Macroscopic and multiple metastases in sentinel lymph node biopsy are respectively associated with poor prognosis in early oral cancer

A multicenter, randomized controlled phase III trial was conducted on sentinel lymph node biopsy (SLNB) and elective neck dissection for T1 (depth of invasion ≥ 4 mm)–T2N0M0 oral cavity squamous cell carcinoma. This study identified factors associated with poor prognosis in patients who underwent SLNB based on a subgroup analysis of this trial. We analyzed 418 sentinel lymph nodes (SLNs) from 132 patients who underwent SLNB. The metastatic SLNs were classified into three categories based on size-isolated tumor cells: < 0.2 mm, micrometastasis: ≥ 0.2 mm and < 2 mm, and macrometastasis: ≥ 2 mm. Three groups were formed based on the number of metastatic SLNs: no metastasis, 1 metastatic node, and ≥ 2 metastatic nodes. The size and number of metastatic SLNs on survival were evaluated using Cox proportional hazard models. Patients with macrometastasis and ≥ 2 metastatic SLNs had worse overall survival (OS) and disease-free survival (DFS) after adjustment for potential confounders (HR for OS: macrometastasis, 4.85; 95% CI 1.34–17.60; ≥ 2 metastatic SLN, 3.63; 95% CI 1.02–12.89; HR for DFS: macrometastasis, 2.94; 95% CI 1.16–7.44; ≥ 2 metastatic SLN, 2.97; 95% CI 1.18–7.51). In patients who underwent SLNB, a poorer prognosis was associated with macrometastasis or having ≥ 2 metastatic SLNs.


Introduction
Oral cavity squamous cell carcinoma (OCSCC) is the most common type of oral cancer. Although it may be curable after early detection and treatment, there is no international consensus addressing cervical lymph nodes (LNs) in T1-T2N0M0 cases. Therefore, a combination of three strategies is currently used: follow-up, elective neck dissection (END), and sentinel LN biopsy (SLNB). In a large randomized controlled trial, D'Cruz et al. found that overall survival (OS) and disease-free survival (DFS) were significantly increased in the END group compared to those in the follow-up group [1]. Further, in a multicenter, randomized controlled phase III trial, we found that the 3-year OS and 3-year DFS of an SLNB group were not inferior to those of an END group [2]. In a letter to the editor [3], to which we published a response [4], Kaul et al. raised the issue of the predictive value of the presence of isolated tumor cells (ITC) and other pathological findings in the survival data and recurrence rates in this patient population. We considered that more detailed histopathological and prognostic studies were needed, which led to the analyses performed here. This study aimed to perform a subgroup analysis of our multicenter, randomized controlled phase III trial [2] to identify factors associated with poor prognosis from the histopathological characteristics of 132 patients in the SLNB group.

Patients
Patients were enrolled from November 2011 to January 2016. The inclusion criteria were T1-T2N0M0 OCSCC (UICC TNM classification 7th edition), no prior treatment, written consent, and ≥ 18 years old. The exclusion criteria were patients with T1 < 4 mm depth of invasion (DOI), history of radiation therapy to the neck, currently pregnant/breastfeeding or planning to conceive, or otherwise deemed ineligible by a physician. Computed tomography (CT) was used as the initial mode of examination for diagnostic imaging of LN metastasis, which was supplemented by ultrasound (US) examination when necessary. Details of the diagnostic criteria of N0 have been described in the appendix of the previous article [2]. In the phase III trial, the patients were randomly assigned to an SLNB group (n = 134) or an END group (n = 137). Two patients in the SLNB group did not undergo SLNB. This subgroup analysis targeted the 132 patients of the SLNB group who underwent SLNB (Table 1).

Sentinel lymph node (SLN) identification
Details are given in phase II [5] and phase III trial protocols [2]. 99m Tc-phytate was used as the radiopharmaceutical. The day before surgery, a total of 1 mL 74 MBq (2 mCi) 99m Tc-phytate was evenly administered to four sites in the mucosa around the tumor using a 27G needle. Lymphoscintigraphy was performed 1-2 h after administration. Whenever possible, single-photon emission computed tomography (SPECT) was performed to create fusion SPECT and CT images. A gamma probe was used to search for SLNs while referencing the lymphoscintigraphy results on the day of surgery.

Histopathologic diagnosis of SLNs
Details are given in phase II [5] and phase III trial protocols [2]. The SLN analysis was performed in two stages. First, 2-mm blocks were created as rapid intraoperative frozen specimens (FS). These blocks were embedded in paraffin, and two 4-μm slices were prepared from each of the block's cut surfaces, which underwent hematoxylin-eosin (HE) and cytokeratin (CK) staining. CK immunostaining was performed using an anti-CK primary antibody (AE1/3; Signet Laboratories, Dedham, MA) and streptavidin-biotin labeling. Metastatic LN size was classified into three groups [6]: ITC (size < 0.2 mm), micrometastasis (size ≥ 0.2 mm and < 2 mm), and macrometastasis (size ≥ 2 mm). In this study, ITC was considered positive for metastasis to avoid disadvantaging the subjects.

