There still remain some controversies surrounding the extent of neck dissection in PTC patients with lateral neck LN metastasis. We found that preoperative US and FNA could accurately predict the distribution of metastatic lymph nodes in the lateral neck in only half of patients.
In the past, the extent of surgery is generally based on the frequency and pattern of abnormal lymphatic spread. Similar to other studies,17-19 we also demonstrated that level Ⅲ and level Ⅳ are the most commonly involved regions. A systematic review of 18 unique studies found that levels IIA, IIB, III, IV, and V had metastatic disease involvement in 53%, 16%, 71%, 66%, and 25% of LNDs, respectively.20 Most of the positive LNs in level V are located in level VB. Based on the low incidence of metastasis, it is widely accepted that routine dissection of levels ⅡB and VB is unnecessary in all PTC patients, particularly in the early stage of lateral metastases. Koo et al.12 reviewed the pathological records of 76 N1b PTC patients. The metastasis rate of level IIB was 11.8% in therapeutic LNDs. Multivariate analysis showed that all lateral neck involvement (levels IIA, III, and IV) was an independent predictive factor of metastasis in level IIB. Similarly, Lim et al.18 studied 70 patients with lateral neck metastasis from PTC and found a 16% occult metastasis at level V; simultaneous level IIA-IV lymphatic metastases were associated with level V metastasis. Our study confirmed that multilevel metastases were a predictor of level ⅡB and ⅤB involvement. More importantly, we could obtain the pathological information preoperatively through multilevel FNA to determine whether these regions should be removed.
US-guided FNA is the most accurate and cost-effective method for evaluating cervical LNs.4 The sensitivity of cytological analysis of FNA samples is approximately 75-85%, with a rate of false-negative results of 6-8%.21,22 To improve the diagnostic performance of FNA-C, Pacini et al.21 suggested measuring Tg in the needle washout fluid in 1992. In his study, the sensitivity of FNA-Tg was 100%. However, the diagnostic FNA-Tg threshold is very difficult to determine.23 Based on our study, the Tg cutoff value for the detection of LN metastases was 1.0 ng/dl, which was consistent with our previous study24 and those of several others.25-27
According to our study, US-guided biopsy is not sufficiently reliable to locate the exact distribution of metastatic LNs in the lateral neck. The possible reason for the discrepancies is that we used metastasis at a “level” instead of metastasis in a “node” as the golden standard. If occult metastatic LNs that could not be detected by US were in the same level as with an FNA-negative LN, the result would be false-negative, hence the decrease in the sensitivity and NPV. It is evident that level Ⅱ LNs in our cohort had a lower accuracy than other levels. This is probably because level Ⅱ LNs are larger in volume, rendering them more difficult to distinguish from abnormal LNs under US. Subclinical occult lateral neck lymph node metastasis was present in 20-69% of patients with PTC with stage N0 disease.28 In patients with unilateral N1b PTC, contralateral lateral LN involvement could be found in about one-third of patients, despite the preoperative US being negative.29 In a study by Noda et al.,30 more than 50% of the involved LNs found in the lateral compartment were microscopic LN metastases. However, it is unclear whether these microscopic metastases would progress to clinically significant disease if not included in the dissection. A prospective controlled study with long-term follow-up is needed to answer this question.
Another limitation of this study was the false-positive levels in the cohort, which may be due to the inaccurate separation of specimen into different levels. In particular, preoperatively positive LNs at level Ⅱ were incorrectly assigned to level Ⅲ in histopathology. Despite the use of anatomical stitch marks, LNs near or along the division lines were sometimes packed into adjacent levels, because of body posture changes.
Despite the limited accuracy of multilevel cervical FNA, preoperative mapping of abnormal LNs may predict the patterns of metastasis and help determine the appropriate surgical strategy. For patients with one or two metastatic levels diagnosed prior surgery, prophylactic ⅡB and VB dissection may be unnecessary, even for positive LNs located at level ⅡA or level Ⅳ. However, when three levels are involved, a more extensive CND, including levels ⅡB and VB, is strongly recommended.