Should an Oral Glucose Tolerance Test Be Performed Routinely in All Renal Transplant Recipients? Kirsten A. Armstrong,* Johannes B. Prins,† Elaine M. Beller,‡ Scott B. Campbell,* Carmel M. Hawley,* David W. Johnson,* and Nicole M. Isbel* Departments of *Nephrology and †Endocrinology and ‡School of Population Health, University of Queensland at Princess Alexandra Hospital, Brisbane, Queensland, Australia Posttransplantation diabetes (PTD) contributes to cardiovascular disease and graft loss in renal transplant recipients (RTR). Current recommendations advise fasting blood glucose (FBG) as the screening and diagnostic test of choice for PTD. This study sought to determine (1) the predictive power of FBG with respect to 2-h blood glucose (2HBG) and (2) the prevalence of PTD using FBG and 2HBG compared with that using FBG alone, in prevalent RTR. A total of 200 RTR (mean age 52 yr; 59% male; median transplant duration 6.6 yr) who were >6 mo posttransplantation and had no known history of diabetes were studied. Patients with FBG 200 mg/dl [11.1 mmol/L]) and impaired glucose tolerance (IGT; 2HBG 140 to 200 mg/dl [7.8 to 11.0 mmol/L]). An abnormal OGTT was reported in 79 (42%) nondiabetic RTR: PTD (n ⴝ 22) and IGT (n ⴝ 57). The optimal FBG that was predictive of PTD was 101 mg/dl (5.6 mmol/L; area under the curve 0.70; sensitivity 64%, specificity 67%, positive predictive value 20%, negative predictive value 93%). The optimal FBG that was predictive of IGT was less well defined (area under the curve 0.54). The prevalence of PTD was higher by OGTT than by FBG alone (17 versus 6%; P < 0.001). FBG may not be the optimal screening or diagnostic tool for PTD or IGT in RTR. Consideration should be given to introducing the OGTT as a routine posttransplantation investigation, although the implications of a pathologic OGTT are still to be determined in this population. Clin J Am Soc Nephrol 1: 100 –108, 2006. doi: 10.2215/CJN.00090605

P

osttransplantation diabetes (PTD) and impaired glucose tolerance (IGT) represent a serious complication of renal transplantation, with incidence figures ranging from 2.5 to 53% (1– 4). Clinical studies have identified that PTD is associated with reduced long-term graft survival, reduced patient survival, and increased cardiovascular morbidity and mortality (2,5–7). PTD incurs significant cost in the treatment of the renal transplant recipient (RTR), over and above the costs inherent to routine care (8). The consequences of IGT in RTR are less well established, but given the association of IGT with the development and progression of diabetes and cardiovascular disease in the general population (9 –12), it is likely that IGT may be important in the pathogenesis of these disease processes in RTR. Standardized screening for PTD and IGT in RTR is essential to diagnose asymptomatic individuals at an early stage with a view to preventing, through tight glucose control, the development and progression of the associated complications. Recently developed consensus guidelines (13) have sug-

Received June 3, 2005. Accepted August 16, 2005. Published online ahead of print. Publication date available at www.cjasn.org. Address correspondence to: Dr. Kirsten Armstrong, Department of Nephrology, Level 2 Ambulatory Renal and Transplant Services Building, Princess Alexandra Hospital, Ipswich Road, Brisbane Qld 4102, Australia. Phone: ⫹61-7-3240-5080; Fax: ⫹60-7-3240-5480; E-mail: [email protected] Copyright © 2006 by the American Society of Nephrology

gested that the diagnosis of PTD and IGT in RTR should be based on currently accepted definitions of diabetes and IGT in the nontransplant population (14,15). Fasting blood glucose (FBG) therefore is recommended as the screening test of choice for PTD in RTR. An FBG ⱖ126 mg/dl (7.0 mmol/L) is diagnostic of PTD, although confirmatory testing should be performed on a separate occasion. The consensus guidelines advise that screening for disorders of glucose homeostasis in established (⬎1 yr) RTR should be performed on an annual basis, in keeping with current American Society of Transplantation recommendations (16). The oral glucose tolerance test (OGTT) is not recommended as a screening tool in RTR, although it is suggested that it should be performed in all RTR with FBG of ⱖ110 mg/dl (6.1 mmol/L) (17). In liver transplant recipients, the OGTT has been advocated as an annual screening tool in all individuals with a normal FBG (18). The diagnostic power of FBG with respect to 2-h blood glucose (2HBG) has never been validated in RTR; thus, the relationship between FBG and 2HBG in RTR is unclear. It is also unknown whether 2HBG provides additional prognostic information to FBG testing in RTR as it does in the general population. The aims of this study were to determine whether 2HBG could be predicted from FBG and to compare the prevalence of PTD using FBG alone with that using an OGTT in a cohort of prevalent RTR. ISSN: 1555-9041/101-0100

Clin J Am Soc Nephrol 1: 100 –108, 2006

Screening for Glucose Intolerance in Renal Transplant Recipients

Materials and Methods Study Design and Population This study was an observational analysis of prevalent RTR from the Princess Alexandra Hospital (Brisbane, Queensland, Australia). All RTR were eligible for inclusion in the study unless they were ⬍6 mo posttransplantation or were known to have either type 1 or type 2 diabetes (defined as the use of insulin or oral hypoglycemic agents or self-reported). A total of 280 eligible patients were approached for inclusion in the study, which was approved by the Human Ethics Committee of the University of Queensland and Princess Alexandra Hospital. Observational data were collected between January 2004 and January 2005.

