This study retrospectively analyzed patients who underwent TT combined with BCLND for PTC at the Department of Thyroid Surgery, West China Hospital of Sichuan University, from December 2017 to November 2023. The study was approved by the Ethics Committee of West China Hospital, Sichuan University (Approval No. 20221525), and registered in the Chinese Clinical Trial Registry (Registration No. ChiCTR2200067079).

Inclusion criteria were: (1) histopathologically confirmed PTC without mixed histologic types; (2) initial surgical procedure including TT and BCLND, with or without lateral neck dissection; (3) detailed operative records indicating the number, location, and intraoperative handling (preservation vs. autotransplantation) of parathyroid glands; (4) no evidence of preoperative parathyroid disease or abnormal serum calcium (Ca) levels; (5) absence of prior neck surgery or radiotherapy; (6) complete follow-up data for ≥6 months, including PTH and Ca levels.

Exclusion criteria included: history of neck irradiation, previous thyroid/parathyroid surgery, concurrent malignancies, chronic organ dysfunction, pregnancy/lactation, incomplete TT+BCLND, or missing key biochemical or follow-up data.

Patients were retrospectively assigned to groups based on operative records. The intraoperative decision for autotransplantation versus in situ preservation was based on the surgeon’s objective assessment of parathyroid gland vascularity during surgery. Baseline clinical data, laboratory results, operative findings, and postoperative outcomes were extracted from electronic medical records and follow-up interviews.

Laboratory reference ranges were based on standards from the Clinical Laboratory Department of West China Hospital: (1) PTH: 1.6–6.9 pmol/L; (2) Ca: 2.1–2.7 mmol/L; (3) Magnesium (Mg): 0.66–1.07 mmol/L; (4) Phosphorus (P): 0.81–1.45 mmol/L; (5) Vitamin D (VitD): ≥30 ng/mL; (6) Thyroglobulin (Tg): <77 ng/mL. THP was defined as a postoperative serum PTH level below the lower normal limit that returned to normal within 6 months (7). PHP was defined as persistently subnormal PTH levels beyond 6 months after surgery (7).

Recurrence was defined as either local recurrence or distant metastasis confirmed by cytology, histopathology, or radiologic imaging. Local recurrence referred to tumor regrowth in the thyroid bed or cervical lymph nodes after surgery, while distant metastasis indicated tumor spread to other tissues or organs following surgery (22).

Logistic regression was used to generate propensity scores based on the following covariates: age, sex, body mass index (BMI), tumor size, capsular invasion, T stage, N stage, scope of lymph node dissection, number of dissected and metastatic lymph nodes, and surgical duration. Patients were matched 1:1 using the nearest-neighbor method without replacement and a caliper width of 0.02 of the standard deviation of the logit of the propensity score. Balance between groups before and after matching was evaluated using standardized mean differences (SMD), with SMD <0.1 considered acceptable.

Continuous variables were expressed as mean ± standard deviation or median (interquartile range) and compared using Student’s t-test or Mann–Whitney U test. Categorical variables were compared using chi-square or Fisher’s exact test. A two-sided p value <0.05 was considered statistically significant. Statistical analysis was performed using SPSS Statistics version 29 and Python 3.10. Univariate analyses (chi-square test or Mann–Whitney U test) were used to explore associations between clinical variables and outcomes, including THP, PHP, and recurrence. Multivariate logistic regression was employed to identify independent predictors after adjusting for potential confounders. L2-regularization was applied to prevent model overfitting and improve stability.

Among 1,099 PTC patients who TT+BCLND at West China Hospital from December 2017 to November 2023, 592 patients were successfully matched using propensity score matching (1:1 ratio), with 296 patients in each group. The distribution of surgical strategies remained relatively stable throughout the study period, with no significant temporal trend detected (chi-square test, p>0.05). Matching was based on variables related to THP, PHP, tumor recurrence/metastasis, and operative time to minimize confounding.

SMD for all covariates after matching were < 0.10, indicating good balance between the two groups ( Supplementary Figure S1 ). Detailed baseline characteristics after PSM are shown in Table 1 . There were no significant differences between groups in terms of age (p=0.92), sex (p=0.86), BMI (p=0.18), tumor location (p=0.05), extent of lymph node dissection (p=0.70), T stage (p=0.45), N stage (p=0.86), or other relevant clinical variables, confirming the effectiveness of the PSM process. Operative time was significantly longer in the “2+2” group than in the preservation group (150 min vs. 123 min, p<0.01). The prevalence of nodular goiter was also higher in the “2+2” group (50.00% vs. 40.88%, p=0.03). Notably, the incidence of THP was significantly higher in the “2+2” group (72.97% vs. 48.31%, p<0.01), whereas PHP occurred significantly less frequently (0.68% vs. 3.72%, p=0.03).

