Sentinel lymph node biopsy in cT1-2N0 minor salivary gland cancer in oral cavity Authors: Peng Li, Xu Zhang, Qigen Fang, Wei Du Published: 2024 ABSTRACT ObjectiveTo evaluate the efficacy of sentinel lymph node biopsy (SLNB) in cT1/2N0 minor salivary gland cancer (mSGC) located within the oral cavity.MethodsA retrospective analysis was conducted on patients diagnosed with cT1/2N0 oral mSGC, who were categorized into two groups based on neck management approaches. The impact of SLNB versus observation on regional control and overall survival was assessed using a Cox model.ResultsA total of 177 patients were included in the study, with 53 cases undergoing SLNB. All patients had at least one sentinel lymph node, with the majority having two sentinel lymph nodes. The sentinel lymph nodes were predominantly situated in level I, followed by level II. Four patients had positive sentinel lymph nodes, all of whom had primary tumors in the tongue or the floor of the mouth, and were classified as cT2 stage. This yielded a sensitivity and specificity of 100%, a false negative rate of 0%, and a negative predictive value of 100% for SLNB in predicting occult metastasis. In terms of regional control, SLNB exhibited a reduced hazard ratio of 0.90 (95% confidence interval: 0.64–0.96) compared to observation. However, SLNB did not confer a superior overall survival benefit compared to observation.ConclusionIn patients with cT1/2N0 oral mSGC, SLNB proved to be both technically feasible and oncologically safe. When contrasted with observation, SLNB was associated with enhanced regional control, particularly recommending its use for cases of cT2 mSGC arising from the tongue or the floor of the mouth. Introduction Salivary gland cancer is a relatively infrequent occurrence among head and neck malignancies, with approximately 5% of cases arising in the minor salivary glands of the oral cavity [1]. The primary course of treatment involves resection of the primary site, with or without neck dissection. According to multiple authoritative guidelines [2–5], therapeutic neck dissection is strongly recommended for patients with clinically positive (cN+) minor salivary gland cancer (mSGC). However, for those with clinically negative (cN0) mSGC, there exists a debate surrounding the optimal approach to neck management. It is noted that occult metastasis presents in 10–25% of cN0 mSGC patients [6, 7], a likelihood escalated in cases characterized by T3/T4 staging, high-grade presentations, perineural invasion (PNI), lymphovascular invasion (LVI), and specific histologic types [8–10]. In such scenarios, elective neck dissection (END) is advisable, offering enhanced regional control and a predicted improvement in overall survival. Nonetheless, the primary challenge lies in the lack of preoperative information regarding these characteristics, often necessitating a secondary operation for conclusive pathology. Consequently, routine END may result in overtreatment for over 70% of patients. Hence, more precise tools are requisite to assess lymph node (LN) status as a substitute for END and observation. Sentinel lymph node biopsy (SLNB) has witnessed widespread adoption in the management of oral cancer since its inception, offering a high level of technical feasibility and oncologic safety [11, 12]. Notably, SLNB has emerged as the preferred choice over END for clinically staged T1-2N0 oral squamous cell carcinoma [13]. However, its application in mSGC has received scant attention. In our knowledge, only five pertinent studies are available, with four focusing on parotid cancer [14–17], and the remaining study presenting SLNB through a singular case report [18], affirming its reliability. Therefore, our objective is to assess the utility of SLNB in cT1/2N0 mSGC located in the oral cavity, specifically delineating the location and quantity of sentinel LNs, and elucidating the long-term oncologic outcomes in patients managed through SLNB. Methods Ethical approval This study received approval from the Henan Cancer Hospital Institutional Research Committee, and all patients provided written informed