Insulin therapy in organ donation and transplantation

Hyperglycaemia is common in hospitalized individuals, and is often caused by physiological stress associated with critical illness or major surgery. Insulin therapy is an established treatment for hyperglycaemia and acute hyperkalaemia, and has also been used for myocardial dysfunction resistant to inotropic support. Insulin is commonly used in both organ donors and transplant recipients for hyperglycaemia, but the underlying knowledge base supporting its use remains limited. Insulin therapy plays an important yet poorly understood role in both organ donation and transplantation. Tight glycaemic control has been extensively studied in critical care over the past 15 years; however, this has not yet translated into the field of transplantation, where patients are more unwell and where improved outcomes remain an ongoing challenge. Insulin therapy and optimization of glycaemic control represent important areas for future hypothesis‐driven research into organ donation and transplantation, such as amelioration of ischaemia‐reperfusion injury, rejection and infection.


| INTRODUCTION
Hyperglycaemia, an abnormally high blood glucose level, occurs either because of reduced insulin secretion or reduced hormone sensitivity.
It is commonly seen in diabetes mellitus, but can also occur in people without diabetes at times of extreme physiological stress. Stress hyperglycaemia is frequently seen in patients with critical illness, functions as a barometer of physiological response, and is associated with increased morbidity and mortality. 1 Organ donors experience severe physiological insults, which lead to either the declaration of brain death or circulatory death following the withdrawal of life-sustaining treatment. Patients who have undergone transplantation surgery are also often at the limits of their physiological reserve and will often require multi-organ supportive M.K.R. and D.v.D. contributed equally to this paper. treatments in a critical care environment. For these reasons, both organ donors and transplant recipients are likely to require insulin to manage glycaemic control, hyperkalaemia or myocardial dysfunction during their inpatient hospital admission. Understanding the physiological effects of insulin therapy in organ donors and transplant recipients is therefore essential to optimize patient outcomes. This review considers the role of insulin therapy in organ donation and transplantation. Medline, EMBASE and Cochrane electronic databases were systematically searched  using the Medical Subject Heading (MeSH) terms: "insulin therapy", "critical care", "organ donation", "transplantation", and "surgery", and these were combined using Boolean operators.

| INSULIN THERAPY AND ORGAN DONATION
In organ donors, the process of brain stem death results in the release of catecholamines and initiation of inflammatory processes affecting all organs. [1][2][3][4][5][6][7] These changes, and potentially the use of corticosteroids in intensive care units (ICUs), lead to hyperglycaemia in~50% of donors, which is usually managed with insulin. 8 Hyperglycaemia stimulates the generation of reactive oxygen species, ultimately leading to inflammation. 9 Hyperglycaemia after brain death, may therefore exacerbate the systemic inflammatory response, adversely affecting organ function.
In the United Kingdom, national donor management guidelines suggest a target blood glucose range of between 4.0 and 10.0 mmol/L, and, when glucose levels exceed 10.0 mmol/L, the commencement of an intravenous insulin infusion at a minimum rate of 1 unit per hour. 10 High-quality research in this area is limited because variation in practice amongst ICU staff means that it is difficult to determine: a) the extent and duration of hyperglycaemia that is tolerated prior to commencing insulin; b) the precise duration of insulin administration; and c) the triggers for ceasing insulin infusion.
Exogenous insulin requirement during organ donation could be the result of any combination of the following: insulin deficiency caused by underlying diabetes (known or unknown); irreversible β-cell death occurring as a consequence of brain death; reversible β-cell dysfunction caused by short-term metabolic stressors, such as hyperlipidaemia; or insulin resistance caused by inflammation, exogenous steroids and endogenous hormones, such as catecholamines, which are elevated in brain death.

| Inflammation and brain death
Insulin is also reported to have anti-inflammatory and anti-thrombotic properties through its action on nitric oxide, which leads to a reduction in reactive oxygen species and platelet aggregation via plasminogen activator inhibitor-1. [11][12][13][14] In a porcine model of brain death, continuous intravenous insulin infusion reduced the expression of pro-inflammatory cytokines (tumour necrosis factor-α [TNF]-α and interleukin [IL]-6) in both the heart and kidney. 15 Similar findings were confirmed in 15 brain dead donors (six treated, nine controls) in a randomized controlled trial of insulin using a hyperinsulinaemicnormoglycaemic clamp. In this instance, serum levels of proinflammatory cytokines (IL-6 and monocyte chemoattractant protein (MCP)-1) were significantly reduced, whilst the anti-inflammatory cytokine IL-10 was significantly elevated after insulin administration 16 ; however the hyperinsulinaemic-normoglycaemic clamp uses a fixed insulin infusion and a variable glucose infusion, and can be challenging to administer because of the intensive nursing and medical resources required. This may constitute a significant barrier to integrating this intervention into routine clinical practice. A more practical alternative to the hyperinsulinaemic-normoglycaemic clamp is highdose insulin delivery using a fixed rate glucose-insulin-potassium infusion. Using this mode of insulin delivery intra-operatively, two randomized controlled trials in patients undergoing coronary artery bypass grafting showed reduction in levels of TNF-α, IL-6 and IL-8, and complement factors 3 and 4. 17,18