In both groups, PTH levels declined significantly after surgery and gradually recovered over time. The most substantial drop occurred on postoperative day 1: in the preservation group, the median PTH level decreased from 5.03 to 1.66 pmol/L, while in the “2+2” group, it dropped from 5.03 to 0.95 pmol/L. PTH levels gradually increased thereafter, approaching preoperative levels by month 12 (see Supplementary Table S1 ). The intergroup difference was statistically significant on postoperative day 1 (p<0.01), but not at subsequent time points. VitD levels showed relatively mild fluctuations in both groups. Slight increases were observed at 1 month postoperatively in both the preservation and “2+2” groups. The change in VitD levels reached statistical significance in the preservation group (p=0.02), but not in the “2+2” group (p=0.29). Ca levels also demonstrated significant postoperative changes, persisting through month 12 (p<0.01). The largest intergroup difference occurred on postoperative day 1.

PTH recovery patterns are illustrated in Supplementary Figure S2 . Briefly, on postoperative day 1, both groups experienced a marked drop, with a greater decrease in the ‘2+2’ group (median: 0.95 vs 1.66 pmol/L, p<0.01). However, the ‘2+2’ group exhibited faster recovery between day 1 and month 1, with both groups converging toward similar levels by month 12 (approximately 4.35 pmol/L). Calcium recovery followed a similar pattern ( Supplementary Figure S3 ), with the ‘2+2’ group showing greater initial decline but faster subsequent recovery. The intergroup difference in Ca levels was statistically significant on day 1 (p<0.01) but diminished thereafter, indicating convergent recovery trajectories.

Patients were divided into a THP group (n=359, 60.64%) and a non-THP group (n=233, 39.36%) based on the occurrence of THP during follow-up. Compared with the non-THP group, patients in the THP group had significantly longer hospital stays (7.00–9.00 vs. 7.00–8.00 days, p<0.01), longer operative times (148.00 vs. 124.00 minutes, p<0.01), and larger tumor diameters (1.10 vs. 1.00 cm, p=0.04) ( Table 2 ). In addition, the THP group had a significantly higher rate of PA (60.17% vs. 34.33%, p<0.01), higher postoperative complication rates (10.03% vs. 4.72%, p=0.03), and higher incidence of PHP (3.62% vs. 0.00%, p=0.01).

No significant differences were observed between the two groups regarding baseline characteristics such as sex, age, BMI, follow-up duration, tumor recurrence/metastasis, radioactive iodine treatment, nodular goiter, capsular invasion, number of metastatic or dissected lymph nodes, lymph node metastasis rate, tumor location, extent of lymph node dissection, or TNM stage (p>0.05).

Univariate logistic regression analysis revealed that PA was significantly associated with an increased risk of THP (OR = 2.889, 95% CI: 2.05–4.072, p<0.01) ( Figure 1 ). Operative time was also a significant risk factor (OR = 1.003, 95% CI: 1.001–1.006, p<0.01). Other variables, including tumor location, sex, age, BMI, capsular invasion, tumor size, and number of dissected lymph nodes, were not significantly associated with THP (p>0.05).

In multivariate logistic regression analysis including age, sex, BMI, PA status, tumor location, operative time, capsular invasion, tumor size, number of dissected lymph nodes, extent of lymph node dissection, T stage, N stage, and AJCC stage, PA remained an independent risk factor for THP (OR = 2.476, 95% CI: 1.600–3.832, p<0.01). Additionally, tumor size was also independently associated with THP (OR = 1.424, 95% CI: 1.014–2.000, p=0.04).

Based on the occurrence of PHP, patients were divided into the PHP group (n=13, 2.20%) and the non-PHP group (n=579, 97.80%). Compared with the non-PHP group, patients in the PHP group had significantly higher THP incidence (100.00% vs. 59.76%, p=0.01), higher postoperative complication rates (38.46% vs. 7.25%, p<0.01), and lower rates of PA (15.38% vs. 49.22%, p=0.03) ( Table 3 ). Additionally, the PHP group had a higher number of metastatic lymph nodes (8.00 vs. 2.00, p=0.02) and dissected lymph nodes (27.00 vs. 14.00, p=0.02). No statistically significant differences were observed between the two groups in terms of T stage, N stage, or AJCC stage. Other variables—including age, BMI, sex, operative time, intraoperative blood loss, length of hospital stay, follow-up duration, tumor size, radioactive iodine treatment, recurrence/metastasis, capsular invasion, tumor location, and comorbidities (e.g., Hashimoto’s thyroiditis, diabetes, hypertension)—showed no significant differences (p>0.05).

Univariate logistic regression analysis indicated that PA was significantly associated with a lower risk of developing PHP (OR = 0.176, 95% CI: 0.039–0.802, p=0.03) ( Figure 2 ). No other variables—including age, sex, BMI, operative time, capsular invasion, tumor size, tumor location, number of dissected lymph nodes, extent of lymph node dissection, T stage, N stage, or AJCC stage—were significantly associated with PHP (p>0.05).

Multivariate logistic regression analysis incorporating age, sex, BMI, PA status, tumor location, operative time, capsular invasion, tumor size, number of dissected lymph nodes, extent of dissection, T stage, N stage, and AJCC stage confirmed that PA remained independently and negatively associated with the occurrence of PHP (OR = 0.139, 95% CI: 0.025–0.772, p=0.02). No other covariates reached statistical significance after adjustment.