This work is licensed under a Abstract
Objective
Patients with differentiated thyroid carcinoma (DTC) classified as low risk of recurrence and an excellent response to initial treatment should have a particularly low risk of recurrence. Our objective was to assess the recurrence rate in these patients to support follow-up de-escalation.
Methods
A multicenter retrospective study was conducted in six Spanish hospitals, including consecutive patients diagnosed with low-risk DTC between 2015 and 2018 with an excellent response to initial therapy.
Results
A total of 379 patients were included, with a median follow-up of 6.7 years. Response to therapy at the last follow-up, according to the 2015 ATA/Momesso criteria, was excellent in 349 patients (92.1%), indeterminate in 29 (7.6%), and biochemical incomplete in 1 (0.3%). The overall incidence of indeterminate response was 11.8 events per 1,000 person-years, peaking at year 3 of follow-up and dropping to nearly zero after year 7. PTC and RAI treatment were associated with a lower probability of developing an indeterminate response (P = 0.02 and P = 0.03, respectively). When applying the 2025 ATA criteria, the excellent response rate increased to 95.8% and neither papillary histology nor RAI treatment was significantly associated (P = 0.07 and P = 0.61, respectively). During follow-up, indeterminate/suspicious findings on ultrasound were observed in 39 patients (10.3%). FNA was performed in 20 of these, none of which found recurrence.
Conclusions
Recurrences in patients with low risk of recurrence and an excellent response to initial therapy are extremely infrequent. These data support follow-up de-escalation in this population.
Significance statement
In this multicenter retrospective study of low-risk DTC patients with an excellent response to initial treatment, no recurrences were observed among 379 patients after a median follow-up of 6.7 years. The overall incidence rate of indeterminate response was 11.8 events per 1,000 person-years, peaking around year 3 after surgery and progressively declining thereafter, approaching zero after year 7. Using the 2025 ATA response-to-therapy criteria, 95.8% of patients maintained an excellent response at the last follow-up, and no variable was associated with the development of an indeterminate response. These findings support follow-up de-escalation.
Keywords: papillary thyroid cancer, follicular thyroid cancer, oncocytic thyroid cancer, differentiated thyroid cancer, iodine 131
Introduction
Thyroid cancer is the most common endocrine malignancy and the seventh most frequent solid tumor among women (1). Although differentiated thyroid cancer (DTC) has an excellent prognosis (over 98% five-year disease-specific survival), recurrences are more frequent (2). The American Thyroid Association (ATA) classification is used to assess the recurrence risk. This classification proposed in 2006 has been updated in each iteration of the guidelines (3). It is based on postoperative clinicopathological features; thus, it is useful at diagnosis but cannot be adapted to the patient’s clinical course (4, 5, 6, 7, 8, 9, 10).
To overcome this limitation, the 2015 ATA guidelines introduced a dynamic risk stratification system that reevaluates response to therapy (RTT) at each visit during follow-up. Imaging studies and serum thyroglobulin (Tg) and antithyroglobulin antibody (TgAb) levels are used to classify patients into four RTT categories: excellent, biochemical incomplete, structurally incomplete, and indeterminate (11).
Patients with low-risk DTC who achieve an excellent response after initial therapy might have a very low probability (<2%) of developing structural disease during long-term follow-up (4, 10, 12). Most recently, the 2025 ATA guidelines have introduced the concept of complete remission, which in these patients would allow discontinuation of imaging after 5–8 years and biochemical monitoring after 10–15 years (12). Nevertheless, the supporting evidence for this recommendation remains of low quality.
The objective of our study was to assess whether patients with low-risk DTC and an excellent response to initial therapy maintain this favorable status in the long term, thus supporting follow-up de-escalation or even early discharge. In addition, we aimed to analyze frequency and yield of follow-up investigations performed in these patients.
Materials and methods
Study design and population
This retrospective cohort study was conducted at six Spanish academic centers after approval by the Hospital Universitario Ramón y Cajal’s Ethics Committee (#292/23), in accordance with the Declaration of Helsinki, with a waiver for written informed consent. Consecutive patients who underwent surgery for DTC between January 1, 2015, and December 31, 2018, at the participating institutions were screened. Patients were eligible for inclusion if they had DTC classified as low risk of recurrence according to the 2015 ATA risk stratification system and demonstrated an excellent initial response to therapy. Tg had to be measured with an ultrasensitive assay (functional sensitivity <0.1 ng/mL) from the time of diagnosis, and a minimum follow-up of 5 years was required for inclusion. Patients who did not meet these criteria were excluded.
The ATA classification was updated in the 2025 ATA guidelines. In particular, the intermediate-risk group was subdivided into two categories: ‘low-intermediate’ and ‘intermediate-high’. Several features, such as multifocality, T3a, or microscopic positive posterior tumor margins (R1), have been changed from low-risk category to one of these two intermediate-risk categories (12). Patients were reclassified post hoc according to these new criteria. Patients with R1 resection were not reclassified into low-intermediate risk if the affected margin was anterior or not specified. Patients with lymph node metastases were only included in the study if the pathology report specified the number (<5) of metastatic nodes and the size of metastatic foci (all ≤ 2 mm). All other N1 cases were excluded. Stage was classified according to the 8th edition of the American Joint Committee on Cancer (AJCC) (13).
We considered initial treatment as the surgical resection, and RAI was administered within the first 6 months after surgery. Initial RTT is assessed during the first 6 months after completing the initial treatment. Initial RTT was defined according to the 2015 ATA guidelines for patients treated with TT and RAI (negative imaging, undetectable TgAb and suppressed Tg < 0.2 ng/mL or stimulated Tg <1 ng/mL, and negative TgAb), whereas the criteria defined by Momesso et al. were used for patients managed exclusively with surgery (negative imaging and undetectable TgAb and suppressed Tg < 0.2 ng/mL after TT or <30 ng/mL after hemithyroidectomy) (8, 11). Final RTT was first assessed using these same criteria and then reevaluated according to the updated 2025 ATA guidelines (12). Changes from the 2015 ATA/Momesso to 2025 ATA criteria are summarized in Table 1.
