A total of 423 patients with PTC who underwent total thyroidectomy and postoperative ¹³¹I ablation at the Affiliated Hospital of Qingdao University from January 2017 to October 2020 were enrolled. The inclusion criteria were as follows: (1) patients aged 20 to 80 years old who underwent total thyroidectomy, (2) postoperative pathological diagnosis confirmed as PTC, (3) patients received ¹³¹I radioiodine therapy at least one time, (4) no distant metastasis was confirmed by imaging and pathology, and (5) patients finished 2 years of follow-up. The participants with the following conditions were excluded: (1) pathologically confirmed as other types of thyroid carcinoma, (2) other underlying diseases including other malignant tumors or autoimmune disorders, and (3) positive thyroglobulin antibody (TgAb). Chest and abdominal computed tomography (CT) scan were performed before ¹³¹I ablation. The initial dosage of ¹³¹I ablation was set based on the recurrence risk stratification according to the 2015 ATA Guidelines (6). For low- and moderate-risk patients, the initial ¹³¹I ablation dose was 50–100 mCi. For high-risk patients with extraglandular invasion (larynx, trachea, esophagus, recurrent laryngeal nerve, striated muscle, etc.), the initial ¹³¹I ablation therapeutic dose was 120–180 mCi (6, 11). The patients achieved the goal of TSH > 30 mU/L after L-evothyroxine withdrawal and followed a low-iodine diet for 3–4 weeks. The patients were administrated L-T4 at day 3 after ¹³¹I ablation. The whole-body scan (Rx-WBS) and single-photon emission computed tomography/computed tomography (SPECT/CT) were performed within 1 week after ¹³¹I ablation (6).

During the 2-year follow-up, the first evaluation was performed at 3 months after initial ¹³¹I ablation therapy, and then dynamic evaluation was performed every 6 months during follow-up. Dynamic evaluation involved both serological and imaging measurements, including serum Tg levels, TgAb, thyroid function especially TSH level, and neck ultrasound. ¹³¹I diagnostic scanning (Dx-WBS), chest CT scan or ¹⁸F-FDG positron emission tomography (PET)-CT scanning, and fine needle aspiration (FNA) biopsy were also used if necessary.

Preablative stimulated thyroglobulin (ps-Tg) was defined as the Tg levels when TSH > 30 IU/mL as patients stopped taking L-T4 before ¹³¹I ablation (6). The diagnosis of Hashimoto’s thyroiditis (HT) was identified based on postoperative pathological analysis. Pre-operative lymph node metastasis was diagnosed by FNA (cytopathological diagnosis of PTC, or elevated Tg washout measurement) before surgery, or postoperative pathological analysis. Postoperative lymph node metastasis referred to the presence of metastatic lymph nodes detected by ¹³¹I-WBS or ultrasound and confirmed by FNA after surgery. Lymph node metastasis rate (LR) referred to the rate of involved lymph nodes, indicating the number of involved lymph nodes/total number of resected lymph nodes, based on postoperative pathological analysis.

In this study, patients with PTC with lymph node metastasis underwent total thyroidectomy according to the 2015 ATA Guidelines (6). Total thyroidectomy was performed for tumor diameter >4 cm. For the tumor diameter <4 cm (including diameter ≤1 cm), total thyroidectomy was performed under the following conditions: (1) bilateral foci; (2) lymph node metastasis in the lateral cervical area or lymph node metastasis number ≥5 or diameter ≥3 cm; (3) extracapsular invasion and metastatic lymph node invasion of surrounding tissues and organs, such as peripheral fat tissue, muscles, trachea, esophagus, laryngeal reentry nerve, and invaded blood vessels; and (4) high-risk factors such as head and neck radiotherapy history over the course of childhood and adolescence, and thyroid carcinoma family history.

The lymph node dissection was performed as follows: (1) central lymph node dissection: the therapeutic and prophylactic central-compartment lymph node dissection is routinely performed by experienced and skilled surgeons and the surrounding tissues were adequately protected; (2) lateral lymph node dissection: for PTC with lateral lymph node metastases confirmed by preoperative FNA or intraoperative frozen pathological examination, therapeutic lateral lymph node dissection was performed. The prophylactic lateral lymph node dissection is not routinely performed.

According to the 2015 ATA Guidelines (6), the response to treatment was categorized into ER, indeterminate response (IDR), biochemical incomplete response (BIR), and structural incomplete response (SIR), based on Tg level and imaging examination. Based on the follow-up at 2 years, the recruited patients were divided into two groups, the ER group and the non-ER group. Patients with IR, BIR, and SIR were assigned as the non-ER group (6). Recurrence risk stratification and TNM staging were determined based on the 2015 ATA Guidelines (6) and the 8th edition TNM staging system of the American Joint Committee on Cancer (AJCC) (12).

Statistical analysis was performed using the SPSS 26.0 software. Categorical variables were presented as frequencies and percentages, and two-group comparison was performed by the χ ² test. Partial distributed data were presented as the median value and comparisons between two groups were performed using the Mann–Whitney U test. Logistic regression analysis was used to analyze the factors affecting the achievement of ER. Prognostic receiver operating characteristic (ROC) curves were applied to obtain the optimal threshold for estimating ER achievement, and the prediction of the cumulative risk of non-ER was performed using the Kaplan–Meier model. The statistical significance was p < 0.05 compared between two groups. The comparison among the three groups and four groups was p < 0.017 and p < 0.008, which were considered statistically significant, respectively.

