Morphological MRI criteria improve the detection of lymph node metastases in head and neck squamous cell carcinoma: multivariate logistic regression analysis of MRI features of cervical lymph nodes

European Radiology, Mar 2009

The aim was to evaluate whether morphological criteria in addition to the size criterion results in better diagnostic performance of MRI for the detection of cervical lymph node metastases in patients with head and neck squamous cell carcinoma (HNSCC). Two radiologists evaluated 44 consecutive patients in which lymph node characteristics were assessed with histopathological correlation as gold standard. Assessed criteria were the short axial diameter and morphological criteria such as border irregularity and homogeneity of signal intensity on T2-weighted and contrast-enhanced T1-weighted images. Multivariate logistic regression analysis was performed: diagnostic odds ratios (DOR) with 95% confidence intervals (95% CI) and areas under the curve (AUCs) of receiver-operating characteristic (ROC) curves were determined. Border irregularity and heterogeneity of signal intensity on T2-weighted images showed significantly increased DORs. AUCs increased from 0.67 (95% CI: 0.61–0.73) using size only to 0.81 (95% CI: 0.75–0.87) using all four criteria for observer 1 and from 0.68 (95% CI: 0.62–0.74) to 0.96 (95% CI: 0.94–0.98) for observer 2 (p < 0.001). This study demonstrated that the morphological criteria border irregularity and heterogeneity of signal intensity on T2-weighted images in addition to size significantly improved the detection of cervical lymph nodes metastases.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://link.springer.com/content/pdf/10.1007%2Fs00330-008-1187-3.pdf

Morphological MRI criteria improve the detection of lymph node metastases in head and neck squamous cell carcinoma: multivariate logistic regression analysis of MRI features of cervical lymph nodes

R. B. J. de Bondt P. J. Nelemans F. Bakers J. W. Casselman C. Peutz-Kootstra B. Kremer P. A. M. Hofman R. G. H. Beets-Tan 0 1 2 3 0 C. Peutz-Kootstra Department of Pathology, Maastricht University Medical Center , Maastricht, The Netherlands 1 J. W. Casselman Department of Radiology, AZ St. Jan Hospital , Bruges, Belgium 2 P. J. Nelemans Department of Epidemiology, Maastricht University Medical Center , Maastricht, The Netherlands 3 B. Kremer Department of Otolaryngology/ Head and Neck Surgery, Maastricht University Medical Center , Maastricht, The Netherlands The aim was to evaluate whether morphological criteria in addition to the size criterion results in better diagnostic performance of MRI for the detection of cervical lymph node metastases in patients with head and neck squamous cell carcinoma (HNSCC). Two radiologists evaluated 44 consecutive patients in which lymph node characteristics were assessed with histopathological correlation as gold standard. Assessed criteria were the short axial diameter and morphological criteria such as - The presence of cervical lymph node metastases is an important prognostic factor in patients with HNSCC as it worsens significantly the treatment outcome [18]. The border irregularity and homogeneity of signal intensity on T2-weighted and contrast-enhanced T1-weighted images. Multivariate logistic regression analysis was performed: diagnostic odds ratios (DOR) with 95% confidence intervals (95% CI) and areas under the curve (AUCs) of receiver-operating characteristic (ROC) curves were determined. Border irregularity and heterogeneity of signal intensity on T2-weighted images showed significantly increased DORs. AUCs increased from 0.67 (95% CI: 0.610.73) using size only to 0.81 (95% CI: 0.750.87) using all four criteria for observer 1 and from 0.68 (95% CI: 0.620.74) to 0.96 (95% CI: 0.940.98) for observer 2 (p < 0.001). This study demonstrated that the morphological criteria border irregularity and heterogeneity of signal intensity on T2-weighted images in addition to size significantly improved the detection of cervical lymph nodes metastases. choice of management depends on the existence and extent of lymph node metastases in the neck. Therapy could consist of surgery, radiation therapy, chemotherapy, or a combined therapy. However, all these therapies have a considerable morbidity and mortality. Therefore, treatment refinements like limitation of the field of radiation or a heterogeneity of the signal intensity on T2-weighted more selective neck dissection are necessary [915]. MRI images and heterogeneity of enhancement on post-contrast is one of the imaging techniques that is used to guide T1-weighted images. treatment decisions, but the ability of