Blood glucose linked to aggressive features in papillary thyroid cancer

In a retrospective study of 1,038 adults with papillary thyroid carcinoma who underwent surgery at the Affiliated Hospital of Qingdao University, elevated fasting blood glucose was associated with both extracapsular extension and lymph node metastasis, while the triglyceride–glucose index was linked to extracapsular extension but not clearly to lymph node spread. The investigators reported that adding fasting glucose to a clinical model modestly improved discrimination for both aggressive pathological outcomes, whereas the triglyceride–glucose index offered limited additional value for predicting lymph node metastasis.

The research, conducted between June 2020 and June 2024 and reported in Frontiers in Endocrinology, included patients with postoperative pathological confirmation of papillary thyroid carcinoma. Eligibility required complete clinical, laboratory and pathological information. The study excluded other thyroid cancers, other malignant tumors, major organ or autoimmune diseases, distant metastases, a body mass index below 18.5 kg/m², and medications that could affect thyroid function, lipid profiles, uric acid or body mass index. Because the analysis was cross-sectional and retrospective, it could identify associations with tumor characteristics but could not establish whether abnormal glucose or lipid metabolism preceded those characteristics or caused them.

The cohort had a mean age of 45.71 years and included 244 men and 794 women. Type 2 diabetes was present in 102 participants, or 9.83% of the cohort, and their mean diabetes duration was 6.13 years. Compared with patients without diabetes, those with diabetes had higher fasting glucose, triglycerides, body mass index and triglyceride–glucose index values, as well as a lower high-density lipoprotein cholesterol concentration. They were also more likely to have primary tumors larger than 1 cm, extracapsular extension, lymph node metastasis, intermediate-to-high recurrence risk and advanced TNM stage III or IV disease. After adjustment for age, sex and thyroid-stimulating hormone, diabetes was associated with odds ratios of 1.618 for a tumor larger than 1 cm, 2.867 for extracapsular extension, 3.090 for lymph node metastasis, 3.157 for intermediate-to-high recurrence risk and 2.313 for advanced TNM stage.

When the researchers examined individual metabolic indicators, fasting blood glucose of at least 6.1 mmol/L was associated with multifocal and bilateral tumors, a maximum tumor diameter greater than 1 cm, extracapsular extension, lymph node metastasis, intermediate-to-high recurrence risk and advanced TNM stage. After adjustment for age, sex and thyroid-stimulating hormone, the corresponding odds ratios ranged from 1.412 for multifocality to 3.163 for extracapsular extension. Triglycerides above 1.92 mmol/L were associated with extracapsular extension, with an odds ratio of 1.611, while uric acid above 420 μmol/L was associated with extracapsular extension and intermediate-to-high recurrence risk. A body mass index of at least 25 kg/m² was associated with several pathological features, including extracapsular extension and lymph node metastasis. No significant associations were found for total cholesterol, low-density lipoprotein cholesterol or high-density lipoprotein cholesterol in these models.

The triglyceride–glucose index, calculated as the natural logarithm of fasting triglycerides in mg/dL multiplied by fasting glucose in mg/dL and divided by two, was analyzed both at a reference threshold and across quartiles. Patients in the highest quartile had a higher proportion of extracapsular extension than those in the lower three quartiles. The third and fourth quartiles also contained more patients with advanced TNM stage than the first and second quartiles. After adjustment for age, sex and thyroid-stimulating hormone, the highest quartile was associated with extracapsular extension compared with the lowest quartile, with an odds ratio of 2.103. The authors described the index as a practical surrogate for insulin resistance, although the study did not directly measure insulin concentrations, clamp-based insulin sensitivity or longitudinal changes in metabolic status.

More extensive multivariable analysis supported an association between elevated fasting glucose and the principal aggressive features. In a model adjusted for age, sex, thyroid-stimulating hormone, body mass index and several lipid and metabolic variables, fasting glucose of at least 6.1 mmol/L was associated with extracapsular extension with an odds ratio of 2.965 and with lymph node metastasis with an odds ratio of 2.516. Elevated triglyceride–glucose index was associated with extracapsular extension, with an odds ratio of 2.296. The fasting-glucose associations persisted after further adjustment for the number, distribution and maximum diameter of primary tumor foci, with an additional adjustment for extracapsular extension in the lymph-node model: the final odds ratios were 2.780 for extracapsular extension and 2.034 for lymph node metastasis. Elevated triglyceride–glucose index remained associated with extracapsular extension, with an adjusted odds ratio of 2.337.

The study also tested whether the metabolic measures improved prediction beyond a baseline model containing demographic, thyroid, anthropometric, lipid and tumor characteristics. For extracapsular extension, the area under the receiver operating characteristic curve was 0.710 for the baseline model, 0.738 after fasting glucose was added and 0.722 after the triglyceride–glucose index was added. The corresponding increases were statistically significant for fasting glucose, at 0.028, but not for the index, at 0.012. For lymph node metastasis, the areas under the curve were 0.712, 0.727 and 0.714, respectively. Adding fasting glucose produced a significant increase of 0.015, whereas adding the triglyceride–glucose index produced a negligible increase of 0.002 that was not statistically significant.

Risk-reclassification measures told a similar story. For extracapsular extension, adding fasting glucose yielded a continuous net reclassification improvement of 0.406 and an integrated discrimination improvement of 0.040, both statistically significant. Adding the triglyceride–glucose index also generated statistically significant values of 0.285 and 0.015 for these measures, but the smaller changes in discrimination indicated limited incremental performance. For lymph node metastasis, the index showed no obvious incremental contribution: its continuous net reclassification improvement was 0.016 and was not significant. Following 1,000 bootstrap resamples, the optimism-corrected areas under the curve for the fasting-glucose model were 0.722 for extracapsular extension and 0.712 for lymph node metastasis. The corresponding values for the triglyceride–glucose model were 0.704 and 0.698, while calibration slopes ranged from 0.881 to 0.902 across the validated models.

These findings should be interpreted as observational rather than causal. The single-center design, retrospective collection of baseline metabolic measurements and lack of external validation restrict generalization. Data on smoking, alcohol consumption, glucose-lowering medication, glycated hemoglobin and inflammatory markers were missing, raising the possibility of residual confounding. The study also used the triglyceride–glucose index as an indirect marker of insulin resistance rather than measuring it directly, and patients with distant metastases or certain major illnesses were excluded. The results therefore do not show that lowering blood glucose, treating diabetes or changing triglyceride levels would alter tumor behavior. The authors called for multicenter prospective cohorts, mechanistic investigations and external validation to determine whether the observed metabolic associations reflect causal relationships and whether metabolic interventions could influence thyroid cancer progression or prognosis.

Subject of Research: Associations of fasting glucose and the triglyceride–glucose index with aggressive pathological features of papillary thyroid carcinoma

Article Title: Associations of fasting blood glucose and the triglyceride–glucose index with aggressive pathological features of papillary thyroid carcinoma

Article References: Xu, X., Gong, K., Li, C., Hu, Y., Liu, C., Guo, P., Zhao, W., Wang, Y., & Dong, B. (2026). Associations of fasting blood glucose and the triglyceride–glucose index with aggressive pathological features of papillary thyroid carcinoma. Frontiers in Endocrinology, 17, Article 1931124. https://doi.org/10.3389/fendo.2026.1931124

Image Credits: AI Generated

DOI: 10.3389/fendo.2026.1931124

Keywords: papillary thyroid carcinoma, fasting blood glucose, triglyceride–glucose index, extracapsular extension, lymph node metastasis, insulin resistance, type 2 diabetes, thyroid cancer, metabolic health, tumor aggressiveness, Associations, fasting