|Year : 2017 | Volume
| Issue : 2 | Page : 442-446
Paralingual and sublingual space invasion in magnetic resonance imaging of squamous cell carcinoma of anterior two-thirds of tongue: Is there a prognostic significance? A prospective evaluation
SC Jayasankaran1, PG Chelakkot2, K Thankappan3, S Iyer3, S Moorthy1
1 Department of Radiology, Amrita Institute of Medical Sciences, Amrita University, Kochi, Kerala, India
2 Department of Radiation Oncology, Amrita Institute of Medical Sciences, Amrita University, Kochi, Kerala, India
3 Department of Head and Neck Surgery, Amrita Institute of Medical Sciences, Amrita University, Kochi, Kerala, India
|Date of Web Publication||21-Feb-2018|
Dr. P G Chelakkot
Department of Radiation Oncology, Amrita Institute of Medical Sciences, Amrita University, Kochi, Kerala
Source of Support: None, Conflict of Interest: None
BACKGROUND: Magnetic resonance imaging (MRI) in tumors of anterior two-thirds of tongue has a significant role in assessing different tumor parameters, and in prognosticating. AIM: This prospective study conducted in a tertiary cancer care center, focused on patients with squamous cell carcinoma of anterior two-thirds of tongue. The significance of invasion of paralingual and sublingual spaces in relation to the pathological grade of these tumors, and its predictive value in pathological nodal involvement were analyzed. MATERIALS AND METHODS: All consecutive patients with the required inclusion criteria were accrued. Imaging was done with 3 Tesla MRI and invasion of sublingual and paralingual spaces were accurately assessed. Data elucidated were tabulated and analysed using IBM SPSS version 20.0. Chi-square test, nonparametric correlation using Spearman's Rho correlation, and two-independent sample test using Mann–Whitney's U-test were used to arrive at correlations between the imaging and histopathological parameters. RESULTS: Sixty-three patients were analyzed. Mean age was 52.3 ± 11.45 years. 74.6% were males. MRI showed sublingual space invasion in 47.6%. 18/28 with and 11/33 without invasion had node positivity. Paralingual space involvement was observed in 31.7% of patients. Thirteen of these and 16/43 with no involvement had positive cervical nodes. No statistically significant correlation was observed. CONCLUSION: This prospective study did not establish any statistically sound correlation, and robust data are lacking to support newer parameters such as sublingual space and paralingual space as probable predictors of cervical nodal involvement, and for prognostication.
Keywords: Magnetic resonance imaging, nodal status, paralingual space, squamous cell carcinoma of anterior two-thirds of tongue, sublingual space, tumor differentiation
|How to cite this article:|
Jayasankaran S C, Chelakkot P G, Thankappan K, Iyer S, Moorthy S. Paralingual and sublingual space invasion in magnetic resonance imaging of squamous cell carcinoma of anterior two-thirds of tongue: Is there a prognostic significance? A prospective evaluation. Indian J Cancer 2017;54:442-6
|How to cite this URL:|
Jayasankaran S C, Chelakkot P G, Thankappan K, Iyer S, Moorthy S. Paralingual and sublingual space invasion in magnetic resonance imaging of squamous cell carcinoma of anterior two-thirds of tongue: Is there a prognostic significance? A prospective evaluation. Indian J Cancer [serial online] 2017 [cited 2021 Jul 31];54:442-6. Available from: https://www.indianjcancer.com/text.asp?2017/54/2/442/225806
| » Introduction|| |
Imaging plays an important role in the management of malignancies, starting with initial assessment of the lesion, accurate staging, deciding the extent of surgical excision and delineating the radiation treatment volume, in the post-treatment assessment of outcomes and in deciding future interventions. Imaging thus remains a mandatory tool in the management of head and neck tumors as anywhere else, and in tongue tumors as well. Among the imaging modalities available, computerized axial tomography (CAT) scan, and magnetic resonance imaging (MRI) are more widely used. An advantage in accurate assessment of extent in tumors of anterior two-thirds of tongue for MRI over CAT scan has been established by Sigal et al.
Tumor extension to sublingual and paralingual spaces of tongue has been documented by earlier authors as a significant prognostic factor, among others. Extension to these spaces are precisely delineated in imaging, but cannot be corroborated on the basis of histopathological evidence.