MRI to discriminate between lymph nodes with and without metastasis is still poor [16]. Materials and methods A commonly used criterion for the assessment of cervical lymph node involvement on MRI is the short Patients axial diameter, and several studies have been undertaken to determine the optimal cutoff size of the short axial diameter This study was approved by the local medical ethics for discrimination between metastatic and non-metastatic committee. Between January 2002 and December 2006 a lymph nodes. On MRI a commonly used size cutoff point is series of 44 consecutive patients [mean age was 61 years a short axial diameter of 10 mm, but a range varying from 9 (range: 4086 years); 11 women and 33 men] with a to 15 mm has been described [1721]. The challenge for HNSCC who underwent a unilateral or bilateral supraradiologists remains the detection of metastases in small omohyoid neck dissection (SOHND) (dissection of levels lymph nodes with a short axial diameter below 10 mm, I-III) or radical (modified) neck dissection [R(M)ND] because mere use of the size criterion will result in (dissection of levels I-V) was studied. All patients misclassification of these nodes as normal. In this respect, underwent MR imaging as part of the routine diagnostic according to the literature, the performance of MRI is still workup. This was performed on a 1.5-T machine poor for detection of lymph node metastases, and this (Gyroscan, Powertrack 6000, Philips, Best, The Netherconcerns especially the detection of metastases in small lands) by using a head-neck coil (Philips, Best, The lymph nodes [16]. Netherlands). Table 1 presents the relevant parameters of In addition to the size criterion, the use of morphological the MRI protocol. criteria might have added value to detect metastatic disease in lymph nodes. The diagnostic value of morphological criteria has not been well evaluated in patients with MR evaluation HNSCC. In patients with rectal cancer, Kim et al. demonstrated that in addition to the size criterion, One general radiologist (observer 1) and one radiologist morphological criteria such as borders and heterogeneity specialized in head and neck imaging (observer 2) could be helpful signs to predict nodal involvement [22]. independently evaluated the MR images retrospectively. The aim of our study was to evaluate the accuracy and Both observers were blinded to each others MR assessadditional diagnostic value of morphological criteria ments, clinical information, and the results of the histoobserved on MRI images, such as border irregularity and logical examination. Table 1 Parameters of the head and neck MRI protocol SPIRa TSE T2-weighted SPIR CEb T1-weighted SPIR CE T1-weighted Field of view (mm) Section thickness (mm) No. of signals acquired Acquisition time (min:s) Fig. 1 Schematic drawing of the four borders for assessment of lymph nodes In the first session, all lymph nodes were determined by position and slice number per neck level. The short axial diameter was recorded per lymph node. All lymph nodes were classified as normal (10 mm) or malignant (>10 mm). After a time interval of 8 weeks, the same MR images were reviewed again by the same radiologists blinded to each others MR assessments, clinical information, and the results of the histological examination. For every lymph node the decision was made whether it was normal or metastatic using three morphologic criteria. The first criterion was border as observed on fat-suppressed T2weighted images (smooth, lobulated, spiculated, indistinct) (Fig. 1 demonstrates the four different borders on the Fig. 2 Transverse turbo spinecho T2-weighted images of the neck. Examples of the four different scores on border regularity. The lymph nodes (arrows) show (A) a smooth border, (B) a lobulated border, (C) a spiculated border, and (D) an indistinct border schematic drawing). Smooth (Fig. 2A) and lobulated (Fig. 2B) borders were considered as indicative of normal lymph nodes. Spiculated (Fig. 2C) and indistinct (Fig. 2D) borders were considered as indicative of metastatic lymph nodes. The second criterion was degree of homogeneity of the signal intensity on contrast-enhanced T1-weighted images. A homogeneous enhancement was considered as normal. A heterogeneous enhancement was considered as indicative of metastatic lymph nodes. The third criterion was degree of homogeneity of the signal intensity on T2weighted images. A homogeneous signal intensity was considered as normal. A heterogeneous signal intensity was considered as indicative of metastatic lymph nodes (Fig. 3). Pathological examination All neck dissections were performed en bloc. Immediately