Table 1.
Changes in response to therapy criteria from 2015 ATA/Momesso to 2025 ATA guidelines. Response to therapy criteria defined in the 2015 ATA guidelines (TT + RAI) and by Momesso et al. (TT and HT). Criteria without arrows are unchanged. Changes introduced in the 2025 ATA guidelines are highlighted in bold.
| TT + RAI | TT | HT* | |
|---|---|---|---|
| ERT | |||
| nsTg | <0.2 | <0.2 → <2.5 | <30.0 → NC |
| sTg | <1.0 | <2.0 → NC | NC |
| TgAb | Negative | Negative → NC | Negative → NC |
| Imaging | Negative | Negative | Negative |
| IRT | |||
| nsTg | 0.2–1.0 | 0.2–5.0 → 2.5–5.0 | NC |
| sTg | 1.0–10.0 | 2.0–10.0 → NC | NC |
| TgAb | Positive stable/declining | Positive stable/declining | Positive stable/declining → NC |
| Imaging | Nonspecific findings | Nonspecific findings | Nonspecific findings → NC |
| BIR | |||
| nsTg | >1.0 | >5.0 | >30.0 → NC |
| sTg | >10.0 | >10.0 → NC | NC |
| TgAb | Positive increasing | Positive increasing | Positive increasing → NC |
| Imaging | Negative | Negative | Negative → NC |
| SIR | |||
| Imaging | Positive | Positive | Positive |
BIR, biochemical incomplete response to therapy; ERT, excellent response to therapy; IRT, indeterminate response to therapy; HT, hemithyroidectomy; NC, variable ‘not contributing’ to the definition of the response to therapy; nsTg, nonstimulated serum thyroglobulin; RAI, radioactive iodine; SIR, structurally incomplete response to therapy; sTg, stimulated serum thyroglobulin; TgAb, antithyroglobulin antibody; TT, total thyroidectomy.
According to the 2025 ATA guidelines, IRT and BIR are not applicable to patients treated with hemithyroidectomy.
Disease recurrence was defined as cytological or histological confirmation following an excellent response to initial therapy, regardless of initial treatment extent (14).
Statistical analysis
Descriptive statistics are summarized as counts and percentages for categorical variables and as mean (standard deviation, SD) or median (interquartile range, IQR) for continuous variables, as appropriate based on data distribution. Baseline comparisons between groups were performed using the Pearson chi-square test or Fisher’s exact test for categorical variables and Student’s t-test or the Mann–Whitney U test for continuous variables, according to distributional assumptions.
Time-to-event analyses were conducted to evaluate the development of an indeterminate or biochemical incomplete response. Time to event was defined as the interval between initial surgery and the first occurrence of the outcome, or the last follow-up for censored observations. Cumulative incidence was estimated using the Kaplan–Meier method and reported as 1 minus the survival function. Incidence rates were calculated assuming a Poisson distribution and expressed as events per 1,000 person-years. Instantaneous hazard rates were estimated using an adaptive kernel-smoothed hazard estimator and are presented graphically.
Univariable Cox proportional hazards models were used to assess the association between clinicopathological variables and the risk of developing an indeterminate/biochemical incomplete RTT. Covariates evaluated included age (modeled per 10-year increase), sex, tumor size, MACIS score, tumor stage, incidental diagnosis, multifocality, thyroid-stimulating hormone (TSH) suppression status, resection status, histological subtype, BRAF V600E mutation status (restricted to patients with available molecular data), initial RAI treatment, and the 2025 ATA risk classification. For categorical predictors in which no events occurred in one or more strata, hazard ratios were not estimated due to complete separation and are reported descriptively. Separate analyses were performed using response-to-therapy classifications according to the 2015 ATA/Momesso criteria and the 2025 ATA guidelines. Given the reduced number of events, bivariate analyses were limited to variables demonstrating significant associations in univariable analyses.
Missing data were not imputed, as missingness was limited to a small number of clinicopathological variables and no missing values were present for the outcomes of interest. All statistical tests were two-sided, and a P value <0.05 was considered statistically significant. Statistical analyses were performed using R software (version 4.5.1; R Core Team, 2025) (15). The Sankey diagrams were created using the online tool SankeyMATIC (16).
Results
Demographic and baseline characteristics
A total of 942 patients were identified from the cohort of patients who underwent thyroid surgery for DTC, of whom 379 met inclusion criteria and were analyzed (Fig. 1). Baseline patient and tumor characteristics are summarized in Table 2. The mean age was 54 years (SD = 13), and most patients were female (82.3%). Papillary thyroid cancer (PTC) accounted for most cases (89.7%), followed by follicular thyroid cancer (7.1%) and oncocytic thyroid cancer (3.2%). Incidental cancer diagnosis was reported in 31.7% of cases. The median tumor size was 0.9 cm (IQR: 0.5–1.6). Most tumors were unifocal (72.0%). Multifocality (n = 106; 28%) was unilateral in 37 patients, bilateral with all foci ≤1 cm in 39 patients, and bilateral with at least one focus >1 cm in 30 patients. Initial treatment was hemithyroidectomy in 36 (9.5%) patients, subtotal thyroidectomy in 1 (0.3%), isthmectomy in 1 (0.3%), TT in 153 (40.3%), and TT plus RAI in 188 (49.6%). The median administered RAI activity was 100 mCi (range: 27–155). Complete resection (R0) was achieved in 96.0% of patients. BRAF V600E mutation status was available for 158 patients (42%), of whom 85 (54%) were positive. Most patients were stage I (95.5%), and 4.5% were stage II. Tumor stage was T1a in 214 patients (56.5%), T1b in 105 (27.7%), T2 in 47 (12.4%), and T3a in 13 (3.4%). Nodal stage was N0/Nx in 358 patients (94.5%), N1a in 18 (4.7%), and N1b in 3 (0.8%). According to the 2025 ATA classification, 121 patients would be reclassified into the following risk categories: 91 (24%) into low-intermediate risk and 30 (8%) into intermediate-high risk (Supplementary Fig. 1 (see section on Supplementary materials given at the end of the article)). The median follow-up time was 6.7 years (IQR: 5.9–7.9).