This prospective study conducted in the Departments of Radiology, Radiation Oncology, Head and Neck Oncology and Pathology at a tertiary cancer care center, focused on patients with squamous cell carcinoma of anterior two-thirds of tongue. The significance of invasion of paralingual and sublingual spaces in relation to the pathological grade of these tumors, and its predictive value in pathological nodal involvement were analyzed.
| » Materials and Methods|| |
All consecutive patients of histopathology proved, nonmetastatic, resectable, squamous cell carcinomas involving anterior two-thirds of tongue, who presented to our tertiary cancer care center in South India, and who were planned for primary surgery and adjuvant treatment were included in the study. Clearance from the Institutional Scientific Review Board and the Institutional Ethical Committee were obtained before accrual. Written informed consent with permission to use data for scientific studies were obtained as an institutional policy from all patients before inclusion in the study. The period of accrual extended from August 2014 to June 2016. Exclusion criteria were inoperable disease, medical comorbidities preventing curative intent management, prior treatment with neoadjuvant chemotherapy, prior treatment with surgery or radiation therapy for head and neck tumors, metastatic disease, recurrent disease and patients planned for palliative management, and those who refused surgery.
Complete clinical evaluation and tissue diagnosis for confirmation were done by the oncology team initially. After this, the patient was taken up for imaging with 3 Tesla superconducting magnet scanner (GE Healthcare, Milwaukee, Wisconsin) using 8 channel head and neck coil. Ample time was given for the biopsy related inflammatory changes to subside so as to ensure precision. Imaging was done as close to the surgery as feasible, and the interval was kept less than a week. T1-weighted spin echo (short echo time/repetition time [TE/TR], 4 mm slice thickness, no inter-slice gap) was taken in coronal plane. T2 sequences (long TE/TR, slice thickness 4 mm, no inter-slice gap) were taken in axial, sagittal and coronal planes. T2 weighted sequences with fat saturation was acquired in coronal plane. Three-dimensional T1-weighted sequences with fat suppression in axial plane were used to acquire the postcontrast images, which were reformatted in sagittal and coronal planes. Planes for reformatting were with reference to the midline septum of the tongue, with 0.37 mm slice thickness. Gadodiamide (Omniscan GE) was used for contrast and was administered intravenously at a rate of 1.5 cc/s up to a maximum dose of 12 cc. Mechanical injector device was used for administering the contrast, and images were acquired immediately after the injection.
All routine tumor parameters such as tumor dimensions, infiltration of adjacent structures, and nodal involvement were assessed. As a study protocol invasion of sublingual and paralingual spaces were also accurately assessed. Sublingual space was defined as the space deep to the intrinsic muscles of tongue, and lying superior to the mylohyoid muscle which homes the sublingual salivary gland. Medial to the sublingual space is the genioglossus muscle. Sublingual space has a high intensity in coronal T1-weighted images. The distance between the tumor and the sublingual space determines the sublingual space invasion [Figure 1]. Involvement of the space was recorded as minus values, with “0” indicating just abutting the space and was considered as involvement. Paralingual space is defined as the space between the genioglossus muscle medially and the other intrinsic muscles of tongue laterally [Figure 2]. In a coronal T1-weighted image, the high-intensity area is seen to extend from the medial border of the sublingual space to the deep lingual artery which runs along the border of genioglossus muscle. This area of high intensity defines the paralingual space. Paralingual space extent was measured as the distance between the paralingual space and the tumor. Invasion of the paralingual space was documented as minus value. Involvement of these spaces being observational variants, to avoid observer dependent variations, same radiologist reported on these for all patients.
|Figure 1: Postcontrast T1-weighted magnetic resonance image showing sublingual space. T-denotes the tumor, *the sublingual space with the sublingual gland and the line denotes the distance from the tumor edge to the sublingual space|
Click here to view
|Figure 2: Postcontrast T1-weighted magnetic resonance image showing paralingual space. T-denotes the tumor, G– the genioglossus muscle, * the paralingual space, and the line denotes the distance from the tumor edge to the paralingual space|
Click here to view
All patients after initial clinical evaluation, imaging, staging, and metastatic work up underwent surgery, which most often was a wide local excision. The adequacy of margin was considered as 1 cm beyond all palpable margins of the lesion including the induration. Cervical nodes were addressed either with ipsilateral or bilateral neck dissection, being either selective or modified radical neck dissection depending on the clinical stage. Resected specimens were evaluated by dedicated head and neck oncopathologist. Specimens were oriented by the surgeon, raw surfaces inked and cut in the craniocaudal plane. Tumor parameters were assessed and reported by dedicated head and neck pathologist as per international guidelines.