| Metabolic activity
Hyperinsulinaemic-normoglycaemic clamps using high-dose insulin therapy have been widely used in cardiac surgery and are reported to improve cardiac function and lower levels of anaerobic metabolism in numerous randomized trials. 17,[19][20][21][22][23][24] Insulin therapy raises the threshold at which metabolic activity shifts from aerobic to anaerobic processes and results in the reduction of lactate prior to, and after, the ischaemic insult of aortic cross-clamping. 23 Directly available cytosolic ATP is also increased after insulin administration, and is believed to confer a protective mechanism by blunting the activation of AMPactivated protein kinases through the phosphorylation of AKT in various cell types. 25,26 These findings have not been validated, however, in human organ donors.

| Donor organ function
The use of insulin in the management of organ donors in intensive care has been considered to be associated with poorer organ function.
In one retrospective observational study, data from routine blood samples from 458 organ donors taken at the final stages prior to organ procurement showed that insulin use is associated with a lower glomerular filtration rate (GFR; insulin use vs no insulin use: 77 ± 55 vs 87 ± 59 mL/min/1.73 m 2 ; P = 0.009). 27 It is unclear from the published literature whether insulin use is a potential marker of the severity of the systemic inflammatory response to brain death or whether poor glycaemic control is the driver behind a potential relationship between insulin use and adverse donor organ function. A more extensive analysis of a larger cohort of kidney transplant recipients (n = 1036), where recipients with and without delayed graft function (DGF) were matched 1:1, showed no significant relationship between donor glucose levels and DGF 28 ; however, the relationship between donor insulin use and DGF remains unclear.

| Donor organ utilization
The relationship between donor insulin use and organ procurement has only been examined on one occasion. Novitzky et al 29 analysed data from 40 124 brain-dead donors in the United Network for Organ Sharing (UNOS) registry where they considered the relationship between various combinations of hormone therapies (thyroid hormone, antidiuretic hormone, corticosteroids and insulin) and organ procurement. Donor insulin use was associated with significantly lower rates of pancreas graft procurement, which suggests that donor insulin therapy could be a surrogate marker of donor pancreatic failure. This could be either attributable to pre-existing diabetes (potentially unrecognized and a contra-indication for pancreas transplantation) or because the surgeon considered donor insulin use a marker of β-cell failure secondary to brain death. Crucially, no information about donor characteristics was provided to describe donor phenotypes requiring insulin and why they are less commonly transplanted compared to donors not using insulin. The analysis also did not include data on post-transplantation function and outcomes.

| INSULIN THERAPY AND CRITICAL CARE
Intensive insulin therapy to manage hyperglycaemia in the critical care environment came to prominence in 2001 following the seminal publication by Van den Berghe et al. 30 In that single-centre randomized trial, 1548 surgical ICU patients received either intensive insulin therapy (aiming for tight glucose control of between 4.4 and 6.1 mmol/L) or conventional treatment (whereby insulin infusion was only commenced if blood glucose values exceeded 11.9 mmol/L, with a target level between 10.0 and 11.1 mmol/L). Intensive insulin therapy resulted in a statistically significant lower mortality rate: 4.6%, vs 8.0% with conventional therapy. The lower mortality rate was most evident in the subgroup of patients with an ICU stay >5 days (20.2% vs 10.6%). That trial included 90 transplant patients, but type of organ transplant received was not reported and the number of deaths was too small to provide a meaningful subgroup analysis. 30 In all patients, there was a 46% reduction in blood-borne infections and a 41% reduction in acute renal failure requiring renal replacement therapy.
Van den Berghe et al 31 repeated the initial intensive insulin therapy trial in a cohort of 1200 medical ICU patients, but no difference in mortality was seen between the two groups. A 9.5% lower mortality rate was seen in patients receiving insulin among the subgroup of 757 patients who stayed in the ICU for ≥3 days; however, mortality in patients receiving insulin was greater in patients staying <3 days. The authors attributed the beneficial effects to the prevention of acquired kidney injury, earlier weaning from mechanical ventilation and earlier discharge from the ICU in patients who received insulin. Insulin therapy has been demonstrated to prevent cellular hypoxia through its protective effects on the endothelium and mitochondria; this may provide a valuable insight into the potential underlying protective mechanisms of action. 32,33 The Normoglycaemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation (NICE-SUGAR) 34  After the publications by Van den Berghe et al 30,31 "tight glycaemic control" for all-comers to ICU was widely adopted, in particular across Europe; however, different approaches to insulin therapy without substrate delivery, as well as a concurrent movement towards early enteral nutrition, may have compounded poor safety profiles of intensive insulin therapy in ICUs, particularly the risk of hypoglycaemia. This was exposed, to some extent, by the conflicting findings of the NICE-SUGAR trial, 34 in which participants received eight times less concomitant intravenous glucose than in the initial trial by Van den Berghe et al. 30 However, in both trials, the incidence of hypoglycaemia was increased in the intensive insulin therapy arm An important message from the NICE-SUGAR trial is that the staff of an ICU should not attempt tight glycaemic control if they are unable to achieve this without increasing the risk of hypoglycaemia. Consequently, tight glycaemic control of critically unwell patients has been largely abandoned, perhaps inappropriately. 35 Intensive insulin therapy is also now an established treatment for the management of profound cardiogenic shock secondary to β-blocker and calcium channel blocker overdose. [36][37][38] More recently, the safety, haemodynamic effects and impact on catecholamine dosage of high-dose insulin therapy in patients with inotropic resistant myocardial dysfunction have been demonstrated in critically ill patients without β-blocker and calcium channel blocker overdose. 39 In multi-organ donors high-dose catecholamine infusions are commonly required to provide inotropic support and maintain cardiovascular stability, but this can have detrimental effects on organ quality. 40 Hormone therapy with insulin infusion may have a role to play in maintaining cardiovascular stability in organ donors without the concomitant adverse effects from catecholamines.

