A Large-Scale Study Identifies a TgAb Interference Threshold for Thyroglobulin Immunoassays Using Segmented Regression and Bootstrap Validation.

BACKGROUND

Thyroglobulin (Tg) measurement is key for differentiated thyroid cancer (DTC) follow-up, but anti-Tg antibodies (TgAbs) can interfere with Tg immunoassays and bias Tg interpretation. Current TgAb cutoffs are heterogeneous, and an assay-specific interference threshold is needed.

METHODS

We analyzed 85 949 euthyroid health-screening records (Roche e801 platform) using segmented linear regression and 1000 bootstrap resamples to estimate a TgAb interference threshold. For external validation, 125 TgAb-positive post-thyroidectomy papillary thyroid carcinoma (PTC) patients were included, and receiver operating characteristic (ROC) analyses were performed. Two ROC analyses were performed: (a) low-TgAb patients without structural recurrence (Group A) vs patients with structural recurrence (Group C), and (b) Groups A + B (no structural recurrence, including persistently elevated TgAb) vs Group C.

RESULTS

The population-derived TgAb interference threshold was 48.1 U/mL (95% CI, 45.1-50.6). Above this level, Tg values shifted toward lower ranges with increased distributional asymmetry. In the PTC validation cohort, ROC-derived thresholds were 45.64 U/mL (A vs C) and 50.14 U/mL (A + B vs C), with areas under the curve (AUCs) of 1.00 and 0.8764, supporting internal stability and clinical concordance.

CONCLUSION

On the Roche cobas e801 platform, a TgAb level near 48 U/mL provides a pragmatic interference threshold for Tg immunoassay interpretation. This threshold stratifies Tg interpretability and is concordant with ROC-derived cutoffs from clinically defined comparisons in a TgAb-positive PTC cohort.