Differentiated thyroid cancer (DTC) is the most prevalent endocrine malignancy, the majority of which are papillary thyroid cancer (PTC) and follicular thyroid cancer (FTC) with an excellent prognosis (1). Currently, total or near-total thyroidectomy, postoperative selective radioactive iodine (RAI/¹³¹I), and long-term thyroid-stimulating hormone (TSH) suppression are commonly used as the first-line clinical treatments for patients with DTC (2). Biochemical measurements in combination with imaging examinations such as cervical ultrasound (US), chest computed tomography (CT), and radioiodine diagnostic whole-body scan (DxWBS) are considered the standard of care for the management of DTC. Thyroglobulin (Tg) is usually used as a sensitive and convenient biochemical tumor marker after initial therapy. Nevertheless, its accuracy may be interfered by anti-thyroglobulin antibody (TgAb) in up to 25% of DTC patients (3). Moreover, a dilemma emerged in such patients as to how to biochemically monitor the disease status during follow-up. In the 2015 American Thyroid Association (ATA) guidelines, measuring both serum TgAb and Tg levels simultaneously is recommended for DTCs during follow-up (4). Theoretically, after the removal of all thyroid tissue by total thyroidectomy and ¹³¹I therapy in patients with DTC, Tg should be cleared, leading to a gradual decrease of TgAb until it disappears (5). Therefore, a falling TgAb trend over time often indicates the remission or improvement of the disease (6). However, as TgAb is a marker reflecting immuno-reaction, depending on the different immuno-backgrounds of patients, it remains quite illusive for us to define and expect the definite change as Tg. Several studies have demonstrated that in DTC patients with TgAb reduced slowly, long-term stable or increasing TgAb levels appear to have a varying degree of increased risk of persistent or recurrent disease (PRD) compared to those exhibiting a considerable reduction (7–13). However, so far, the dynamic changes of TgAb at postoperative follow-up and TgAb quantitative analysis in relation to the clinical outcomes of TgAb-positive DTC patients after RAI therapy remain not yet well addressed. Our objective was to evaluate the clinical value of change in TgAb as a predictor of disease status during the follow-up of patients after thyroid ablation.

The management of DTCs with positive TgAb has been a challenge due to its measuring interference of Tg in up to one-fourth of the patients. The focus of this study was to explore the clinical value of TgAb change during follow-up to predict different disease status in those TgAb-positive DTC patients and estimate whether it might be used as an indicator for physicians to timely adjust the follow-up management. Different from the median age at diagnosis of 51 years among thyroid cancer patients revealed by Surveillance, Epidemiology, and End Results (SEER) statistics and the male-to-female incidence ratio of approximately 1:3, in our cohort with TgAb-positive, the patients were much younger, with a median diagnosis age of 34 years old, and have a much lower male-to-female incidence ratio as 1:4.66 (41 vs. 191) (17, 18). These discrepancies may suggest that female patients are more likely to suffer autoimmune thyroiditis (AT) and a tendency of early onset of thyroid cancer derived from this immuno-inflammatory background, which are quite similar as the findings shown by others (19, 20). A recent prospective cohort study that included 9,851 patients with thyroid nodules also noted an increased risk of malignancy associated with Hashimoto’s thyroiditis (HT), which suggested that thyroiditis might lead to tumor formation in a similar way that chronic inflammation of many tissues leads to cancer development (21). We thus assumed that young people and women may be more susceptible to underlying thyroid disorders (e.g., Hashimoto’s thyroiditis), which might accelerate the development and lead to their earlier onset of DTC by the change of thyroid function, particularly elevated TSH and inflammatory stimulation that adversely affects normal thyroid tissue. All these may remind us that thyroiditis and TgAb-positive might be a potential pre-cancerous sign which should be taken more seriously, especially in young and female patients. Studies have been performed to evaluate the tumor characteristics and prognosis of DTC patients with Hashimoto’s thyroiditis. A multicenter study including 301 intrathyroidal PTC assessed the relationship between HT and disease outcome. The results showed that HT was detected in 42.5% of patients and was associated to female gender, smaller tumor size, lower incidence of aggressive PTC variants, and less frequent radio-iodine administration. Besides these, HT was associated with a significantly higher clinical remission rate, and recurrence-free survival (RFS) was significantly longer in PTCs with HT compared to non-HT tumors (22). Another multicenter study prospectively collected 4,233 DTC patients, of whom 36.7% had autoimmune thyroiditis. By comparing the clinical characteristics and outcomes of patients with or without AT at 1-year follow-up, they observed that AT patients were significantly younger and had a smaller and bilateral tumor. In addition, AT patients had more biochemical persistence disease, but there was no significant association between the presence of AT and the structural persistence disease, the potential explanation being the ability of the residual thyroid tissue of AT patients to continuously secrete TgAb, thus leading to an incomplete biochemical response (23). These results indicated that the presence of HT might only cause serological abnormalities in TgAb during clinical follow-up, but not related to the persistence or occurrence of structural disease, and might even be associated with better clinico-pathological features and prognosis. In our study, a higher percentage of 54.31% in TgAb-positive DTC patients with HT was present, illustrating that some of TgAb are probably induced by HT. We used structural persistent/recurrent disease as primary endpoint, and the result of the univariate logistic regression analysis showed no significant association between Hashimoto’s thyroiditis and structural persistent/recurrent DTC in the follow-up (P = 0.877). Our study mainly predicted structural disease by serum TgAb changes, and since both suspected structural lesions and unsatisfactory biochemical levels provide evidence of disease persistence/recurrence, it is noteworthy that in the clinical follow-up of DTC we should not only monitor the occurrence of structural disease by imaging techniques but also focus on the serological changes at the same time for a more comprehensive assessment of persistent/recurrent disease.

