# Evaluation of three inflammation-associated blood indices for predicting malignancy in thyroid nodules

## TL;DR

This study evaluates three blood-based inflammation indices to see if they can help predict thyroid cancer before surgery.

## Contribution

The study is the first to assess the diagnostic value of SII, SIRI, and PIV in thyroid nodules.

## Key findings

- Only the Systemic Immune-Inflammation Index (SII) differed significantly between benign and malignant thyroid tumors.
- SII remained an independent predictor of malignancy in multivariate analysis.
- SII had limited diagnostic accuracy (AUC = 0.597) but may be useful as a low-cost adjunct in preoperative models.

## Abstract

Differentiated thyroid carcinoma (DTC) is the most common endocrine
malignancy and usually has a favorable prognosis. However, both diagnostic
and prognostic evaluations currently rely mainly on postoperative
histopathological results. Systemic inflammation-based indices — such as the
Systemic Immune-Inflammation Index (SII), Systemic Inflammation Response
Index (SIRI), and Pan-Immune Inflammation Value (PIV) — have recently
emerged as potential biomarkers in various cancers. This study aimed to
evaluate the diagnostic and prognostic utility of these indices in patients
undergoing thyroid surgery.

This retrospective study included 554 patients who underwent total
thyroidectomy between 2014 and 2021. Tumors were categorized as benign or
malignant according to final histopathology. SII, SIRI, and PIV were
calculated from preoperative complete blood counts. Multivariate logistic
regression was performed and included age, sex, thyroid-stimulating hormone
(TSH) level, glycated hemoglobin (HbA1c) level, and diabetes status.
Receiver operating characteristic (ROC) analysis was used to determine
diagnostic performance.

Among 554 patients, 366 had benign and 188 had malignant tumors. Among the
systemic inflammatory markers, only the SII differed significantly between
groups (p = 0.002) and remained an independent predictor of malignancy in
multivariate analysis (OR = 0.85 per 100-unit increase, p = 0.007). ROC
analysis revealed an AUC of 0.597, with 65.8% sensitivity and 58.2%
specificity. None of the indices demonstrated prognostic value in the
subgroup analyses.

The SII demonstrated independent but clinically limited diagnostic value in
differentiating malignant from benign thyroid lesions. Although its accuracy
was poor (AUC <0.6), the SII may serve as a low-cost adjunct within
multivariable preoperative models, particularly in indeterminate cytology
cases.