Ultrasonographic evaluation of small cervical lymph nodes in head and neck cancer

Clinical, ultrasound and histopathological correlation of clinically N0 neck nodes in patients with cancers of the pharynx and larynx

Background: The presence of metastatic cervical adenopathy is essential for treatment planning and prognosis assessment. Treatment of patients with head and neck cancer with clinically negative cervical lymphadenopathy (N0) remains controversial. Neck palpation, as the method used in tumor, node, metastasis (TNM) staging, has limitations and can provide false negative results in some cases. Lymph node metastases are associated with a reduced survival rate but at the same time, neck dissection for the patient with N0 neck is not without risks or complications. Objectives: In prospective study, we compared palpation, ultrasonography (US) examination of the neck and histopathological examination in patients with cancers of the pharynx and larynx. Patients, Materials and Methods: Forty-six patients with cancers of the pharynx and larynx that presented with a N0 neck were prospectively analyzed. They were divided in two groups: 23 patients operated with an external approach including the control of the lymph node areas, and a second group of 23 patients operated using endoscopy and carbon dioxide (CO₂) laser, no neck dissection – “watchful waiting policy”. All patients have had a flexible endoscopy of the pharynx and larynx, US of the neck and all received surgical treatment for their primary tumor. Imaging was performed in selected cases. All the removed lymph nodes were sent for histopathology. US was also used as a follow-up method. The US features of the examined lymph nodes were: diameters [longitudinal (L) and transverse (T)]; the ratio of the two diameters (L/T); shape; lymph node area; central hypodensity; regular/irregular margins; aspect (homogeneous or not). Results: US has detected 25 lymph nodes in the open surgery group and intraoperatively, we excised 31 (sensitivity of 80.6%). Ten lymph nodes showed metastases, with 100% accuracy of US, which have been confirmed both pathologically and immunohistochemically. US in the second group – patients treated with CO₂ laser – detected at four patients 10 cervical lymph nodes that did not presented any malignant features. At recurrence alone, the US confirmed 100% presence of nodes metastases. Conclusions: US was superior to palpation and this method can be recommended as a diagnostic tool in preoperative assessment of patients without palpable metastasis (N0).

The outcomes and decision-making process for neck lymph nodes with indeterminate fine-needle aspiration cytology

Objectives

This study aims to propose a cytological classification, to evaluate predictive factors of the final malignancy, and to suggest a proper management strategy for neck lymph nodes (LNs) with indeterminate cytology.

Methods

Patients who had neck lymphadenopathy with indeterminate cytology between 2007 and 2017 were analyzed retrospectively in a tertiary medical center. Cytological classification was conducted according to the cytological descriptions. We examined the clinical characteristics according to the final diagnosis of the neck lymphadenopathy.

Results

According to the final diagnoses, there were 142 malignant and 95 benign neck LNs among 237 patients. Multivariate analyses using a stepwise logistic regression model showed that cytological classification [p < 0.001, OR = 5.67 (3.48–9.23)], prior history of malignancy [p = 0.01, OR = 2.97 (1.26–6.99)], long axis [p = 0.01, OR = 3.06 (1.33–7.06)], short-to-long axis (S/L) ratio [p = 0.047, OR = 2.15 (1.01–4.57)] and internal echogenicity [p = 0.01, OR = 2.72 (1.26–5.86)] were independent predictors of malignancy.

Conclusions

In patients who have neck LNs with indeterminate cytology, a cytological classification and four other predictors (prior history of malignancy, long axis ≥ 1.93 cm, S/L ratio ≥ 0.64 and heterogeneity of internal echogenicity) are statistically associated with the risk of malignancy and helpful in guiding further management.

Ultrasonographic Evaluation of Cervical Lymph Nodes in Thyroid Cancer

Objective To determine what ultrasonographic features can identify metastatic cervical lymph nodes, both preoperatively and in recurrences after complete thyroidectomy. Study Design Prospective. Setting Outpatient clinic, Department of Head and Neck Surgery, School of Medicine, University of São Paulo, Brazil. Subjects and Methods A total of 1976 lymph nodes were evaluated in 118 patients submitted to total thyroidectomy with or without cervical lymph node dissection. All the patients were examined by cervical ultrasonography, preoperatively and/or postoperatively. The following factors were assessed: number, size, shape, margins, presence of fatty hilum, cortex, echotexture, echogenicity, presence of microcalcification, presence of necrosis, and type of vascularity. The specificity, sensitivity, positive predictive value, and negative predictive value of each variable were calculated. Univariate and multivariate logistic regression analyses were conducted. A receiver operator characteristic (ROC) curve was plotted to determine the best cutoff value for the number of variables to discriminate malignant lymph nodes. Results Significant differences were found between metastatic and benign lymph nodes with regard to all of the variables evaluated ( P < .05). Logistic regression analysis revealed that size and echogenicity were the best combination of altered variables (odds ratio, 40.080 and 7.288, respectively) in discriminating malignancy. The ROC curve analysis showed that 4 was the best cutoff value for the number of altered variables to discriminate malignant lymph nodes, with a combined specificity of 85.7%, sensitivity of 96.4%, and efficiency of 91.0%. Conclusion Greater diagnostic accuracy was achieved by associating the ultrasonographic variables assessed rather than by considering them individually.

