Perioperative glucose monitoring with continuous glucose monitors identifies risk factors for post-transplant diabetes mellitus in kidney transplant recipients

Abstract

Post-transplantation diabetes mellitus (PTDM) negatively affects graft and patient survival after kidney transplantation (KT). This prospective study used continuous glucose monitoring (CGM) to evaluate perioperative blood glucose dynamics, identify PTDM risk factors, and compare predictive accuracy with capillary blood glucose monitoring (CBGM) in 60 non-diabetic living-donor KT recipients. Patients underwent 2-week pre- and postoperative CGM, including routine CBGM during their in-hospital stays. PTDM-related risk factors and glucose profiles were analyzed with postoperative CGM and CBG. PTDM developed in 14 (23.3%) patients and was associated with older age, male sex, higher baseline HbA1c, high-density lipoprotein cholesterol, and 3-month cumulative tacrolimus exposure levels. Male sex and postoperative time above the range (TAR) of 180 mg/dL by CGM were PTDM-related risk factors in the multivariate analysis. For predictive power, the CGM model with postoperative glucose profiles exhibited higher accuracy compared with the CBGM model (areas under the curves of 0.916, and 0.865, respectively). Therefore, we found that male patients with a higher postoperative TAR of 180 mg/dL have an increased risk of PTDM. Postoperative CGM provides detailed glucose dynamics and demonstrates superior predictive potential for PTDM than CBGM.

Keywords: Blood glucose; Continuous glucose monitoring; Diabetes mellitus; Follow up; Kidney transplantation; Postoperative care.

Fig. 1

Daily glucose variation in the preoperative and postoperative non-PTDM and PTDM groups was assessed using CGM. (A) Preoperative daily glucose variation in the non-PTDM group. (B) Postoperative daily glucose variation in the non-PTDM group. (C) Preoperative daily glucose variation in the PTDM group. (D) Postoperative daily glucose variation in the PTDM group. CGM continuous glucose monitoring, PTDM post-transplantation diabetes mellitus.

Fig. 2

Perioperative daily variation of glucose levels in the entire patient cohort was detected using a continuous glucose monitoring device in the preoperative 2-week period (A) and postoperative 2-week period (B). The solid black lines in each graph represent the mean glucose levels of the cohort.

Fig. 3

Comparison of preoperative and postoperative glucose levels in the PTDM and non-PTDM groups as assessed using continuous glucose monitoring. (A) Preoperative and postoperative mean daily peak glucose levels in the non-PTDM and PTDM groups. (B) Preoperative and postoperative mean daily glucose levels in the non-PTDM and PTDM groups. (C) Preoperative and postoperative mean daily nadir glucose levels in the non-PTDM and PTDM groups. Pre-op preoperative, post-op postoperative, PTDM post-transplantation diabetes mellitus.

Fig. 4

Postoperative changes in the metabolic indices HbA1c, HDL, HOMA-IR, HOMA-B, Triglyceride/HDL ratio, and body weight during the follow-up period. *P < 0.05, **P < 0.01. CGM continuous glucose monitoring, HOMA-B homeostasis model assessment of beta-cell function, HOMA-IR homeostasis model assessment of insulin resistance, HbA1c hemoglobin A1c, HDL high-density lipoprotein, PTDM post-transplantation diabetes mellitus.

Fig. 5

Receiver operating characteristics curve analysis for evaluation of the predictive performance of post-transplantation diabetes mellitus occurrence using postoperative CBGM and CGM models. AUC area under the curve, CBGM capillary blood glucose monitoring, CGM continuous glucose monitoring.