Effect of post-transplant glycemic control on long-term clinical outcomes in kidney transplant recipients with diabetic nephropathy: A multicenter cohort study in Korea

Abstract

Purpose: Diabetic nephropathy is the leading cause of end stage renal disease. The number of kidney transplantation (KT) due to diabetic nephropathy is increasing and there is debate on glycemic control after KT. In this study, we used a multi-center database to determine the relationship between post-transplant glycemic control and the outcomes of KT in patients with diabetic nephropathy.

Methods: We conducted a retrospective chart review of kidney transplant recipients (KTRs) with diabetic nephropathy from three tertiary hospitals to analyze the association between post-transplant glycemic control and the clinical outcomes of graft failure, including patient death and biopsy-proven acute rejection (BPAR). We assessed time-averaged glucose level and hemoglobin A1c (HbA1c) for 36 months after KT.

Results: Among 3,538 KTRs, a total of 476 patients received kidney transplantation because of diabetic nephropathy. Mean time-averaged glucose and HbA1c levels were 147 ± 46 mg/dl and 7.7 ± 1.5%, respectively. Patients with diabetic nephropathy had poor graft and patient survival rate compared with non-diabetic nephropathy. Among KTRs with diabetic nephropathy, the highest quartile of time-averaged glucose was related to poor graft outcomes and the 3rd quartile of time-averaged HbA1c was associated with significantly better graft outcomes than the 1st, 2nd or 4th quartiles. There were no significant differences in the risk of BPAR across the 4 quartiles of glucose and HbA1c.

Conclusions: Strict glycemic control before KT might not be related to successful outcomes but poor glycemic control after KT is associated with poor graft outcomes. There was no significant relationship between pre- or post-transplant glycemic control and BPAR.

Fig 1

Number (A) and proportion (B) of patients with diabetic nephropathy among total kidney transplantations from 1997 to 2011 in three hospitals (SNUH, AMC and KNUH). DN, diabetic nephropathy; SNUH, Seoul National University Hospital; AMC, Asan Medical Center; KNUH; Kyungpook National University Hospital.

Fig 2

Patient survival (A) and graft survival (B) for kidney transplant patients. DN, diabetic nephropathy.

Fig 3. Transition of post-transplant glycemic control by serum glucose level and HbA1c.

Fig 4. Kaplan-Meier estimates according to quartiles of glucose and HbA1c.

Graft survival included graft failure and patient death with functioning graft.

Fig 5

Hazard ratios of graft failure by serum glucose using standard Cox proportional hazards regression (A) and a time-averaged model (B). Hazard ratios of graft failure by HbA1c using standard Cox proportional hazards regression (C) and a time-averaged model (D). Model 1 is adjusted for age and gender. Model 2 is adjusted for age, gender, comorbidities (hypertension, ischemic heart disease), donor age, donor type, baseline hemoglobin level, and BPAR.

Fig 6

Hazard ratios of BPAR by serum glucose using standard Cox proportional hazards regression (A) and a time-averaged model (B). Hazard ratios of BPAR by HbA1c using standard Cox proportional hazards regression (C) and a time-averaged model (D). Model 1 is adjusted for age and gender. Model 2 is adjusted for age, gender, comorbidities (hypertension, ischemic heart disease), donor age, and donor type.