Contralateral Nodal Metastasis from Tongue Malignancy

Abstract

Oral cavity is the most common site of cancer in India, tongue forms the most common single subsite. Neck node involvement plays an important prognostic role in disease-free survival and overall survival. Most of the nodal metastasis are occult and only an elective neck dissection can disclose the nodal metastasis, which is accepted only when the risk of metastasis exceeds 15–20%. Contralateral nodal metastasis(CLNM) are rarely analyzed due to sparse event.Retrospective analysis of data of patients who underwent surgery for carcinoma tougue with a curative intend between April 2017 and March 2019 were collected. Factors influencing CLNM in carcinoma tongue were studied.The study included 149 patients, with 102 men and 47 women. The mean age of study subjects was 51.5 ± 11.8 years. Thirteen (8.7%) patients had CLNM. With respect to tumor, advanced stage(p < 0.001), floor of mouth involvement(p = 0.001) and lesion crossing midline(p < 0.001) had an increased risk of CLNM. Extranodal extension(ENE)(p < 0.001) and lymphnode ratio(p = 0.002) in ipsilateral node ratio influenced CLNM. Multivariant analysis showed that in addition to lesion crossing midline(p = 0.03), ENE played significant role in CLNM (p < 0.001), with a risk of 12.8 times compared to no ENE. Suspicion of CLNM either clinically/radiologically was a good predictor of metastasis(p < 0.001).CLNM in tongue cancer was significantly influenced by lesion crossing midline and ENE. Contralateral neck dissection should be performed in any case with clinical/radiological suspicion of nodal involvement and above mentioned risk factors.

Introduction

Oral squamous cell carcinoma(OSCC) accounts for over 200,000 newly diagnosed cases of cancer worldwide annually. It is highly prevalent in developing countries like India, due to rampant use of tobacco products, where it accounts for almost one-third of the cancer. Tongue forms the most commonly affected single subunit in OSCC [1, 2].

OSCC has a poor prognosis due to high locoregional recurrence. Neck node involvement in OSCC plays an important prognostic role in disease-free survival and overall survival of the patient, wherein neck node involvement will reduce survival by 50%, being said contralateral nodal metastasis still has a poor prognosis. Most of the nodal metastasis from OSCC are occult and only an elective neck dissection can disclose the nodal metastasis, in most cases. Neck dissection has its own morbidities and so elective treatment of neck is accepted only when the risk of metastasis exceeds 15–20% [3, 4]. This makes the surgeons consider neck dissection based on individual case by case scenario, especially for contralateral neck, in OSCC.

Contralateral nodal metastasis (CLNM) includes (1) the presence of significant contralateral lymph node on initial presentation (2) occult contralateral neck lymph node metastasis diagnosed after neck dissection (3) contralateral neck nodal relapse [5]. CLNM is far less common than ipsilateral nodal metastasis (ILNM) making a debate between surgery and observation. CLNM in OSCC occurs within the first 2 years of the primary surgery in 89.9% of the patients and its incidence varies between 0.9% and 36% [6]. When detected early this would be a potentially preventable cause of recurrence in OSCC. As per Kurita et al. [7] CLNM in tongue cancers was noted in 15.4%, whereas Gracia et al. noted CLNM in 4.4% of their tongue cancers [8].

Most of the studies are done in the western population where the incidence of oral cancers is less common and most of the studies had included all the subsites of the oral cavity. So this study was intented to evaluate factors influencing CLNM from a single subsite of oral cavity i.e., tongue. Further effectiveness in the diagnosis of CLNM preoperatively is also studied.

Methodology

This was a retrospective analysis carried out in a tertiary care center spanning the period from April 2017 to March 2019. Institutional review board permission was obtained prior to the study. Patients newly diagnosed with tongue OSCC who were treated with curative intend were included in the study. Second Primary/recurrent tumor of tongue and patients who expired in the immediate perioperative period were excluded. One hundred and forty nine patients who met inclusion criteria were evaluated further.