Application of Computer-Aided Diagnosis to the Sonographic Evaluation of Cervical Lymph Nodes

We initiated an observer study to evaluate a computerized system developed in our previous study for automatic extraction of 10 features and estimation of the malignancy probability of cervical nodes in sonograms. In the present study, five expert radiologists and five resident radiologists interpreted the sonograms of 178 nodes. The malignancy rating and patient management recommendation (biopsy or follow-up) were made without and then with the computer aid. Under these two reading conditions, the performances of radiologists and agreement among a group of radiologists were evaluated by using the receiver operating characteristic (ROC) analysis and the κ statistic, respectively. With the computer aid, the performances of radiologists improved significantly, as indicated by the increase in the area under the ROC curve (Az) from 0.843 to 0.896 (p = 0.031) and from 0.705 to 0.822 (p < 0.001), for the expert and resident groups, respectively. Agreement among all 10 radiologists improved from slight to moderate as indicated by an increase in the κ value from 0.195 to 0.421 (p < 0.001). The average performance of residents with aid (Az = 0.822) was close to that of experts without aid (Az = 0.843). Results indicate that computer-aided diagnosis is useful to improve radiologist performance (especially that of inexperienced radiologists) in the ultrasonographic evaluation of cervical nodes and to reduce variability among radiologists.

Real-time and computerized sonographic scoring system for predicting malignant cervical lymphadenopathy

BACKGROUND

To identify malignant cervical lymphadenopathy, we established a real-time, computerized scoring system based on sonographic findings and demographic data.

METHODS

One hundred eight patients with neck lymphadenopathy, receiving ultrasonography and ultrasound-guided fine-needle aspiration (US-FNA), were used to construct a predictive model. This model was validated by another independent patient cohort.

RESULTS

A predictive scoring scale was proposed by multivariate logistic regression analysis: 0.06 x (age) + 4.76 x (S/L ratio) + 2.15 x (internal echo) + 1.80 x (vascular pattern). Cervical lymphadenopathy was regarded as malignant with a score >or=7. The formula was programmed into a synchronized, computerized sonographic reporting system. Prospective validation of this predictive tool showed excellent sensitivity (100%), specificity (88.0%), and overall accuracy (90.1%).

CONCLUSION

A real-time and practical sonographic scoring system was built and validated to provide the physician prompt and reliable probability guidance for performing US-FNA cytology in managing cervical lymphadenopathy.

Computer-aided diagnosis of cervical lymph nodes on ultrasonography

A computerized system differentiating cervical lymph nodes on ultrasonography as malignant or benign was developed from a database of 210 cases. Ten quantitative features representing sonographic features of size, margin, nodal border, shape, medulla ratio, medulla distribution, echogenicity, echogeneity, vascular density, and vascular pattern, were respectively calculated under the node contour segmented by an improved snake model. A rough margin based support vector machine was trained to distinguish between malignant and benign nodes using the 10 computerized features. The receiver operating characteristic (ROC) analysis was used to evaluate the performance. The developed system showed the normalized area under the ROC curve (Az, which is used as a summarized measure of the accuracy, ranges from 0.5 to 1.0) of 0.892. Compared with the radiologist's performance of Az of 0.784 this system has the potential to be an aid to radiologists in the task of distinguishing between malignant and benign cervical nodes on ultrasonography.

Ultrasonographic feature selection and pattern classification for cervical lymph nodes using support vector machines

A rough margin based support vector machine (RMSVM) classifier was proposed to improve the accuracy of ultrasound diagnoses for cervical lymph nodes. Thirty-six features belonging to 10 kinds of ultrasonographic characteristics were extracted for each of 110 lymph nodes in ultrasonograms. Comparison studies were done for three classifiers--the classical support vector machine (SVM), the general regression neural network and the proposed RMSVM, with or without the feature selection by the recursive feature elimination (RFE) algorithm, respectively, based on SVMs and the mean square error discriminant. It was indicated by experimental results that all classifiers benefited from the feature selection. The best classification performance was obtained by the RMSVM using thirteen features selected by the RMSVM based RFE, which yielded the normalized area under the receiver operating characteristic curve (A(z)) of 0.859. Compared with the radiologist's performance of A(z) of 0.787, the developed computer-aided diagnosis algorithm has the potential to improve the diagnostic accuracy.

