Koos Boeve

177 General discussion and future perspectives GENERAL DISCUSSION Current evidence and consensus for neck staging in early stage (cT1- 2N0) OSCC using the SLNB procedure A major problem in early stage (cT1-2N0) oral squamous cell carcinomas (OSCC) is that eventually23-37%of thesepatients arediagnosedwithoccultmetastases [1]. Conventionally, neck staging was done with an elective neck dissection (END) or a watchful waiting strategy in these patients [2]. In 1994, Weiss et al. proposed that if the risk of occult metastasis was more than 20%, an END was recommended over a watchful waiting [2,3]. That 20% risk was mainly based on T status and anatomical location [4,5]. After multiple studies showed a relationship between tumour infiltration depth and the risk of occult metastasis, tumour infiltration depth with a 4 mm cut-off was added for neck strategy selection [4]. In 2015 a prospective randomized controlled trial analysed 255 watchful waiting and 245 elective neck dissection cases and found a higher five year overall survival and disease free survival for the END (80% and 70% respectively) compared to watchful waiting (68% and 46% respectively) [6] and concluded that END was superior over watchful waiting in early stage OSCC. A major consequence of subjecting all patients to an END is that in the majority (~75%) of the early stage OSCC cases no neck metastasis will be present, while patients are at risk for developing surgery induced morbidities such as shoulder dysfunction [2]. A new era with less invasive neck staging in early stage OSCC started with the introduction of the sentinel lymph node biopsy (SLNB) procedure. Although the SLNB is still an invasive procedure, it is minor surgery compared to the END, which is reflected by a smaller incision (48 mm versus 92 mm [7]), shorter hospital stay (1 versus 3 days [8]) and lower complication rates. In a study with 33 patients staged by SLNB and 29 by END, all 15 postoperative complications appeared in the END group (bleeding n = 5, nerve injury = 8, infection n = 1 and tracheotomy n = 1) and no complications were seen in the SLNB group [7]. Two other studies reported better postoperative shoulder functions and also a smaller scar: 84.2 mm and 73.9 mm for the SLNB group, compared to a scar of 183.3 mm and 171.5 mm in the END group [9,10]. Although the less invasive character is important, correct staging of the neck is the main reason for the use of the SLNB procedure. A meta-analysis using 66 studies in 2017 reported a high pooled sensitivity of 87% and pooled negative predictive value (NPV) of 94% for the SLNB procedure in detecting occult metastasis in early stage OSCC [1]. Limitation of that meta-analysis was the heterogeneity in study protocols with differences in experience of the surgeons, reference standard for the negative SLNB (clinical follow- up or END), preoperative imaging procedures (with or without Single Photon Emission Computed Tomography (SPECT)-CT) and pathological assessment (with or without step- serial-sectioning and additional keratin immunohistochemistry (IHC)) [1]. In chapter 3 we