Elective neck dissection and its extent in salivary gland cancers: A dilemma

Abstract

Salivary gland cancers (SGC) represent a heterogenous and rare group of cancers. The necessity of a neck dissection in a clinically node negative neck (cN0) for SGC has been a topic of long-standing debate, and there is no consensus whether in such a scenario a neck dissection should be performed and if performed then what should be the extent of dissection. The article by Westergaard–Nielsen titled “Surgical treatment of the neck in patients with salivary gland carcinoma” attempts to simplify this debate. We have read this article with a keen interest as this is a common clinical scenario faced by most clinicians. Patients with SGC are evaluated for disease extent with various imaging modalities such as an ultrasonography (USG), a contrast-enhanced computed tomography scan (CECT), a magnetic resonance imaging scan (MRI) and occasionally with a positron emission tomography computed tomography scan (PET-CT). A meta-analysis by Liao et al compared the various imaging modalities for evaluation of neck lymph node metastasis in patients with cN0 head and neck cancer. This study showed a pooled specificity of 93% (87%–97%), 81% (64%–91%), 87% (77%–93%) and 78% (71%–83%) for CT, MRI, PET, and US, respectively. Similarly, Chaukar et al in their paper demonstrated a high accuracy and specificity for both USG and CECT scans for detection of nodal metastasis in clinically node negative neck. The accuracy and specificity of the USG and CECT were higher for the Neck Nodal stations of Level II and Level III. Although these data were primarily given for oral cancers (OC) with a cN0, it can be extrapolated to SGC as nodal metastases from SGC and OC are seen at similar nodal stations.3–5 Therefore, it is unusual that the authors reported a 38% falsepositive rate in necks, deemed clinically and radiologically positive (cN+). Traditionally an elective neck dissection (END) is done in cN0 SGC in the presence of higher grade tumor or in advanced stage tumors (T3–T4). In the primary article, 324 patients did not undergo a neck dissection, a significantly high number of patients in this cohort had either advanced tumors (n = 81) or high-grade histology (n = 47). We believe that if these patients had undergone an appropriate END as indicated, the rate of occult metastasis for the study would vary significantly. In the discussion section, the authors note that occult metastasis from parotid carcinomas (PC) is generally seen to involve the Level II and Level III nodal stations and recommend a Level II and Level III neck node dissection for PC. Stodulski et al in a retrospective analysis of 66 patients with PC noted 20% and 16% occult nodal metastasis at Level I and Level V, respectively. Other authors such as Chisholm et al have noted lymph node metastasis from PC at Level Ib, Level IV and Level V in addition to the Level II and Level III nodal stations, with a significant number of patients also demonstrating skip metastasis to Level V. Therefore, with the heterogeneity of nodal involvement from PC as seen in the various articles, a Level II/III nodal clearance alone may not suffice in all cases. It would be prudent to consider additional adjuncts such as frozen section (FS) or sentinel node biopsy (SNB) to confirm nodal metastasis and subsequently modify the extent of nodal dissection. It is noteworthy that in a study on 198 patients who underwent SNB for PC by Pan et al, 33% of occult nodal metastasis were detected with a SNB, and these patients were treated with a comprehensive neck dissection. In addition, FS has been found to be more specific and accurate as compared to a preoperative fine needle aspiration cytology (FNAC) to provide a definitive diagnosis (with regard to the grade of tumor) in SGC, which would accurately guide a surgeon about the extent of the surgical procedure. FS has shown an accuracy of 98% for a diagnosis in SGC, whereas FNAC has relatively lesser accuracy for malignancies and their grade. With variability in its accuracy, it is difficult to bank upon grading on FNAC to decide about neck dissection. As of now, there are no standard treatment protocols for the treatment of cN0 SGCs. Most of the literature on this subject is retrospective in nature. Given the heterogeneity of these cancers, a randomized controlled trial which compares END with Observation for cN0 SGC may be challenging. In the meanwhile, we will continue to bank on evidence generated by good-quality retrospective and prospective studies. Received: 7 May 2021 Accepted: 3 June 2021