Clinical Data Demographic data, including age, race, gender, cause of kidney disease, duration of current transplant, type of transplant (living or cadaveric), weight, height, waist circumference, and current use and dose of immunosuppressive agents were documented.

Screening Protocol Eligible patients had a preliminary blood test to measure blood glucose after an 8-h overnight fast (screening FBG). They then were invited to attend for an OGTT on a separate occasion. An OGTT was performed according to 1999 World Health Organization (WHO) specifications (14). FBG was measured at baseline (time 0 FBG). All patients with time 0 FBG ⬍126 mg/dl (7.0 mmol/L) then were given a drink that contained 75 g of dextrose monohydrate dissolved in 250 ml of water. A second blood sample was taken 2 h later for analysis of 2HBG. In patients with time 0 FBG ⱖ126 mg/dl (7.0 mmol/L), the test was not performed. A midstream urine sample was tested for the presence of glucose.

Biochemical Analysis All samples for glucose tolerance testing (screening FBG, time 0 FBG, and 2HBG) were collected in sodium fluoride/potassium oxalate tubes. Glucose was analyzed using the glucose hexokinase enzymatic method (Roche, Sydney, Australia). The between-run coefficients of variation for FBG in our laboratory are 2.6% at 63 mg/dl (3.5 mmol/L) and 1.64% at 284 mg/dl (15.8 mmol/L; Hitachi Modular D; Roche, Sydney, Australia).

Diagnosis and Management of PTD and IGT Two different screening and diagnostic methods were used to define PTD and IGT in the study cohort: (1) FBG: PTD was diagnosed when FBG was ⱖ126 mg/dl (ⱖ7.0 mmol/L), and (2) OGTT: Patients were classified according to time 0 FBG and 2HBG as follows: normal glucose tolerance (time 0 FBG ⬍110 mg/dl [6.1 mmol/L] and 2HBG ⬍126 mg/dl [7.0 mmol/L]), IGT (time 0 FBG ⬍126 mg/dl [7.0 mmol/L] and 2HBG 140 to 200 mg/dl [7.8 to 11.0 mmol/L]) or PTD (time 0 FBG ⱖ126 mg/dl [7.0 mmol/L] or 2HBG ⱖ200 mg/dl [11.1 mmol/L]). All patients with an abnormal OGTT or time 0 FBG ⱖ126 mg/dl [7.0 mmol/L] were seen by a renal dietician and received advice on lifestyle changes including exercise and weight maintenance. Those with PTD were referred for diabetes education. Initiation of hypoglycemic agents and/or referral to an endocrinologist was at the discretion of the treating nephrologist.

Statistical Analyses Statistical analyses were performed using standard statistical software (SPSS Version 11.5; North Sydney, Australia). Results are expressed as mean ⫾ SD, median (interquartile range [IQR]), and fre-

101

quencies (percentages), depending on the data type. Comparisons of means were made where appropriate, using the paired and unpaired samples t test. P ⬍ 0.05 were considered statistically significant. Intrapatient variability of FBG was assessed by performance of two measurements of FBG (screening FBG and time 0 FBG). Aim 1. Receiver operator characteristic (ROC) analyses were performed to determine the optimal level of time 0 FBG that was predictive of (1) PTD (defined as 2HBG ⱖ200 mg/dl [11.1 mmol/L]) and (2) IGT (defined as 2HBG between 140 and 200 mg/dl [7.8 to 11.1 mmol/L]). For each ROC analyses, we evaluated (1) the area under the curve (AUC), (2) the time 0 FBG value at the point on the ROC curve closest to the ideal of 100% sensitivity and 100% specificity, and (3) the positive predictive value (PPV) and negative predictive value (NPV) of the time 0 FBG level associated with the optimal sensitivity and specificity. Comparison of the mean AUC for screening FBG and time 0 FBG was performed using the t test. Aim 2. The prevalence of PTD using FBG (time 0 FBG ⱖ126 mg/dl [7.0 mmol/L]) was compared with the prevalence of PTD using the OGTT (time 0 FBG ⱖ126 mg/dl [7.0 mmol/L] or 2HBG ⱖ200 mg/dl [11.1 mmol/L]). An analysis of the agreement between the two classifications of PTD was performed using the ␬ test.

Results In this section, FBG refers to time 0 FBG unless otherwise stated.

Clinical Characteristics A total of 200 attended for an OGTT (study population). The baseline characteristics of the study population (n ⫽ 200) are shown in Table 1. Eligible patients who did not attend for an OGTT (n ⫽ 80) were younger (44 ⫾ 13 yr) than the study population (52 ⫾ 12 yr; P ⬍ 0.001) but otherwise well matched. In the study population, 117 (59%) were male and 184 (92%) were white. Median transplant duration of current transplant was 6.6 yr (2.8 to 12.9 yr). A total of 177 (89%) patients were on a stable regimen of either cyclosporin (n ⫽ 139; 2.2 mg/kg per d [IQR 1.6 to 2.7 mg/kg per d]) or tacrolimus (n ⫽ 38; 0.07 mg/kg per d [IQR 0.03 to 0.11 mg/kg per d]), and 162 (81%) patients were on a stable dose of prednisolone (0.08 mg/kg per d [range 0.06 to 0.11 mg/kg per d]). Mean screening FBG was 101 ⫾ 20 mg/dl (5.6 ⫾ 1.1 mmol/L), and mean time 0 FBG was similar (101 ⫾ 18 mg/dl [5.6 ⫾ 1.0 mmol/L]; P ⫽ 0.74). The intraindividual coefficient of variation between screening FBG and time 0 FBG was 10.6% (95% confidence interval [CI] 7.6 to 21.4%). An OGTT was performed in 188 (94%) of the study population. Mean 2HBG was 137 ⫾ 45 mg/dl (7.6 ⫾ 2.2 mmol/L). The remaining 12 patients had PTD by definition (FBG ⱖ126 mg/dl [7.0 mmol/L]) and did not proceed to an OGTT. An abnormal OGTT was reported in 79 (42%) patients; 57 fulfilled 1999 World Health Organization diagnostic criteria for IGT and 22 for PTD. The univariate correlates of a pathologic OGTT are shown in Table 2. Neither the use nor the dose of corticosteroids and calcineurin inhibitors was associated with a pathologic OGTT. Fourteen (18%) patients with an abnormal OGTT had glucosuria: 10 had PTD, and four had IGT. Four (33%) patients with FBG ⱖ126 mg/dl (7.0 mmol/L) had glucosuria. No patients with normal glucose tolerance had glucosuria. The