Figure 1.
Flow chart of patient selection.
Table 2.
Comparison of baseline characteristics by response to therapy (excellent vs indeterminate/BIRg). Data are presented as mean ± SD, n (%), or median (IQR).
| Characteristics | Total patients | 2015 ATA GL + Momesso et al. | 2025 ATA GL | ||||
|---|---|---|---|---|---|---|---|
| Excellent | Indeterminate/BIR | P value1 | Excellent | Indeterminate/BIR | P value1 | ||
| n | 379 | 349 | 30 | 363 | 16 | ||
| Age, years | 54 ± 13 | 54.5 ± 13.5 | 52.0 ± 10.8 | 0.2 | 54.3 ± 13.5 | 54 ± 10.1 | 0.9 |
| Sex | 0.1 | 0.09 | |||||
| Female | 312 (82.3) | 284 (81.6) | 28 (90.3) | 296 (81.5) | 16 (100) | ||
| Male | 67 (17.7) | 65 (18.6) | 2 (6.7) | 67 (18.5) | 0 (0) | ||
| Histology | 0.005 | 0.036 | |||||
| PTC | 340 (89.7) | 317 (90.8) | 23 (76.7) | 328 (90.4) | 12 (75.0) | ||
| FTCa | 27 (7.1) | 20 (5.7) | 7 (23.3) | 23 (6.3) | 4 (25.0) | ||
| OTCb | 12 (3.2) | 12 (3.4) | 0 (0.0) | 12 (3.3) | 0 (0.0) | ||
| Tumor size | 0.9 (0.5–1.6) | 0.9 (0.5, 1.6) | 0.8 (0.5, 1.2) | 0.3 | 0.9 (0.5, 1.6) | 1.1 (0.5, 1.4) | 0.9 |
| MACIS | 4.6 (3.9–5.5) | 4.7 (3.9, 5.5) | 4.3 (3.8, 5.1) | 0.2 | 4.6 (3.9, 5.5) | 4.5 (3.9, 5.2) | 0.8 |
| T stage (AJCC 8th) | 0.6 | 0.8 | |||||
| T1a | 214 (56.5) | 195 (55.9) | 19 (63.3) | 206 (56.8) | 8 (50.0) | ||
| T1b | 105 (27.7) | 99 (28.4) | 6 (20.0) | 100 (27.5) | 5 (31.3) | ||
| T2 | 47 (12.4) | 42 (12.0) | 5 (16.7) | 44 (12.1) | 3 (18.7) | ||
| T3a | 13 (3.4) | 13 (3.7) | 0 (0) | 13 (3.6) | 0 (0) | ||
| N stage (AJCC 8th) | 1 | 0.6 | |||||
| N0a/b/x | 358 (94.5) | 329 (94.3) | 29 (96.7) | 343 (94.5) | 15 (93.8) | ||
| N1a | 18 (4.7) | 17 (4.9) | 1 (3.3) | 17 (4.7) | 1 (6.2) | ||
| N1b | 3 (0.8) | 3 (0.9) | 0 (0) | 3 (0.8) | 0 (0) | ||
| DTC stage (AJCC 8th) | 0.4 | 1 | |||||
| I | 362 (95.5) | 332 (95.1) | 30 (100.0) | 346 (95.3) | 16 (100.0) | ||
| II | 17 (4.5) | 17 (4.9) | 0 (0) | 17 (4.7) | 0 (0) | ||
| Incidentally diagnosed | 0.8 | 0.6 | |||||
| No | 259 (68.3) | 239 (68.5) | 20 (66.7) | 249 (68.6) | 10 (62.5) | ||
| Yes | 120 (31.7) | 110 (31.5) | 10 (33.3) | 114 (31.4) | 6 (37.5) | ||
| Multifocality | 0.4 | 0.9 | |||||
| Absent | 273 (72) | 248 (71.1) | 25 (83.3) | 260 (71.6) | 13 (81.3) | ||
| Unilateral | 37 (9.8) | 35 (10.0) | 2 (6.7) | 36 (9.9) | 1 (6.2) | ||
| Bilateral | 69 (18.2) | 66 (18.9) | 3 (9.7) | 67 (18.5) | 2 (12.5) | ||
| Treatment with RAI | 0.03 | 0.6 | |||||
| Noe | 191 (50.4) | 170 (48.7) | 21 (70.0) | 184 (50.7) | 7 (43.8) | ||
| Yes | 188 (49.6) | 179 (51.3) | 9 (30.0) | 179 (49.3) | 9 (56.3) | ||
| Resection status | 0.6 | 1 | |||||
| R0 | 364 (96.0) | 334 (96.0) | 30 (100.0) | 348 (96.1) | 16 (100.0) | ||
| R1c | 14 (3.7) | 14 (4.0) | 0 (0 | 14 (3.9) | 0 (0) | ||
| Unknown | 1 (0.3) | ||||||
| Tumor BRAF V600E mutation | 0.6 | 0.7 | |||||
| Negative | 73 (19.3) | 65 (45.5) | 8 (53.3) | 68 (45.6) | 5 (55.6) | ||
| Positive | 85 (22.4) | 78 (54.5) | 7 (46.7) | 81 (54.4) | 4 (44.4) | ||
| Not evaluated | 221 (58.3) | ||||||
| TSH suppressed | 0.8 | 0.2 | |||||
| Yesd | 94 (24.8) | 86 (24.6) | 8 (26.7) | 88 (24.2) | 6 (37.5) | ||
| No | 285 (75.2) | 263 (75.4) | 22 (73.3) | 275 (75.8) | 10 (62.5) | ||
| Risk reclassification ATA 2025 | 0.1 | 0.1 | |||||
| Low | 258 (68.1) | 233 (66.8) | 25 (83.3) | 245 (67.5) | 13 (81.3) | ||
| Low-intermediate | 91 (24.0) | 88 (25.2) | 3 (10.0) | 90 (24.8) | 1 (6.3) | ||
| Intermediate-high | 30 (7.9) | 28 (8.0) | 2 (6.7) | 28 (7.7) | 2 (10.5) | ||
19 minimally invasive; 8 encapsulated angioinvasive <4 vessels.