All patients included in the study had multidisciplinary tumor board discussion prior to the definitive line of management. They underwent wide excision of the lesion with neck dissection as per the surgeon's clinical judgment. Reconstruction was done either by primary closure, nasolabial flap, submental flap, infrahyoid flap, pectoralis major myocutaneous flap radial artery free flap or anterolateral thigh flap. Pathological margin less than 0.1 cm were considered involved margin, 0.2–0.5 cm as close margin and greater than 0.5 as clear margins. Following surgery adjuvant radiotherapy was planned based on advanced pathological stage or based on risk factors. Pathologically staging was done according to the American Joint Committee on Cancer (AJCC) 8th staging system for cancer of the oral cavity. Contralateral neck node involvement was not considered on staging, for an effective statistical analysis.

On histopathological examination, following tumor-related factors were considered: pathological tumor stage, grade of differentiation, depth of invasion, perineural invasion, lymphovascular emboli, primary resection margin. Lymph nodes were analyzed for nodal stage, extracapsular extension and lymph node ratio(defined as the ratio between number of pathological cervical lymph nodes and the total number of resected cervical lymph nodes[3].

Statistical Analysis

Association between clinicopathological features and CLNM was performed using contingency tables and chi-square tests. The p-value ≤ 0.05 was considered significant. Independent variables used were all clinicopathological parameters and CLNM was considered as the dependent variable. The categorical variables were reported using frequency and or proportions and continuous variables were represented using mean and standard deviation.

Results

The study included 102 men and 47 women, with a median followup of 12 months. The median age of the study population was 51 years (range: 25 years–84 years). In the study population, 26 (17.5%) patients underwent neoadjuvant chemotherapy(NACT) followed by surgery due to advanced nature of the disease.

Sixty-three (42.3%) patients belonged to early pathological stage(Stage I/II), 83(55.7%) belonged to advanced pathological stage (Stage III/IV). Five (3.3%) of the 26 patients who received NACT had a complete chemo response, in the final histopathological specimen.

All patients included in the study underwent neck dissection, ipsilateral neck was addressed in all as extended supraomohyoid neck dissection or modified radical neck dissection and bilateral neck dissection was done in 27 patients(18.1%). 65 patients(43.6%) had ipsilateral nodal involvement and CLNM was noted in 13(8.7%) patients. Nine (33.3%) of the 13 patients had CLNM during neck dissection.

Univariate analysis (Table 1) showed that extranodal extension in ipsilateral node, tumor involving floor of mouth(FOM), and those tongue tumors crossing midline, independently had a high statistically significant impact on CLNM. Other factors that influenced CLNM were advanced pathological T stage (p < 0.001), advanced N stage (p < 0.001), advaned composite stage (p = 0.011), increase in depth of tumor invasion (p = 0.012), ipsilateral nodal ratio more than 0.09 (p = 0.002), extra nodal extension(ENE) in ipsilateral node (p < 0.001) and the type of reconstruction (p = 0.001). Multivariant analysis (Table 2) of all these most significant tumor-related factors indicated ENE had a very significant statistically impact on CLNM. ENE in the ipsilateral neck node had 12.8 times increased risk of CLNM (95% CI: 3.087–52.905, p < 0.001). Lesion crossing midline had a 6.3 times (95% CI: 1.201–33.219, p = 0.030) risk of CLNM.

Table 1.