102

Clinical Journal of the American Society of Nephrology

Clin J Am Soc Nephrol 1: 100 –108, 2006

Table 1. Baseline characteristics of study population and patients who did not attend for OGTTa Variable

Demographic age (yr) male gender, n (%) white race, n (%) cadaveric transplant, n (%) transplant duration (yr) body mass index (kg/m²) Cause of renal failure glomerulonephritis, n (%) adult polycystic kidney disease, n (%) reflux nephropathy, n (%) analgesic nephropathy, n (%) hereditary, n (%) miscellaneous, n (%) uncertain cause, n (%) Immunosuppression cyclosporin, n (%) tacrolimus, n (%) mycophenolate mofetil, n (%) azathioprine, n (%) prednisolone, n (%)

Study Cohort (n ⫽ 200)

Nonattenders (n ⫽ 80)

P

52 ⫾ 12 117 (59) 184 (92) 144 (72) 6.6 (2.8 to 12.9) 27.2 ⫾ 5.3

44 ⫾ 13 49 (61) 49 (95) 62 (78) 6.6 (2.6 to 12.1) 26.5 ⫾ 4.7

⬍0.001 0.67 0.71 0.99 0.96 0.32

65 (33) 27 (13) 25 (12) 10 (5) 10 (5) 26 (13) 37 (19)

28 (35) 4 (5) 18 (23) 2 (3) 2 (3) 9 (11) 17 (20)

0.68 0.04 ⬍0.001 0.35 0.35 0.47 0.60

139 (70) 38 (19) 87 (44) 73 (37) 162 (81)

61 (76) 9 (11) 44 (55) 20 (25) 66 (83)

0.26 0.11 0.08 0.07 0.77

Values are expressed as mean ⫾ SD for normally distributed data, median (interquartile range) for nonnormally distributed data, and frequencies (%) for categorical data. OGTT, oral glucose tolerance test. a

sensitivity and the specificity of glucosuria as a screening test in this population were 17 and 100%, respectively.

FBG and Prediction of PTD

Twenty-two (12%) RTR had 2HBG ⱖ200 mg/dl (11.1 mmol/L). Mean FBG in these patients was 106 ⫾ 14 mg/dl (5.9 ⫾ 0.8 mmol/L), and mean 2HBG was 227 ⫾ 25 mg/dl (12.6 ⫾ 1.4 mmol/L). There was no correlation between FBG and 2HBG (r ⫽ ⫺0.26, P ⫽ 0.24) in these patients. The ROC curve for FBG predicting 2HBG ⱖ200 mg/dl (11.1 mmol/L) is shown in Figure 1a.

The AUC was 0.70 (95% CI 0.57 to 0.83) and was similar to that for screening FBG (AUC 0.75; 95% CI 0.65 to 0.87; P ⫽ 0.71; Figure 1b). Table 3 shows the sensitivity and the specificity for different levels of FBG predictive of 2HBG ⱖ200 mg/dl (11.1 mmol/L). Optimal sensitivity and specificity for FBG predicting 2HBG ⱖ200 mg/dl (11.1 mmol/L) were obtained at an FBG of 101 mg/dl (5.6 mmol/L; 64 and 67%, respectively). At this level, the PPV was poor (20%), although the NPV was more acceptable (93%). Instead of optimizing sensitivity and specificity, an alternative screening strategy could be adopted whereby sensitivity rather

Table 2. Univariate correlates of abnormal glucose tolerance in renal transplant recipients who underwent glucose tolerance testing (n ⫽ 188)a

a

Variable

Normal Glucose Tolerance (n ⫽ 109)

Abnormal Glucose Tolerance (n ⫽ 79)

P

Age (yr) Body mass index (kg/m²) Waist circumference (cm) History of dyslipidemia, n (%) Family history of diabetes, n (%) Pre-obesity, n (%) Central obesity, n (%) FBG (mg/dl; mmol/L) Glucosuria, n (%)

50 ⫾ 13 26.5 ⫾ 5.4 95 ⫾ 16 57 (52) 25 (23) 60 (55) 46 (42) 94 ⫾ 9; 5.2 ⫾ 0.5 0 (0)

54 ⫾ 11 28.0 ⫾ 4.9 100 ⫾ 14 55 (70) 34 (43) 56 (71) 46 (58) 101 ⫾ 13; 5.6 ⫾ 0.7 14 (18)

0.03 0.05 0.03 0.02 0.003 0.03 0.03 ⬍0.001 ⬍0.001

FBG, fasting blood glucose.