consent for their involvement in medical research prior to commencing treatment. The methods adhered to pertinent guidelines and regulations. Study design To address our aims, a retrospectively analysis of patients with surgically treated oral mSGC from January 2005 to December 2023 was performed. Enrolled patients had to meet specific criteria, including primary disease staging as cT1/2N0, and undergoing either observation or SLNB followed by END or not, with a requisite follow-up period of at least 2 years. Patients with a history of prior cancer were excluded, and their demographic, pathological, treatment, and follow-up information was documented. At our center, the selective use of SLNB for patients with cT1/2N0 oral mSGC began in January 2015, which was mostly determined based on the surgeon’s preference and experience, and information on these patients was prospectively collected but retrospectively analyzed. Subsequently, all enrolled patients were grouped based on whether they underwent SLNB. Study variable Clinical staging was determined as cT1/2 and cN0 according to the 8th AJCC system, and pathological grade was stratified into low, intermediate, and high categories [19]. LVI was considered positive if cancer cells were present within the lymphatics, while PNI was deemed positive if cancer cells had invaded a nerve. Additionally, extra-nodal extension was defined as the presence of cancer cells outside the lymph node capsule. The primary outcome variable comprised of 5-year regional control (RC) and 5-year overall survival (OS), with the time of RC calculated from the date of surgery to the date of the first neck recurrence or last follow-up, and the time of OS calculated from the date of surgery to the date of death or last follow-up. The secondary outcome variable included the location and number of sentinel LNs, as well as the sensitivity and specificity of SLNB. Treatment principle Patients underwent ultrasound and imaging examinations, including CT and/or MRI. For cT1-2 mSGC, neck management included observation and SLNB. If a metastatic sentinel lymph node was confirmed through pathology, dissection of levels I-III/IV was performed. Adjuvant radiation was recommended for cases with identified high-risk factors such as PNI, LVI, and more. Additionally, the presence of extra-nodal extension or positive margins warranted consideration of adjuvant chemotherapy (platinum- or paclitaxel-based) to intensify treatment and enhance efficacy. Routine follow-ups were scheduled every three to six months for the first two years, followed by six to 12-month intervals thereafter. SLNB procedure The SLNB procedure adhered to established protocols as previously outlined [11]. The day before surgery, submucosal injection of 99mTc-nanocolloid was carried out around the perimeter of the primary tumor site. Dynamic lymphoscintigraphy was performed in anterior and lateral views, along with single photon emission CT imaging. Gamma-probe guidance and methylene blue staining facilitated the precise identification of the sentinel lymph node on the skin surface, further validated through real-time assessment utilizing frozen section analysis. During the frozen section procedure, two consecutive slices of fixed LNs were procured from the maximum cross-section and subjected to hematoxylin and eosin staining, with or without the addition of immunohistochemistry. Thereafter, these lymph nodes were sent for postoperative pathological assessment. Following the processes of dehydration and fixation, the samples were again sectioned at the maximum cross-section to produce one slice designated for hematoxylin and eosin staining, accompanied by or without immunohistochemistry. Statistical analysis A multiple imputation approach was employed to address missing data patterns for pathologic grade, PNI, and LVI. Missing rates among the variables were 15.4% for pathologic grade, 10.5% for PNI, and 9.2% for LVI [20, 21]. Following this, the Chi-square test was utilized to compare clinicopathologic variables between the observation and SLNB groups. Additionally, a propensity score-matching