10 minimally invasive; 2 encapsulated angioinvasive <4 vessels.
5 anterior margin; 1 posterior margin 1; 8 NA.
Median TSH suppression time, years (range, IQR): 2 (1–7, 1–3).
Surgical treatment consisted of total thyroidectomy in 153 patients, hemithyroidectomy in 36, subtotal thyroidectomy in 1, and isthmectomy in 1 patient.
Fisher’s exact test; Mann–Whitney U test; Pearson’s chi-squared test; Student’s t-test.
BIR, biochemical incomplete response; RTT, response to therapy; ATA GL, American Thyroid Association Guidelines; SD, standard deviation; PTC, papillary thyroid carcinoma; FTC, follicular thyroid carcinoma; OTC, oncocytic thyroid carcinoma; IQR, interquartile range; RAI, radioactive iodine.
As defined by the 2015 ATA GL and Momesso criteria or by 2025 ATA GL
Recurrence and final treatment response to therapy
No structural recurrences were observed during follow-up in our cohort (Table 3). At the last follow-up, 349 patients (92.1%) had an excellent response, 29 (7.6%) had an indeterminate response, and 1 (0.3%) had a biochemical incomplete response according to the 2015 ATA/Momesso criteria. Patients with an indeterminate/biochemical incomplete RTT (23 PTCs and 7 FTCs) had elevated Tg (n = 15), with a median Tg of 0.34 ng/mL (range: 0.2–1.74), positive TgAb (n = 11), or un-biopsied nonspecific imaging findings (n = 7), with some patients presenting several of these factors. According to the 2025 ATA criteria, 25 patients were classified as low risk (n = 25), 3 as low-intermediate risk, and 2 as intermediate-high risk. According to the 2025 ATA RTT criteria, 363 patients (95.8%) maintained an excellent RTT, 15 (3.9%) had an indeterminate RTT, and 1 (0.3%) had a biochemical incomplete RTT at the last follow-up (Table 3). Patients with an indeterminate/biochemical incomplete RTT (12 PTCs and 4 FTCs) had elevated Tg (n = 4), positive TgAb (n = 8), or un-biopsied nonspecific imaging findings (n = 7), with some patients presenting two of these factors. According to the 2025 ATA criteria, 13 were classified as low risk (n = 13), 1 as low-intermediate risk, and 2 as intermediate-high risk.
Table 3.
Response to therapy at last follow-up. Values are presented as n (%) or median (IQR).
| 2015 ATA GL and Momesso criteria | 2025 ATA GL | |
|---|---|---|
| Response to therapy at last follow-up | ||
| Excellent | 349 (92.1) | 363 (95.8) |
| Indeterminate | 29 (7.6) | 15 (3.9) |
| Biochemical incomplete | 1 (0.3) | 1 (0.3) |
| Structural incomplete | 0 (0) | 0 (0) |
| Reason for indeterminate/BIR, n | ||
| Basal Tg | 15 | 4 |
| Anti-Tg antibody trend | 11 | 8 |
| Indeterminate not-biopsied US finding | 7 | 7 |
| Median time to development of IRT or BIR (years) | 3.17 (2.3-4.3) | 3.39 (3.0-4.0) |
ATA GL, American Thyroid Association Guidelines; BIR, biochemical incomplete response; IQR, interquartile range; IRT, indeterminate response to therapy; Tg, thyroglobulin; US, ultrasound.
Time-to-event analysis and predictors of indeterminate/biochemical incomplete RTT during follow-up
The cumulative incidence of developing an indeterminate/biochemical incomplete RTT according to the 2015 ATA/Momesso criteria was 8.2% (95% CI: 5.3–10.9) after a median follow-up of 6.7 years. The overall incidence rate was 11.8 events per 1,000 person-years (95% CI: 8.3–17.2), with the smoothed instantaneous hazard peaking at 18.3 events per 1,000 person-years around the third year of follow-up and progressively declining thereafter, approaching zero beyond seven years (Fig. 2). Using the 2025 ATA RTT criteria, the cumulative incidence of developing an indeterminate/biochemical incomplete response was 4.3% (95% CI: 2.2–6.4), with an incidence rate of 6.1 events per 1,000 person-years (95% CI: 3.5–10.0; Fig. 3).
Figure 2.
Cumulative incidence (A) and instantaneous hazard rate (B) of indeterminate or biochemical incomplete response during follow-up according to the ATA 2015 guidelines and the Momesso criteria. The rug marks indicate event times.
Figure 3.
Cumulative incidence (A) and instantaneous hazard rate (B) of indeterminate or biochemical incomplete response during follow-up according to the ATA 2025 guidelines the Momesso criteria. The rug marks indicate event times.
Only two factors influenced the development of an indeterminate/biochemical incomplete RTT under the 2015 ATA/Momesso criteria. Rates were higher in follicular carcinomas (P = 0.005) and were lower following RAI treatment (P = 0.03). No other basal characteristics were found to have an impact on the risk of developing an indeterminate/biochemical incomplete RTT (Table 2).
In univariable Cox proportional hazards analyses based on the 2015 ATA/Momesso criteria, papillary histology was associated with a significantly lower hazard of developing an indeterminate/biochemical incomplete RTT (HR: 0.38, 95% CI: 0.16–0.90; P = 0.02), as was initial RAI treatment (HR: 0.41, 95% CI: 0.19–0.90; P = 0.03). In a bivariate Cox model including both variables, papillary histology and initial RAI treatment remained independently associated with a lower risk of developing an indeterminate/biochemical incomplete RTT (Table 4). However, when the analysis was repeated using the 2025 ATA RTT criteria, neither RAI treatment (HR: 1.29, 95% CI: 0.48–3.47; P = 0.61) nor papillary histology (HR: 0.35, 95% CI: 0.11–1.08; P = 0.07) was significantly associated with the development of an indeterminate/biochemical incomplete response.