Clin J Am Soc Nephrol 1: 100 –108, 2006

Screening for Glucose Intolerance in Renal Transplant Recipients

103

Figure 2a. The AUC was 0.54 (95% CI 0.45 to 0.62) and was similar to that for screening FBG (AUC 0.57; 95% CI 0.48 to 0.66; P ⫽ 0.68; Figure 2b). Table 4 shows the sensitivity and the specificity for different levels of FBG predictive of IGT. The FBG predictive of IGT, with optimal sensitivity and specificity, was not clearly identified from the ROC analysis. One (25%) in four RTR with IGT had FBG ⬍90 mg/dl (5.0 mmol/L).

Comparison of PTD Prevalence Using FBG versus OGTT Twelve (6%) patients had PTD by FBG, and 34 (17%) RTR had PTD by OGTT (P ⬍ 0.001). Specifically, significantly more patients received a diagnosis of PTD from 2HBG compared with those who received a diagnosis from FBG (65 versus 35%; P ⬍ 0.001). Only 12 (35%) RTR with PTD fulfilled both FBG and OGTT diagnostic criteria. Concordance between the two classifications of PTD was moderate (␬ ⫽ 0.47). If the diagnosis of PTD had been based on FBG alone, then 65% of RTR with PTD would have been undetected.

Comparison of Different Screening Strategies for PTD and IGT in RTR If an OGTT were reserved for RTR with impaired fasting glycemia (FBG 110 to 124 mg/dl [6.1 to 6.9 mmol/L]), then 20% of RTR would require an OGTT and only 40% of patients with PTD and 16% with IGT would receive a diagnosis. If the threshold of FBG that is indicative of impaired fasting glycemia were reduced to 101 mg/dl (5.6 mmol/L), then 40% would require an OGTT and 64% patients with PTD and 47% with IGT would receive a diagnosis. Alternatively, if an FBG level of 90 mg/dl (5.0 mmol/L) were used as a threshold for performing an OGTT, then 70% of RTR would require the test and 86% of RTR with PTD and 75% with IGT would be identified. Only by screening all RTR with an OGTT, irrespective of FBG, was diagnosis of PTD and IGT maximized (Figure 3).

Glycosylated Hemoglobin Figure 1. Receiver operator characteristic (ROC) curve for fasting blood glucose (FBG) predicting posttransplantation diabetes (PTD) using time 0 FBG (a) and screening FBG (b). The area under the curve (AUC) was similar for both (0.70 versus 0.75; P ⫽ 0.71).

than specificity is optimized. Use of a lower cutoff of FBG of 90 mg/dl (5.0 mmol/L) as a threshold for performing an OGTT improved sensitivity to 86%, allowing diagnosis of five (23%) more patients with PTD but at the expense of lower specificity (27%).

FBG and Prediction of IGT Fifty-seven (29%) patients had IGT. Mean FBG in RTR who received a diagnosis of IGT was 100 ⫾ 11 mg/dl (5.6 ⫾ 0.6 mmol/L), and mean 2HBG was 160 ⫾ 18 mg/dl (8.9 ⫾ 1.0 mmol/L). There was no correlation between FBG and 2HBG (r ⫽ 0.05, P ⫽ 0.71). The ROC curve for FBG predicting 2HBG between 140 and 200 mg/dl (7.8 and 11.1 mmol/L) is shown in

Forty-seven (59%) patients with an abnormal OGTT and 7 (58%) with FBG ⱖ126 mg/dl (7.0 mmol/L) were also tested for glycosylated hemoglobin (HbA1c%). There was no correlation between FBG or 2HBG level and HbA1c% in patients with a pathologic OGTT (r ⫽ 0.23, P ⫽ 0.13; and r ⫽ 0.05, P ⫽ 0.72, respectively); however, there was a highly significant correlation between HbA1c% and FBG level in patients with FBG ⱖ126 mg/dl (7.0 mmol/L; r ⫽ 0.84, P ⫽ 0.02).

Cost Analysis Two hundred RTR were tested with an FBG at a laboratory cost of US$6.40 per test. Twelve patients received a diagnosis of PTD by FBG; thus, the laboratory cost per case of PTD diagnosed was US$107.10. A total of 188 patients were tested with an OGTT at a laboratory cost of US$7.70 per test. Twenty-two patients received a diagnosis of PTD by OGTT; thus, the laboratory cost per case diagnosed by this method was US$65.90, representing a decrease in cost per case diagnosed of 39% compared with FBG alone as a diagnostic test. By including patients with IGT in the cost analysis (n ⫽ 57), the cost per case of abnormal glucose tolerance by OGTT was US$18.90, repre-

104

Clinical Journal of the American Society of Nephrology

Clin J Am Soc Nephrol 1: 100 –108, 2006

Table 3. Relationship among sensitivity, specificity, and different levels of FBG in renal transplant recipients with PTDa FBG (mmol/L)

No. of RTR in FBG Category

No. of RTR with PTD

Sensitivity (%)

Specificity (%)

ⱖ117 (6.5) ⱖ110 (6.1) ⱖ108 (6.0) ⱖ104 (5.8) ⱖ103 (5.7) ⱖ101 (5.6) ⱖ99 (5.5) ⱖ97 (5.4) ⱖ95 (5.3) ⱖ94 (5.2) ⱖ90 (5.0) ⱖ81 (4.5) ⱖ72 (4.0)

14 25 26 43 55 69 81 95 109 117 141 180 188

9 9 10 10 12 14 14 14 17 17 19 22 22

40 40 45 45 55 64 64 68 77 77 86 100 100

97 92 91 81 74 67 60 48 45 40 27 8 1

RTR, renal transplant recipients; PTD, posttransplantation diabetes. Patients with FBG ⱖ126 mg/dl (7.0 mmol/L; n ⫽ 12) are not included. a

senting a total decrease in cost per case diagnosed of 82%, compared with FBG alone with the benefit of 100% cases of PTD and IGT being identified.