Demographic, clinical, imaging, and histopathological data elucidated were tabulated and analyzed using IBM SPSS version 20.0. Statistical tools such as Chi-square test and nonparametric correlation using Spearman's Rho correlation were used to arrive at correlations between the imaging and histopathological parameters. A correlation of up to 0.2 was considered negative, 0.2–0.4 mild, 0.4–0.6 moderate, 0.6–0.8 good, and 0.8-1.0 as very good correlation. Two independent sample test using Mann–Whitney's U-test was carried out to assess the association between the parameters. Asymptotic significance (two-tailed) of <0.05 was considered as statistically significant.
| » Results|| |
During the study, 88 cases of malignancies of the anterior two-thirds of tongue were seen at the institute. Among these, 63 patients were eligible for inclusion in the study. Mean age at presentation was 52.3 ± 11.45 years (range: 18–74). Gender wise, a male preponderance with 74.6% was observed. Moderately differentiated squamous cell carcinomas formed the majority with 76.2%, while well differentiated accounted for 11.1% and poorly differentiated 12.7%. Among the 63, two patients refused surgery, of whom, one was taken up for radical brachytherapy. Wide local excision was done in 87.3% of patients; 4.8% underwent hemiglossectomy and 3.2% total glossectomy. Supraomohyoid ipsilateral neck dissection was done in 95.2%. Dissection was done up to level four nodal station on the ipsilateral side in 88.9% of patients. Contralateral nodal dissection was considered only in nine patients. Pathological evaluations showed 46% of patients to be node positive. Patient details are tabulated in [Table 1].
Imaging with MRI showed sublingual space invasion in 47.6% of patients. Among these 28 patients, 18 (64.3%) showed cervical node metastasis, and among the 33 who did not have sublingual space invasion, 11 (33.3%) had node positivity [Figure 3]a. Paralingual space involvement was observed in 31.7% of patients. Among these, 13 (65%, 2 had no surgery) had positive cervical nodes [Figure 3]b. Paralingual space involvement was not seen in 43 patients and among this, 16 had positive nodes. Okura et al. had suggested an invasion of more than 5.2 mm into the paralingual space to be a significant predictor of node negativity. However, this failed to show any statistically significant correlation (Spearman correlation: 0.335, P = 0.008).
|Figure 3: Graph depicting relation between nodal status and invasion of; (a) Sublingual space and (b) Paralingual space|
Click here to view
The statistical assessment showed a moderate correlation between sublingual and paralingual space invasions with a correlation coefficient of 0.568 (P < 0.001), using the Spearman's Rho test. No correlation was noted between the grade and invasion of sublingual space (R = 0.15, P = 0.233), and paralingual space (R = 0.09, P = 0.482) [Figure 4]a and b]. Mann-Whitney U-test also showed no significant association between the invasion of sublingual space (asymptotic sig [two-tailed] 0.139), or paralingual space (asymptotic sig [two-tailed] 0.922) and the pathological tumor grade.
|Figure 4: Graph depicting relation between tumor differentiation and invasion of; (a) Sublingual space, and (b) Paralingual space|
Click here to view
Statistical evaluation using Spearman's Rho analysis failed to show any significant correlation between the invasion of sublingual space and pathological cervical node positivity (R = 0.422, P < 0.001). Invasion of paralingual space also failed to show any statistically significant correlation between pathological cervical node positivity (R = 0.363, P = 0.004), in the nonparametric correlation. In contrast, a significant correlation was observed between node negativity and sublingual space invasion with node negativity having a high mean rank (asymptotic sig [two-tailed]-0.001; U-237.5). With paralingual space invasion also node negativity showed a high mean rank with significance (asymptotic sig [two-tailed] 0.007; U-276.0)
Nonparametric correlation analysis using Spearman's Rho showed a mild correlation between pathological tumor differentiation and the cervical nodal involvement (R = 0.393, P = 0.002), and Mann–Whitney U-test showed a high mean rank for node positivity (asymptotic sig [2-tailed]-0.015; U-377.0) [Figure 5].