Table 4.
Factors associated with loss of excellent response to therapy: univariable and bivariate analysis.
| Variable | 2015 ATA GL + Momesso | 2025 ATA GL | ||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Univariable analysis | ||||
| Age (per 10 years) | 0.88 (0.67–1.15) | 0.34 | 0.98 (0.68–1.42) | 0.93 |
| Male sex (vs female) | 0.33 (0.08–1.37) | 0.12 | Not estimable‡ | - |
| Tumor size (per cm) | 0.83 (0.56–1.22) | 0.35 | 0.92 (0.57–1.49) | 0.75 |
| MACIS (per point) | 0.75 (0.52–1.09) | 0.14 | 0.90 (0.56–1.46) | 0.67 |
| Incidental diagnosis (yes vs no) | 1.1 (0.51–2.35) | 0.81 | 1.32 (0.48–1.46) | 0.6 |
| Multifocality (yes vs no) | 0.5 (0.19–1.31) | 0.16 | 0.59 (0.48–3.62) | 0.41 |
| No TSH suppression (vs suppressed) | 0.90 (0.40–2.03) | 0.80 | 0.55 (0.20–1.50) | 0.24 |
| BRAF V600E positive (vs negative)* | 0.75 (0.27–2.07) | 0.58 | 0.68 (0.18–2.53) | 0.57 |
| Papillary histology (vs non-papillary) | 0.38 (0.16–0.88) | 0.02 | 0.35 (0.11–1.08) | 0.07 |
| Initial RAI treatment (yes vs no) | 0.41 (0.19–0.90) | 0.03 | 1.29 (0.48–3.47) | 0.61 |
| Risk reclassification ATA 2025 | ||||
| Low-intermediate (vs low) | 0.33 (0.10–1.09) | 0.07 | 0.21 (0.03–1.63) | 0.14 |
| Intermediate-high (vs low) | 0.68 (0.16–2.88) | 0.60 | 1.34 (0.30–5.92) | 0.70 |
| Bivariate analysis | ||||
| Papillary histology (vs non-papillary) | 0.31 (0.13–0.74) | 0.01 | 0.36 (0.11–1.13) | 0.08 |
| Initial RAI treatment (yes vs no) | 0.36 (0.16–0.79) | 0.01 | 1.16 (0.43–3.15) | 0.77 |
BRAF analyses restricted to patients with available mutation status.
HR could not be estimated because no events were observed in males.
ATA GL, American Thyroid Association Guidelines; CI, confidence interval; HR, hazard ratio; RAI, radioactive iodine.
Additional RAI treatment during follow-up
Four patients (1.1%) received RAI therapy after the first year of follow-up. Three of them had not received RAI as part of their initial management. Of the four patients, two received RAI despite maintaining an excellent response: one at 440 days after surgery due to patients’ preference and the other at 1,229 days after surgery due to treating physician’s recommendation based on an indeterminate lymph node on ultrasound with benign cytology. The third patient received RAI at 1,807 days after surgery due to rising Tg levels, which reached 1.74 ng/mL before treatment. According to the 2025 ATA RTT criteria, this patient would still meet criteria for an excellent RTT despite rising thyroglobulin levels. One patient who received RAI as part of the initial treatment received a second dose at 2,079 days due to the development of a biochemical incomplete RTT and died 160 days later from an acute myeloid leukemia.
Biochemical and imaging assessments during follow-up
During follow-up, 3,655 Tg–TgAb measurements (median: 9/patient; 1.3/patient/year of follow-up (IQR: 1.1–1.6)) and 1,364 neck ultrasounds (median: 3/patient; 0.5 patient/year of follow-up (IQR: 0.3–0.7)) were performed. Among the 379 patients, 40 suspicious or indeterminate findings on ultrasound were identified in 39 patients (10.3%) during follow-up. These included nodules in the remnant thyroid lobe (n = 7) or in the thyroid bed (n = 13) and suspicious cervical lymph nodes (n = 20). Half of the findings (n = 20) led to a fine-needle aspiration (FNA) cytology, whereas the rest were monitored. FNA was performed on 14 cervical lymph nodes, 3 nodules in the thyroid bed, and 3 nodules in the remnant thyroid tissue. Cytology results were benign in 18 (90.0%) and nondiagnostic in 2 (10.0%). One of the patients with indeterminate sonographic findings was treated with RAI despite having benign cytology and undetectable Tg. No further treatment was deemed necessary in the rest (Supplementary Fig. 2).
Discussion
In this multicenter retrospective cohort of 379 patients with low-risk DTC, as defined by the 2015 ATA classification, and an excellent response to initial treatment, no recurrences were observed after a median follow-up of 6.7 years. The cumulative incidence rate of an indeterminate/biochemical incomplete RTT was 11.8/1,000 person-years, and the risk of developing this response declined substantially after the third year of follow-up, becoming virtually negligible beyond year seven. Furthermore, 40 indeterminate sonographic findings were found during follow-up, leading to biopsy in 20, none of which found recurrence. These data support follow-up de-escalation as proposed in the 2025 ATA guidelines (12).
Previous large prospective randomized trials have confirmed the estimated risks of recurrence for low-risk DTC (<5%) after 5 years of follow-up regardless of RAI treatment or RAI-administered activities (17, 18, 19, 20). These studies, however, have not stratified recurrence risk by RTT, which seems to influence it (12). Patients achieving an excellent RTT, regardless of the ATA risk category, have an estimated 1–4.7% risk of structural recurrence (4, 5, 8, 21, 22). Thus, in patients with low-risk DTC and an excellent response to initial therapy recurrences are expected to be very infrequent (<2%) (12, 23). Our study supports this hypothesis and adds a longitudinal perspective to the RTT in these patients, showing that the incidence of new indeterminate/biochemical incomplete RTT decreases substantially after the third year of follow-up, becoming negligible after year 7. This finding strongly supports follow-up de-escalation after 3–5 years and follow-up discontinuation after 7–10 years in low-risk DTC with a persistent excellent RTT.