Discussion Principal Findings This study has produced several important findings. First, previously undiagnosed PTD and IGT, as determined from oral glucose tolerance testing, are extremely common after renal transplantation, occurring in up to 50% of RTR beyond the first 6 mo posttransplantation. Second, FBG alone may not be a suitable screening test for PTD and IGT in RTR because its ability to predict an abnormal 2HBG is poor. Finally, an FBG ⱖ126 mg/dl (7.0 mmol/L) may not be the optimal diagnostic test for PTD, because 65% of RTR with PTD diagnosed from 2HBG level will go undetected unless glucose tolerance testing is performed routinely.

PTD and IGT in RTR PTD is a significant complication of renal transplantation, and its incidence continues to rise (19,20). The development and progression of PTD and its associated complications may be delayed by readily available interventions that are proved to be beneficial, such that early detection is likely to be important. IGT is also common posttransplantation with reported prevalence rates as high as 35% at 10 wk posttransplantation, 31 to 45% at 1 yr posttransplantation, and 13% at 6 yr posttransplantation (1,21,22). In the general population, asymptomatic IGT is an independent risk factor for all-cause mortality and of cardiovascular mortality in some studies (9 –11,23–25). It is also associated with later development of type 2 diabetes (9). The relationship between IGT and mortality in RTR is less well established, although a causal link with the development of PTD is recognized (22). Whether intervention prevents the progression of IGT to PTD has yet to be established, although

preliminary studies in the general population are encouraging (26 –29). The detection of IGT and PTD in asymptomatic RTR at an early stage by an appropriate screening strategy is clearly justified (30). Because the pathogenesis of PTD and IGT is similar to the pathogenesis of type 2 diabetes and IGT in the general population (31), it seems reasonable to apply screening and diagnostic guidelines that are applicable to the general population.

Screening and Diagnostic Recommendations for PTD and IGT in RTR Although current recommendations are that FBG is the screening and diagnostic test of choice for PTD and IGT in RTR, it is acknowledged that the OGTT may provide useful supplementary information to FBG testing (13). In the general population, there is evidence from several epidemiologic trials that up to 50% of patients whose diabetes is diagnosed by 2HBG will have FBG ⬍126 mg/dl (7.0 mmol/L) (12,23,32,33), and a similar prevalence was observed in this small study. In nontransplant subjects, the 2HBG is more sensitive at diagnosing IGT than FBG (34) and is also a more sensitive predictor of cardiovascular mortality and all-cause mortality than FBG (10,35,36). The prognostic implications of a pathologic OGTT in RTR are currently uncertain and await further study.

Relationship between FBG and 2HBG in RTR The suitability of FBG as a screening and diagnostic test for PTD and IGT can be determined by examining the predictive power of FBG with respect to 2HBG. In the general population, the predictive power of FBG has been reported in several studies (37–39). In a study of 1034 subjects, FBG was found to predict accurately 2HBG ⱖ200 mg/dl (11.1 mmol/L) with optimal sensitivity and specificity (⬎80%) obtained at FBG of 103 mg/dl (5.7 mmol/L) (37). In another study, by taking a screening FBG level of 99 mg/dl (5.5 mmol/L) as a threshold for

Clin J Am Soc Nephrol 1: 100 –108, 2006

Screening for Glucose Intolerance in Renal Transplant Recipients

105

detect a common problem that can be treated and that if left untreated results in considerable morbidity and mortality (42). PTD and IGT are disease processes in which screening clearly is justified; however, the screening test of choice remains debatable. Although FBG is reproducible with a reported intraindividual coefficient of variation of 6.4 to 11.4% (43,44), its suitability in terms of sensitivity, specificity, and predictive value is questionable, particularly with regard to IGT. This was evident from our study in which FBG was nondiscriminatory at predicting IGT. Nonetheless, it is a relatively inexpensive test (US$6.40/test) that causes minimal inconvenience to the patient. The OGTT is more time-consuming to perform, is inconvenient for the patient, and is less reproducible (45) with a higher intraindividual coefficient of variation (14.3 to 16.7%) (43,44). Despite these limitations, however, the OGTT is a more sensitive test for diagnosing PTD and IGT than FBG with minimal additional laboratory expense (US$7.70/test).

Comparison of Diagnostic Strategies for PTD in RTR The prevalence of PTD using FBG alone as a diagnostic test was significantly less (6%) than that using an OGTT (17%) in this study. Our findings are congruent with results from studies in the general population in which prevalence of diabetes has been compared using different diagnostic criteria (32,35,46).

Limitations

Figure 2. ROC curve for FBG predicting impaired glucose tolerance (IGT) using time 0 FBG (a) and screening FBG (b). The AUC was similar for both (0.57 versus 0.54; P ⫽ 0.68).

performing an OGTT, 90% of patients with newly diabetes and 78% of patients with dysglycemia were detected (38). In RTR, although 2HBG has been used to assess both the prevalence of PTD and IGT and changing glucose tolerance over time (1,22,40,41), no studies to date have examined the relationship between FBG and 2HBG. In this study, the FBG that gave the optimal sensitivity and specificity for predicting PTD was 101 mg/dl (5.6 mmol/L). For IGT, the predictive power of FBG was poor, and an optimal level at which sensitivity and specificity were maximized was not clearly defined.

