# A predictive algorithm for the optimal daily dosage of thiamazole to control cats with hyperthyroidism

**Authors:** Pak-Kan Tang, Nicola Lötter, Rebecca F Geddes, Rosanne E Jepson, Yu-Mei Chang, Harriet Syme, Jonathan Elliott

PMC · DOI: 10.1093/jvimsj/aalag009 · Journal of Veterinary Internal Medicine · 2026-02-03

## TL;DR

This study created a tool to predict the best starting dose of thiamazole for cats with hyperthyroidism based on blood test results.

## Contribution

A validated predictive algorithm for thiamazole dosage in hyperthyroid cats using baseline TT4 and creatinine levels.

## Key findings

- Higher TT4 and lower creatinine at diagnosis predict the need for >5 mg thiamazole daily.
- The algorithm achieved 91.1% accuracy in predicting the correct dosage.
- The model showed high specificity (96.2%) and moderate sensitivity (84.2%).

## Abstract

Hyperthyroidism is the most common endocrinopathy in cats and is frequently managed using anti-thyroid medication.

To develop and validate an algorithm to predict the optimal starting daily dose of thiamazole required to control hyperthyroidism in cats.

One hundred eighty-eight client-owned cats with hyperthyroidism for algorithm development (2011-2021) and 45 hyperthyroid cats to validate the algorithm (2022-2024).

Retrospective case-control study. Cats with hyperthyroidism controlled medically using thiamazole within a year since diagnosis were enrolled. Controlled dose of thiamazole was categorized into “≤5 mg” or “>5 mg.” Binary logistic regression was performed to explore predictors associated with thiamazole dose. The performance of the final multivariable model in prediction was assessed by receiver operating characteristic (ROC) curve analysis. A cohort of cats subsequently diagnosed with hyperthyroidism and managed chronically with thiamazole were used to test algorithm performance.

At hyperthyroidism diagnosis, baseline plasma total thyroxine (TT4); (odds ratio [OR] 1.29 [95% CI, 1.19-1.42] per 10 nmol/L; P < .001) and creatinine concentrations (OR 0.83 [95% CI, 0.7-0.96] per 0.1 mg/dL; P = .02) were independent predictors for higher thiamazole dose (>5 mg). The area under the ROC curve was 0.92 (95% CI, 0.88-0.96). In the test cohort, 26 cats controlled on ≤ 5 mg and 19 required >5 mg thiamazole. The predictive model had overall accuracy of 91.1%, sensitivity of 84.2%, and specificity of 96.2%.

Hyperthyroid cats with higher plasma TT4 and lower creatinine concentrations at diagnosis are likely to require >5 mg total daily dose of thiamazole to achieve euthyroidism.

## Linked entities

- **Chemicals:** thiamazole (PubChem CID 1349907), thyroxine (PubChem CID 853), creatinine (PubChem CID 588)
- **Diseases:** hyperthyroidism (MONDO:0004425)

## Full-text entities

- **Genes:** Renin [NCBI Gene 101081695], ALB (albumin) [NCBI Gene 213] {aka FDAHT, HSA, PRO0883, PRO0903, PRO1341}, thyroglobulin [NCBI Gene 101088550], TPO (thyroid peroxidase) [NCBI Gene 7173] {aka MSA, TDH2A, TPX}, GPT (glutamic--pyruvic transaminase) [NCBI Gene 2875] {aka AAT1, ALT, ALT1, GPT1, SGPT}, ALPP (alkaline phosphatase, placental) [NCBI Gene 250] {aka ALP, PALP, PLAP, PLAP-1}, TPO [NCBI Gene 101098706]
- **Diseases:** IRIS (MESH:C000719191), renal dysfunction (MESH:D007674), muscle condition (MESH:D009135), thyrotoxicosis (MESH:C566386), goiter (MESH:D006042), death (MESH:D003643), hypothyroidism (MESH:D007037), endocrinopathy (MESH:C567425), weight loss (MESH:D015431), tachycardia (MESH:D013610), Hyperthyroid (MESH:D006980), gland (MESH:D000307), Azotemic CKD (MESH:D051436)
- **Chemicals:** creatinine (MESH:D003404), calcium (MESH:D002118), Anti-thyroid medication (-), thyroxine (MESH:D013974), carbimazole (MESH:D002231), potassium (MESH:D011188), sodium (MESH:D012964), urea (MESH:D014508), I131 (MESH:C000614965), tyrosine (MESH:D014443), Thiamazole (MESH:D008713), cholesterol (MESH:D002784), iodide (MESH:D007454), chloride (MESH:D002712), phosphate (MESH:D010710), bilirubin (MESH:D001663), aldosterone (MESH:D000450)
- **Species:** Felis catus (cat, species) [taxon 9685], Homo sapiens (human, species) [taxon 9606], Canis lupus familiaris (dog, subspecies) [taxon 9615]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12866908/full.md

## Figures

4 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12866908/full.md

## References

38 references — full list in the complete paper: https://tomesphere.com/paper/PMC12866908/full.md

---
Source: https://tomesphere.com/paper/PMC12866908