Univariate analysis

Variable	No. of patients(n = 149)(%)	No. of contralateral nodal recurrence(n = 13)(%)	p value
Sex			0.549
Male	102(68.5)	8(61.5)
Female	47(31.5)	5(38.5)
Age(median = 51 years)			0.294
Less than median	71(47.6)	8(61.5)
More than median	78(52.4)	5(38.5)
Staging
Pathological T stage			< 0.001
T1,T2	95(63.8)	2(15.4)
T3,T4	49(32.9)	9(69.2)
Complete NACT tumor response	5(3.3)	2(15.4)
Pathological N stage			< 0.001
N0	83(55.7)	2(15.4)
N1	16(10.7)	0(0)
N2	29(19.5)	3(23.1)
N3	21(14.1)	8(61.5)
Composite stage			0.011
Early stage (1/II)	63(42.3)	1(7.7)
Advanced stage(III/IV)	82(55.0)	10(76.9)
Complete NACT tumor response	5(3.3)	2(15.4)
Pathology
Tumor grade			0.093
Not ascertained	5(3.3)	2(15.4)
Well	20(13.4)	1(7.7)
Moderate/poor	124(83.2)	10(76.9)
Perineural invasion			0.071
Present	79(53)	10(76.9)
Absent	70(47)	3(23.1)
Lymphovascular emboli			0.221
Present	52(34.9)	7(53.8)
Absent	97(65.1)	6(46.2)
Tumor depth			0.012
0.1–0.5 cm	53(35.6)	1(7.7)
0.6–1.0 cm	48(32.2)	3(23.1)
More than 1 cm	48(32.2)	9(69.2)
Primary resection margin in frozen			0.179
Clear(> 0.5 cm)	72(48.3)	5(38.5)
Close(0.2 to 0.5 cm)	70(46.9)	6(46.2)
Involved(< 0.1 cm)	7(4.7)	2(15.4)
I/L Node ratio			0.002
< 0.09	114(76.5)	5(38..5)
> 0.09	35(23.5)	8(61.5)
Extracapsular spread in I/L node			< 0.001
Present	25(16.8)	9(69.2)
Absent	124(83.2)	4(30.8)
Clinical/surgical
I/L clinical neck node			0.081
Involved	57(38.3)	8(61.5)
Not involved	92(61.7)	5(38.5)
C/L clinical neck node			< 0.001
Involved	8(5.4)	7(53.8)
Not involved	141(94.6)	6(46.2)
Floor of mouth			0.001
Involved	39(26.2)	9(69.2)
Not involved	110(73.8)	4(30.8)
Lesion crossing midline			< 0.001
Present	33(24.1)	9(69.2)
Absent	116(75.9)	4(30.8)
Base of tongue involvement			0.054
Involved	18(12.1)	4(30.8)
Not involved	131(87.9)	9(69.2)
Reconstructions			0.001
Primary closure	85(57)	3(23)
Pedicled flap	14(9.4)	5(38.5)
Local/Free flap	50(33.6)	5(38.5)
Therapy
NACT			0.103
Given	22(14.8)	4(30.8)
Not given	127(85.2)	9(69.2)
Radiotherapy			0.185
Given	110(73.8)	12(92.3)
Not given	39(26.2)	1(7.7)

Table 2.

Multivariate analysis

	P Value	Hazard ratio	Confidence interval
			Min	Max
FOM involvement	0.493	1.795	0.338	9.540
Crosing midline	0.030	6.318	1.201	33.219
ENE	< 0.001	12.780	3.087	52.905

Imaging along with clinical examination were able to predict CLNM(p < 0.001) with high accuracy compared to ipsilateral neck node as shown in (Table 3 and Fig. 1).

Table 3.

Prediction of significant neck node with clinical examination and imaging

	Prevalence	Sensitivity	Specificity	PPV	NPV	Accuracy	Occult metastasis missed
Ipsilateral neck	43.62	50.77	71.43	57.89	65.42	62.42	42.1
Contralateral neck	8.72	53.85	99.26	87.50	95.74	95.30	12.5

Fig. 1.

Detecting a metastatic node by clinicoradiological correlation preoperatively

Discussion

Contralateral neck are usually addressed in tongue malignancy, when the lesion is crossing midline, the base of tongue is involved, tip of tongue involvement, anterior FOM involvement, significant Ia nodal involvement and when a significant contralateral neck node is detected preoperatively.

Spiro et al. in 1973 reported the importance of elective neck dissection in OSCC where they noted an occult metastasis in 30% of patients [9]. In our study we noted that 65 patients(43.6%) had ipsilateral nodal metastasis on the final histopathological specimen. Among them 32 patients(21.5%) had an occult metastasis that was not detected by clinical examination and magnetic resonant imaging(MRI) evaluation. This strongly proves that ipsilateral neck dissection is mandatory, especially in tongue cancers.