|Figure 5: Graph depicting the relation between the tumor differentiation and the nodal status|
Click here to view
| » Discussion|| |
Tumors of anterior two-thirds of tongue, unlike other subsites of oral cavity, tend to behave differently, that the quest for understanding the biology has triggered much research. Clinically tumors of the tongue tend to be either infiltrative or proliferative in their growth pattern or expansive or reductive in their size. Accurate pretreatment evaluation of the tumor extent is mandatory for planning the primary treatment and possible adjuvants. Becker and Zaidi have opined the superiority of MRI over other imaging modalities in delineating the accurate extent of these tumors. The recent (2016) version of American Joint Committee on cancer staging system (AJCC) has made significant changes in the tumors of oral cavity as compared to the 7th edition, by incorporating depth of invasion as one of the deciding parameters.
Prognostic significance has been attributed to tumor parameters such as depth of invasion, infiltration of muscles, and extension to sublingual and paralingual spaces by Okura et al., The importance of tumor thickness has been addressed by many authors , and the incorporation of this parameter in the tumor staging is ample evidence of its significance.
Anatomically, sublingual space is one of the suprahyoid spaces of the head and neck. It is an inverted “V” shaped space, with its apex pointing anteriorly, and is between the muscles of the tongue superiorly and anterior one-third of mylohyoid inferolaterally. Mylohyoid muscle intervenes between the sublingual space and the submandibular space. The sublingual spaces of both sides communicate anteriorly through a small isthmus just below the frenulum, and posteriorly, it freely communicates with submandibular and the parapharyngeal spaces along the posterior free border of mylohyoid muscle. Medially, it is bounded by the geniohyoid and genioglossus complex, and laterally the mylohyoid. Its anterior border is formed by the medial surface of mandible and superior border by the mucosa of the floor of mouth and the intrinsic muscles of tongue. The hyoglossus, styloglossus, and palatoglossus muscle form a lateral muscle bundle and are a surgical landmark. This space contains the lingual nerve and vessels, glossopharyngeal and hypoglossal nerves, glossopharyngeal and hypoglossal nerves, sublingual gland and duct, deep part of submandibular gland, and Wharton's duct. On imaging sublingual space is seen superiorly and medially to mylohyoid muscle and lateral to genioglossus. While in MRI it is an hyperintense area, it has a low-density appearance in computed tomography.
Lesions extending to the sublingual space can infiltrate the submandibular space through the posterior aspect of the space which offers the least resistance to the advancing tumor, due to the absence of an effective facial barrier. On the other hand, if the lesion is adjacent to a defect in the mylohyoid, which most often is in the mid-portion of the muscle, it herniates through the defect. Sublingual space extension is a radiological observation and is a feature not discernible in histopathology examination for corroboration.
Our analysis showed a significant correlation between sublingual space invasion and cervical nodal involvement, with node negativity showing a high mean rank. However, there was no significant association between sublingual space invasion and tumor grade. Yet, in view of the least resistance offered by the space to an advancing aggressive lesion, whether this significance is due to the sublingual space involvement per se, or to the resultant increased depth of tumor invasion needs to be clarified.
The paramedian septum described by Abd-El-Malik in 1939 corresponds to the paralingual space. He had described this as a strong septum between genioglossus medially and the hyoglossus and styloglossus laterally. The intrinsic muscles of tongue, namely the transverse and vertical muscles lie dorsal to the space, and the inferior longitudinal, hyoglossus and styloglossus muscles laterally. It has a triangular outline with its apex anteriorly and corresponds to the extent of genioglossus. The posterior aspect is broader and continues as hyoglossal membrane attaching to the hyoid bone. Lingual artery and vein and a short part of hypoglossal nerve are contained in the substance of this space. Paralingual space is a radiological observation, and pathological correlation is difficult for ascertaining involvement.
Okura et al. in their study had suggested that there is a significant correlation between the distance of paralingual space involvement and nodal metastasis, and also that it was an inverse correlation. They projected a probable risk of 20% for a paralingual invasion of 5.2 mm and had suggested addressing the neck if imaging tumor thickness was >9.7 mm or if paralingual space invasion was <5.2 mm. They suggested a cut-off of 5.2 mm as significant with 82% specificity and 70% sensitivity. However, in their analysis, an increase in paralingual space invasion to 6.5 mm dropped the risk of nodal involvement to 10%, while a lesser involvement of 3 mm showed 50% node positivity. They in their earlier paper had also shown that  all twenty patients with paralingual extension of >5.3 mm were node negative, and all six patients who had paralingual space <0 mm had positive lymph nodes. In our study, twenty patients had involvement of paralingual space <0 mm, among whom 13 were node positive; while 16 patients with no paralingual space involvement had node positivity, and no statistical significance was obtained (R = 0.335, P = 0.008). Our analysis also showed a significant correlation for node negativity and paralingual space invasion (P = 0.007). Although our cohort was sufficiently large compared to other studies, we failed to obtain a significant correlation between the invasion of paralingual space and cervical nodal involvement.