FBG or 2HBG as a Screening Test for PTD or IGT in RTR The ideal screening situation involves the use of an inexpensive, noninvasive, easily reproducible test that has a high level of sensitivity, specificity, and predictive value and that will

There were several limitations to this study. The study population was predominantly white, which may have affected the reported prevalence of PTD and IGT by oral glucose tolerance testing. More than 80% of the population was on corticosteroids, which may have had an impact on OGTT results. Care should be taken not to overinterpret these results, which may not be applicable to nonwhite transplant subjects or to populations in whom maintenance therapy with corticosteroids is less common. Nevertheless, it is evident that the prevalence of a pathologic OGTT is common in white RTR on a steroid-based regimen. Any potential limitations imposed by ethnicity or an immunosuppression regimen could be overcome easily by conducting a multicenter study, with particular attention paid to the effect of these possible confounders on FBG and 2HBG levels. Concurrent testing with HbA1c% in all study participants could have provided useful information, particularly because in those who had a pathologic OGTT and were tested for HbA1c% (47%), there was no correlation between 2HBG and HbA1c%. Further studies should address the relationship between HbA1c% and 2HBG in the context of both normal and abnormal glucose tolerance. An additional limitation of the OGTT is that it is less reproducible than an FBG. However, this could be addressed by reassessing glucose homeostasis with a second OGTT performed on a separate occasion. Finally, approximately 30% of RTR who were requested to have an OGTT did not attend for the test. These patients were younger and more likely to be in full-time employment and therefore less able to attend for testing, which highlights an important practical issue with regard to performing any timeconsuming test in a clinical setting. However, we are confident that if the test were implemented as part of necessary, routine

106

Clinical Journal of the American Society of Nephrology

Clin J Am Soc Nephrol 1: 100 –108, 2006

Table 4. Relationship between sensitivity, specificity, and different levels of FBG in renal transplant recipients with impaired glucose tolerancea FBG (mmol/L)

No. of RTR in FBG Category

No. of RTR with IGT

Sensitivity (%)

Specificity (%)

ⱖ117 (6.5) ⱖ110 (6.1) ⱖ108 (6.0) ⱖ104 (5.8) ⱖ103 (5.7) ⱖ101 (5.6) ⱖ99 (5.5) ⱖ97 (5.4) ⱖ95 (5.3) ⱖ94 (5.2) ⱖ90 (5.0) ⱖ81 (4.5) ⱖ72 (4.0)

14 25 26 43 55 69 81 95 109 117 141 180 188

3 9 10 16 17 23 27 29 32 37 43 55 57

5 16 18 28 30 40 47 51 56 65 75 97 100

86 82 82 76 74 70 63 56 48 42 31 7 1

a

IGT, impaired glucose tolerance. Patients with FBG ⱖ126 mg/dl (7.0 mmol/L; n ⫽ 12) are not included.

Figure 3. Graph showing percentage of total patients with PTD or IGT that would be detected in each category of FBG if an oral glucose tolerance test were performed (patients with FBG ⱖ126 mg/dl [7.0 mmol/L; n ⫽ 12] are not included).

posttransplantation care, rather than as an option for the patient within the context of a clinical study, the attendance rate may be greatly improved.

Conclusion The results of this study suggest that despite its simplicity and easy applicability, FBG alone may not be the optimal screening or diagnostic tool for PTD, owing to its lack of sensitivity and specificity. These findings confirm current recommendations that an OGTT should be performed in all RTR with normal or impaired fasting glycemia, although the frequency of routine OGTT testing remains uncertain and abnormal results may need to be interpreted with caution in patients

who are steroid dependent. Consideration could be given to adopting a similar OGTT screening policy in RTR to what has been recommended in the posttransplantation treatment of liver transplant recipients, namely, a 6-monthly OGTT in any patient with a impaired fasting glycemia and an annual OGTT otherwise (18). However, the results from this small study would need to be validated in much larger, heterogeneous populations before such a policy could be widely recommended. The prognostic implications of a pathologic OGTT when the FBG is normal are still to be determined. It is also unknown whether targeting of an elevated 2HBG will improve prognosis in RTR, and randomized, controlled clinical intervention trials are required to address this. It seems reasonable, however, on the basis of presented evidence that consideration should be given to introducing the OGTT as a routine posttransplantation investigation in all RTR. This will maximize detection of asymptomatic individuals at an early stage with a view to preventing the progression of the associated complications.

Acknowledgments K.A.A. was supported in this work by project grant 219285 and a Clinical Centre of Research Excellence Award from the National Health and Medical Research Council of Australia. We acknowledge the assistance of the staff and patients in the renal unit at the Princess Alexandra Hospital, Brisbane.