Rapid intraoperative diagnosis of FS of SLNs and postoperative response
SLN detection, SLN resection, and intraoperative pathological diagnosis of FS were performed. In patients with metastatic SLNs, therapeutic neck dissection (ND) was performed in one stage, either level I-IV or I-V. Details of the neck dissection procedure have been described in the protocol of the previous article [2]. If metastases are found intraoperatively, in the level I-IV, ipsilateral neck dissection (level I-IV) is done, and in the level V, ipsilateral neck dissection (level I-V) is done according to the protocol. If there were no metastatic SLNs in the intraoperative pathological diagnosis of the FS, only SLNB was performed. Moreover, supraomohyoid neck dissection (level I-III) was performed in cases requiring pull-through resection of the primary tumor. If the intraoperative pathological diagnosis on the FS detected metastatic SLNs, either level I-IV or I-V ND was also performed in the cases requiring pull-through resection of the primary lesion. When the intraoperative pathological diagnosis of the FS indicated no metastasis and postoperative HE or CK staining did, ND was performed in two stages within 6 weeks of the initial surgery.

Postoperative adjuvant treatment
In patients with extranodal invasion of metastatic LNs, chemoradiotherapy (radiotherapy) was given as adjuvant therapy within 6 weeks of surgery. Whether to administer chemotherapy was left to the discretion of each institution.
In patients with positive resection margins, reoperation, chemoradiotherapy, or radiotherapy was performed at the institution's discretion.

Size and number of metastatic SLNs
FS and HE/CK staining were evaluated for the SLNB. We evaluated the largest metastatic SLN and the number of metastases in HE/CK staining. The largest metastatic SLNs were categorized into four groups and analyzed: no metastasis, ITC, micrometastasis, and macrometastasis. The number of metastases was classified into three groups: no metastasis, 1 metastasis, and ≥ 2 metastases. The differences in the assessments of metastatic SLN size between FS and HE/CK staining were also investigated. Positive rates of ITC, micrometastasis, and macrometastasis were compared between FS and HE/CK staining.

Statistical analysis
The primary endpoint of this study was OS, defined as the interval between the date of SLNB and the date of death from any cause or last follow-up date. The secondary endpoint was DFS (defined as the interval between the date of SLNB and the date of diagnosis of recurrence). Patients who were not followed up were treated as censored. Kaplan-Meier product-limit method and univariate and multivariate Cox proportional hazards models were performed to evaluate the prognostic impact of the size and number of metastatic SLNs. The measure of association in this study was hazard ratio (HR) with a 95% confidence interval (CI). Confounding variables considered in multivariate analyses were age (≤ 63 vs. > 63 years), sex (male vs. female), primary site (tongue vs. other), resection method (transoral vs. pull-through), and pathological T classification (CIS or 1 vs. 2 vs. 3 or 4a). All statistical analyses were performed using STATA version 16 (Stata Corp., College Station, TX, USA). All tests were two-sided, and values of P < 0.05 were considered statistically significant.

Study design and ethics
This study was a subgroup analysis of a multicenter, randomized controlled phase III trial [2] Figure 1 shows the courses of treatment for the 134 patients in the SLNB group in the multicenter, randomized controlled phase III trial [2]. This subgroup analysis included 418 SLNs obtained from 132 patients; the median number of SLNs per patient was 3 (range 1-8).  were obtained from evaluating up to 3 or 8 SLNs when discriminating between SLN-positive and SLN-negative cases. Table 3 shows the number of positive metastases from HE/ CK staining of SLNs. Among the 416 SLNs stained with HE/CK, 67 were evaluated as metastatic. Most of them were located on the affected side level II (30 SLNs). Four metastatic LNs were found on the healthy side. The positive metastasis rate in HE/CK staining was 16.1%.   Table 5 shows the number and rate of positive SLNs in FS and HE/CK staining for ITC, micrometastasis, and macrometastasis. The detection rate of ITC was significantly higher with HE/CK staining than with FS (p = 0.020). However, the detection rate of macrometastasis was almost the same for HE/CK staining and FS.