after the resection, the surgeon positioned the neck dissection specimen on a schematic drawing of the neck in real proportions and fixed it with needles (Fig. 4). The pathologist manually identified and localized the lymph nodes per neck level in the specimen. The short axial diameter of all lymph nodes was recorded. Subsequently, lymph nodes were fixed, sectioned, and hematoxylineeosine (HE) stained, and the presence of tumor in each lymph node was examined microscopically. The pathological results were used as the reference standard. Fig. 3 Transverse turbo spin-echo T2-weighted image of the neck showing an example of heterogeneous signal intensity (SI) in a lymph node in level II on the right side. The lymph node (arrowhead) with a homogenous SI, although a short axial diameter of 13 mm, showed no metastasis, whereas the smaller lymph node (arrow) (short axial diameter 9 mm) with a heterogeneous and eccentric area of low SI (small arrow) revealed metastasis at the pathologic examination Fig. 4 Schematic drawing of the different levels in the neck Matching MR assessment to pathological examination The neck was subdivided according to the six different levels in agreement with the classification of the American Joint Committee on Cancer (AJCC); this classification was used by the radiologists when viewing the MR images and used by the pathologist when interpreting the specimen after neck dissection [23]. The results of the measurements on MRI were compared with the results of the pathologic examination of the neck dissection specimens. By recording the combination of the short axial diameter and the exact location of each lymph node, it was possible to perform a topographical correlation for each lymph node per neck level. Inter-observer agreement with respect to the evaluation of the criterion nodal size, and the criteria border irregularity, homogeneity of contrast enhancement on T1weighted images, and homogeneity of signal intensity on T2-weighted images was expressed by Cohens kappa coefficient (). For each lymph node, the scores for the MRI criteria were correlated with the outcome of the histological examination of the neck dissection specimen: presence or fied) neck dissections (in 2 patients a bilateral neck absence of lymph node metastasis. Diagnostic perfor- dissections). Distribution of the dissected neck levels was mances of nodal size and the new criteria for the diagnosis as follows; 32 level I, 41 level II, 34 level III, 25 level of metastasis were evaluated using diagnostic parameters, IV, and 8 level V. such as sensitivity, specificity, and diagnostic odds ratio At pathological examination a total of 261 lymph nodes (DOR). In diagnosis, one always has to compromise was found: 71 lymph nodes with metastasis and 190 between sensitivity and specificity: as sensitivity increases, without metastasis (prevalence = 27.2%). In all levels specificity will decrease and vice versa. The DOR is a together, a total of 111 lymph nodes was found in level measure of diagnostic performance, which incorporates II, 52 lymph nodes in level I, 65 lymph nodes in level III, sensitivity and specificity and thus captures the trade-off 25 lymph nodes in level IV, and 8 lymph nodes in level V. between these measures. A DOR = 1 indicates that the test Lymph node metastases were present in 33 patients (patient under study has no diagnostic value; a DOR >1 indicates prevalence = 80.5%). The majority (82%) of 261 assessed that the test under study (in this case, the new MRI criteria) lymph nodes was scored as having a size <10 mm. The has the ability to discriminate between lymph nodes with average pathologic yield per patient was 11.5 lymph nodes. and without metastasis. The larger the DOR is, the larger At assessment of the MR images of the dissected neck the discriminative ability [24]. levels, a total of 360 lymph nodes was detected. Of these lymph Multivariate logistic regression analysis was used to nodes, 99 were shown to have a short axial diameter of 3 mm. evaluate whether a criterion, when used in combination Because these lymph nodes were not detected at pathological with the other criteria, can be used as an independent examination, they were excluded from further analysis. indicator for differentiating metastatic from non-metastatic A total of 261 lymph nodes, which was found at lymph nodes. The independent contribution of each pathological examination, was matched to lymph nodes criterion to the diagnostic performance is expressed as that were detected on MRI by the two observers. the DOR, which can be derived by exponentiation of the For observer 1, scores were lacking in 1 lymph node