method in a 1:1 ratio using the nearest neighbor approach was executed to minimize any differences in clinicopathologic factors between the two groups [22]. The impact of different neck management strategies on RC and OS was analyzed using univariate analysis and a Cox model, expressed as hazard ratios (HR) and corresponding 95% confidence intervals (CI). Descriptive methods were utilized to present the results for the secondary outcome variable. All statistical analyses were performed using the R 3.4.3 software package, with statistical significance defined as a p-value less than 0.05. Results Baseline data A total of 177 patients participated in the study, comprising 77 males and 100 females, with a mean age of 50 ± 10 years. The primary tumor sites were distributed as follows: palate (87 patients), buccal region (50 patients), and tongue or the floor of the mouth (TFOM) (40 patients). The most prevalent pathological type was mucoepidermoid carcinoma (n = 98), followed by adenoid cystic carcinoma (n = 58), in the remaining patients, there were six instances of ductal carcinoma, five instances of adenocarcinoma not otherwise specified, four instances of myoepithelial carcinoma, three instances of acinic cell carcinoma, two instances of pleomorphic low-grade adenocarcinoma, and one instance of epithelial-myoepithelial carcinoma. Clinical tumor stages were observed as cT1 in 87 patients and cT2 in 90 patients. Notably, PNI was present in 24 patients, and LVI in 19 patients. Histologic grading revealed low grade in 56 patients, intermediate grade in 80 patients, and high grade in 41 patients, with negative margins successfully achieved for all patients. Adjuvant radiation was administered to 45 patients, while no patients received adjuvant chemotherapy. SLNB was performed in 53 patients, primarily among those presenting with a cT2 tumor (Table 1, p = 0.047). Subsequently, the tumor stage variable was incorporated into propensity score matching. Recognizing the potential impact of adjuvant radiation in eliminating any unrecognized microscopic disease, patients with negative SLNB results who underwent adjuvant radiation were excluded from the analysis. This adjustment led to a final cohort comprising a total of 92 patients (46 in each group), with both groups demonstrating comparable distributions across all variables (Table 1). Subsequently, the rates of RC and OS were compared between these two groups. Table 1Baseline data of enrolled patientsVariableBefore propensity score-matchingAfter propensity score-matchingObservation(n = 124)SLNB*(n = 53) p Observation(n = 46)SLNB(n = 46) p Age <5060282524 ≥5064250.58821220.834Sex Male54231820 Female70300.98528260.672Site Palate60272024 Buccal34161413 TFOM^30100.7351290.661Pathology& MEC69292326 ACC40181616 Others1560.972750.776Tumor stage cT167201818 cT257330.04728281.000PNI% Yes18623 No106470.57044431.000LVI# Yes15422 No109490.43844441.000Grade Low/intermediate98383435 High26150.28912110.810Adjuvant radiation No90424242 Yes34110.351441.000* SLNB: sentinel lymph node biopsy^ TFOM: tongue and the floor of the mouth& MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma% PNI: perineural invasion# LVI: lymphovascular invasion Table 1Baseline data of enrolled patientsVariableBefore propensity score-matchingAfter propensity score-matchingObservation(n = 124)SLNB*(n = 53) p Observation(n = 46)SLNB(n = 46) p Age <5060282524 ≥5064250.58821220.834Sex Male54231820 Female70300.98528260.672Site Palate60272024 Buccal34161413 TFOM^30100.7351290.661Pathology& MEC69292326 ACC40181616 Others1560.972750.776Tumor stage cT167201818 cT257330.04728281.000PNI% Yes18623 No106470.57044431.000LVI# Yes15422 No109490.43844441.000Grade Low/intermediate98383435 High26150.28912110.810Adjuvant radiation No90424242 Yes34110.351441.000* SLNB: sentinel lymph node biopsy^ TFOM: tongue and the floor of the mouth& MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma% PNI: perineural invasion# LVI: lymphovascular invasion Baseline data of enrolled patients * SLNB: sentinel lymph node biopsy ^ TFOM: tongue and the floor of the mouth & MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma % PNI: perineural invasion # LVI: lymphovascular invasion SLNB Of the 53 patients who underwent SLNB, a sentinel LN was identified in all cases, yielding a detection rate of 100%. Specifically, one sentinel LN was detected in 17 patients, two sentinel LNs in 30 patients, and three sentinel LNs in 6 patients. Furthermore, in 50 patients, the sentinel LN was located in only one level, with level I involved in 29 patients, level II in 17 patients, and level III in 4 patients, with no involvement of level IV or V. In the remaining 3 patients, two had sentinel LNs locating in level I and II, and one patient had a sentinel LN locating in level II and III. Notably, positive sentinel LNs (micrometastasis) were found in 4 patients intraoperatively, all of whom had a primary site of TFOM, a cT2 stage, and level I involvement (Table 2). Moreover, in these 4 patients, metastatic focus was not noted in non-sentinel LNs. Importantly, no regional recurrence was observed in patients with negative SLNB during the follow-up period, reflecting a 100% sensitivity and specificity, a false negative rate of 0%, and a negative predictive value of 100% for SLNB in predicting occult metastasis. Table 2Detailed information of the patients with positive sentinel lymph nodesCaseNumber/location of sentinel LNSite$Pathology%Tumor stagePNI^LVI&Grade12/level I;TFOMMECcT2NoNoHigh21/level I;TFOMDCcT2YesNoHigh32/level I;TFOMMECcT2NoNoIntermediate42/level I; and IITFOMMECcT2NoNoIntermediate$ TFOM: tongue and the floor of the mouth% MEC: mucoepidermoid carcinoma, DC: duct carcinoma, SmCC: small cell carcinoma^ PNI: perineural invasion& LVI: lymphovascular invasion Table 2Detailed information of the patients with positive sentinel lymph nodesCaseNumber/location of sentinel LNSite$Pathology%Tumor stagePNI^LVI&Grade12/level I;TFOMMECcT2NoNoHigh21/level I;TFOMDCcT2YesNoHigh32/level I;TFOMMECcT2NoNoIntermediate42/level I; and IITFOMMECcT2NoNoIntermediate$ TFOM: tongue and the floor of the mouth% MEC: mucoepidermoid carcinoma, DC: duct carcinoma, SmCC: small cell carcinoma^ PNI: perineural invasion& LVI: lymphovascular invasion Detailed information of the patients with positive sentinel lymph nodes $ TFOM: tongue and the floor of the mouth % MEC: mucoepidermoid carcinoma, DC: duct carcinoma, SmCC: small cell carcinoma ^ PNI: perineural invasion & LVI: lymphovascular invasion Univariate analysis After a follow-up with median time of 4.6 (range: 0.9–15) years, variables of primary site, tumor stage, and pathological grade were significantly associated with RC and OS (all p < 0.05). When compared to observation, SLNB demonstrated an association with improved RC (Fig. 1, p = 0.014) rather than OS (Fig. 1, p = 0.884). Subsequently, these significant factors were included in a multivariable Cox model to confirm their independent relationships. Fig. 1Comparison of regional control (RC) and overall survival (OS) in patients treated by sentinel lymph node biopsy (SLNB) and observation Comparison of regional control (RC) and overall survival (OS) in patients treated by sentinel lymph node biopsy (SLNB) and observation Multivariable analysis In terms of RC (Table 3), SLNB exhibited a decreased HR of 0.90 (95% CI: 0.64–0.96) compared to observation, signifying a significant difference (p = 0.039). Other independent variables influencing RC included TFOM site (HR: 2.13, 95%CI: 1.18–6.92), cT2 stage (HR: 2.11, 95%CI: 1.22–5.48), and high grade (HR: 2.78, 95%CI: 1.25–8.98). Regarding OS (Table 4), predictors for worse OS were TFOM site (HR: 2.25, 95%CI: 1.28–8.19), cT2 stage (HR: 1.90, 95%CI: 1.19–4.73), and high grade (HR: 2.54, 95%CI: 1.32-8.00). Table 3Prognostic factors for regional control in patients with oral minor salivary gland cancerVariableUnivariateMultivariableHR [95%CI] p HR [95%CI] p Age <50 ≥501.86 [0.42–4.19]0.623Sex Male Female2.04 [0.59–6.34]0.587Site Palaterefref Buccal1.78 [0.72–4.99]0.2341.91 [0.66–6.86]0.321 TFOM^2.19 [1.27–7.14]0.0292.13 [1.18–6.92]0.016Pathology& MECref ACC1.04 [0.29–4.88]0.582 Others1.65 [0.39–6.45]0.590Tumor stage cT1refref