Article information

Korean J Otorhinolaryngol-Head Neck Surg. 2026;.kjorl-hns.2026.00171

Publication date (electronic) : 2026 July 15

doi : https://doi.org/10.3342/kjorl-hns.2026.00171

Hyeon A Lee, Ha Gyeong Lee, Chang Myeon Song, Kyung Tae, Yong Bae Ji

Department of Otolaryngology-Head and Neck Surgery, College of Medicine, Hanyang University, Seoul, Korea

이현아, 이하경, 송창면, 태경, 지용배

한양대학교 의과대학 이비인후-두경부외과학교실

Address for correspondence Yong Bae Ji, MD, PhD Department of Otolaryngology-Head and Neck Surgery, College of Medicine, Hanyang University, 222 Wangsimni-ro, Seongdong-gu, Seoul 04763, Korea Tel +82-2-2290-8585 Fax +82-2-2293-3335 E-mail jyb20000@hanyang.ac.kr

Received 2026 March 12; Revised 2026 April 24; Accepted 2026 May 6.

Abstract

Introduction

Thyroid papillary carcinoma accounts for the majority of thyroid malignancies and is generally associated with a favorable prognosis. Despite the excellent long-term survival rates in papillary thyroid carcinoma (PTC), cervical lymph node metastasis is common, with reported rates ranging from approximately 20% to 90% depending on tumor size and the characteristics of the study population [1]. The most common site of metastasis is the central lymph node compartment, which is generally associated with a less favorable prognosis [2,3], as reflected in the American Joint Committee on Cancer (AJCC) 8th edition staging system [4]. Most surgeon agree that the patients with clinically suspicious lymph node metastasis should undergo therapeutic central compartment neck dissection in addition to thyroidectomy.

In addition, microscopic metastasis in the central lymph nodes may be present even in patients without clinically apparent nodal disease. So, central lymph node dissection has been recommended even in case with no clinically apparent lymph node metastasis. However, the indication and the extent of prophylactic central neck dissection (pCND) have been a subject of debate for several decades [5-8]. The current guidelines recommend considering pCND in patients with advanced primary tumors (T3 or T4) or high-risk PTC [5,9,10]. However, there is insufficient clear evidence regarding this approach.

Proponents of pCND argue that PTC has a high incidence of lymph node metastasis, and therefore, the removal of micro-metastases in the central compartment lymph nodes could prevent recurrence and improve survival [6,11]. Additionally, it offers the advantage of avoiding the need for additional surgery in the future, which is associated with a higher complication rate when performed due to recurrence [7]. Furthermore, there is a growing need for a pathological report on lymph node status to accurately determine the cancer stage and make decisions regarding the necessity of postoperative radioactive iodine (RAI) therapy [12].

However, pCND typically carries the disadvantage of potentially causing postoperative complications such as permanent hypoparathyroidism and damage to the recurrent laryngeal nerve [6,13]. Some argue that it remains unclear whether micro-metastases detected through pCND actually have a significant impact on recurrence and survival [14,15]. Additionally, there are concerns that tumor upstaging due to pCND may result in unnecessary RAI therapy [8].

To minimize the surgical morbidity and clarify the impact of pCND on prognosis, it is crucial to address the current gap in research on the role of pCND. However, there is a little study about the role of pCND in locally advanced PTC cases. Therefore, our study aims to assess the role of pCND in clinically node-negative (cN0) T3/T4 PTC cases.

Subjects and Methods

Results

Patients’ characteristics and pattern of central lymph node metastasis

A total of 1050 cN0 PTC patients were enrolled in this study. According to 8th AJCC staging system, 950 patients were classified as T1/2, and 100 patients were classified as T3/T4. The clinicopathological, and tumor characteristics of both groups are presented in Table 1. Between the two groups, there were no significant differences in age, sex, multiplicity, bilaterality, N category, the extent of CND and follow-up period. However, T3/T4 group showed statistically significant higher in the prevalence of gross ETE, LVI, postoperative RAI therapy, mean administered activity of RAI and recurrence. There were 4 deaths (0.4%) in the T1/T2 group and 1 death (1.0%) in the T3/T4 group, but there were no deaths due to thyroid cancer.

Table 1.

Clinicopathologic characteristics of patients with clinically node-negative papillary thyroid carcinoma

In the T1/T2 group, 200 patients did not undergo CND, while 750 patients underwent CND, including 308 bilateral pCND. In the T3/T4 group, 17 patients did not undergo pCND, while 83 patients underwent pCND including 34 bilateral pCND. When comparing the two groups among the 833 patients who underwent pCND, occult central lymph node metastasis (CLNM) was not significantly different (37.1% in the T1/T2 group vs. 43.4% in the T3/T4 group, p=0.283). When examining the rate of bilateral LN metastasis in patients who underwent bilateral pCND with the tumor confined to one side, bilateral central compartment metastasis rate was also not different 3.9% in the T1/T2 group vs. 2.9% in the T3/T4 group, p>0.999). The average number of removed LNs and metastatic LNs showed no significant difference between the two groups (p=0.538 and p=0.157, respectively). There were 8 (0.8%) in the T1/T2 group and 5 (5.0%) in the T3/T4 group of recurrence during mean follow up of 85.7 months and 99.3 months, respectively.