Discussion
The first SLNB for oral cancer was reported in 1996 [7]. SLNB aims to reduce complications from unnecessary ND by identifying cases in which this procedure can be omitted. A review by de Bree et al. [8] described studies demonstrating the efficacy of SLNB, which included fewer complications, less cervical, and shoulder dysfunction, lower costs, and shorter hospital stays in SLNB groups compared to those in the END groups. They concluded that SLNB is a better option than END for OCSCC, excluding floor of mouth cancer, which is prone to the "shine-through" phenomenon. In our phase III trial, postoperative neck function was better in the SLNB group [2]. The present subgroup analysis showed that having larger macrometastases and ≥ 2 metastases are associated with worse prognosis after adjustment for potential confounders. Considering this, we suggest that the case criteria for SLNB can be further defined. A relationship between ITC and lower survival rates has been reported for breast cancer [9,10]. During the planning of this study, the clinical significance of ITC in head and neck cancer was unclear. An association between ITC and prognosis was later reported in head and neck cancer as well [11][12][13]. Broglie et al. [11] found that 38% of 111 patients with early oral and oropharyngeal squamous cell carcinoma were positive for SLN metastasis, and 20% had ITC. Moreover, they reported that the disease-specific survival rate was significantly lower in the ITC-positive group than in the SLN-negative group. The SENT trial [12] reported statistically significant differences in survival between ITC, micrometastasis, and macrometastasis in early-stage OCSCC. Pedersen et al. [13] also reported that in OCSCC, disease-specific survival was lower in patients with ITC or micrometastasis than in SLN-negative cases. These reports [11][12][13] suggest that in patients with OCSCC, OS differs between patients with SLN metastasis negative, ITC, micrometastases, and macrometastases, although the magnitude of the differences varied between studies. In our analysis, OS and DFS studies did not reveal significant differences, although a trend was observed. Moreover, the multivariate Cox regression analysis showed a significantly poorer prognosis in the macrometastasis group than in the no-metastasis group for both OS and DFS. However, there was no difference in prognosis between the ITC and no-metastasis groups. The French Senti-MERORL trial [14] is a multicenter, randomized, open-label prospective equivalence study on SLNB vs. END. This trial showed that patients with ITCs had better OS than patients with micro and macro metastasis. A global consensus has not been reached on the relationship between ITC and prognosis. Therefore, further research is needed on the relationship between ITC and survival rates. Meanwhile, for the relationship between the number of metastases and survival rate, we compared 3 groups-no metastases, 1 metastasis, ≥ 2 metastases-and found that ≥ 2 metastases was significantly associated with worse DFS, which is similar to the results of the SENT trial [12]. Thus, OS and DFS were significantly poorer in patients with ≥ 2 metastases in both the univariate and multivariate analyses.
Chone et al. [15] reported that it is useful to add immunostaining to the assessments of SLNs. In their study, CK staining was performed on negative SLNs in HE staining, which additionally identified metastatic SLNs in 3.8% of the samples. They concluded that CK and other types of immunostaining are important tools for reducing false-negative results. In the present study, 390 SLNs underwent CK staining. The positive rate of ITC was significantly higher with HE/CK staining than with FS (p = 0.020). Consequently, ITC should be evaluated perioperatively with FS and postoperatively with HE/CK staining. In contrast, the positive rate for macrometastasis was similar between FS and HE/CK staining (8.1% and 8.2%, respectively). This indicates that for relatively large lesions (≥ 2 mm), which were evaluated as negative during surgery, the probability of a diagnosis changing to metastasis in postoperative HE/CK staining is low. Based on this, the clinical question should be, "Should additional ND be performed when ITC is positive in postoperative HE/CK staining?" In the French Senti-MERORL trial [14], neck node and locoregional recurrence rates were not different in the ITC group compared to those of the no-metastasis group. Therefore, it suggested that ITC does not seem to require ND. Den Toom et al. [16] reported that metastases to non-SLN were found in 31% of SLN-positive cases. Metastasis to non-SLN was observed in 13% of ITC, 20% of micrometastasis, and 40% of macrometastasis. Their report concludes that it is important to classify ITC, micrometastasis, and macrometastasis. In the future, a randomized controlled trial with an additional ND group and a follow-up group should be performed to examine patients with positive ITC identified after surgery.
This subgroup analysis was based on data from a multicenter, randomized controlled phase III trial of 16 institutions in Japan, which provides a very high level of evidence. However, only 45 patients were SLN positive, and ITC was only found in 2.4% of the total number of SLNs, even when using HE/CK staining, which represents an insufficient number of cases to provide reliable evidence. The 2014 revision to the NCCN guidelines added SLNB to diagnose stage I/ II oral cancer [17]. In Japan, SLNB is covered by health insurance for breast cancer and malignant melanoma. SLNB may be incorporated into standard treatment regimens once more evidence of early-stage oral cancer becomes available worldwide.
In conclusion, poor prognosis is a factor in patients with OCSCC receiving SLNB, including the presence of macrometastasis or having more than two metastases. The assessment of ITC in these patients requires assessing perioperative FS and performing postoperative HE/CK staining.