Pathological Risk Factors for Occult Nodal Metastasis in Early-Stage Squamous Cell Carcinoma of the Oral Cavity

Abstract

Occult neck metastasis is the presence of metastasis in the cervical lymph nodes that cannot be radiologically or clinically identified. Presence of metastasis in any neck node can have a significant impact on overall survival of patients with oral squamous cell carcinoma (OSCC). Our aim was to analyze the correlation of various histopathological parameters with occult nodal metastasis in early-stage OSCC and to obtain an optimal DOI cut-off value for predicting its increased risk. We conducted a retrospective study on patients who reported to our institute with clinical stage I and II OSCC. The patients having well-differentiated and moderately differentiated OSCC were included. Association of various histopathological parameters with occult nodal metastasis was assessed using statistical analysis. A total of 102 patients of early-stage well-differentiated and moderately differentiated OSCC with clinically negative necks who underwent elective neck dissection at our institute from the year 2018 to 2023 were enrolled in the study. Depth of invasion (DOI), perineural invasion (PNI), worst pattern of invasion (WPOI), and grade of tumor differentiation were the histopathological parameters entered into the univariate regression analysis as predictive variables, and they were found to be predictors of occult nodal metastasis. An optimal DOI cut-off value of 5.5 mm was obtained for predicting the increase in the risk of occult nodal metastasis. DOI, PNI, WPOI, and grade of tumor differentiation are predictors of occult nodal metastasis. There is a need for searching methods for preoperative and intraoperative detection of all these histopathological factors so that unnecessary elective neck treatment can be avoided.

Introduction

Cancers of the lip and oral cavity are commonly found in Southern Asia (e.g., India and Sri Lanka) as well as Pacific Islands (Papua New Guinea, with the highest incidence rate worldwide in both sexes), and it is also the leading cause of cancer death among men in India and Sri Lanka [1].

Cervical lymph node metastasis represents an ominous prognostic factor in head and neck squamous cell carcinoma (HNSCC) [2]. It is a biologic phase in the pathogenesis of the disease with far greater significance than any local geographic size, extent, or duration of the primary tumor [3]. We need to decide whether a patient with a clinically negative neck and a known primary oral squamous cell carcinoma should undergo any elective neck treatment or not, or should the treatment be delayed until a node is detectable, although it is reasonable to perform elective neck treatment in those cases in which the patient is unavailable for regular follow-up [4, 5].

Due to increased risk of nodal metastasis, even in clinically negative necks, most patients with tumor staged as T2 or larger undergo some form of elective neck treatment. The drawback of this policy is that the patients who do not harbor metastasis are also subjected to the morbidity of unnecessary treatment [6]. Hence, there is a need to explore the risk of occult nodal metastasis in early-stage oral cancers with clinically N0 neck so that unnecessary neck dissections can be avoided. The eighth edition of the American Joint Committee on Cancer (AJCC) staging system has revised the rules on the staging of oral squamous cell carcinoma (OSCC), improving the prognostic accuracy [7]. However, AJCC system still needs improvement and there is a need to find new prognostic markers in order to better stratify OSCC patient [8].

Hence, we conducted a study to analyze the correlation of various histopathological factors with occult nodal metastasis in early-stage squamous cell carcinoma (SCC) of the oral cavity and to obtain an optimal cut-off value of DOI for increased risk of occult nodal metastasis.

Materials and Methods

A retrospective study was conducted at Rohilkand Cancer Institute, Bareilly, Uttar Pradesh, India. Patients presenting to our institute from the year 2018 to 2023 with a diagnosis of well-differentiated and moderately differentiated OSCC staged I and II with clinically negative necks were included. For assessment of cervical lymph node metastasis, clinical examination and radiological investigation (CT or MRI scan) were done. Cases of poorly differentiated squamous cell carcinoma were excluded because they are already known to have a markedly high metastatic rate. Age, gender, and subsites, and pathological factors such as depth of invasion(DOI), perineural invasion (PNI), lymphovascular invasion (LVI), worst pattern of invasion (WPOI), and nodal metastasis were recorded. Occult nodal metastasis was considered if detected histopathologically, but not clinically or radiologically.

DOI was measured from the level of the basement membrane of the closest adjacent normal mucosa and then a “plumb line” was drawn from this plane to the deepest point of tumor invasion, as per AJCC 8th edition [7]. The patient cohort was divided into three groups: DOI ≤ 5 mm (D1), 6 to 10 mm (D2), and > 10 mm (D3). PNI was identified as a nerve being surrounded or infiltrated by the tumor. LVI was identified as tumor invasion within arterial, venous, or lymphatic channels [9]. WPOI was classified according to Brandwein-Gensler et al. as follows [10]:

• Type I: Tumor invasion in a broad pushing manner.