Factors influencing ipsilateral nodal metastasis are extensively studied and has been well documented. However, factors that are associated with contralateral nodal metastasis, in tongue cancers are sparse and most of them are Western data. Tongue has rich lymphatics so the location of the lesion plays a key role in CLNM. Lesion crossing midline, base of tongue involvement, floor of mouth involvement and advanced tumor(more than T3) are the tumor-related factors associated with CLNM [4, 5, 7, 10, 11]. Ipsilateral nodal metastasis based on number and involved level had also been known to be associated with increase in risk of CLNM [7, 8]. Histological grading of tumour, lymphocytic invasion and perineural invasion [7, 8, 12] are noted histologic factors associated with CLNM in OSCC. As per study by Kaya et al. elective ipsilateral neck dissection previously, will promote CLNM [10]. It was noted that though tumor stage, FOM involvement and lesion crossing midline independently highly influenced CLNM, multivariate analysis noted lesion crossing midline influenced CLND by 6.3 times. Both ENE and high node ratio in ipsilateral neck highly influenced CLNM independently, but multivariate analysis showed that only the presence of ENE in ipsilateral node increased the risk of CLNM by 12.8 times.

Study by Bikramjit et al. done in India, included patient who underwent bilateral neck dissection for locally advanced tongue cancers which were crossing midline. The study detected CLNM in 29% and ipsilateral nodal metastasis in 28.4%. The study had not included ENE as an influencing factor [13]. Our study included both early and advanced stage tongue cancer also noted 43.6% ipsilateral nodal metastasis and it was also noted that, in patient who underwent bilateral neck dissection, CLNM was noted in 33.3%. It means that there is a high risk of nodal metastasis in Indian population, due to different tumor biology, compared to western population.

ENE is a common phenomenon in head and neck cancer patients and is considered to be a sign of poor prognosis due to increased incidence of locoregional recurrence and survival drops down by 50 percent, compared to no ENE. Presence of ENE has necessitated the use of concurrent chemoradiotherapy in the postoperative setting and the survival benefits of the same have been well documented[14].

Poor prognosis in CLNM is as equivalent to ENE. A study by Spiro et al. on 1069 patients with oral cavity and oropharyngeal squamous cell carcinoma noted 5- year survival rate in patients with ipsilateral nodal metastasis was 28%, compared to 8% in cases of bilateral metastasis [15]. Study by Koo et al. in OSCC also showed the survival fell down by near 50% when there is a CLNM[16]. Current study also showed a close association between ENE and CLNM, thus making a strong association with survival. It is well known that recurrent primary lesions and cervical lymph-node relapse affects the prognosis and decreases the survival rate of patients with OSCC. Thus CLNM after definitive surgery might prompt an aggressive treatment with concurrent chemoradiotherapy, as both ENE and CLNM grossly reduce the survival equally, the need for this has to be futher evaluated.

Lim et al. compared elective contralateral neck dissection versus observation for contralateral neck in early stage tongue cancer and noted that routine bilateral neck dissection is not associated with increase in overall survival [11]. Prevalance of CLNM was also low(8.7%). Thus prophylactic contralateral neck dissection is not preferred in early malignancy and in those who does not have above mentioned risk factors.

Clinical examination and MRI together had an accuracy of 62.42% to predict ipsilateral nodal involvement and on comparison the accuracy to diagnose contralateral significant node was 95.30%, with a high negative predictive value 95.74%. So contralateral neck dissection should be always considered in case of a significant node that was noted in MRI imaging, in addition to other clinical judgements.

Conclusion

CLNM is highly influenced by the presence of ENE the risk of which increases by 12.8 times. In addition to lesion crossing midline, contralateral neck dissection is preferred in the case of an ipsilateral node with ENE and any significant contralateral node that was noted in MRI. Need for concurrent chemoradiotherapy in CLNM needs to be further studied.

Funding

None.

Compliance with Ethical Standards

Conflict of Interest

None.