cT21.82 [1.09–4.35]0.0342.11 [1.22–5.48]0.019PNI% Noref Yes2.17 [0.70–7.54]0.555LVI# Noref Yes1.90 [0.52–5.48]0.437Grade Low/intermediaterefref High3.33 [1.47–8.89]< 0.0012.78 [1.25–8.98]0.002Neck management Observationrefref SLNB*0.87 [0.53–0.97]0.0140.90 [0.64–0.96]0.039Adjuvant radiation Noref Yes0.99 [0.59–8.17]0.773^ TFOM: tongue and the floor of the mouth& MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma% PNI: perineural invasion# LVI: lymphovascular invasion* SLNB: sentinel lymph node biopsy Table 3Prognostic factors for regional control in patients with oral minor salivary gland cancerVariableUnivariateMultivariableHR [95%CI] p HR [95%CI] p Age <50 ≥501.86 [0.42–4.19]0.623Sex Male Female2.04 [0.59–6.34]0.587Site Palaterefref Buccal1.78 [0.72–4.99]0.2341.91 [0.66–6.86]0.321 TFOM^2.19 [1.27–7.14]0.0292.13 [1.18–6.92]0.016Pathology& MECref ACC1.04 [0.29–4.88]0.582 Others1.65 [0.39–6.45]0.590Tumor stage cT1refref cT21.82 [1.09–4.35]0.0342.11 [1.22–5.48]0.019PNI% Noref Yes2.17 [0.70–7.54]0.555LVI# Noref Yes1.90 [0.52–5.48]0.437Grade Low/intermediaterefref High3.33 [1.47–8.89]< 0.0012.78 [1.25–8.98]0.002Neck management Observationrefref SLNB*0.87 [0.53–0.97]0.0140.90 [0.64–0.96]0.039Adjuvant radiation Noref Yes0.99 [0.59–8.17]0.773^ TFOM: tongue and the floor of the mouth& MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma% PNI: perineural invasion# LVI: lymphovascular invasion* SLNB: sentinel lymph node biopsy Prognostic factors for regional control in patients with oral minor salivary gland cancer ^ TFOM: tongue and the floor of the mouth & MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma % PNI: perineural invasion # LVI: lymphovascular invasion * SLNB: sentinel lymph node biopsy Table 4Prognostic factors for overall survival in patients with oral minor salivary gland cancerVariableUnivariateMultivariableHR [95%CI] p HR [95%CI] p Age <50 ≥501.49 [0.50–4.44]0.632Sex Male Female1.19 [0.63–3.67]0.481Site Palaterefref Buccal1.55 [0.71–4.62]0.2781.82 [0.68–4.87]0.715 TFOM^2.28 [1.24–7.43]0.0102.25 [1.28–8.19]0.012Pathology& MECref ACC1.22 [0.61–3.86]0.639 Others1.81 [0.49–6.11]0.427Tumor stage cT1refref cT21.46 [1.19–4.16]0.0241.90 [1.19–4.73]0.011PNI% Noref Yes2.24 [0.75–6.38]0.490LVI# Noref Yes1.55 [0.63–5.29]0.571Grade Low/intermediaterefref High2.90 [1.41–6.88]0.0032.54 [1.32−8.00]0.002Neck management Observationref SLNB*0.87 [0.55–5.88]0.884Adjuvant radiation Noref Yes0.92 [0.64–5.60]0.334^ TFOM: tongue and the floor of the mouth& MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma% PNI: perineural invasion# LVI: lymphovascular invasion* SLNB: sentinel lymph node biopsy Table 4Prognostic factors for overall survival in patients with oral minor salivary gland cancerVariableUnivariateMultivariableHR [95%CI] p HR [95%CI] p Age <50 ≥501.49 [0.50–4.44]0.632Sex Male Female1.19 [0.63–3.67]0.481Site Palaterefref Buccal1.55 [0.71–4.62]0.2781.82 [0.68–4.87]0.715 TFOM^2.28 [1.24–7.43]0.0102.25 [1.28–8.19]0.012Pathology& MECref ACC1.22 [0.61–3.86]0.639 Others1.81 [0.49–6.11]0.427Tumor stage cT1refref cT21.46 [1.19–4.16]0.0241.90 [1.19–4.73]0.011PNI% Noref Yes2.24 [0.75–6.38]0.490LVI# Noref Yes1.55 [0.63–5.29]0.571Grade Low/intermediaterefref High2.90 [1.41–6.88]0.0032.54 [1.32−8.00]0.002Neck management Observationref SLNB*0.87 [0.55–5.88]0.884Adjuvant radiation Noref Yes0.92 [0.64–5.60]0.334^ TFOM: tongue and the floor of the mouth& MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma% PNI: perineural invasion# LVI: lymphovascular invasion* SLNB: sentinel lymph node biopsy Prognostic factors for overall survival in patients with oral minor salivary gland cancer ^ TFOM: tongue and the floor of the mouth & MEC: mucoepidermoid carcinoma, ACC: adenoid cystic carcinoma % PNI: perineural invasion # LVI: lymphovascular invasion * SLNB: sentinel lymph node biopsy Discussion Our principal discovery was the reliability of SLNB in detecting occult metastasis and significantly enhancing 5-year RC by reducing neck recurrence by 10% (95% CI: 4-36%) in cT1-2N0 mSGC located in the oral cavity. Although the improvement in OS was not notable, our findings