This was a multicenter, randomized, blank-controlled study (ChiCTR2300068502 Http://www.chictr.org.cn/). This study was conducted in accordance with the Declaration of Helsinki and approved by the institutional review board. Written informed consent was obtained from all patients.

Based on the different resection ranges of the primary lesions, the clinical trials were divided into two subgroups: unilateral lobectomy and total thyroidectomy.

Patients satisfying the following criteria were recruited in the unilateral lobectomy group: (1) preoperative and intraoperative diagnosis of untreated, unilateral, and single-focal papillary thyroid carcinoma; (2) no cervical lymph node metastasis on imaging examination, and no indication of lateral neck lymph node dissection; (3) unilateral lobectomy plus isthmectomy with ipsilateral central lymph node dissection; (4) age 18–70 years; (5) results of preoperative laboratory blood test–neutrophils (ANC) ≥1.5×10⁹/L, platelets (PLT) ≥100×10⁹/L, total bilirubin (TBI) ≤2 times the upper limit of normal (ULN) (2 mg/dL), alanine aminotransferase (ALT) ≤2 times the ULN, and aspartate aminotransferase (AST) ≤2 times the ULN; and (6) signed informed consent.

Patients who met the following criteria were recruited into the total thyroidectomy group: (1) preoperatively and intraoperatively diagnosed with untreated, unilateral, and single-focal papillary thyroid carcinoma; (2) no enlarged lymph nodes on physical examination; and lateral cervical lymph node metastasis on imaging examination; (3) total/near total thyroidectomy, ipsilateral central and lateral lymph node dissection; (4) age 18–70 years; (5) preoperative laboratory blood test results (ANC≥1.5×10⁹/L, PLT≥100×10⁹/L, TBI ≤ 2 times ULN (2 mg/dL), ALT ≤ 2 times ULN, AST ≤ 2 times ULN; and (6) signed informed consent.

Patients were excluded if they met one or more of the following criteria: (1) concomitant Hashimoto’s thyroiditis; (2) active stage of liver disease or abnormal liver function, with ALT, AST, and TBIL≥2 times the ULN; (3) renal function impairment, Cr≥2 times the ULN or BUN≥2 times the ULN; (4) a history of alcohol, drug, or substance abuse and disturbance of blood coagulation; and (5) a history of serious uncontrolled medical disease, recent myocardial infarction, or acute infection.

The primary endpoints included parathyroid hormone (PTH) levels, number of lymph nodes detected, number of tiny lymph nodes detected, and recognition and retention of the parathyroid glands. The secondary endpoint was recognition and protection of the recurrent laryngeal nerve. PTH levels were measured preoperatively, intraoperatively, and on postoperative days 1 and 1 month.

All statistical analyses were performed using SAS version 9.4 statistical professional analysis software. Measurements were reported as the mean standard deviation. Continuous measurements were compared using the t-test, and count data were analyzed using the chi-squared test. The Wilcoxon rank-sum test was used for intergroup comparisons. The safety analysis was based on descriptive analytics. The adverse events and the incidence in each group were listed in the table and compared using the chi-square test or Fisher’s exact test. Statistical significance was set at P value < 0.05.

A total of 569 patients with PTC from seven hospitals across China were enrolled. According to the different resection ranges of the primary lesion, they were assigned to two subgroups: the unilateral lobectomy group (UL, 283 patients) and total thyroidectomy group (TT, 286 patients), who were randomly assigned to the carbon nanoparticles group (A group) or control group (B group), respectively ( Figure 1 ). The LNs in the A group were stained black by nanoparticles ( Figure 2 ). As shown in Table 1 , no significant differences were detected between the two groups regarding baseline characteristics.

In the UL group, there were no significant differences between the carbon nanoparticles and control groups in terms of preoperative (P=0.1582), intraoperative (P=0.9048), and postoperative day one (P=0.7747) and first month (P=0.3840) serum PTH levels, Table 2 .

In the TT group, there was no significant difference between the carbon nanoparticles and control groups in terms of preoperative (P=0.6870), intraoperative (P=0.6719), and postoperative day one (P=0.4255) serum PTH levels. However, the mean serum PTH in carbon nanoparticles group was significantly higher than that of the control group (25.92 ± 19.76 mmol/L vs. 21.29 ± 19.68 mmol/L, P=0.0368) on first month post-operation, Table 2 .

In the UL group, there was no significant difference between the carbon nanoparticles and control groups in terms of preoperative (P=0.4237) and first month post-operation (P=0.428) serum Ca²⁺ levels. While, there was no significant difference between the carbon nanoparticle and control groups in terms of preoperative (P=0.6148) and first month post-operation (P=0.99) serum Ca²⁺ levels in the TT group, Table 3 .

In UL group, the number of LNs dissected from the carbon nanoparticles group was significantly higher than that of the control group (8.02 ± 7.70 vs 5.37 ± 5.63, P< 0.0001). The detection rate of tiny lymph nodes in the carbon nanoparticles group was higher than that in the control group (P=0.0268). The number of tiny lymph nodes was also higher in the carbon nanoparticles group; however, the difference was not statistically significant (4.99 ± 3.80 vs. 3.93 ± 3.23, P=0.0823). Furthermore, significantly more ≤2 mm lymph nodes were detected in the carbon nanoparticles group than that of the control group (2.45 ± 3.14 vs. 1.45 ± 2.55, P=0.0034), Table 4 .

In the TT group, the number of dissected LNs was higher in the carbon nanoparticles group; however, the difference was not statistically significant (40.82 ± 21.61 vs. 38.20 ± 19.14, P=0.4346). In addition, the numbers and detection rate of tiny lymph nodes was also higher in the carbon nanoparticles group; however, the difference was not statistically significant (80.15% vs. 78.47%, P=0.7296;12.15 ± 8.96 vs 10.45 ± 7.73, P=0.1750). Furthermore, the number of ≤2 mm lymph nodes was also higher in the carbon nanoparticles group, and the difference was not statistically significant (9.77 ± 9.34 vs 8.24 ± 7.96, P=0.2449), Table 4 .

No drug-related AE or complications due to the injection of carbon nanoparticles were recorded in this study. There were no significant differences between the two groups in recurrent laryngeal nerve injury (three cases vs. one case) in the UL group. There were no significant differences between the two groups in recurrent laryngeal nerve injury (five vs. four cases) in the TT group. There were no significant differences between the carbon nanoparticles group and control groups regarding the parathyroid gland preserved in situ in the UL and TT groups, Table 5 .