No recurrences were observed after almost 7 years of follow-up. However, 3,655 biochemical assessments (1.3 assessments/patient/year) and 1,364 neck US examinations (1 every 2 years/patient) were done. Indeterminate or suspicious findings on ultrasound were reported in 10.3% of patients, and half of these findings led to unnecessary invasive procedures. All these findings ultimately proved to be false positives, which is consistent with previous reports (24, 25, 26, 27, 28). An increased use of postoperative neck ultrasound has been associated with a higher rate of subsequent treatments without improving overall survival (29). Seejore et al. reported that in all cases of recurrence, elevations in Tg or TgAb levels preceded radiological evidence of disease, underscoring the value of biochemical surveillance and further supporting a selective, risk-adapted approach to imaging (30). Grani et al. proposed that low- and intermediate-risk DTC patients with undetectable Tg, absent TgAb, and negative cervical ultrasound be monitored with clinical evaluations and unstimulated Tg measurements alone, reserving repeat imaging for patients with rising Tg or TgAb levels (31). Our study supports that ultrasound follow-up might be unnecessary if an excellent response is achieved after initial treatment by thyroglobulin measurement, particularly if it remains undetectable or stable.
The 2025 ATA guidelines recommend for the first time to stop ultrasound follow-up after 5–8 years and thyroglobulin measurement after 10–15 years in patients with low-risk DTC and an excellent response to therapy. Consistent with these findings, the temporal pattern of our cohort also suggests that late biochemical events are uncommon, as the development of an indeterminate/biochemical incomplete response progressively declined after the third year of follow-up, with only two patients developing it beyond 60 months. Similarly, Attia et al. reported that biochemical evidence suggestive of recurrence occurred at a median of 5 years after initial treatment, supporting the notion that the likelihood of new events markedly decreases beyond this point, suggesting that follow-up de-escalation could be safely considered after 10 years (32). Our cohort included only patients classified as low risk according to the 2015 ATA guidelines. However, 24% would be categorized into the low-intermediate-risk category and 8% into the intermediate-high-risk (8%) category by the 2025 ATA risk stratification system. We found no differences in outcomes in those patients. Thus, our study suggests that follow-up de-escalation might also be appropriate for patients with low-intermediate-risk DTC and an excellent RTT.
Histology and RAI treatment were found to influence the risk of developing an indeterminate/biochemical incomplete RTT during follow-up according to the 2015 ATA/Momesso criteria. However, when applying the 2025 ATA guidelines criteria, neither RAI treatment nor papillary histology was associated with a lower risk of developing an indeterminate/biochemical incomplete RTT. The 2025 ATA criteria do not take into account Tg levels or Tg antibodies in patients treated with hemithyroidectomy and allow higher postsurgical Tg levels to consider an excellent RTT in patients treated with TT without RAI. Of note, 143 patients (38% of our cohort) had solitary, node-negative papillary thyroid microcarcinomas.
Our multicenter study includes a well-characterized cohort of patients with low-risk DTC and an excellent RTT followed up for a clinically relevant period. However, its retrospective design may have introduced information bias due to incomplete medical records. Although pathology reports contained the variables required to correctly classify patients according to the 2015 ATA recurrence risk classification, they were not standardized across centers. The location of involved margin (R1) was not systematically reported in pathology reports and thus missing in several patients. This information could increase the number of patients classified into the low-intermediate risk of the 2025 ATA classification but would not change the main outcomes of the study. In patients treated with hemithyroidectomy, the Momesso criteria were used to assess RTT. Thus, only patients with postsurgical thyroglobulin values <30 ng/mL (i.e. excellent RTT) were included. The clinical utility of such a serum thyroglobulin threshold has been questioned by several studies, and the 2025 ATA guidelines do not rely on thyroglobulin levels to assess RTT after hemithyroidectomy. Unfortunately, we did not collect data from patients with higher postsurgical thyroglobulin levels, limiting the applicability of our findings to those patients. Of note, serum thyroglobulin persisted <30 ng/mL and ultrasound remained negative in all patients treated with hemithyroidectomy during the follow-up. Only three of these patients developed an indeterminate RTT due to positive TgAb that remained stable during the subsequent follow-up, which is no longer considered an indeterminate RTT according to the 2025 ATA criteria. In addition, half of the patients received RAI, although this treatment is no longer routinely recommended for low-risk DTC. This likely reflects differences in local treatment protocols across participating centers during the study period. However, it is also possible that patients treated with radioactive iodine had a perceived higher recurrence risk based on the presence of certain clinicopathological features (33). Therefore, our findings should be interpreted considering this treatment heterogeneity. All patients were followed up using standardized dynamic risk stratification criteria, allowing for consistent assessment of outcomes across centers. Nonetheless, minor differences in follow-up protocols and ultrasound interpretation among centers could have influenced the frequency and type of complementary tests performed. Anti-Tg antibody measurements were not standardized across centers. Thus, different methods and cutoff values were used. Nonetheless, we collected negative/positive and title trends (increasing/stable/decreasing). Different ultrasound systems were also used in each center. Thyroglobulin was measured using an ultrasensitive assay (functional sensitivity: 0.1 ng/mL) in all centers, ensuring precise biochemical monitoring and accurate classification of RTT. However, the very low number of recurrences and the indeterminate/biochemical incomplete RTT found in our study limit the ability to perform multivariate analyses or to identify independent predictors of late events. In addition, patients with less than five years of follow-up were excluded from the analysis. However, the Spanish National Healthcare System provides universal care to all inhabitants, and patients usually keep follow-up in the center in which they are operated on, facilitating follow-up and minimizing the potential impact of this limitation. In our cohort, less than 20% of patients were excluded due to truncated follow-up before 5 years from diagnosis.
