Ratio of triglyceride to high-density lipoprotein cholesterol and risk of major cardiovascular events in kidney transplant recipients

The study design was approved by the Institutional Review Board of Seoul National University Hospital (no. H-1803-044-926) and complied with the Declaration of Helsinki. This retrospective, observational study included patients who underwent renal transplantation at Seoul National University Hospital between 2000 and 2017. KT recipients with serum lipid profiles performed 1 year after transplantation were enrolled. There were no age restrictions, and patients who had undergone a second renal transplantation or simultaneous transplantation of other organs were also included. The patients who underwent renal transplantation at another hospital were excluded.

All data were obtained from the hospital’s electronic medical records. Demographic characteristics including age, sex, height, weight, and body mass index (BMI) were collected. The type of renal replacement therapy before transplantation was reviewed, and the duration of dialysis was calculated on the basis of the date of dialysis initiation. The donor type (living or deceased) was also recorded. The donor and recipient’s ABO and human leukocyte antigen (HLA) typing data were collected and compared to identify ABO incompatibility and the number of HLA mismatches. As our center utilizes a standardized, maintenance immunosuppressant regimen of steroids, mycophenolic acid, and calcineurin inhibitors with only a few exceptions, the prescription rate of steroids or mycophenolic acid was over 90% in the total cohort. However, the use of calcineurin inhibitors differed between patients; therefore, the type of calcineurin inhibitor was collected and analyzed. In addition, the use of mammalian target of rapamycin inhibitors was collected within 1 year after transplantation. The presence of comorbidities such as hypertension, diabetes mellitus (DM), and CVD before transplantation was analyzed. A history of CVD was defined as any history of heart failure, coronary artery disease, or cerebrovascular disease before transplantation. In addition, the duration of statin treatment after transplantation was reviewed. In our study, statin use was defined as the prescription of any 3-hydroxy-3-methyl-glutanyl-coenzyme A reductase inhibitor for at least 3 months during the first year after transplantation. In addition, use of fibrates and omega-3-acids has been defined as the prescription of any medication containing fibrates or omega-3-acids, respectively, for at least 3 months during the first year after transplantation. And steroid pulse therapy for rejection after transplantation was considered to be a factor affecting the lipid profile, so we reviewed the history of IV steroid pulse treatment within 1 year in study participants.

Total cholesterol, triglyceride, HDL-C, and LDL-C levels 1 year after renal transplantation were reviewed, and the TG/HDL-C ratio was calculated. In our center, patients are advised to measure the lipid profile after at least 12 h of fasting. Each lipid profile was categorized into five groups according to the values.

The main outcome was the association between TG/HDL-C and MACE. MACE was defined as heart failure, coronary artery disease documented by coronary angiography, and cerebrovascular disease confirmed by imaging. All patients were followed until March 2018.

The secondary outcome was the association of TG/HDL-C with both graft and patient survival. Graft failure was defined as a return to dialysis or kidney re-transplantation. Mortality data were obtained from the National Database of Statistics Korea.

Biopsy findings during the first year after transplantation were also collected from the medical records. In our center, protocol biopsies were performed at time zero (post-reperfusion), on the 10th day, and 1 year after transplantation only if the patient agreed. Additional kidney biopsies were performed if graft function deteriorated or any suspicious symptoms or signs of rejection were observed. All pathologic findings were examined by two experienced nephropathologists.

A total of 1301 KT recipients were enrolled in the final analysis. The mean patient age was 40.9 ± 16.2 years, and men comprised 61.7% of the study population. The statin prescription rate was 30.3%. The patients showed total cholesterol level of 177.3 ± 33.2 mg/dL, TG of 128.5 ± 75.0 mg/dL, and HDL-C of 56.7 ± 16.7 mg/dL at 1 year after transplantation. LDL-C was also reviewed but only 785 of the 1301 (60.3%) patients had results, and the mean level of LDL-C was 96.4 ± 27.5 mg/dL. Table 1 reports the baseline characteristics of all patients.

The median follow-up duration after renal transplantation in the total study population was 7.4 years (interquartile range 4.4–11.1 years). During the follow-up period, 80 (6.2%) patients developed MACE, which included 38 of unstable anginas, 9 of MIs, 19 of heart failures, 18 of cerebral infarcts, and 4 of cerebral hemorrhages. The total number of events was 89.