Risk factors of CLNM and risk of recurrence in cN0 T1-T4 PTC

Table 2 shows factors related to CLNM in cN0 T1-T4 PTC. In multivariate analysis, only younger age (<55), male gender, LVI, and tumor size (>20 mm) were significantly correlated with CLNM (p<0.001, p=0.014, p<0.001, and p=0.001, respectively).

Table 2.

Logistic regression analysis of related factors of central lymph node metastasis in clinically node-negative T1–T4 PTC patients

Using Cox proportional hazards regression model for all cN0 PTC patients, LVI, T3/T4 categories, Gross ETE, and CLNM were significantly associated with recurrence in univariate analysis (Table 3). However, in multivariate analysis, only gross ETE and CLNM was associated with recurrence (p=0.014, p=0.049, respectively).

Table 3.

Cox regression analysis for risk of recurrence in clinically node-negative T1–T4 PTC patients

Surgical outcomes and risk of recurrence in cN0 T3/T4 PTC

The surgical outcomes, postoperative RAI and the pattern of recurrence in cN0 T3/T4 group are listed according to the pCND in Table 4. Postoperative RAI therapy was performed more in the pCND group compared to the non-pCND group (59.8% vs. 88.0%, p=0.008). However, there were no significant differences in RAI dose and stimulated Tg between two groups (p=0.196, p=0.492, respectively). Regarding the size of the metastatic deposits, the pathological records for the size of occult CLNM were available for 12 out of 36 patients in the T3/T4 pCND group. Among these, all three patients in the pCND group who experienced recurrence were identified as having macrometastases (>2 mm).

Table 4.

Surgical outcomes of clinically node-negative T3/T4 PTC according to pCND

In the comparison of postoperative complications, preoperative vocal fold palsy was observed in one case in the non-pCND group and five cases in the pCND group. After excluding cases with preoperative vocal fold palsy, no statistically significant differences were observed in both transient and permanent vocal fold palsy. Although there were no statistical difference in the other postoperative complications between the two groups, transient hypoparathyroidism was significantly higher in the pCND group (p=0.015).

The specific patterns of recurrence in the cN0 T3/T4 group were further analyzed. All five recurrence cases involved the lateral LNs (level II, III, or IV), while no central LN recurrences were observed in either group. In the pCND group, one of the four recurrent cases had suspected recurrence at both the operative bed and a lateral LN on US; this case was duplicated in the site-specific analysis. Specifically, lateral neck recurrences were identified at right levels III and IV (two cases), left levels II and III (one case), and right level III (two cases). Furthermore, there were no cases of distant metastasis or death due to thyroid cancer. The 5-year RFS rate was not significantly different between the two groups (100% in the non-pCND group vs. 95.5% in the pCND group, p=0.754) (Fig. 1).

Fig. 1.

Kaplan–Meier curves for recurrence-free survival of clinically node-negative T3/T4 PTC patients. PTC, papillary thyroid carcinoma; pCND, prophylactic central neck dissection.

In cN0 T3/T4 PTC patients, univariate Cox proportional hazards regression analysis identified male gender as the only variable reaching statistical significance; however, no other clinicopathologic factors were significantly associated with recurrence (Table 5).

Table 5.

Cox regression analysis for risk of recurrence in clinically node-negative T3/T4 PTC

Discussion

The controversies regarding pCND has been persisted for several decades. Over the years, a transition from an era advocating aggressive pCND to the current practice of narrowing indication and extent of surgery, although it is clear that CLNM has a negative impact on prognosis with decreased. CLNM in differentiated thyroid cancer (DTC) patients is a significant risk factor for lateral LN metastasis and has been shown to increase the risk of recurrence [17-19] and associated with decreased overall survival (OS) [20].

Consequently, for cN1 PTC patients, therapeutic CND is widely accepted as the standard treatment in conjunction with thyroidectomy. However, the management of central LNs in cN0 PTC patients remains a subject of ongoing debate.

Several studies have demonstrated pCND may lead to reduced recurrence and mortality rates [6], improved staging accuracy (e.g., converting Nx to N1a, upstaging patients over 55 years of age from AJCC stage I to stage II) [6,11], lower Tg levels [21], enhanced suitability for RAI treatment [22], and a reduced need for reoperative surgery [11,13]. So, some authors have advocated pCND given the importance of achieving complete surgical resection to improve recurrence and survival outcomes, while there is a lack of randomized controlled data to confirm the impact of routine pCND on recurrence or survival rates in DTC.

