We retrospectively collected the clinico-pathologic data of 3,392 patients diagnosed with PTC who underwent thyroidectomy in our center from January 2008 to December 2015. The exclusion criteria consisted of history of thyroidectomy (n=38), concurrent medullary or anaplastic thyroid carcinoma (n=4), and follow-up loss (n=14). Finally, 3,336 patients were enrolled in this study. Of them, 206 patients underwent prophylactic/therapeutic bilateral central compartment neck (level VI; pretracheal, prelaryngeal, and paraesophageal LNs) dissection and therapeutic lateral cervical neck (levels IIa, III, IV, and Vb, not including the sternocleidomastoid, jugular vein, and accessory nerve) dissection (LCND) after lateral LNM was diagnosed on the basis of the preoperative examination findings. These patients were divided into the PTMC and non-PTMC groups. All patients were diagnosed using US-guided FNA biopsy for thyroid nodules before surgery. To confirm the presence of LNM, real-time US and contrast-enhanced CT were performed; all images were assessed by 1 of 3 radiologists who specialize in diseases of the neck and thyroid with >5 years of experience. FNA was performed on the suspicious lateral neck LN for metastasis assessment on US, which was defined as follows: hyper-echogenicity (higher than the surrounding muscles), cystic change, loss of fatty hilum, calcification, round shape, and abnormal vascular pattern [21-23]. If the cytology for FNA revealed metastatic LNs or thyroglobulin in the LNs with a washout fluid level of ≥10 ng/dL, the condition was diagnosed as a metastatic LN. If not visible on US but strongly suspected on CT, the selective node that was grossly enlarged or had a suspicious feature was excised; intraoperative frozen biopsy before LCND then followed. The suspicious CT features were as follows: presence of calcifications, central necrosis or cystic change, and heterogeneous cortical enhancement or enhancement more than that of the adjacent muscle [21-23]. As recommended in the ATA guidelines , prophylactic LCND was not performed. This study was approved by the Institutional Review Board of Kangbuk Samsung Hospital (KBSMC 2017-04-037).
2.1. Clinico-pathologic analysis
We compared the incidence of lateral LNM between the PTMC and non-PTMC groups and analyzed their clinico-pathologic factors. In addition, the prognostic difference between the 2 groups was indirectly analyzed by comparing the differences in the factors affecting the prognosis, which were demonstrated in previous studies [19, 20, 24]. Ito et al.  reported that patients under the N1b stage who underwent LCND had a high recurrence rate if they were older than 55 years or had a metastatic node size of ≥3 cm. In other patients, the LNR was helpful in predicting the prognosis in addition to the number and size of metastatic LNs. Lee et al.  reported that the LNR (levels II-VI) had a cutoff value of 0.5, which was associated with patients’ prognosis. Park et al.  reported that the loco-regional recurrence-free survival was affected by the presence of >6 metastatic nodes and by a cutoff value of 0.22 for the LNR. Based on these previous results, the differences between the 2 groups were compared using the clinico-pathologic data of the patients in this study.
2.2. Statistical analysis
All statistical analyses were performed using R version 3.3.2 [22, 23, 25-28]. For comparison between the PTMC and non-PTMC groups, the t-test for continuous variables and Fisher’s exact test or Chi-square test for categorical variables were used.