• Type II: Tumor invasion with broad pushing “fingers,” or separate large tumor islands, with a stellate appearance. It can have high nuclear grade despite the low-grade pattern of invasion.

• Type III: Invasive islands of tumor greater than 15 cells per island.

• Type IV: Invasive tumor islands that are smaller than 15 cells per island. This includes single cell invasion. It also includes strands of tumor cells in a single-cell filing pattern, regardless of island size.

• Type V: 1 mm or greater of normal tissue between tumor satellite islands at interface.

WPOI Types I–III were grouped together as cohesive patterns, while WPOI Type IV and Type V were grouped as infiltrative and dispersive patterns, respectively.

Age, gender, tumor subsite, and the histopathological parameters including DOI, PNI, WPOI, and tumor differentiation grade were the predictor variables. The primary outcomes of the study were the correlation of various histopathological factors with occult nodal metastasis and to obtain an optimal cut-off value of DOI for increased risk of occult nodal metastasis. The secondary outcome was the correlation of age, gender, and subsite with occult nodal metastasis. Correlation of DOI with all other predictive variables was the tertiary outcome.

Statistical Analysis

Statistical analysis was performed using the IBM SPSS Statistics version 23 software. Univariate analysis using Pearson chi-square test was used to correlate predictive variables with occult nodal metastasis. This test was also used to correlate DOI with other predictive variables. A receiver operator curve (ROC) analysis was also applied to determine the optimal DOI cut-off value for predicting the increase in occult nodal metastasis, for all subsites combined. All tests were 2-sided and P-value < 0.05 was considered significant.

Results

A total of 102 patients were eligible for enrolment in our study and hence subjected to analysis. Out of these, 84.3% were males and 16% were females, with a mean age of 44 ± 12 years. Buccal mucosa was the most common subsite to be involved by SCC followed by oral tongue (30.4%), retromolar trigone (4%), lower lip (4%), lower alveolus (4%), palate (4%), upper alveolus (1%), and upper lip (1%). After END, pathologically proven occult nodal metastasis was seen in 26% of total patients, and in 25% and 36% cases of buccal mucosa and oral tongue, respectively.

Predictors of Occult Nodal Metastasis

Age (P = 0.391), gender (0.230), and subsite (0.650) had no significant association with occult nodal metastasis on univariate analysis (Table 1). However, using this test, DOI (P = 0.036), PNI (P = 0.007), and WPOI (P = 0.043) were found to be predictors of occult nodal metastasis. The risk was the highest in cases with WPOI Type V (62.5%), followed by Type IV (23%) and Type I–III (21.7%), respectively. Grade of differentiation (P = 0.016) also showed significant association with occult nodal metastasis, with greater risk in grade I OSCC (37%) as compared to grade II (16%); LVI had no significant association with occult nodal metastasis (P = 0.097), however, only 3% patients in our study had LVI present in their tumor specimen (Table 2).

Table 1.

Comparison of patient characteristics and tumor subsite between lymph node positive and negative cases

Parameter	Lymph node metastasis n = 26	No lymph node metastasis n = 76	P value
Age groups			0.391
20–29 years	28.6%	71.4%
30–39 years	26.5%	73.5%
40–49 years	36.7%	63.3%
50–59 years	17.6%	82.4%
60–69 years	10.0%	90.0%
70–79 years	0.0%	100.0%
Gender			0.230
Male	23.3%	76.7%
Female	37.5%	62.5%
Anatomic location			0.650
Buccal mucosa	24.5%	75.5%
Tongue	35.5%	64.5%
Retromolar trigone	0.0%	100.0%
Lower alveolus	25.0%	75.0%
Upper alveolus	0.0%	100.0%
Palate	0.0%	100.0%
Lower lip	25.0%	75.0%
Upper lip	0.0%	100.0%

Table 2.