Postoperative Tg is a well-established marker for monitoring residual or recurrent disease in DTC patients. The Tg level at the time of RAI administration serves as an early prognostic indicator [23], and several studies have shown that TT with pCND results in lower postoperative preablative Tg levels compared to TT alone [21], guiding both risk stratification and patient selection for RAI therapy [22]. Supporters of pCND note that postoperative complications are not significantly increased, as confirmed by a recent systematic review of randomized controlled trials reporting similar complication rates in cN0 PTC patients treated with TT±pCND [15]; complications are often attributable to thyroidectomy itself, and recurrent laryngeal nerve dysfunction is usually temporary. Importantly, recurrence risk correlates with the number of metastatic LNs at initial presentation [19,24,25], with five or more CLNMs (<2 mm) or macrometastases (>2 mm) associated with higher recurrence [3], a finding incorporated into the 2025 ATA risk stratification. Despite advances in high-resolution US and CT, preoperative evaluation of the central compartment remains unreliable [26,27], making appropriate risk assessment for CLNM difficult without pCND.

Wada, et al. [28] conducted a retrospective study comparing disease-free survival (DFS) and disease-specific survival between 42 patients who underwent therapeutic central neck dissection for clinically evident LN metastases and 92 patients who underwent prophylactic dissection. They found that microscopic CLNM were not significantly associated with recurrence or survival, whereas macroscopic metastases were significantly correlated with poorer RFS [28]. Moreno, et al. [29] retrospectively analyzed 135 patients with PTC who underwent TT with or without central neck dissection. They found that preoperative central neck US detected only macroscopic nodal disease, and that patients with negative central neck US findings had excellent long-term regional control and OS, regardless of the presence of microscopic metastases on pathology or the patient’s age. In contrast, macroscopic nodal metastases identified preoperatively were significant predictors of recurrence and disease-specific survival, indicating that pCND did not significantly impact long-term outcomes in patients without clinically evident nodal disease. Consistent with these findings, Randolph, et al. [3], in a review of the literature, concluded that the prognostic impact of nodal metastases is largely determined by the size and number of metastatic LNs and the presence of extranodal extension, with micrometastases (<2 mm) having limited influence on recurrence, while macrometastases (>2 mm) and extranodal extension are associated with higher recurrence risk and poorer outcomes.

Systematic reviews show conflicting result on the impact of pCND; some studies indicate the pCND does not influence recurrence [14,15], while others suggest it may reduce the risk of recurrence [30,31]. However, the extent to which pCND reduces the risk of recurrence may be related to the increased use of RAI ablation, making its precise impact still unclear. There is little evidence to suggest that pCND improves OS in cN0 PTC. In terms of recurrence, DFS and RFS, several retrospective studies have found no significant difference between patients undergoing TT with or without pCND [32-34]. Costa, et al. [32] retrospectively analyzed 244 cN0 PTC patients and found no significant difference in DFS between patients who underwent TT with pCND and those who had TT alone. Similarly, Zuniga and Sanabria [33] reviewed 266 cN0 PTC patients and reported that prophylactic dissection did not significantly alter recurrence rates during long-term follow-up. More recently, Yoo, et al. [34] analyzed a total of 477 cN0 PTC patients and applied propensity score matching to create two well-balanced groups of 135 patients each, observing no significant difference in RFS between the pCND and nonpCND groups. This is further supported by the prospective randomized controlled trial of 181 patients, which demonstrated that pCND did not significantly affect recurrence or long-term outcomes compared with TT alone [35].

While pCND may enhance the accuracy of staging, it can lead to upstaging, resulting in more patients undergoing RAI therapy and receiving higher doses of RAI [36]. Especially for T3 and T4 PTC, which are already considered candidates for RAI in several guidelines, the significance of precise staging in locally advanced PTC might be limited. Moreover, it’s conceivable that many cases of micro-metastasis could be effectively managed by RAI treatment. In our study, RAI therapy was more frequently administered in the pCND group compared to the non-pCND group. This discrepancy likely reflects the upstaging effect (pN0 to pN1a) achieved through the pathologic identification of occult metastases by pCND. While it is possible that more frequent RAI administration could have influenced the favorable outcomes in the pCND group, it is noteworthy that the non-pCND group also demonstrated excellent regional control despite receiving RAI therapy less frequently. These observations may suggest that, in the context of cN0 T3/T4 PTC, current baseline management strategies might provide sufficient regional control even without the routine implementation of pCND. Consequently, the incremental benefit of performing pCND to identify occult metastases and potentially escalate RAI therapy might be relatively modest in this specific clinical setting.

