[Detection and assessment of cervical lymph node metastases in head-neck tumors. A comparison of methods]

Current Applications of Artificial Intelligence to Classify Cervical Lymph Nodes in Patients with Head and Neck Squamous Cell Carcinoma-A Systematic Review

Locally-advanced head and neck squamous cell carcinoma (HNSCC) is mainly defined by the presence of pathologic cervical lymph nodes (LNs) with or without extracapsular spread (ECS). Current radiologic criteria to classify LNs as non-pathologic, pathologic, or pathologic with ECS are primarily shape-based. However, significantly more quantitative information is contained within imaging modalities. This quantitative information could be exploited for classification of LNs in patients with locally-advanced HNSCC by means of artificial intelligence (AI). Currently, various reviews exploring the role of AI in HNSCC are available. However, reviews specifically addressing the current role of AI to classify LN in HNSCC-patients are sparse. The present work systematically reviews original articles that specifically explore the role of AI to classify LNs in locally-advanced HNSCC applying Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines and the Study Quality Assessment Tool of National Institute of Health (NIH). Between 2001 and 2022, out of 69 studies a total of 13 retrospective, mainly monocentric, studies were identified. The majority of the studies included patients with oropharyngeal and oral cavity (9 and 7 of 13 studies, respectively) HNSCC. Histopathologic findings were defined as reference in 9 of 13 studies. Machine learning was applied in 13 studies, 9 of them applying deep learning. The mean number of included patients was 75 (SD ± 72; range 10-258) and of LNs was 340 (SD ± 268; range 21-791). The mean diagnostic accuracy for the training sets was 86% (SD ± 14%; range: 43-99%) and for testing sets 86% (SD ± 5%; range 76-92%). Consequently, all of the identified studies concluded AI to be a potentially promising diagnostic support tool for LN-classification in HNSCC. However, adequately powered, prospective, and randomized control trials are urgently required to further assess AI's role in LN-classification in locally-advanced HNSCC.

Assessment of the Diagnostic Accuracy of Baseline Clinical Examination and Ultrasonographic Imaging for the Detection of Lymph Node Metastasis in Patients With High-risk Cutaneous Squamous Cell Carcinoma of the Head and Neck

IMPORTANCE

Patients with cutaneous squamous cell carcinoma (SCC) of the head and neck may develop lymph node metastasis; therefore, additional workup of the regional lymph nodes in these patients should be considered. However, there is uncertainty regarding the value of baseline ultrasonographic imaging in addition to clinical examination for the detection of metastasis.

OBJECTIVE

To assess the diagnostic accuracy of clinical examination and baseline ultrasonography for the detection of metastasis among patients with high-risk cutaneous SCC of the head and neck and to assess the accuracy of ultrasonography when baseline clinical examination produces negative results.

DESIGN, SETTING, AND PARTICIPANTS

This diagnostic study was conducted among a retrospective cohort of 233 patients with 246 high-risk cutaneous SCC tumors of the head and neck. The study included all patients with high-risk cutaneous SCC of the head and neck who received clinical examination and baseline ultrasonographic imaging of their lymph nodes at the Erasmus Medical Center Cancer Institute, a tertiary referral hospital for patients with skin cancer in Rotterdam, the Netherlands, between January 1, 2015, and December 31, 2017. Data were analyzed from October 13, 2020, to September 29, 2021.

MAIN OUTCOMES AND MEASURES

Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. Fine-needle aspiration cytologic biopsy and 6 months of follow-up per patient were used as the reference standards.

