression, prolonged hospitalization, and infection can worsen in the first year even with improved nerve conduction velocities, so does not always reflect neuropathy (143).
Diabetic autonomic neuropathies take longer to develop, are more variable from person to person, and can be much more difficult to quantitate. Reported prevalence of autonomic neuropathy at time of pancreas transplant varies from 76–100% depending on the population being studied (simultaneous pancreas-kidney vs. pancreas transplant-alone candidates) as well as what testing is performed (77, 144). Having autonomic neuropathy already marks an individual at higher risk for mortality as discussed above (74, 76, 77). Autonomic neuropathies can also complicate posttransplant care. For example, gastroparesis can change the timing of drug absorption so that assumptions about optimal dosing or timing of immunosuppression medication concentrations may not be correct. Cyclosporine, in particular, can slow gastric emptying even in nondiabetic renal failure patients so it may not be an ideal choice for the patient diagnosed with prolonged gastric emptying times (145). MMF can cause diarrhea, including an enterocolitis-associated diarrhea, so it may not be the drug of choice in individuals with preexisting diabetic diarrhea or other autoimmune diarrhea such as sprue or inflammatory bowel disease (146).
Whether or how much pancreas transplant alters the course of autonomic neuropathies is still controversial. Improvements are more modest, if present at all, or may not be observed in all individuals in the same study. These differences between studies and between individuals within a given study may reflect the severity of dysfunction at time of transplant, as suggested by the far greater cardiac autonomic neuropathy of pancreas-kidney recipients at the outset compared with nondiabetic kidney recipients in one report. Thus, although both groups improved, the rate of change or final function achieved over time could not reasonably be compared (147). However, the differences between studies may also reflect the amount of time needed to observe changes because these types of neuropathies often take a long time to develop. In support of this last possibility, peripheral neuropathy scores clearly improved within 12–24 months after pancreas transplant, but only mild improvements in autonomic neuropathy scores were observed after 42 months, and these changes were not significantly different (137). Additional variables may alter the results of testing used to define these neuropathies, independent of the impact of changes in glucose concentration. This is particularly true of gastric emptying studies in which recent immobilization and pharmacological agents, including narcotics, cannibinoids, clonidine, and immunosuppressant medications, can affect gastric emptying times (145, 148, 149, 150).
Patients with prolonged diabetes often manifest hypoglycemia unawareness with decreased symptoms and epinephrine output in response to hypoglycemia. These are both considered forms of autonomic neuropathy. After pancreas transplant, symptomatic response to hypoglycemia is normalized. Epinephrine response to hypoglycemia is also improved but does not return to normal (104, 105).
Heart rate variation, gastric emptying time, and skin temperature regulation have all been reported to improve after simultaneous pancreas-kidney transplant (70, 120, 138, 151). In some cases, new methods of measuring change had to be developed and validated to evaluate these responses (152). Symptoms sometimes improve independently of documented objective changes, but whether these reflect changes that current methods are not yet sensitive enough to identify is unknown (151). Improvements do not occur in all recipients, or in all studies (140, 151). For example, cardiorespiratory responses did not improve in any recipient 2–4 yr after pancreas transplant (153), and heart rate variation, one method of quantifying cardiac autonomic neuropathy, showed no improvement by 25 months after transplant or between serial examinations at 20 and 43 months after transplant in another (154). However, improvements in autonomic neuropathies may require more time. For example, R-R interval variation was not changed after 2 yr (138), whereas after 10 yr the same investigators described greater improvements after simultaneous pancreas-kidney transplant than kidney transplant alone (70).
In summary, improvements in sensory and motor neuropathy occur after both simultaneous pancreas-kidney and kidney transplant alone. However, greater improvements have been reported after simultaneous pancreas-kidney transplant with ongoing improvements up to 10 yr after transplant. Autonomic neuropathies may take longer to improve, 10 yr or more, and may be only partially reversible or not reversible at all in some cases. Yet some autonomic neuropathy parameters are improved in some studies, particularly hypoglycemia awareness, autonomic response to hypoglycemia, and cardiac autonomic neuropathies.
F. Microangiopathy
Microangiopathy represents the effects of long-term diabetes (and hyperglycemia) on vascular endothelium, as well as neurovascular response, with implications for healing response. Simultaneous pancreas-kidney transplant improves fluorescein escape rate in the nailfold capillary microcirculation, conjunctival microcirculation as assessed by intravital microscopy, vascular reactivity after suprasystolic occlusion, and microcirculation as assessed by laser Doppler and videophotometric capillaroscopy, with responses greater than after kidney transplant alone when such a comparison was made (155, 156, 157). However, at least one group did not find any difference in a cross-sectional study between pre- and post-simultaneous pancreas-kidney transplant recipients (158). As microangiopathy can also be caused by both tacrolimus and cyclosporine (159, 160, 161) and viral infection (162), not all individuals may see benefit after transplant, even with normoglycemia. In summary, improved glucose control, as with simultaneous pancreas-kidney transplant, can improve vascular reactivity and microvascular integrity and responses, but other factors after transplant may prevent or minimize these improvements in some.