underscore the efficacy of SLNB. Most notably, the majority of sentinel LNs were situated in level I;, closely followed by level II. This study potentially represents an exploratory investigation elucidating the application of SLNB in oral mSGC, boasting high feasibility and promising outcomes, thereby advocating for the preference of SLNB over observation in the management of oral mSGC. Squamous cell carcinoma was the predominant pathologic type in oral cancer, and a widely accepted practice involves obtaining a consensus for END in early stage oral squamous cell carcinoma, once confirmed via preoperative incision biopsy [23]. However, despite the relative rarity of oral mSGC, several authors had aimed to elucidate its oncologic outcomes. Gu et al. [6] conducted a retrospective analysis of 268 cN0 mSGC patients, revealing that occult metastasis occurred in 24.6% of the population, independently influenced by pathologic grade and tumor stage. Additionally, Brajkovic et al. [9] reported that their cohort of 88 mSGC patients achieved a cumulative 5-year OS rate of 82%, primarily determined by advanced tumor stage and high grade. Similar findings were also corroborated by a large-scale study conducted at the MSKCC center [24]. Nevertheless, these studies had raised concerns about the potential of END to yield both survival benefits and overtreatment. The primary dilemma remained the absence of accessible tools for detecting occult metastasis in cT1/2 mSGC. The use of SLNB had been widely adopted in oral malignancies, and its reliability had been well established [11–14]. However, its application in oral mSGC had been scarcely reported until our current study. To the best of our knowledge, SLNB was only cited in a single study through a case report, wherein SLNB revealed no occult metastasis in a 24-year-old male with adenocarcinoma of the minor salivary gland, thereby obviating the need for END. There was no evidence of recurrent disease after 4 years of follow-up, and this study initially highlighted the indication for such a procedure. Notably, our study demonstrated the successful implementation of SLNB in mSGC, achieving a detection rate of 100%. This notably high detection rate could be attributed to the fact that SLNB had been routinely performed at our cancer center since January 2015, affording us ample experience to surmount the learning curve. Our findings were consistent with other reports, which had found SLNB to be a reliable technique for identifying sentinel LNs in other head and neck cancers [25]. The location of the sentinel LN bared significance in determining the appropriate neck management strategy. Theoretically, in cases where the tumor was located in the oral cavity, cancer cells may disseminate to the submandibular LN via the buccal or mouth floor lymphatic system. On the other hand, when the primary tumor arises from TFOM, cancer cells might potentially migrated to the deep cervical LNs through either the parapharyngeal or retropharyngeal lymphatic systems. Nevertheless, the lymphatic drainage patterns of the sentinel LNs in mSGC had been rarely explored, and might be influenced by the primary site. The palate constituted the most common primary site in our study, yet the location of the sentinel LNs remains unclear. Boeve et al. [26] may be the first researchers to shed light on this matter, having successfully detected and harvested sentinel LNs at cervical levels I, II, or III in their cohort of 11 patients, mirroring our own observations. However, a disparity exists between their study and ours, as they reported detecting parapharyngeal sentinel LNs in two patients, which was not observed in our investigation. A recent publication [27] also observed that aberrant lymphatic drainage to the contralateral neck was more prevalent in maxillary malignancy, a phenomenon not evident in our current research. These differences could be ascribed to the fact that our study exclusively enrolled patients with cT1/2 tumors, whereas the majority of their subjects presented with more advanced stages of disease. In advanced cases, involvement of