RESULTS

Among 233 patients (176 men [75.5%]; median age, 79.1 years [IQR, 71.5-83.7 years]; data on race and ethnicity were not collected) with 246 high-risk cutaneous SCC tumors of the head and neck, 20 metastases were cytologically confirmed at baseline, and 2 metastases were detected during 6 months of follow-up, yielding a 9% metastasis rate. The sensitivity of clinical examination was 50% (95% CI, 28%-72%), and the specificity was 96% (95% CI, 93%-98%). The PPV and NPV were 55% (95% CI, 36%-72%) and 95% (95% CI, 93%-97%), respectively. In the total cohort, ultrasonography had a sensitivity of 91% (95% CI, 71%-99%) and a specificity of 78% (95% CI, 72%-83%), with a PPV of 29% (95% CI, 23%-35%) and an NPV of 99% (95% CI, 96%-100%). In the group of patients with negative results at baseline clinical examination, 9 of 11 metastases were detected by ultrasonography, with 82% sensitivity (95% CI, 48%-98%); specificity was 79% (95% CI, 73%-84%), PPV was 17% (95% CI, 12%-23%), and NPV was 99% (95% CI, 96%-100%).

CONCLUSIONS AND RELEVANCE

In this diagnostic study, among a cohort of patients with high-risk cutaneous SCC of the head and neck, baseline ultrasonography was more sensitive than clinical examination alone for the detection of lymph node metastasis. In the setting of a negative result at baseline clinical examination, ultrasonography had high sensitivity for detecting nodal metastases, but this sensitivity should be evaluated against the high rate of false-positive findings.

The diagnostic value of ultrasonography to detect occult lymph node involvement at different levels in patients with squamous cell carcinoma in the maxillofacial region

The purpose of this study was to evaluate ultrasonography (US) and computed tomography (CT) in detecting lymphnodes of the neck affected with squamous cell carcinoma. From 1987 to 1999 the data from 203 untreated patients with a diagnosis of cancer in the maxillofacial have been investigated. Of these, 115 had a primary squamous-cell carcinoma. US diagnosis was made by an oral- and maxillofacial surgeon experienced in US of the head and neck. CT diagnosis was made by a well-experienced radiologist. The following lymph node levels were assesed: level I (submental and submandibular lymphnodes), level II (lymphnodes distal to level I and confined to the region above the skin crease at or just below the level of the thyroid notch), level III (lymphnodes distal to level II and confined to the anterior cervical triangle including those deep to the sternocleidomastoid muscle), and level IV (lymphnodes distal to level III and confined to the posterior cervical triangle). For all levels US yielded a sensitivity of 71%, and a specificity of 87%, while CT showed a sensitivity of 32% and a specificity of 96%. The sensitivity of US decreased from level I to level IV, whereas the specificity increased from level I to level IV. For lymphnode levels I and II US may be useful for the detection of local metastases while for the other levels the application of advanced techniques of US may have to be investigated.

Contrast-enhanced color Doppler sonography for evaluation of enlarged cervical lymph nodes in head and neck tumors

OBJECTIVE

We determined whether contrast-enhanced color Doppler sonography can differentiate benign from malignant enlarged cervical lymph nodes in head and neck tumors.

SUBJECTS AND METHODS

Ninety-four enlarged lymph nodes in 39 adult patients (32 men and seven women; age range, 30-81 years) were examined with B-mode sonography and with unenhanced and contrast-enhanced color Doppler sonography. All patients had carcinoma of the oral cavity. Histologically, lymphadenitis was found in 57 nodes and metastases in 37 nodes. Geometric dimension, texture, and margin of the node and detection and location of vessels were noted. Histology and imaging findings were correlated.

RESULTS

The transverse-to-longitudinal diameter ratio in combination with texture and margin analysis resulted in a correct diagnosis in only approximately 79% of the nodes. With contrast-enhanced color Doppler sonography, 86% of nodes showed vessels, and 28% of nodes showed vessels with this technique exclusively. Characteristic configurations were identified: hilar vessels with branching indicated lymphadenitis (sensitivity, 98%; specificity, 100%), and predominantly peripheral vessels indicated metastases (100%, 98%). These findings changed the diagnosis in 13 nodes, changed the therapy in four patients, and led to an incorrect diagnosis in one patient.

CONCLUSION

Enlarged lymph nodes can be characterized as metastatic or inflammatory with high diagnostic accuracy on the basis of their vascular architecture as seen on contrast-enhanced color Doppler sonography.