Comparison of different histopathological parameters between lymph node positive and negative cases

Parameter	Lymph node metastasis n = 26	No lymph node metastasis n = 76	P value
DOI			0.036
≤ 5 mm(D1)	16.4%	83.6
> 5 and ≤ 10 mm (D2)	37.0%	63.0%
> 10 mm (D3)	0.0%	100.0%
PNI			0.007
Present	53.3%	47.0%
Absent	21.0%	79.3%
LVI			0.097
Present	67%	33.3%
Absent	24.2%	76%
WPOI			0.043
Type I–III	22.0%	78.0%
Type IV	23.0%	77.0%
Type V	63.0%	38.0%
Grade of tumor
Grade I	37.0%	63.0%	0.016
Grade II	16.1%	79.7%

Cut-Off Value of DOI

The DOI was ≤ 5 mm in 54% patients, > 5 and ≤ 10 mm in 44% patients, and > 10 mm in 2% patients and had significant association with occult nodal metastasis (P = 0.036). A ROC-analysis was performed to determine the optimal DOI cut-off value for predicting the increase in occult nodal metastasis per mm DOI. A cut-off value of 5.5 mm showed a sensitivity of 65% and specificity of 60% for subsites overall (AUC = 0.645, P = 0.028) (Fig. 1). This value was obtained from the ROC curve using intersection method and by choosing the sensitivity value that was closest to the AUC. This value also gave the maximum sum of sensitivity and specificity [11, 12]. Incidence of occult nodal metastasis in cases with DOI ≥ 6 mm was 39% (18/46) while that in cases with DOI < 6 mm was 15% (6/39).

Fig. 1.

ROC to find optimal DOI cut-off value for predicting increased risk of occult nodal metastasis in subsites overall

Association of DOI with Other Factors

DOI did not have any significant association with age (P = 0.409), gender (P = 0.750), and subsite (0.595). There was no significant association with PNI (P = 0.228), WPOI (P = 0.289), or tumor grade (0.157) as well.

Discussion

The general issue of END is less controversial than it was in the past [13]. Few studies have been conducted to compare patients treated by END vs. clinical observation alone and have given different conclusions. Vandenbrouck et al. conducted a randomized control trial for OSCC staged T1N0, T2N0, and T3N0 and recommended to delay neck dissection until a node is detectable, although it is reasonable to perform elective neck dissection in those cases in which the patient is unavailable for regular follow-up [5]. In our study, stage T3N0 was excluded because most of these cases underwent regional or free flap reconstruction that mandated neck dissection. In 2004, Duvvuri et al. in their nonrandomized retrospective review of squamous cell carcinoma of the oral cavity and oropharynx staged TXN0 found the neck control rate as 65% and 84% for the observation and the END, cohort respectively [14]. D’Cruz et al. in their prospective, randomized, controlled trial in patients with lateralized stage T1 or T2 oral squamous-cell carcinomas found that elective node dissection (ipsilateral neck dissection at the time of the primary surgery) resulted in higher rates of overall and disease-free survival than did the therapeutic node dissection (watchful waiting followed by neck dissection for nodal relapse) [15].

Most of the studies done in the past are exclusively on early-stage tongue cancers. Kligerman et al. in their prospective randomized study on tongue cancers staged T1T2N0 concluded that neck dissection is mandatory in the early stage of OSCC, because of better survival rates compared to resection alone, and in particular, patients with tumor thickness > 4 mm treated with resection plus elective supraomohyoid neck dissection had significant benefit on disease-free survival [16]. In one study where the squamous cell carcinomas of the tongue staged T1–T3N0 were treated by peroral partial glossectomy alone spread to the cervical lymph nodes occurred subsequently in 29.1, 42.9, and 77.0% of patients with T1, T2, and T3 lesions, respectively [17]. Fakih et al. also conducted a prospective, randomized trial in early-stage SCC of the oral tongue and concluded that routine concurrent elective neck dissection is not warranted in all patients and that the patients with tumor depth less than 4 mm should be observed at 2-month intervals, while those with a tumor depth of greater than 4 mm, should undergo an interval elective neck dissection 8 to 12 weeks after surgery of the primary tumor [18]. Yuen et al. conducted a prospective randomized study on stage I to II oral tongue carcinoma and recommended close follow-up for detection of early salvageable local or nodal recurrences irrespective to the choice of observation or prophylactic neck dissection for treatment of clinically negative neck. For those patients who can be followed-up closely, both elective neck dissection and observation have similar treatment results [19].