In cases whose thyroid tissue and Tg antigen are removed by surgery and RAI ablation without persistent or recurrent disease, TgAb concentrations have always been found to decrease over months to years and eventually disappear (24). Therefore, the normal upper limit of 115 IU/mL obviously cannot be regarded as the negative cutoff of TgAb in those with thyroid removal by surgery and RAI ablation. Thus, in this study, a cutoff value of 40 IU/mL was taken, as it has been applied in several studies to qualitatively diagnose TgAb status (positive or negative) (14, 15). Numerous studies have demonstrated that the trend of TgAb after initial treatment might serve as a surrogate predictor of a disease-free status of no residual disease (especially when TgAb is remarkably decreasing) or the presence of structural disease (especially when TgAb is de novo or gradually increasing) (6–11, 13, 25, 26). Meanwhile, there is still controversy concerning the association between TgAb change and DTC disease status. Studies have shown that the trends of TgAb levels could not predict the disease status of DTCs, manifested as either a similar proportion of recurrent cases in the negative trend (slope < 0) group to that in the positive (slope > 0)/no trend (slope close to 0) group based on linear regressions of TgAb values from the surgery date until the end of follow-up or no statistically significant association between de novo TgAb development and structural recurrence (27, 28). So far, few studies have quantitatively analyzed the dynamic changes of TgAb in adult DTCs and mostly limited to analyze its association with the efficacy of RAI ablation. Thus, in this study, TgAb changes were used and quantitatively assessed to associate with disease status in terms of recurrence/persistence or not, while the well-accepted response to therapy assessment was not used since TgAb-positive DTCs are poorly defined by such. After adjusting for other possible clinical predictors (age at diagnosis, gender, histology, maximal tumor size, bilaterality, multifocality, Hashimoto’s thyroiditis, capsular invasion, N stage, and cumulative RAI dose), we obtained two independent variables including ΔTgAb and LNM rate, which could sensitively reflect DTC disease status during follow-up. ΔTgAb was identified to be negatively associated with persistent/recurrent DTC, while LNM rate was positively associated with such. Our finding suggests that ΔTgAb, an index derived from dynamic serial TgAb monitoring, could be an objective marker to sensitively reflect the disease status over time. As for LNM rate, although it reflects more about the thorough LN assessment at surgery, it also correlates with the degree of lymph node invasion. A high LNM rate denotes that the patient has a more aggressive tumor with a large number of lymph node involvement and may be relatively susceptible to the presence of PRD during follow-up, implying that more active management such as RAI therapy should be considered in such patients. From the ROC curve analysis, a high predictive value for DTC disease status was revealed among all of the independent risk factors including ΔTgAb, LNM rate, and their combination. Though the AUC for the combination was the highest and significantly greater than that for LNM rate, there was no significant difference between the combination and ΔTgAb alone. Therefore, ΔTgAb alone could be applied as a convenient and user-friendly marker during follow-up, holding a high diagnostic value with sensitivity and specificity of 89.13% and 88.27%, respectively. It is impressive that the NPV of ΔTgAb in our study was up to 96.93%, suggesting that ΔTgAb could be used as an excellent negative indicator for disease-free status to avoid over-management in those with ΔTgAb >47.0%, which appeared more objectively than the cutoff of 50% raised arbitrarily by experience in several studies (7, 9, 10, 29, 30).