the soft palate may occur, with the parapharyngeal lymph node being a well-known metastatic site for oropharyngeal squamous cell carcinoma. Midline crossing also represents a risk factor for bilateral lymph node metastasis. Of particular interest, our observations revealed that positive sentinel LNs were exclusively detected in cases of cT2 mSGC originating from the TFOM, situated in level I;, and displaying a tendency towards high grade. This noteworthy finding could potentially be attributed to a similar mechanism observed in oral squamous cell carcinoma, where squamous cell carcinoma in the TFOM had a higher propensity for LN metastasis and a worse prognosis compared to tumors in other subsites [28]. Furthermore, our findings suggested that occult metastasis was more likely to manifest in cT2 TFOM mSGC cases. In such instances, SLNB may be warranted for patients displaying these defining characteristics. Conversely, for those without the aforementioned adverse features, close observation may be deemed an appropriate course of action. The unresolved question of whether SLNB conferred a survival advantage compared to observation in mSGC remained of paramount significance. Previous literature had compared the prognosis between SLNB and END or observation in head and neck cancer. A meta-analysis of six prospective studies [29] yielded comparable results for the two management strategies in terms of regional recurrence, 5-year disease-free survival, and 5-year OS. Furthermore, fewer adverse events were observed in the SLNB arm compared to the END arm in cN0 oral cancer. Another systematic review found no statistically significant difference between END and SLNB with regard to OS, disease-free survival, and nodal recurrence. Additionally, SLNB was determined to be non-inferior to END in terms of disease-specific survival [12]. Both studies affirmed the safety and precision of SLNB as a method for staging cN0 oral cancer patients, with outcomes that were on par with END in terms of survival and nodal recurrence. Regarding the comparison between SLNB and observation, our previous study, which enrolled 9804 cN0 high-risk head and neck cutaneous squamous cell carcinoma patients [30], revealed that SLNB was an independent predictor for improved disease-specific survival. In patients presenting with two or three high-risk factors, SLNB was associated with better disease-specific survival, but similar OS when compared to observation. However, in patients exhibiting four high-risk factors, SLNB demonstrated significantly improved disease-specific survival and OS compared to observation. These studies collectively confirmed that SLNB conferred a survival benefit compared to observation in head and neck squamous cell carcinoma, with SLNB exhibiting non-inferior prognosis to END. However, whether this phenomenon existed in oral mSGC remained unknown due to the significant differences in biological behavior between squamous cell carcinoma and adenocarcinoma. Our study was the first to observe that, compared with observation, SLNB enhanced regional control but did not alter the mortality risk in oral mSGC. This finding was entirely plausible for several reasons. Firstly, LN status served as a crucial prognostic factor, and delayed detection of occult metastasis decreased regional control. Secondly, all instances of neck recurrence were salvaged in the observation group, with the primary cause of mortality being distant metastasis rather than locoregional recurrence. Limitation in current study must be acknowledged, first, this was a retrospective study, there was inherent selective bias; second, our sample size was small, and it might decrease our statistic power; third, external validation was required before clinical application. In summary, for cT1/2N0 oral mSGC, SLNB demonstrated both adequate technical feasibility and oncologic safety, with the majority of sentinel LNs located at level I;. When compared with observation, SLNB was linked to improved regional control, it is strongly recommended for cases of cT2 mSGC originating from the tongue or the floor of the mouth.