Sonographic feature extraction of cervical lymph nodes and its relationship with segmentation methods

OBJECTIVE

The purpose of this study was to extract quantitative features for characterization of cervical lymph nodes on sonographic images and analyze the effect of a semiautomated segmentation method on the feature extraction.

METHODS

Contours of 186 cervical lymph nodes on sonographic images were separately delineated by 2 radiologists (R1 and R2) and a semiautomated segmentation method. For each node, 10 kinds of sonographic features (including 3 parameters of size; 12 parameters of margin; 4 parameters of nodal border; 10 parameters of echogeneity; and 1 parameter of shape, echogenicity, medulla ratio, medulla distribution, vascular density, and vascular pattern, respectively) were quantified by a computerized scheme based on the segmented contour. Correlations between the quantitative parameter and the radiologists' consensus grading were computed to assess the effectiveness of these parameters. Concerning the 14 best correlated parameters, the effect of the segmentation stage on the feature extraction was estimated by comparing the parameter values calculated under different segmentations in terms of relative ultimate measurement accuracy.

RESULTS

Good correlations between the computerized scheme and radiologists were seen in features of size, nodal border, shape, echogenicity, medulla ratio, medulla distribution, vascular density, and vascular pattern, whereas 10 of 12 parameters of margin features and 8 of 10 parameters of echogeneity features showed poor correlations. Paired t tests comparing the relative ultimate measurement accuracy computed using the R1-R2 and the R1-computer pairing showed no significant difference on 11 parameters for the 14 parameters analyzed.

CONCLUSIONS

The computerized feature parameters may be used as assisted indices for evaluating cervical lymphadenopathies from sonographic images. The semiautomated segmentation method satisfied the accuracy requirement of the feature extraction.

Ultrasonography of the medial iliac lymph nodes in the dog

Sixty-one medial iliac lymph nodes of 38 different dogs (eight with adenocarcinoma of the apocrine glands of the anal sac, 13 with multicentric lymphoma, six with multicentric lymphoma but in clinical remission, and 11 control dogs) were evaluated to assess the ability of ultrasound to identify and interrogate these lymph nodes across the different groups and to differentiate these groups using different sonographic parameters. Ultrasound proved to be useful to assess canine medial iliac lymph nodes. An increase in size or number of detected lymph nodes or finding rounder or heterogeneous lymph nodes could differentiate lymph nodes of dogs of the control group from lymph nodes of dogs with lymphoma or an adenocarcinoma of the apocrine glands of the anal sac. Subcategories of malignancy could not be differentiated. More studies need to be performed, both with patients with reactive lymph nodes and also focusing on other canine superficial lymph nodes, before generalizing the results of this study to other areas or diseases.

Heme oxygenase-1 expression predicts cervical lymph node metastasis of tongue squamous cell carcinomas

Heme oxygenase-1 (HO-1) is known as a stress-inducible protein. The present study was designed to investigate the relationship between HO-1 expression levels and clinical features of tongue cancer by using HO-1 responsiveness to stress as a clinical indicator. One-hundred and twelve biopsy samples from tongue squamous cell carcinomas were analyzed semiquantitatively by immunohistochemistry. Correlations between the expression level of HO-1 and the clinical features of tumors were statistically analyzed. Fifty-four cases with surgical confirmation of lymph node metastasis were examined for the association between cervical lymph node metastasis (pN) and other clinical features, including the HO-1 expression level, using logistic regression. The low HO-1 expression group contained significantly more undifferentiated samples (P=0.04) and pN positive cases (P=0.01) by univariate analysis. The low HO-1 expression group (odds ratio=8.49; 95% confidence interval=1.64-44.09, P=0.01) and an endophytic shape (odds ratio=16.79; 95% confidence interval=1.77-159.53, P=0.01) were significantly associated with an increased risk of developing lymph node metastasis by multivariate analysis. Low HO-1 expression was associated with lymph node metastasis. The expression profile suggests HO-1 could be used clinically as a marker for tumors possessing the potential for lymph node metastasis. This method could prove useful as an adjuvant method to detect lymph node metastasis and may help reduce the number of surgeries by indicating when surgery is unnecessary.

Volumetric and histologic responses to radiotherapy or radiochemotherapy of metastatic cervical lymph nodes of oral squamous cell carcinoma

PURPOSE

The relationship between the volumetric and histologic responses of metastatic cervical nodes to radiotherapy or chemotherapy in the oral and maxillofacial region is unclear. In this study, we evaluated the correlation between the initial volume and regression rate of metastatic nodes with their histologic response to preoperative radiotherapy or radiochemotherapy.

