# Development and internal validation of a TLR2-based nomogram for diagnosing pulmonary infection in type 2 diabetes

## TL;DR

This study developed a TLR2-based diagnostic model to help identify pulmonary infections in patients with type 2 diabetes, showing strong accuracy and potential clinical use.

## Contribution

The novel contribution is the development and validation of a TLR2-based nomogram for early diagnosis of pulmonary infection in type 2 diabetes patients.

## Key findings

- Patients with T2DM and pulmonary infection had higher TLR2/4 mRNA, inflammatory markers, and diabetes-related risk factors.
- The developed nomogram achieved excellent discrimination with an AUC of 0.987 and strong internal validation performance.
- The model showed good calibration and clinical benefit across a wide range of threshold probabilities.

## Abstract

This study aimed to characterize the expression levels and identify the risk factors associated with Toll-like receptor 2 and 4 (TLR2/4) mRNA in peripheral blood mononuclear cells of patients with type 2 diabetes mellitus (T2DM) complicated with pulmonary infection, and to develop and internally validate a nomogram-based diagnostic model.

A total of 239 patients with T2DM admitted to our hospital between January and August 2025 were selected. Based on the presence of concurrent pulmonary infection at admission, they were divided into the T2DM group (n=128) and the T2DM with pulmonary infection group (n=111). TLR2/4 mRNA expression levels, general characteristics, and peripheral blood inflammatory markers were compared between the two groups. Predictors were identified using LASSO regression and logistic regression to construct a discriminant model, with receiver operating characteristic (ROC) curves plotted. Internal validation employed 10-fold cross-validation and bootstrap-based optimism correction (B = 1000). Model performance was assessed via Hosmer-Lemeshow tests and decision curve analysis (DCA).

Patients with T2DM and pulmonary infection exhibited significantly elevated levels of fasting blood glucose, inflammatory markers (WBC, NEUT, hsCRP, PCT, ESR), and TLR2/4 mRNA expression, as well as higher rates of invasive procedures, compared with the T2DM group (all P < 0.05). Using LASSO feature selection followed by multivariable logistic regression, a diagnostic nomogram was developed incorporating TLR2, IL-6, TNF-α, ESR, age, and diabetes duration. The nomogram demonstrated excellent discrimination, with an apparent AUC of 0.987. Internal validation confirmed robust performance, yielding a 10-fold cross-validation AUC of 0.980 ± 0.006 and a bootstrap optimism-corrected AUC of 0.980 (B = 1,000).The Hosmer–Lemeshow test indicated good calibration (P > 0.05). DCA showed substantial net clinical benefit across threshold probabilities ranging from 0.10 to 0.60.

TLR2, IL-6, TNF-α, ESR, age, and duration of diabetes can serve as a combined biomarker panel to aid in the early diagnosis of pulmonary infection in T2DM patients at hospital admission. The proposed nomogram demonstrates strong diagnostic performance and potential clinical utility.