| INSULIN THERAPY AND MAJOR SURGERY
Diabetes and dysglycaemia are associated with postoperative morbidity and mortality in patients undergoing major surgery and subsequent admission to critical care. 41 Optimizing glycaemic control in critical care and defining the parameters of a safe, yet efficacious, target glucose range may be important. Clear guidelines for the use of variable rate intravenous infusion in medical patients exist 42 ; however, the paucity of evidence relating to clinical care and outcomes in surgical patients outside of critical care means that optimal care remains poorly defined.
A recent Cochrane review (2012) 43 of peri-operative glycaemic control considered the association of tight glycaemic control with infections, all-cause mortality and hypoglycaemic episodes as primary outcomes. It also examined cardiovascular events, renal failure and length of ICU and hospital stay, health-related quality of life, economic costs, weight gain, and mean blood glucose during intervention as secondary outcomes. Twelve randomized controlled trials were included in the review, with a total of 694 patients randomized to intensive insulin therapy and 709 patients randomized to conventional glycaemic control. The review did not report any improvement in outcomes associated with intensive insulin therapy, although this may reflect the fact that it was underpowered because of a small number of reported events (Table 1).

| INSULIN THERAPY AND TRANSPLANTATION
Patients undergoing transplantation have end-stage organ failure, but must have sufficient objectively assessed physiological reserve to withstand the demands of major surgery and postoperative immunosuppression. Organ transplantation therefore requires a multidisciplinary approach because the pre-transplant physiological status, operative procedure, and post-transplantation risks from immunosuppression create a complex high-risk patient cohort that requires specialist input to optimize outcomes. The direct application of findings from key insulin therapy trials in critical care to the transplant population without a robust evidence base and consideration given to all of these factors would therefore be inappropriate. Similarly, findings from other studies and meta-analyses [43][44][45][46][47] should be used to inform future studies within transplantation but, in the absence of robust scientific evidence, should not be directly applied without due rigour.

| Liver transplantation
A retrospective case series of 184 liver transplant recipients reported that patients with poorly controlled glucose levels (mean glucose > 8.3 mmol/L) had a higher 1-year mortality rate than those receiving insulin therapy with tightly controlled glucose levels in a retrospective case series (21.9% vs 8.8% respectively). 48 Posttransplantation infections were also lower in the tight glycaemic control cohort (48% vs 40%). Such findings have also been confirmed by other studies. 49,50 These improved outcomes in insulin-treated patients may be attributable to the multi-modal effects of insulin on the oxidative stress, coagulation, and inflammatory pathways. 51 This effect may be potentiated in immunosuppressed patients with the additional burden of an ischaemia-reperfusion injury. 52 In live donor liver transplantation, insulin therapy has also been reported to increase graft volume and reduce hepatic dysfunction following liver resection, which leads to better outcomes. 53,54 When investigated in a randomized controlled trial, insulin therapy was not associated with any difference in DGF between those treated with intensive intravenous insulin vs standard subcutaneous insulin therapy (18% vs 24%; P = 0.46). 57 However, given the target glucose range was similar in both groups, it is not surprising that no difference was seen between the treatment arms. The evidence regarding the T A B L E 1 Summary of key results of a Cochrane review analysing the association of adverse outcomes with tight peri-operative glycaemic control 43

Outcomes
Intensive insulin therapy, n/N Conventional glycaemic control, n/N Risk ratio (95% CI) P relationship between glycaemic control and acute rejection in both patients with and without diabetes is similarly conflicting. 55,57,58

| Islet transplantation
In islet cell transplantation, peri-transplant insulin therapy has been associated with higher rates of insulin independence after islet transplantation but these observational data are far from conclusive regarding clinical benefits of intervention in this setting. 59,60 Nonetheless, intensive insulin therapy to achieve normoglycaemia in the peritransplant period has become standard of care. This is partly because islet function is delayed for >10 days after transplantation, and also because there are theoretical benefits in that insulin has been shown to reduce glucotoxicity during the delicate engraftment process. 61 However, the optimal target blood glucose range in patients after islet cell transplantation remains poorly defined.

CONFLICTS OF INTEREST
None declared.

AUTHOR CONTRIBUTIONS
All authors contributed to the conception of the idea for the article,

ETHICAL APPROVAL
Ethical approval was not required for the present paper.