References 1. Hjelmesaeth J, Hartmann A, Kofstad J, Egeland T, Stenstrom J, Fauchald P: Tapering off prednisolone and cyclosporin the first year after renal transplantation: The effect on glucose tolerance. Nephrol Dial Transplant 16: 829 – 835, 2001 2. Kasiske BL, Snyder JJ, Gilbertson D, Matas AJ: Diabetes mellitus after kidney transplantation in the United States. Am J Transplant 3: 178 –185, 2003

Clin J Am Soc Nephrol 1: 100 –108, 2006

Screening for Glucose Intolerance in Renal Transplant Recipients

3. Montori VM, Basu A, Erwin PJ, Velosa JA, Gabriel SE, Kudva YC: Posttransplantation diabetes: A systematic review of the literature. Diabetes Care 25: 583–592, 2002 4. Cosio FG, Pesavento TE, Osei K, Henry ML, Ferguson RM: Post-transplant diabetes mellitus: Increasing incidence in renal allograft recipients transplanted in recent years. Kidney Int 59: 732–737, 2001 5. Cosio FG, Pesavento TE, Kim S, Osei K, Henry M, Ferguson RM: Patient survival after renal transplantation: IV. Impact of post-transplant diabetes. Kidney Int 62: 1440 – 1446, 2002 6. Lindholm A, Albrechtsen D, Frodin L, Tufveson G, Persson NH, Lundgren G: Ischemic heart disease—Major cause of death and graft loss after renal transplantation in Scandinavia. Transplantation 60: 451– 457, 1995 7. Revanur VK, Jardine AG, Kingsmore DB, Jaques BC, Hamilton DH, Jindal RM: Influence of diabetes mellitus on patient and graft survival in recipients of kidney transplantation. Clin Transplant 15: 89 –94, 2001 8. Woodward RS, Schnitzler MA, Baty J, Lowell JA, LopezRocafort L, Haider S, Woodworth TG, Brennan DC: Incidence and cost of new onset diabetes mellitus among US wait-listed and transplanted renal allograft recipients. Am J Transplant 3: 590 –598, 2003 9. Jarrett RJ, Keen H, McCartney P: The Whitehall Study: Ten year follow-up report on men with impaired glucose tolerance with reference to worsening to diabetes and predictors of death. Diabet Med 1: 279 –283, 1984 10. DECODE Study Group, the European Diabetes Epidemiology Group: Glucose tolerance and cardiovascular mortality: Comparison of fasting and 2-hour diagnostic criteria. Arch Intern Med 161: 397– 405, 2001 11. Meigs JB, Nathan DM, Wilson PW, Cupples LA, Singer DE: Metabolic risk factors worsen continuously across the spectrum of nondiabetic glucose tolerance. The Framingham Offspring Study. Ann Intern Med 128: 524 –533, 1998 12. Shaw JE, Hodge AM, de Court, Chitson P, Zimmet PZ: Isolated post-challenge hyperglycaemia confirmed as a risk factor for mortality. Diabetologia 42: 1050 –1054, 1999 13. Davidson J, Wilkinson A, Dantal J, Dotta F, Haller H, Hernandez D, Kasiske BL, Kiberd B, Krentz A, Legendre C, Marchetti P, Markell M, van der Woude FJ, Wheeler DC: New-onset diabetes after transplantation: 2003 international consensus guidelines. Proceedings of an international expert panel meeting. Barcelona, Spain, 19 February 2003. Transplantation 75: S3–S24, 2003 14. Definition, Diagnosis and Classification of Diabetes Mellitus and Its Complications: Report of a WHO Consultation. Part 1: Diagnosis and Classification of Diabetes Mellitus, Geneva, World Health Organization, 1999 15. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 26: S5, 2003 16. Kasiske BL, Vazquez MA, Harmon WE, Brown RS, Danovitch GM, Gaston RS, Roth D, Scandling JD, Singer GG: Recommendations for the outpatient surveillance of renal transplant recipients. American Society of Transplantation. J Am Soc Nephrol 11[Suppl 15]: S1–S86, 2000 17. Wilkinson A, Davidson J, Dotta F, Home PD, Keown P, Kiberd B, Jardine A, Levitt N, Marchetti P, Markell M, Naicker S, O’Connell P, Schnitzler M, Standl E, Torregosa JV, Uchida K, Valantine H, Villamil F, Vincenti F, Wissing M: Guidelines for the treatment and management of new-

18.

19.

20.

21.

22.

23.

24. 25.

26.

27.

28.

29.

30.

31.

32.

107

onset diabetes after transplantation. Clin Transplant 19: 291–298, 2005 Marchetti P: New-onset diabetes after liver transplantation: From pathogenesis to management. Liver Transpl 11: 612– 620, 2005 Moore R, Boucher A, Carter J, Kim SJ, Kiberd B, Loertscher R, Mongeau JG, Prasad GV, Vautour L: Diabetes mellitus in transplantation: 2002 consensus guidelines. Transplant Proc 35: 1265–1270, 2003 Reisaeter AV, Hartmann A: Risk factors and incidence of posttransplant diabetes mellitus. Transplant Proc 33: 8S– 18S, 2001 Nam JH, Mun JI, Kim SI, Kang SW, Choi KH, Park K, Ahn CW, Cha BS, Song YD, Lim SK, Kim KR, Lee HC, Huh KB: Beta-cell dysfunction rather than insulin resistance is the main contributing factor for the development of postrenal transplantation diabetes mellitus. Transplantation 71: 1417– 1423, 2001 Hagen M, Hjelmesaeth J, Jenssen T, Morkrid L, Hartmann A: A 6-year prospective study on new onset diabetes mellitus, insulin release and insulin sensitivity in renal transplant recipients. Nephrol Dial Transplant 18: 2154 –2159, 2003 Saydah SH, Loria CM, Eberhardt MS, Brancati FL: Subclinical states of glucose intolerance and risk of death in the US. Diabetes Care 24: 447– 453, 2001 Harris MI: Impaired glucose tolerance in the US population. Diabetes Care 12: 464 – 474, 1989 Tominaga M, Eguchi H, Manaka H, Igarashi K, Kato T, Sekikawa A: Impaired glucose tolerance is a risk factor for cardiovascular disease, but not impaired fasting glucose. The Funagata Diabetes Study. Diabetes Care 22: 920 –924, 1999 Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M: Acarbose for prevention of type 2 diabetes mellitus: The STOP-NIDDM randomised trial. Lancet 359: 2072–2077, 2002 Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, Nathan DM: Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 346: 393– 403, 2002 Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX, Hu ZX, Lin J, Xiao JZ, Cao HB, Liu PA, Jiang XG, Jiang YY, Wang JP, Zheng H, Zhang H, Bennett PH, Howard BV: Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care 20: 537–544, 1997 Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P, Keinanen-Kiukaanniemi S, Laakso M, Louheranta A, Rastas M, Salminen V, Uusitupa M: Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 344: 1343–1350, 2001 Markell M: New-onset diabetes mellitus in transplant patients: Pathogenesis, complications, and management. Am J Kidney Dis 43: 953–965, 2004 Hjelmesaeth J, Hagen M, Hartmann A, Midtvedt K, Egeland T, Jenssen T: The impact of impaired insulin release and insulin resistance on glucose intolerance after renal transplantation. Clin Transplant 16: 389 –396, 2002 Dunstan DW, Zimmet PZ, Welborn TA, de Court, Cameron AJ, Sicree RA, Dwyer T, Colagiuri S, Jolley D,