A total of 423 patients with PTC with lymph node metastasis were enrolled, with age ranging from 20 to 78 years (average age, 43.82 ± 11.85 years old). The male-to-female ratio was 1:2.18. The follow-up time was 24–32 months. The maximum diameter of tumor foci was 0.1–5.5 cm, averaging 1.57 ± 1.06 cm. The number of foci was one to eight (average foci number was 1.81 ± 1.09). The number of lymph node metastasis was 1–36, with a median of 4. Lymph node metastasis rate (% of involved metastatic lymph nodes/total number of resected lymph nodes) was 5%–100%, with a median of 40%. There were 226 cases (53.43%) with extracapsular invasion, including 146 cases (34.52%) in the strap muscles, 44 cases (10.43%) with recurrent laryngeal nerve invasion, 18 cases (4.25%) with tracheal invasion, 6 cases (1.41%) with esophageal invasion, 6 cases (1.41%) with blood vessel (internal jugular vein, common carotid artery, and transverse cervical artery) invasion, and 6 cases (1.41%) with surrounding soft tissue invasion. The clinicopathologic features are shown in Table 1 .

Based on the response to ¹³¹I ablation after 2 years of follow-up, 314 patients with PTC achieved ER with an ER rate of 74.2%. The patients with PTC were divided into ER and non-ER groups. The results showed that the proportion of female patients, unifocal, primary foci ≤1 cm in diameter, no extracapsular invasion, combination with HT, low ps-Tg level (≤3.87 ng/mL), and low risk of recurrence risk stratification were significantly higher in the ER group than in the non-ER group (all p < 0.05). However, age, laterality of tumor foci, and TNM staging did not show significant difference between two groups (all p > 0.05) ( Table 2 ).

Comparison of the pathologic characteristics of metastatic lymph nodes between ER and non-ER groups was analyzed. The ER group had a lower number and rate of lymph node metastases, less postoperative lymph node metastases, and a significantly lower proportion of lymph nodes located in the central and cervical lateral region than the non-ER group (all p < 0.05) ( Table 3 ).

In this study, the ER group had an increased rate of HT. Therefore, we further analyzed the relationship between HT and clinicopathological features. The results showed that female patients (female patients 84.50% vs. male patients 15.50%) and non-extracapsular invasion (non-extracapsular invasion 61.97% vs. extracapsular invasion 38.03%) in the HT group (71 cases, 16.78%) were significantly higher than those in the non-HT group (352 cases, 83.22%) (all p < 0.05), while age, number of primary tumor foci, the maximum diameter of primary foci, unilateral/bilateral foci, number of lymph node metastases, LR, and postoperative lymph node metastasis showed no difference between HT and non-HT groups.

To analyze the relationship between frequency/dosage of ¹³¹I ablation and ER, we found that the ER rate in the 1 time ¹³¹I ablation therapy group was higher than that in the 2 times and ≥3 times ¹³¹I ablation therapy group (all p < 0.017). In addition, the ER rate in the low initial ¹³¹I ablation dose ≤100 (50–100) mCi group was significantly higher than that in the high-dose >100 (120–180) mCi group (p < 0.05). We further divided the patients into four groups based on the total dose of ¹³¹I therapy. The results showed that the ER rate of the ¹³¹I therapy total dose 50–100 mCi group was significantly higher than that of the 120–150 mCi group, as well as the >200–(220–600) mCi group (all p < 0.008) ( Table 4 ).

Multivariate logistic regression analysis was performed to analyze the factors associated with ER after ¹³¹I therapy. The results revealed that the maximum diameter of tumor foci (≤1 cm), unifocal, combination with HT, lymph node metastases rate (≤40%), no postoperative lymph node metastasis, lower level of ps-Tg (≤3.87 ng/mL), and the frequency of ¹³¹I therapy (one time) were independent risk factors that positively correlated with the ER achievement (OR: 1.744, 3.114, 3.920, 4.018, 2.074, 9.767, and 49.491, respectively; all p < 0.05). There was no correlation between ER and the number of metastatic lymph nodes, laterality of lymph node metastasis, extracapsular invasion, or initial dose and total dose of ¹³¹I therapy (all p > 0.05) ( Table 5 ).

The predictive value of ps-Tg level and LR in ER achievement was analyzed, respectively. The results showed that the maximum area under the curve (AUC) of the ROC curve was 0.821 (95% CI 0.777–0.865) and 0.746 (95% CI 0.691–0.800). The best cutoff values were 4.625 ng/mL for ps-Tg and 50.50% for LR, with corresponding sensitivities of 84.4% and 63.3%, specificities of 67.8% and 79.6%, PPVs of 92.6% and 86.2%, and negative predictive values of 48.7% and 51.8%, respectively. Results indicate that ps-Tg ≤ 4.625 ng/mL and LR ≤ 50.50% are effective factors for predicting ER achievement ( Figure 1 ).

The ps-Tg value and LR were analyzed for the joint prediction of the cumulative risk for non-ER. The results revealed that regardless of whether the ps-Tg value was at low level (≤4.625 ng/mL) or high level (>4.625 ng/mL), the cumulative risk of non-ER elevated with the increase of LR, especially for the high-level ps-Tg group, whereas at the same LR, the cumulative risk of non-ER was higher in the high-level ps-Tg group than the low-level ps-Tg group (both p < 0.05) ( Figure 2 ).