# Construction and validation of a risk prediction model for hyperamylasemia after kidney transplantation

## TL;DR

This study creates a risk prediction model to identify kidney transplant patients likely to develop hyperamylasemia, helping doctors monitor and manage risks more effectively.

## Contribution

The study introduces a novel nomogram for predicting hyperamylasemia after kidney transplantation, validated internally with clinical data.

## Key findings

- The nomogram incorporates six factors including WBC count, tacrolimus concentration, and donor age to predict hyperamylasemia risk.
- The model achieved an AUC of 0.73 in both training and validation cohorts, indicating moderate predictive accuracy.
- Calibration plots and decision curve analysis confirmed the model's clinical utility and reliability.

## Abstract

Kidney transplantation (KT) is the preferred treatment for patients with end-stage renal disease (ESRD); however, postoperative hyperamylasemia (HA) remains common and has been associated with acute rejection (AR), infection, and impaired graft function. Early identification of HA risk factors is essential to improve outcomes of kidney transplant recipients (KTR). This study aimed to develop and internally validate a novel nomogram for predicting the risk of HA after KT, thereby supporting personalized monitoring, prevention and intervention strategies.

We retrospectively analyzed KTR treated at the Transplant Medicine Institution of the Second Affiliated Hospital of Guangxi Medical University from July 2021 to June 2022. Based on admission dates, patients were assigned to a training cohort (n=243, July 2021 to March 2022) and a validation cohort (n=107, April 2022 to June 2022). In the training cohort, risk factors of HA were identified using logistic regression, Lasso regression and clinical consideration. Subsequently, a nomogram was developed to predict HA risk in patients who underwent KT based on the identified variables. Model performance was evaluated using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA).

A total of 350 KTR and their corresponding 182 donors were enrolled in this study. The nomogram incorporated six predictive factors: recipient preoperative white blood cell (WBC) count, induction, tacrolimus (FK506) trough concentration, AR, donor age, and donor total bilirubin (TBIL) level according to results of logistic regression, Lasso regression and clinical consideration. The nomogram showed moderate predictive performance, with an area under the ROC curve (AUC) of 0.730 (Youden index = 0.683) in the training cohort and 0.731 (Youden index = 0.767) in the validation cohort. Furthermore, calibration plots indicated close agreement between predicted and actual outcomes, and DCA confirmed net clinical benefit across a range of threshold probabilities.

A novel nomogram was established to predict HA after KT, which may support early risk stratification and personalized management of KTR. External multicenter validation is needed before clinical implementation.