When Guidelines Meet Reality: The Combined Impact of Assay Variability and Prescribing Differences on TSH Management...

Patients with differentiated thyroid cancer (DTC) receive thyroxine substitution targeting thyroid-stimulating hormone (TSH) levels based on their treatment response category. However, variations in prescribing and inter-assay TSH variability may result in over or undertreatment.

We measured TSH in 220 consecutive DTC patients using three automated immunoassay platforms (Elecsys, Atellica, Alinity). Each patient was assigned to a response-to-therapy category (Excellent Response [ER], Indeterminate Response [IndR], Biochemical Incomplete Response [BIR], Structural Incomplete Response [SIR]) by an experienced thyroid oncologist. We defined recommended TSH targets according to the American Thyroid Association (ATA) 2015 guidelines and the response-adapted ATA 2025 framework that allows progressive relaxation of TSH suppression in patients with ER while maintaining tight suppression in those with persistent disease. Analytical agreement between assays was assessed using Passing-Bablok regression and Bland-Altman analysis. Clinical appropriateness was evaluated by classifying each measured TSH value as below, within, or above the recommended range for that patient's response category.

The three immunoassays demonstrated high analytical agreement with only minor biases unlikely to affect clinical interpretation. However, significant deviations from guideline-defined TSH targets were observed. Among ER patients, 37% remained oversuppressed despite the absence of active disease. Conversely, in IndR or BIR patients, 76% had TSH levels above the recommended range, indicating undersuppression where residual disease could not be excluded. SIR patients were generally managed appropriately. The ATA 2025 framework reclassified more ER patients as appropriately managed, but undersuppression persisted in non-ER patients.

Guidelines are not uniformly applied in thyroxine dosing for DTC patients. TSH immunoassays have achieved adequate analytical performance. The focus must now shift toward addressing clinical, educational, and systemic factors that prevent optimal levothyroxine management.