G. Macrovascular disease risk factors and events
Immunosuppression agents used in organ transplantation contribute to weight gain, dyslipidemia, increased blood pressure, and insulin resistance after transplant. In fact, each combination of immune suppression agents may have different effects. Thus, normalization of glucose alone may not be enough to assume cardiovascular outcomes improve in this high-risk population if other risk factors worsen. Variable prevalence of genetic, ethnic, or behavioral variables that impact cardiovascular risk, and differences in procedure and immunosuppression protocols used may also contribute to differences reported in cardiovascular risk factors and outcomes between populations and centers. With these possibilities in mind, the data of how each risk factor might be affected by pancreas transplant will be reviewed.
1. Diet and weight.
In the one study of diet before and after pancreas transplant, neither total calories nor distribution of calories between carbohydrates, fats, and protein changed after successful pancreas transplant (163). Most studies report minimal, if any, effects of pancreas transplant on weight, 0–1 kg/yr (163, 164), although greater weight gain can be observed in individuals with preexisting obesity and in centers with greater ethnic diversity where risk of insulin resistance may also be greater (165). Type 2 diabetes patients who receive simultaneous pancreas-kidney transplant have not only a greater body mass index at time of transplant, but are also more likely to experience weight gain after transplant although this has not been well documented (31).
2. Hypertension.
Blood pressure after transplant is affected by a variety of factors. First, the immunosuppressive drugs themselves contribute to hypertension. Cyclosporine and corticosteroids have the greatest negative impact (166, 167, 168). Second, the transplant procedure can affect blood pressure. Blood pressure decreases after simultaneous pancreas-kidney transplant with BD, even with cyclosporine- and corticosteroid-based immunosuppression, in part, because of the salt and water loss that accompanies BD of the exocrine duct (66, 163, 169). In fact, after pancreas transplant with ED, blood pressure may not improve at all (170). Blood pressure can increase after pancreas transplant alone, even with BD, in a study using cyclosporine-based immunosuppression and in which renal insufficiency resulted (171). However, in another study of solitary pancreas transplant in which PVD and tacrolimus-based immunosuppression were used, renal function was unchanged, and blood pressure was improved (172). Thus, potential variables that impact direction of blood pressure change after pancreas transplant could include the procedure (simultaneous pancreas-kidney transplant vs. pancreas transplant alone), exocrine duct management (BD vs. ED), renal function or renovascular complications of the surgery, and choice of immunosuppression medications. Finally, differences in blood pressure can be "iatrogenic" if medications are changed or terminated, and blood pressure goals are not monitored.
3. Lipids and lipoproteins.
Multiple variables can also impact lipid concentrations before and after pancreas transplant. Most of the commonly used immunosuppressant medications, cyclosporine, tacrolimus, sirolimus, and corticosteroids, have adverse effects on lipid profile in many patients, particularly those predisposed to dyslipidemia. The two agents with the least impact on lipid metabolism are MMF and azathioprine. Many centers are using more sirolimus (rapamycin) in combination with calcineurin inhibitors without corticosteroids. However, sirolimus itself, or in combination with a calcineurin inhibitor or steroids, may be associated with greater dyslipidemia than other choices (173, 174, 175, 176, 177). Diabetic patients may even be more susceptible to sirolimus-associated dyslipidemia (177).
Simultaneous pancreas-kidney transplant, which restores insulin secretion and improves renal function, does improve lipid profile. In general, most studies observed lower fasting triglycerides, increased high-density lipoprotein (HDL), increased low-density lipoprotein (LDL) particle size, increased lipoprotein lipase activity, and improved postprandial lipemia compared with kidney transplant alone (62, 165, 178, 179, 180, 181, 182). One study reported no improvement in elevated triglycerides after pancreas transplant performed with SVD, but considerable weight gain was also reported at this center after transplant, which might explain these differences (60).
Lipoprotein composition may not be normal even if fasting lipids improve and may differ between those receiving pancreas transplant performed with PVD vs. SVD. As described previously, SVD causes systemic hyperinsulinemia, which can increase CET activity in other settings. In pancreas transplant recipients, increased CET activity increases unesterified cholesterol in the lipoprotein surface composition of HDL2, HDL3, and LDL, that is not observed with PVD (53, 60, 62, 165). Although greater improvements in LDL and VLDL subcomponents are reported after PVD, compared with SVD, the groups were not completely matched for other potential variables that could affect lipoprotein profile, including immunosuppressant doses and serum creatinine (60).
In contrast to simultaneous pancreas-kidney transplant, solitary pancreas transplant may not always improve fasting lipids. Triglycerides were higher 1 yr after pancreas transplant alone in one study using predominantly cyclosporine-based immunosuppression and SVD, which correlated with both immunosuppression dose and a decrease in renal function after transplant (171). However, in a more recent study using tacrolimus-MMF-based immunosuppression and PVD
Top
Abstract
I. Introduction