PATIENTS AND METHODS

The volume of 54 metastatic nodes in 32 patients with squamous cell carcinoma in the oral and maxillofacial region was measured by ultrasonography before and after preoperative therapy, and the rate of the volume change was calculated. All surgically removed nodes were histologically classified as poor, good, or complete response according to their histologic features.

RESULTS

There was no significant difference in initial volume among the 3 response groups. Good and complete response nodes showed a significant increase in regression rate compared with poor response nodes. All 11 nodes showing no regression were poor response nodes, and 7 with a regression rate of more than 90% were good or complete response nodes. The remaining 36 nodes (regression rate, 0% to 90%) represented all 3 types of histologic response. Of these, 7 of 9 complete response nodes were found in 5 patients who received combination chemotherapy consisting of 5-fluorouracil, leucovorin, and cisplatin.

CONCLUSIONS

The initial nodal volume before therapy is not a good indicator for the response to radiotherapy and/or chemotherapy. A regression rate of more than 90% may be a useful predictor of the effectiveness of preoperative treatments, but it was difficult to define the cutoff values in regression rates for differentiating types of histologic response.

Follow-up ultrasonography for late neck metastases of head and neck cancer

Ultrasonographic examinations to detect late cervical lymph node metastases were performed at follow-up examinations for 52 patients with head and neck cancer who had not received neck dissection. Using diagnostic criteria we established previously, we accurately diagnosed 24 nodes in 10 patients as late metastases; 21 of these were nonpalpable. Of 12 nodes in 8 patients that were diagnosed as nonmetastatic using ultrasound (US) before radical treatment of the primary tumor, 11 showed an increase in the minimal diameter, 9 showed an internal echo pattern shift from homogeneous to heterogeneous and, in 7, the echogenic hilus changed from present to absent. A diagnostic equation was used to calculate a predictive value lambda for 146 nodes. Of these, 5 of 25 with a lambda value of -1 <or= lambda, and 7 of 121 with a lambda value of < -1 were late neck metastases. The incidence of late metastasis was significantly higher in nodes whose value was -1 <or= lambda. Thus, US examination appears useful for the follow-up of cervical lymph nodes in patients with oral and maxillofacial cancer.

Ultrasonographic assessment for response to radiochemotherapy of metastatic cervical lymph nodes in head and neck cancer: usefulness of grey-scale and color doppler sonography

To predict the response of lymph node metastasis to preoperative radiochemotherapy sonographically, the correlation between ultrasonographs and histologic features was retrospectively examined in 43 metastatic cervical lymph nodes from 24 patients with squamous cell carcinoma in the oral and maxillofacial region. Ultrasonographs were compared among poor-, good-, and complete-response lymph nodes. Before radiochemotherapy, hypoechoic internal echo and intranodal blood perfusion demonstrated many complete-response nodes; in contrast, most poor-response nodes showed peripheral blood perfusion and an avascular pattern, but did not have specific internal echo intensity. Complete-response nodes showed a significant reduction in their maximum and minimum diameters after radiochemotherapy. These results indicate that ultrasonography is useful for predicting the response of cervical lymph node metastasis to radiochemotherapy.

A probability prediction rule for malignant cervical lymphadenopathy using sonography

BACKGROUND

Our purpose was to weigh various sonographic parameters as predicting malignant cervical lymphadenopathy and build a reliable prediction rule.

METHODS

One hundred and eighty-nine cervical lymph node lesions from 125 consecutive patients were used for building the prediction model. Sonographic variables, including 15 morphologic features of B-mode, 5 vascular parameters of color Doppler mode, along with age and sex, were analyzed with multivariate logistic regression to evaluate the joint effect of a set of independent variables. A prediction rule for malignant lymphadenopathy was established, and prospective validation was assessed on a new group consisting of 100 lymph nodes from another 60 consecutive patients.

RESULTS

The association of heterogeneous content, long transverse diameter, pathologic vascular pattern, high vascular density, and older age provided the most robust prediction value. Scoring scale was designed as 1x (age) + 2x (vascularity index) + 3x (short axis) + 4x (vascular pattern) + 4x (internal echo) according to the parameter estimates of multivariate logistic regression analysis. Cut-off value of score >==10 as malignancy resulted in 89.2% sensitivity and 85.2% specificity. Prospective validation also showed satisfactory results (sensitivity, 82.9%; specificity, 86.2%).

CONCLUSIONS

By measuring only 4 sonographic parameters and age, this prediction rule could provide the physician a nonconfusing and reliable probability reference for managing cervical lymphadenopathy.