108

33.

34.

35.

36.

37.

38.

Clinical Journal of the American Society of Nephrology

Knuiman M, Atkins R, Shaw JE: The rising prevalence of diabetes and impaired glucose tolerance: The Australian Diabetes, Obesity and Lifestyle Study. Diabetes Care 25: 829 – 834, 2002 Taubert G, Winkelmann BR, Schleiffer T, Marz W, Winkler R, Gok R, Klein B, Schneider S, Boehm BO: Prevalence, predictors, and consequences of unrecognized diabetes mellitus in 3266 patients scheduled for coronary angiography. Am Heart J 145: 285–291, 2003 Genuth S, Alberti KG, Bennett P, Buse J, DeFronzo R, Kahn R, Kitzmiller J, Knowler WC, Lebovitz H, Lernmark A, Nathan D, Palmer J, Rizza R, Saudek C, Shaw J, Steffes M, Stern M, Tuomilehto J, Zimmet P: Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care 26: 3160 –3167, 2003 Barzilay JI, Spiekerman CF, Wahl PW, Kuller LH, Cushman M, Furberg CD, Dobs A, Polak JF, Savage PJ: Cardiovascular disease in older adults with glucose disorders: Comparison of American Diabetes Association criteria for diabetes mellitus with WHO criteria. Lancet 354: 622– 625, 1999 Will new diagnostic criteria for diabetes mellitus change phenotype of patients with diabetes? Reanalysis of European epidemiological data. DECODE Study Group on behalf of the European Diabetes Epidemiology Study Group. BMJ 317: 371–375, 1998 Botas P, Delgado E, Castano G, Diaz dG, Prieto J, DiazCadorniga FJ: Comparison of the diagnostic criteria for diabetes mellitus, WHO-1985, ADA-1997 and WHO-1999 in the adult population of Asturias (Spain). Diabet Med 20: 904 –908, 2003 Simmons D, Thompson CF, Engelgau MM: Controlling the diabetes epidemic: How should we screen for undiagnosed diabetes and dysglycaemia? Diabet Med 22: 207–212, 2005

Clin J Am Soc Nephrol 1: 100 –108, 2006

39. Carnevale Schianca GP, Rossi A, Sainaghi PP, Maduli E, Bartoli E: The significance of impaired fasting glucose versus impaired glucose tolerance: Importance of insulin secretion and resistance. Diabetes Care 26: 1333–1337, 2003 40. Hjelmesaeth J, Hartmann A, Kofstad J, Stenstrom J, Leivestad T, Egeland T, Fauchald P: Glucose intolerance after renal transplantation depends upon prednisolone dose and recipient age. Transplantation 64: 979 –983, 1997 41. Mathew JT, Rao M, Job V, Ratnaswamy S, Jacob CK: Posttransplant hyperglycaemia: A study of risk factors. Nephrol Dial Transplant 18: 164 –171, 2003 42. Nielsen C, Lang RS: Principles of screening. Med Clin North Am 83: 1323–1337, v, 1999 43. Feskens EJ, Bowles CH, Kromhout D: Intra- and interindividual variability of glucose tolerance in an elderly population. J Clin Epidemiol 44: 947–953, 1991 44. Mooy JM, Grootenhuis PA, de Vries H, Kostense PJ, PoppSnijders C, Bouter LM, Heine RJ: Intra-individual variation of glucose, specific insulin and proinsulin concentrations measured by two oral glucose tolerance tests in a general Caucasian population: The Hoorn Study. Diabetologia 39: 298 –305, 1996 45. Ko GT, Chan JC, Woo J, Lau E, Yeung VT, Chow CC, Cockram CS: The reproducibility and usefulness of the oral glucose tolerance test in screening for diabetes and other cardiovascular risk factors. Ann Clin Biochem 35: 62– 67, 1998 46. Gabir MM, Hanson RL, Dabelea D, Imperatore G, Roumain J, Bennett PH, Knowler WC: The 1997 American Diabetes Association and 1999 World Health Organization criteria for hyperglycemia in the diagnosis and prediction of diabetes. Diabetes Care 23: 1108 –1112, 2000