Conclusion
Recurrences in patients with low risk of recurrence and an excellent RTT are extremely infrequent. An indeterminate/biochemical incomplete RTT may develop during the initial years of follow-up, with incidence peaking at year 3 and becoming very unusual beyond year 5. Our findings support follow-up de-escalation in this population, a recommendation that may also be extended to patients classified as low-intermediate risk according to the 2025 ATA classification system. Cervical ultrasound seems unnecessary after unsuspicious initial postsurgical evaluation in this set of patients provided an excellent RTT is maintained according to biochemical follow-up.
Supplementary materials
Declaration of interest
The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
C Garcia-Regal and P Valderrabano were partly funded with a grant from the Scientific Foundation of the Spanish Association Against Cancer (FCAECC) (Award number: CLSEN258522VALD). C Garcia-Regal, J Quiñones-Silva, N Brox-Torrecilla, A Martinez-Lorca, I Ruz Caracuel and P Valderrábano participate as coinvestigators in a project funded by the Community of Madrid (P2022/BMD-7379-iTIRONET-CM). While these authors received no direct financial support or salary from this grant, this publication has been funded through this project.
Author contribution statement
CG-R: conceptualization (supporting); investigation (equal); methodology (supporting); project administration (supporting); visualization (equal); writing – original draft (lead); writing – review & editing (equal). JQ-S: formal analysis (lead); validation (lead); visualization (equal); writing – review & editing (equal). NB-T: conceptualization (supporting); writing – review & editing (equal). AM-L: conceptualization (supporting); writing – review & editing (equal). IRC: conceptualization (supporting); investigation (equal); writing – review & editing (equal). HGPA: conceptualization (supporting); investigation (equal); writing – review & editing (equal). SAM: conceptualization (supporting); investigation (equal); writing – review & editing (equal). AMGC: conceptualization (supporting); writing – review & editing (equal). MRC: conceptualization (supporting); writing – review & editing (equal). PLP: conceptualization (supporting); writing – review & editing (equal). BPG: conceptualization (supporting); writing – review & editing (equal). MIGGM: conceptualization (supporting); writing – review & editing (equal). CFC: investigation (equal); writing – review & editing (equal). MJPC: investigation (equal); writing – review & editing (equal). VAL: investigation (equal); writing – review & editing (equal). MPF: investigation (equal); writing – review & editing (equal). MGGdA: investigation (equal); writing – review & editing (equal). PIB: investigation (equal); writing – review & editing (equal). PV: conceptualization (lead); funding acquisition (lead); methodology (lead); project administration (lead); supervision (lead); writing – original draft (supporting); writing – review & editing (equal).
Acknowledgments
This study was developed by members of the Thyroid Cancer Working Groups of the Endocrinology and Nutrition Spanish Society (SEEN) and the Society of the Endocrinology, Diabetes and Nutrition of the Community of Madrid (SENDIMAD).
References
- 1.Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025. 75 10–45. ( 10.3322/caac.21871) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.National Cancer Institute Cancer of the thyroid – cancer stat facts. SEER 2026. (https://seer.cancer.gov/statfacts/html/thyro.html) [Google Scholar]
- 3.Cooper DS, Doherty GM, Haugen BR, et al. Management guidelines for patients with thyroid nodules and differentiated thyroid cancer: The American thyroid association guidelines taskforce. Thyroid 2006. 16 109–142. ( 10.1089/thy.2006.16.109) [DOI] [PubMed] [Google Scholar]
- 4.Pitoia F & Jerkovich F. Dynamic risk assessment in patients with differentiated thyroid cancer. Endocr Relat Cancer 2019. 26 R553–R566. ( 10.1530/ERC-19-0213) [DOI] [PubMed] [Google Scholar]
- 5.Tuttle RM, Tala H, Shah J, et al. Estimating risk of recurrence in differentiated thyroid cancer after total thyroidectomy and radioactive iodine remnant ablation: using response to therapy variables to modify the initial risk estimates predicted by the new American thyroid association staging system. Thyroid 2010. 20 1341–1349. ( 10.1089/thy.2010.0178) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pacini F, Basolo F, Bellantone R, et al. Italian consensus on diagnosis and treatment of differentiated thyroid cancer: joint statements of six Italian societies. J Endocrinol Investig 2018. 41 849–876. ( 10.1007/s40618-018-0884-2) [DOI] [PubMed] [Google Scholar]
- 7.Momesso DP & Tuttle RM. Update on differentiated thyroid cancer staging. Endocrinol Metab Clin N Am 2014. 43 401–421. ( 10.1016/j.ecl.2014.02.010) [DOI] [PubMed] [Google Scholar]
- 8.Momesso DP, Vaisman F, Yang SP, et al. Dynamic risk stratification in patients with differentiated thyroid cancer treated without radioactive iodine. J Clin Endocrinol Metab 2016. 101 2692–2700. ( 10.1210/jc.2015-4290) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jeon MJ, Kim WG, Park WR, et al. Modified dynamic risk stratification for predicting recurrence using the response to initial therapy in patients with differentiated thyroid carcinoma. Eur J Endocrinol 2014. 170 23–30. ( 10.1530/EJE-13-0524) [DOI] [PubMed] [Google Scholar]