Although not showing high correlation (R = 0.314, P = 0.002), a positive relation between poorer tumor differentiation and nodal involvement was noted with high mean rank for node positivity (asymptotic sig [two-tailed]-0.015; U-377.0). Furthermore, more than the paralingual space, depth of invasion showed a significant correlation with nodal positivity in Okura et al. analysis, with an increase in depth showing increasing nodal involvement. A tumor thickness of 7.1 mm showed a 10% probability of nodal involvement while 14.2 mm increased it to 50%, in their analysis. Hence it would appear that depth of invasion is a more robust surrogate for nodal involvement than extension to either paralingual space or sublingual space. This stand validated by the fact that depth of invasion has been incorporated in the 2016 version of AJCC staging system for tumors of anterior two-thirds of tongue.
| » Conclusion|| |
This prospective study in a cohort of 63 patients that aimed to assess the significance of invasion of sublingual and paralingual space as a predictor of cervical nodal metastasis in tumors of anterior two-thirds of tongue did not establish any statistically sound correlation. As observed in other studies, robust data are lacking to support newer parameters such as sublingual space and paralingual space as probable predictors of cervical nodal involvement, and for prognostication.
We would like to acknowledge The Department of Pathology, Amrita Institute of Medical Sciences, Kochi in processing of tumour specimens and timely interpretation of histopathology slides and their valuable inputs.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| » References|| |
Sigal R, Zagdanski AM, Schwaab G, Bosq J, Auperin A, Laplanche A, et al
. CT and MR imaging of squamous cell carcinoma of the tongue and floor of the mouth. Radiographics 1996;16:787-810.
Okura M, Iida S, Aikawa T, Adachi T, Yoshimura N, Yamada T, et al
. Tumor thickness and paralingual distance of coronal MR imaging predicts cervical node metastases in oral tongue carcinoma. AJNR Am J Neuroradiol 2008;29:45-50.
Becker M, Zaidi H. Imaging in head and neck squamous cell carcinoma: The potential role of PET/MRI. Br J Radiol 2013;87:1-15.
Edge SB, Compton CC. The American Joint Committee on Cancer: The 7th
edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol 2010;17:1471-4.
Lydiatt WA, Patel SG, O'Sullivan B, Brandwein MS, Ridge JA, Migliacci JC, et al
. Head and neck cancers – Major changes in the AJCC 8th
edition Cancer staging manual. CA Cancer J Clin 2017;67:122-37.
Okura M, Sawai NY. Management for the N0 Neck of SCC in the Oral Cavity. In: Prof Raja Kummona, editor. Neck Dissection - Clinical Application and Recent Advances. 1st
ed. ISBN: 978-953-51-0104-8, InTech.2012. p. 113-24.
Clark JR, Naranjo N, Franklin JH, de Almeida J, Gullane PJ. Established prognostic variables in N0 oral carcinoma. Otolaryngol Head Neck Surg 2006;135:748-53.
Fan S, Tang QL, Lin YJ, Chen WL, Li JS, Huang ZQ, et al
. A review of clinical and histological parameters associated with contralateral neck metastases in oral squamous cell carcinoma. Int J Oral Sci 2011;3:180-91.
Meesa IR, Srinivasan A. Imaging of the oral cavity. Radiol Clin North Am 2015;53:99-114.
Arya S, Chaukar D, Pai P. Imaging in oral cancers. Indian J Radiol Imaging 2012;22:195-208.
] [Full text]
Otonari-Yamamoto M, Nakajima K, Tsuji Y, Otonari T, Curtin HD, Okano T, et al
. Imaging of the mylohyoid muscle: Separation of submandibular and sublingual spaces. AJR Am J Roentgenol 2010;194:W431-8.
Abd-El-Malek S. Observations on the morphology of the human tongue. J Anat 1939;73(Pt 2):201-10.3.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]