Lymph node metastases: CT and MRI

Imaging is playing a major role in the assessment of cervical lymphadenopathy. In head and neck malignancies, imaging can be helpful for staging, and sometimes in differentiating different types of metastases, such as squamous cell carcinomas, non-hodgkins disease and thyroid carcinomas. This article on imaging of cervical lymph node metastases will describe both radiological and clinical aspects. Computed tomography (CT) and magnetic resonance (MR) are widely used for primary tumor and nodal imaging. However, very seldom these modalities have clinical consequences for the management of the neck, such as a wait-and-see policy if no nodes are depicted. This is caused by the limited accuracy of both modalities caused by the fallibility of radiologic criteria for metastases. Ultrasound (US) is hampered by similar morphologic criteria, and only US-guided fine needle aspiration cytology (FNAC) can offer additional cytologic criteria which are more reliable.

Diagnostic evaluation of the neck

Modern imaging for neck node metastases has increased our ability to predict the presence of palpably occult metastases and predict the resectability of large metastases. The accuracy of CT scans, MR imaging, but especially US-FNAC, has increased to such an extent that clinical consequences are being attached to negative radiologic findings in a clinically N0 neck. As not all palpably occult metastases can be detected, especially when they are micrometers, follow-up should be strict, with use of US-FNAC in the neck observed instead of treated electively.

Contrast-enhanced computed tomography of freely transplanted jejunal patches after surgery in the oral cavity and oropharynx

Thirty-two patients with jejunal patches in the oral cavity and oropharynx were examined with contrast-enhanced computed tomography (CT). The morphology of the intestinal wall, the mesenteric fatty tissue, mesenteric lymph nodes and the region of the anastomosis were evaluated and correlated with clinical findings. The size and number of mesenterial lymph nodes, which were found in 16 cases, did not correlate with tumour recurrence. The benign alterations of the intestinal wall ranged from flat surfaces with no enhancement to significant enhancement with persisting folds. In cases with persisting intestinal folds, separation of tumour recurrence was achieved by identification of a typical double-layer configuration of the enhancing mucosa, and the restriction of alterations to the intestinal wall compartment. Tumour recurrences occurred at patch margins and were reliably distinguished from normal patches. For differentiation of findings, an exact localization of jejunum-patch compartments and margins with contrast-enhanced CT was, therefore, necessary.

Differentiation between benign and metastatic cervical lymph nodes with ultrasound

OBJECTIVE

The differentiation between benign and metastatic lymph nodes with ultrasound (US) is based primarily on the evaluation of size, shape, margin and internal echo structure. The aim of this study is to determine whether these parameters are reliable indicators and to correlate internal echo structure and histopathological findings.

MATERIALS AND METHODS

Seventy-one nodes in 21 patients with pathologically proven oral squamous cell carcinoma were examined. The shortest diameter, the short/long diameter ratio (S/L ratio), margins and internal echo structure of the lymph node were evaluated by US. The internal echo structure was divided into six patterns: homogeneous hypoechoic, homogeneous hyperechoic, heterogeneous, eccentric hyperechoic, centric hyperechoic and anechoic pattern. In addition, internal echo structure was correlated with histopathological findings.

RESULTS

In 71.4% of the metastatic nodes, the shortest diameter was more than 10 mm and the S/L ratio was higher than that of benign nodes (average 0.71). Eleven (84.6%) of the 13 lymph nodes with irregular margins were metastatic. Heterogeneous and anechoic patterns were observed in metastatic nodes, whereas homogeneous hypoechoic and eccentric hyperechoic patterns were present in benign nodes. On ultrasonography with the corresponding histopathological findings, echogenic areas in the homogeneous hyperechoic, heterogeneous and centric hyperechoic patterns of metastatic nodes proved to be necrosis or fibrosis. Eccentric hyperechoic areas in benign nodes corresponded to the hilus and surrounding fatty tissue.

CONCLUSIONS

The shortest diameter, S/L ratio, margin and internal echo structure were considered to be critical indicators to differentiate between benign and metastatic nodes. Secondary changes caused by tumour infiltration, necrosis, or fibrosis should be assessed when metastatic lymph nodes are differentiated from benign ones by internal echo structure.