Additionally, there are several studies that present opposing results to the notion that pCND lowers postoperative Tg levels. In a retrospective study, pCND did not impact the postablative sTg levels or the number of athyroglobulinemia cases [36]. In the study by Lang, et al. [21], preablative sTg levels and the presence of athyroglobulinemia were observed to be higher in the pCND group, but these differences lost statistical significance 6 months after surgery. These findings suggest that adding pCND may have little impact on the microscopic residual disease burden.

Over time, the role of pCND has been re-estimated, as reflected in several guidelines, which no longer recommend pCND for low-risk PTC. According to the 2009 ATA guidelines, pCND could be performed in patients with papillary carcinoma. However, the updated 2025 ATA guidelines specify that pCND should exclusively be contemplated for cases who have cNo patients with advanced primary tumors (T3 or T4), or if the information will be used to plan further treatment [5]. This recommendation is graded as conditional and supported by low-certainty evidence.

The majority of participants in past studies focused on T1/T2, and due to changes in AJCC staging system, many patients who were previously classified as pT3 have now been reclassified as pT1/T2. This transition has resulted in limited research on advanced PTC. We used the T category according to the 8th AJCC staging system to investigate the role of pCND in T3/T4 PTC.

In cN0 T3/T4 PTC, the occult metastasis rate was not statistically different from T1/T2 group, although it is slightly higher (37.1% in T1/T2 and 43.4% in T3/T4 PTC). Recurrence rates and 5-year RFS according to the pCND were also not significantly different between the two groups. Additionally, our study found no significant differences in stimulated-Tg levels, which are used as markers for predicting recurrence or persistent disease.

To better understand these findings, we evaluated the size of the metastatic deposits. Although our study was limited by the retrospective availability of LN size data for all cases, the findings from the available subset align with recent consensus: all confirmed recurrences in the pCND group occurred in patients with macrometastases (>2 mm). This observation suggests that while pCND may identify occult metastases, its impact on reducing recurrence might be clinically negligible for patients with only micrometastases.

Another striking finding in our study was that all regional recurrences occurred exclusively in the lateral neck compartment, with no central compartment recurrences observed even in the non-pCND group. The absence of central recurrence in patients who did not undergo formal pCND can be attributed to several factors. First, postoperative RAI therapy, which was administered to a substantial portion (59.8%) of the non-pCND group, may have effectively eradicated any occult microscopic residual disease in the central basin [5,23]. Second, the biological behavior of some locally advanced PTCs may favor “skip metastases,” where tumor cells drain directly into the lateral neck nodes, bypassing the central compartment entirely [37,38]. Lastly, a meticulous TT inherently involves the incidental removal of pericapsular soft tissue and microscopic lymphatic networks. Even if distinct LNs are not pathologically identified or harvested, the clearance of this adjacent tissue may provide sufficient local disease control in true cN0 patients. Consequently, our findings suggest that for cN0 T3/T4 PTC patients, diligent preoperative and postoperative surveillance of the lateral neck may yield greater clinical benefit than performing routine prophylactic central clearance.

Therefore, the benefits of recommending pCND in T3/T4 PTC in several guidelines are questionable. Based on the results of our study, pCND may have little effect on recurrence even in T3/T4 PTC with increased hypoparathyroidism as a complication. So, pCND may not be routinely performed in T3/T4 PTC.

There were some limitations to this study. Firstly, the limited number of patients who underwent bilateral pCND hinders an accurate evaluation of the occult contralateral CLNM rate and the recurrence rates according to pCND. Secondly, RAI therapy was significantly more common in the group that underwent pCND, which could confound the evaluation of recurrence rates by potentially influencing outcomes. Furthermore, although T3 and T4 tumors may have different prognostic characteristics, we had to combine them for the survival analysis because the extremely low number of recurrence events in each separate subgroup limited the statistical power of the models.

Nevertheless, to our knowledge, our study is the largest single- center cohort study utilizing the AJCC 8th Edition staging system focused on the role of pCND in cN0 T3/T4 PTC. Notably, by excluding microscopic ETE from the T3 category in accordance with this modern staging context, our research provides a more precise evaluation of true locally advanced disease compared to previous studies. Further investigation through large-scale prospective research will be important to confirm our findings and provide more robust evidence regarding the necessity of pCND in T3/T4 PTC.

In conclusion, occult metastasis rate was not different between the T3/T4 group than T1/T2 group in PTC. There were also no significant differences in recurrence rate and 5-year RFS according to pCND in cN0 T3/T4 PTC cases. Furthermore, there was no recurrence in the central LN and all of recurrence occurred at the lateral lymph compartment. So, the impact of pCND on recurrence in T3/T4 PTC patients might be limited with increased complication. Routine pCND might be unnecessary even in cN0 T3/T4 PTC.

Article information Continued