According to a retrospective study of 824 DTC patients by Kim et al. (7), serum TgAb levels measured at 6–12 months after remnant ablation could predict recurrence in TgAb-positive patients. We compared different TgAb decrease rates and status in PFS by 1 year as a boundary and found that patients with ΔTgAb >47.0% in the first year had a longer PFS than those with ΔTgAb ≤47.0%. Another result that we obtained from the survival curve was that patients who turned TgAb-negative in the first year had a longer PFS than those who did not turn negative. These results implied that the decrease rate and status of TgAb in the first year after RAI treatment could serve as early indicators for predicting disease-free survival in TgAb-positive patients. Of the 135 patients who dropped to below 40 IU/mL at final follow-up, the median time to achieve negative conversion was 8 months, which was similar to our prior study (31). For those with positive TgAb and failed to turn negative, one potential explanation might be the relatively higher level of TgAb prior to ¹³¹I therapy, which may take a longer time to achieve negative conversion; thus, a longer follow-up for such patients in this study is still ongoing. For patients with rapidly rising or even consistently positive but stable TgAb, one recently published consensus warns of the risk that such patients may carry a high suspicion of PRD (32). A meta-analysis including 34 studies also reported that patients with persistent/increasing TgAb levels had a higher risk of PRD (12), which was also the case in our study. Nearly 90% (89.29%, 25/28) of the 28 patients with increasing TgAb levels were proved to be PRD. Thus, in patients with rising TgAb, more active surveillance, including both serological and imaging diagnostic procedures, is required. It is worth noting that from the Kaplan–Meier curves, we can see as long as 50 months of PFS even in patients with a slow decrease of TgAb, reminding that a more patient and longer follow-up is needed for such patients. It is noteworthy that, in the present study, the value of ΔTgAb in identifying persistent/recurrent disease of DTC seems less promising, with a PPV less than 70% (66.12%). This may be due to the indolent nature of DTC and the relatively short follow-up time of our study, which might have limited its power to identify PRD. Additionally, 21 of 46 (45.65%) PRD patients had decreasing TgAb levels, and five (10.87%) turned out TgAb-negative before the latest visit. One potential explanation is the time delay of TgAb change, with PRD occurring within a relatively short follow-up time after treatment while TgAb is still in a period of decrease. Alternatively, some PRD patients may present an iodine refractory status of tumor dedifferentiation, which reduces the capacity of the lesion to secrete Tg, resulting in TgAb maintained at a low level (33). Under this circumstance, additional imaging such as US, CT, or ¹³¹I WBS and PET/CT may be more complementary in identifying those with persistence or recurrence. Besides these, in this study, patients with TgAb levels consistently above the upper limit of detection were noticed not only in the PRD group (3/7) but also in the NED group (4/7), which has also been reported by Chiovato (6). We propose that it may be due to several scenarios as follows: there may still be undetected tumor foci or related to immuno-thyroiditis background or the existence of long-lived memory cells that continue to produce TgAb in the patient’s body (3). In such cases, a longer follow-up and a high-sensitivity examination for the detection of potential foci are necessary.

Several limitations in our study should be mentioned. First, the number of PRD cases was limited, which may result in a low PPV, and a larger sample size is required to confirm our findings in future studies. Second, due to ethical concerns, some of the PRD cases, particularly distant metastases, were diagnosed merely based on typical imaging features and anatomical changes in their follow-up, without having been pathologically confirmed. Additionally, the follow-up period of our study was relatively short, a more long-term follow-up is required to verify the predictive value of ΔTgAb over time.

In summary, our study suggested that ΔTgAb could serve as a valuable indicator of disease status in DTC patients with positive postoperative TgAb. ΔTgAb of >47.0% is conducive to identify those with NED to obviate their overtreatment. The decrease rate and negative conversion of TgAb in the first year were good predictors of disease-free survival in patients.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

The studies involving humans were approved by Ethical Committee of Peking Union Medical College (PUMC) Hospital. The studies were conducted in accordance with the local legislation and institutional requirements. The human samples used in this study were acquired from primarily isolated as part of your previous study for which ethical approval was obtained. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.

YZ: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. ZM: Investigation, Data curation, Writing – review & editing, Visualization, Validation, Supervision, Methodology, Formal analysis, Conceptualization. DL: Investigation, Writing – review & editing, Visualization, Validation, Methodology, Formal analysis, Data curation. TZ: Writing – review & editing, Conceptualization, Visualization, Validation, Methodology, Formal analysis, Data curation. XZ: Writing – review & editing, Supervision, Validation, Investigation, Data curation, Conceptualization. DS: Writing – review & editing, Validation, Supervision, Formal analysis, Methodology, Investigation. YS: Writing – review & editing, Supervision, Conceptualization, Methodology, Data curation. JL: Writing – review & editing, Visualization, Methodology, Validation, Supervision. YL: Methodology, Writing – review & editing, Validation, Supervision, Conceptualization.