Since lymph node metastasis is an important prognostic factor for patients with OSCC, END is recommended based on the risk of occult metastasis. However, due to risks and morbidity associated with the procedure (e.g., damage of the accessory nerve, dysfunction of the trapezius muscle and shoulder impairment), reliable predictors are needed, for identification of patients with N0 necks, to reduce the number of unnecessary ENDs. DOI has been reported to be as predictor for occult nodal metastasis [20]. However, the existing data is heterogeneous, and there are huge differences between studies regarding the accuracy of the DOI measurement and its definition since it has been confused with the tumor thickness (TT) [21]. As stated by Moore et al. [22], DOI and TT are not the same, and a distinction has to be made. DOI means the extent of cancer growth into the tissue beneath an epithelial surface. In cases in which the epithelium is destroyed, some investigators prefer reconstructing a surface line and then measuring from this line. On the other hand, TT concerns the entire tumor mass [22, 23]. The eighth edition of the AJCC guidelines defines the DOI as the distance from the level of the basement membrane of the closest adjacent normal mucosa to the deepest point of tumor invasion [7]. Most of the studies demonstrate that an optimal cut-point for detection of occult nodal metastasis has also been done exclusively on early-stage oral tongue cancers. Muhammad et al. found a significant increase in the risk of occult nodal metastasis in early-stage oral tongue SCC with a DOI > 5 mm [21]. Tam et al. demonstrated an optimal DOI cut-off value of 7.25 mm for early-stage lateral oral tongue cancer [24]. Faisal et al. reported DOI > 10 mm to be significantly associated with decreased 5-year survival and increased risk of occult metastasis in early-stage oral tongue cancer. At the same time, they found significant risk of occult metastasis (> 20%) with DOI < 5 mm, thus recommending consideration of END even in cases having DOI < 5 mm in case of oral tongue [25]. Tongue tumors are associated with increased risk due to contraction of tongue muscles promoting entry of cancer cells into the lymphatics [26]. Kakuguchi et al. found DOI > 5 mm to be a predictor of occult nodal metastasis [27].

Apart from determining the cut-off value for oral tongue specifically, some studies have also determined it for the subsites overall. van Lanschot et al. found an optimal DOI cut-off value of 4 mm for subsites overall, and after subsite stratification, they found an optimal value of DOI ≥ 7 mm, ≥ 5 mm, and DOI ≥ 4 mm for buccal mucosa tumors, tongue tumors, and floor of the mouth tumors, respectively [20]. Aaboubout et al. reported a cut-off value of DOI as 4.3 mm for overall subsite, considering the 20% risk. For oral tongue, the cut-off value ranged between 3 and 4 mm [28]. Verma et al. found a cut-off depth of 6 mm in subsites overall [29]. Our study was also based on the cut-off value for overall subsites, and it was found to be 5.5 mm (P = 0.028). Thus, we found DOI as a predictor of occult nodal metastasis in early-stage OSCC. Kane et al. also found a significant association between DOI and occult nodal metastasis [30]. There are very few studies that did not report a significant association [9, 31, 32]. In a study about computer-assisted decision-analysis using data from retrospective series, it was found that a risk of occult neck metastases (in a palpatory-negative neck) above 20% was indicative of elective neck treatment, either surgery or irradiation, whereas observation was indicated when the risk was less than 20%, reserving the treatment for recurrence [33]. In some of the studies mentioned above, the decision on END has been based on this 20% criteria (20, 24, 28]. Aaboubout et al. determined the decision for intervention by the side effects of surgery (END) and radiotherapy at that time and assumed that, 25 years later, the treatment modalities will be substantially improved and suggested that a risk lower than 20% should be taken into consideration when deciding for END [28]. DOI can be accurately determined intraoperatively with the help of frozen section [34, 35]. Studies have also been conducted to describe methods to measure preoperative DOI in oral tongue SCC [36]. Lee and Choi in their systematic review and meta-analysis found excellent correlations of radiological DOI with pathological DOI when measured by ultrasound, computed tomography, and magnetic resonance imaging [37].

Apart from DOI, association of other histopathological factors with occult nodal metastasis was also analyzed. One of these is perineural invasion (PNI), which is a tropism of tumor cells for nerve bundles in the surrounding stroma. It is a form of tumor spread exhibited by neurotropic malignancies that correlates with aggressive behavior, disease recurrence, and increased morbidity and mortality [38]. Aaboubout et al. [28], Verma et al. [29], Lakhera et al. [32], Kane et al. [30], and Amaral et al. [31] did not find significant association of PNI with occult nodal metastasis in the subsites overall. However, we found significant association in our study. Similarly, Chatterjee et al. also found a significant association in early-stage oral tongue and buccal mucosa SCC [9]. Faisal et al. [25] and Kakaguchi et al. [27] found a significant association in early-stage oral tongue SCC. Since PNI is also a minor prognostic factor, preoperative and intraoperative detection of this pathological factor is required. Podoplanin expression has been found to be as a reliable marker of PNI, even in pre-operative biopsy specimens [39]. For intraoperative detection of PNI, currently, the surgeons do visualization of gross neural invasion. Frozen section histologic analysis can also be performed to determine PNI intraoperatively, but this requires sectioning the nerve and thereby renders it nonfunctional. A novel imaging modality known as Dynamic Optical Contrast Imaging (DOCI) not only guides cancer resection and margins during surgery but also accurately identifies perineural invasion, thus allowing the surgeon to identify the nerves that need to be sacrificed and decide for reconstructive options that are available once the affected nerve portion is removed. It is label-free and does not require the administration of any contrast agent [40].