- 10.Valerio L, Dalmiglio C, Maino F, et al. Dynamic risk stratification integrated with ATA risk system for predicting long-term outcome in papillary thyroid cancer. Cancers 2023. 15 4656. ( 10.3390/cancers15184656) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Haugen BR, Alexander EK, Bible KC, et al. 2015 American thyroid association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American thyroid association guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid 2016. 26 1–133. ( 10.1089/thy.2015.0020) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ringel MD, Sosa JA, Baloch Z, et al. 2025 American Thyroid Association Management Guidelines for Adult Patients with Differentiated Thyroid Cancer. Thyroid 2025. 35 841–985. ( 10.1177/10507256251363120) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Tuttle RM, Haugen BR, Shah J, et al. Thyroid-differentiated and anaplastic carcinoma. In AJCC Cancer Staging Manual, 8th edn. New York, NY: Springer International Publishing, 2017. [Google Scholar]
- 14.Van De Berg DJ, Rodriguez Schaap PM, Jamaludin FS, et al. The definition of recurrence of differentiated thyroid cancer: a systematic review of the literature. Thyroid 2024. 34 1324–1334. ( 10.1089/thy.2024.0271) [DOI] [PubMed] [Google Scholar]
- 15.R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing 2025. (https://www.r-project.org/) [Google Scholar]
- 16.Bogart S. SankeyMATIC: A Sankey diagram builder for everyone. 2015. (https://sankeymatic.com/)
- 17.Mallick U, Newbold K, Beasley M, et al. Thyroidectomy with or without postoperative radioiodine for patients with low-risk differentiated thyroid cancer in the UK (IoN): a randomised, multicentre, non-inferiority trial. Lancet 2025. 406 52–62. ( 10.1016/S0140-6736(25)00629-4) [DOI] [PubMed] [Google Scholar]
- 18.Leboulleux S, Bournaud C, Chougnet CN, et al. Thyroidectomy without radioiodine in patients with low-risk thyroid cancer: 5 years of follow-up of the prospective randomised ESTIMABL2 trial. Lancet Diabetes Endocrinol 2025. 13 38–46. ( 10.1016/S2213-8587(24)00276-6) [DOI] [PubMed] [Google Scholar]
- 19.Dehbi HM, Mallick U, Wadsley J, et al. Recurrence after low-dose radioiodine ablation and recombinant human thyroid-stimulating hormone for differentiated thyroid cancer (HiLo): long-term results of an open-label, non-inferiority randomised controlled trial. Lancet Diabetes Endocrinol 2019. 7 44–51. ( 10.1016/S2213-8587(18)30306-1) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schlumberger M, Leboulleux S, Catargi B, et al. Outcome after ablation in patients with low-risk thyroid cancer (ESTIMABL1): 5-year follow-up results of a randomised, phase 3, equivalence trial. Lancet Diabetes Endocrinol 2018. 6 618–626. ( 10.1016/S2213-8587(18)30113-X) [DOI] [PubMed] [Google Scholar]
- 21.Lee Y, Cho JW, Hong SJ, et al. Dynamic risk stratification in papillary thyroid carcinoma measuring 1 to 4 Cm. J Surg Oncol 2018. 118 636–643. ( 10.1002/jso.25182) [DOI] [PubMed] [Google Scholar]
- 22.Jeon MJ, Kim M, Park S, et al. A follow-up strategy for patients with an excellent response to initial therapy for differentiated thyroid carcinoma: less is better. Thyroid 2018. 28 187–192. ( 10.1089/thy.2017.0130) [DOI] [PubMed] [Google Scholar]
- 23.Pałyga I, Rumian M, Kosel A, et al. The frequency of differentiated thyroid cancer recurrence in 2302 patients with excellent response to primary therapy. J Clin Endocrinol Metab 2024. 109 e569–e578. ( 10.1210/clinem/dgad571) [DOI] [PubMed] [Google Scholar]
- 24.Peiling Yang S, Bach AM, Tuttle RM, et al. Frequent screening with serial neck ultrasound is more likely to identify false-positive abnormalities than clinically significant disease in the surveillance of intermediate risk papillary thyroid cancer patients without suspicious findings on Follow-Up ultrasound evaluation. J Clin Endocrinol Metab 2015. 100 1561–1567. ( 10.1210/jc.2014-3651) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Seo Y, Tiu R, Stahl K, et al. Limited utility of routine surveillance ultrasound in differentiated thyroid cancer patients with undetectable postoperative thyroglobulin levels. Clin Endocrinol 2025. 102 600–606. ( 10.1111/cen.15198) [DOI] [PubMed] [Google Scholar]
- 26.Yang SP, Bach AM, Tuttle RM, et al. Serial neck ultrasound is more likely to identify false-positive abnormalities than clinically significant disease in low-risk papillary thyroid cancer patients. Endocr Pract 2015. 21 1372–1379. ( 10.4158/EP15851.OR) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ryoo I, Kim DW, Lee CY, et al. Analysis of postoperative ultrasonography surveillance after total thyroidectomy in patients with papillary thyroid carcinoma: a multicenter study. Acta Radiologica 2018. 59 196–203. ( 10.1177/0284185117700448) [DOI] [PubMed] [Google Scholar]
- 28.Hekimsoy İ, Güven M, Tokaç RH, et al. Contribution of neck ultrasonography to patient management in patients with differentiated thyroid carcinoma with excellent response to therapy. Endocrine 2024. 88 135–142. ( 10.1007/s12020-024-04147-x) [DOI] [PubMed] [Google Scholar]
- 29.Banerjee M, Wiebel JL, Guo C, et al. Use of imaging tests after primary treatment of thyroid cancer in the United States: population based retrospective cohort study evaluating death and recurrence. BMJ 2016. 354 i3839. ( 10.1136/bmj.i3839) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Seejore K, Mulla O, Gerrard GE, et al. Outcomes of 756 patients with differentiated thyroid cancer and excellent response to treatment: an evidence‐based paradigm for long‐term surveillance strategies. Clin Endocrinol 2022. 96 395–401. ( 10.1111/cen.14549) [DOI] [PubMed] [Google Scholar]
- 31.Grani G, Ramundo V, Falcone R, et al. Thyroid cancer patients with no evidence of disease: the need for repeat neck ultrasound. J Clin Endocrinol Metab 2019. 104 4981–4989. ( 10.1210/jc.2019-00962) [DOI] [PubMed] [Google Scholar]
- 32.Attia A, Touma E, Lussey-Lepoutre C, et al. Consideration of early dynamic risk stratification to guide discharge from oncologic Follow-Up in patients with differentiated thyroid cancer. Thyroid 2024. 34 1465–1475. ( 10.1089/thy.2024.0119) [DOI] [PubMed] [Google Scholar]
- 33.Garcia-Regal C, Quiñones-Silva J, Martinez-Lorca A, et al. Determinants of physicians’ risk-benefit perceptions of adjuvant radioactive iodine for differentiated thyroid cancer. Endocrine 2026. 91 259. ( 10.1007/s12020-026-04717-1) [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.