Then, we explored risk factors for development of MACE after renal transplantation. In the univariate analysis, age over 60 years, male sex, BMI over 25 kg/m², deceased donor, prior history of CVD, hemodialysis as the RRT modality before transplant, duration of RRT over 1 year, high numbers of HLA mismatch, a history of diabetes and statin usage were related to an increased HR of MACE. However, after adjustment for multiple variables including lipid profiles, aged over 60 years and a history of diabetes showed a statistically significant elevation of the HR for MACE (Table 2).

For the lipid profile, cubic spline analysis was performed on the continuous variable before conducting cox regression on the categorical variable (Fig. 1). In the cubic spline analysis, among the lipid profiles, TC, TG, and HDL-C showed no significant tendency for risk of MACE according to lipid levels. Only TG/HDL-C showed significant correlation with MACE risk and a significant increase in MACE risk at high TG/HDL-C values. To perform cox regression analysis for these lipid profiles, the lipid profile was divided into five groups according to the values. The first quintile was defined as the reference group for the analysis, but in the case of TG/HDL-C, the second quintile was defined as the reference group because the spline curve showed J shape. In univariate and multivariate analyses, the fourth and fifth quintiles of TG/HDL-C showed a significant elevation of the HR compared to the reference group [fourth quintile: aHR, 3.38; 95% confidence interval (CI) 1.44–7.95; p = 0.005; fifth quintile: aHR, 2.67; 95% CI 1.13–6.30; p = 0.025]. Moreover, the first quintile showed marginal difference of MACE risk compared to the reference group (aHR, 2.49; 95% CI 1.01–6.23; p = 0.047). The cubic spline curves and results of cox regression are shown in Fig. 1 and Table 2, respectively.

Nelson–Aalen cumulative hazard curves for quintiles of TG/HDL-C also showed a significant difference of cumulative risk at the fourth and fifth quintiles compared to the second quintile (log rank p = 0.003 for the fourth quintile, p = 0.004 for the fifth quintile). Figure 2 shows the cumulative risk for MACE in the quintiles.

When comparing the baseline characteristics of the TG/HDL-C quintiles, there were significant differences in the proportion of men, BMI, prior history of CVD and DM, and statin use. Higher TG/HDL-C quintiles had a higher ratio of men, a higher BMI, and a more frequent history of CVD and DM. Statin prescription rate was the lowest in the third quintile. Although there was no statistical significance, the statin prescription rate was higher in the fourth and fifth quintiles than in the first and second quintiles. Table 3 shows the demographic and clinical characteristics of each quintile of TG/HDL-C.

Subgroup analysis was performed to evaluate the confounding effect of multiple baseline characteristics (Fig. 3). For the subgroup analysis, we divided TG/HDL-C into two groups, the second quintile or lower and greater than the second quintile. This criterion is due to the lowest MACE risk by TG/HDL-C in the second quintile. When comparing the MACE risk between the two categories of TG/HDL-C, age, sex, and BMI showed no significant difference of risk between the subgroups. In contrast, the group with a history of diabetes showed no difference in risk according to TG/HDL-C, whereas the group with no history of diabetes showed an elevated risk of MACE. Similarly, the subgroups with a history of hypertension, no history of CVD, and statin usage were related to the risk of MACE; the counterparts of these subgroups showed no significant risk elevation.

To determine the effect of TG/HDL-C on graft pathology, 1-year protocol biopsy results were collected. In our center, protocol biopsy was recommended to KT recipients 1 year after transplantation and performed if informed consent could be obtained. Approximately one-third of KT recipients (n = 387, 29.7%) underwent allograft biopsy 1 year after transplantation. Among these recipients, logistic analysis showed no significant relationship between TG/HDL-C and 1-year post-transplant interstitial fibrosis and tubular atrophy (IFTA) or T-cell mediated rejection (Table 4). Additionally, the cox regression analysis was performed on all study patients for graft and patient survival. We found lower risk of graft failure in the first quintile of TG/HDL-C compared to reference group with marginal significance (aHR, 0.48; 95% CI 0.22–1.03; p = 0.058, Table 4).