LVI has been defined by Shariat et al. as the unequivocal presence of tumor cells within an endothelium lined space without underlying muscular walls. No attempt was made to differentiate between vascular and lymphatic vessels because of the difficulty and lack of reproducibility when using routine light microscopic examination [41]. Hence, it includes lymphatic invasion, vascular invasion, or lymph-vascular invasion [42]. Chatterjee et al. [9], Verma et al. [29], and Lakhera et al. [32] found significant association between LVI and occult nodal metastasis. Faisal et al. [25] and Kakaguchi et al. [27] did not find any significant association between the two. We also could not find LVI as a predictor of occult nodal metastasis, but we cannot comment on this association because only 3% cases of OSCC showed the presence of LVI in our study.

Pattern of invasion is another histopathological factor that was analyzed, and the one classified according to Brandwein-Gensler et al. was followed in our study [10]. In accordance with Verma et al., we found WPOI as a predictor of occult nodal metastasis, with significant increased risk in cases with WPOI Type V [29]. Chatterjee et al. [9] and Kakuguchi et al. [27] found significant association of both WPOI Type IV and V with risk of occult nodal metastasis in early-stage OSCC. Lakhera et al. also found a significant correlation between increasing WPOI grade and the nodal stage [32]. Frozen section analysis of oral cancer specimens has also been attempted for intraoperative assessment of WPOI. Beute et al. conducted a study in patients with primary T1/T2/cN0 OSCC undergoing resection and END and assessed the concordance between intraoperative and final pathology WPOI determination and found that WPOI Type V may be accurately determined intraoperatively [43].

Grade of tumor differentiation also showed significant association with occult nodal metastasis, with increased incidence in cases with well-differentiated OSCC. Cases of poorly differentiated OSCC were excluded because we would not recommend to skip any elective neck treatment in these tumors as they are already known to have a markedly high metastatic rate. However, previous studies had included grade III tumors in their analysis. Some studies found a significant association between tumor grade and occult nodal metastasis [27, 31, 32], while others did not [28–30].

Occult nodal metastasis cannot be clinically or radiologically identified [26]. Small sample size to find the pathological risk factors for occult nodal metastasis in OSCC is a limitation of our study. Hence, more and more studies with a larger sample size, and with subsite stratification to find the optimal DOI cut-off value for different subsites, especially the tongue, are required to be conducted in the future. More and more methods are being looked for accurate detection of pathological risk factors preoperatively and intraoperatively [34–37, 39, 40, 43].

Conclusion

Based on the findings of our study, DOI, PNI, WPOI, and tumor grade are predictors of occult nodal metastasis in early-stage OSCC. We found an optimal DOI cut-off value of 5.5 mm for predicting the increased risk of occult nodal metastasis in the overall subsites. Cases without PNI were at least risk, and those with WPOI Type 5 were at highest risk of occult nodal metastasis. There is a need for searching methods for preoperative and intraoperative detection of these risk factors so that unnecessary elective neck treatments can be avoided and the patient can be saved from the morbidity of these treatments and also get a better quality of life. These methods should enable us to preserve the neck treatment for cases with DOI ≥ 5.5 mm and those showing presence of PNI or WPOI Type V, provided that the tumor is well- or moderately differentiated squamous cell carcinoma only, and there is surety for the patient to turn up for regular follow-up, so that neck treatment can be considered later if recurrence or late cervical lymph node metastasis occurs. Studies should be conducted to compare the association of grade I and II OSCCs with occult nodal metastasis and to also analyze the association of lymphovascular invasion with occult nodal metastasis. Going forward, we recommend a prospective multi-center study with a larger sample size to analyze the association of all these histopathological parameters with occult nodal metastasis in early-stage OSCC.

Declarations

Ethical Approval

The study was approved by the Institutional Review Board of Rohilkhand Medical College and Hospital, Bareilly, Uttar Pradesh, India.

Informed Consent

Not required.

Conflict of Interest

The authors declare no competing interests.