Radioimmunoscintigraphy of squamous cell carcinoma in the head and neck region

Radioimmunoscintigraphy (RIS) with a monoclonal antibody (SQ 174, Biomira, Inc, Canada) is introduced as a new diagnostic method for detection of squamous cell carcinoma in the head and neck region. RIS could detect eight of the 10 primary tumors. One patient had histopathologically proven nodal disease. This lymph node was seen in RIS. No false positive results were seen in RIS, whereas sonography, computed tomography, and magnetic resonance imaging showed several false positive results.

Modern imaging techniques and ultrasound-guided aspiration cytology for the assessment of neck node metastases: a prospective comparative study

Although palpation has proved to be an unreliable staging procedure, the indications for and the implications of more reliable radiologic staging methods for the neck in patients with a primary squamous cell carcinoma of the head and neck remain controversial. Only a very accurate imaging technique can replace neck dissection in clinical NO disease. This study compares the value of palpation with computed tomography (CT), magnetic resonance imaging (MRI) and ultrasound (US) with or without guided aspiration cytology for neck node staging. One hundred and thirty-two patients with squamous cell carcinoma of the head and neck were examined radiologically before undergoing a total of 180 neck dissections as part of their treatment. CT, US and MRI proved to be significantly more accurate than palpation for cervical lymph node staging. The accuracy of US-guided aspiration cytology was significantly better than of any other technique used in this study. Modern imaging techniques are essential for appropriate assessment of neck node metastases. In view of advances in the accuracy of contemporary imaging, the need for elective treatment of the neck requires reappraisal.

The role of flexible endosonography in diagnostic imaging of carcinomas of the oral cavity and oropharynx

Currently-used imaging methods often fail to depict carcinomas of the oral cavity and oropharynx properly. Scanning these tumours with recently developed digitally-guided transducers of 5 and 7.5 MHz, we found that diagnosis and assessment are significantly facilitated by flexible endosonography. The demonstration of T1- and T2-tumours was shown to be facilitated by this technique as compared to other imaging methods. The assessment of large tumours is improved in special areas. Nonetheless, we recommend that in such cases flexible endosonography be supported by CT or MRI for a complete demonstration of the extent of the tumour.

Lymph node staging in patients with clinically negative neck examinations by ultrasound and ultrasound-guided aspiration cytology

To assess the usefulness of ultrasound and ultrasound-guided fine-needle aspiration cytology in detecting occult cervical lymph node metastases, 107 patients with squamous cell carcinoma of the head and neck, who underwent 132 elective neck dissections, had preoperative ultrasound examination. During the assessment of the last 54 patients, who underwent 70 elective neck dissections, ultrasound-guided aspiration cytology was available. Although ultrasound was able to detect lymph node metastases in the majority of patients, the accuracy of this technique never exceeded 70% (93 of 132 procedures). With ultrasound-guided aspiration cytology, accuracy was 89% (62 of 70 procedures). This latter technique seems to be the modality of choice for the assessment of the clinically negative neck. Its use obviates the need for elective neck dissection, done because of the high risk of occult metastasis.

Computed tomography in nasopharyngeal carcinoma: Part I: T-stage conversion with CT-staging

Two hundred seventeen consecutive patients were treated with radiotherapy alone, with curative intent, from 1970 to 1985 at the Radiotherapy Unit of the University and Hospital of Florence. The distribution according to T and N staging with polytomography was compared to patients (106 out of 217) who had CT scans done at presentation. T1 cases were less frequent (6.6% vs 27%) in the CT-staged series, whereas T3 showed a higher incidence (30.2% vs 12.6%). The advantages of CT over conventional tomography were quantitated in a subset of 97 patients who underwent both staging procedures. Site-by-site, CT displayed a higher percentage of involvement than polytomography: parapharyngeal spread 18% vs 2%, oropharynx 16% vs 8%, choanae and nasal cavities 28% vs 13%, ethmoid and maxillary sinus 29% vs 13%. Information provided by CT caused a T-stage conversion in 23 out of 97 cases (23%): 4 out of 11 T1, 16 out of 44 T2, 3 out of 16 T3.