corresponding regression coefficient, where 95% con- metastasis, leaving 260 lymph nodes for analysis: 70 lymph fidence intervals (95% CI) are used to indicate whether nodes with metastasis and 190 without metastasis. For observer the DOR is significantly different from DOR = 1. Multi- 2, complete data were available for all 261 lymph nodes. variate logistic regression analysis was also used to Observer agreement was = 1.00 for size, = 0.62 for examine which of the new MRI criteria were most border irregularity, = 0.51 for signal intensity on contrastpredictive of the presence of metastases and whether the enhanced T1-weighted images, and = 0.51 for signal use of these new criteria in addition to size significantly intensity on T2-weighted images. improved the detection of cervical lymph nodes metas- Table 2 shows the sensitivity, specificity, and diagnostic tases in HNSCC. odds ratio (DOR) with 95% confidence interval (95% CI) for The dependent variable in these models was the presence the diagnosis of metastasis for the criteria border (abnormal or absence of metastasis according to the pathological versus normal) on T2-weighted images, signal intensity on examination. The models incorporated as independent contrast-enhanced T1-weighted images (heterogeneous vervariable size as well as the three new MRI criteria. All sus homogeneous), and signal intensity on T2-weighted criteria were entered as dichotomous variables, where images (heterogeneous versus homogeneous) for the two abnormal and normal results were coded as 1 and 0, observers. With respect to the new criteria, the highest DORs respectively. are found for border irregularity and signal intensity on T2 Predicted probabilities of metastasis from these models weighted images. The lowest DORs are found for the were used to calculate the area under the curve (AUC) with criterion contrast enhancement on T1-weighted images. 95% CI as measure of diagnostic performance. The AUC Table 3 shows the results for lymph nodes that were of the model incorporating both the new criteria and the smaller than 10 mm. The DORs associated with the new size criterion was compared with the AUC of the model criteria are smaller compared to the DORs for all lymph that incorporated only the size criterion. The difference in nodes, but are still elevated, indicating that the morphoAUCs for the different models was tested using the method logic criteria are also useful for the detection of metastases described by Hanley et al., which accounts for the fact that in small lymph nodes. the AUCs are derived from the same sample of patients Table 4 shows the results from multivariate logistic [25]. A p-value <0.05 was considered to be statistically regression analyses. The DORs associated with border significant. irregularity were 2.61 (95% CI: 1.126.08) and 66.2 (95% CI: 20.4 -217) for observer 1 and 2, respectively; for heterogeneity on T2-weighted images DORs were 2.97 Results (95% CI: 1.426.18) and 22.6 (95% CI: 6.4080.1). The DORs associated with heterogeneity on T1-weighted In 44 patients a neck dissection was performed: 9 images are not significantly different from DOR = 1. supraomohyoid neck dissections and 37 radical (modi- Another observation was that for observer 2, the DOR Table 2 Sensitivity, specificity, and diagnostic odds ratio (DOR) with 95% confidence interval (95% CI) for the diagnosis of metastasis for the MR criteria: size, border irregularity, signal intensity (SI) on contrast-enhanced (CE) T1-weighted images and SI on T2-weighted images for all lymph nodes Sensitivity Specificity DOR (95% CI) Sensitivity Specificity DOR (95% CI) 8.16 (4.0916.3) 42 8.68 (4.6916.1) 87 4.81 (2.658.71) 61 6.99 (3.8312.7) 93 Short axial diameter (mm) >10 versus 10 Borders on T2-weighted images spiculated/ indistinct versus smooth/lobulated SI on CE T1-weighted images heterogeneous 71 versus homogeneous SI on T2-weighted images heterogeneous versus homogeneous Table 3 Sensitivity, specificity, and diagnostic odds ratio (DOR) with 95% confidence interval (95% CI) for the diagnosis of metastasis for the MR criteria: size, border irregularity, signal intensity (SI) on contrast-enhanced (CE) T1-weighted images and SI on T2-weighted images for lymph nodes <10 mm in short axial diameter Lymph nodes <10 mm Sensitivity Specificity DOR (95% CI) Sensitivity Specificity DOR (95% CI) Borders on T2-weighted images spiculated/ indistinct versus smooth/lobulated SI on CE T1-weighted images heterogeneous 53 versus homogeneous SI on T2-weighted images heterogeneous versus homogeneous 4.61 (2.129.99) 78 2.52 (1.255.08) 42 4.24 (2.068.71) 88 Table 4 Results of multivariate logistic regression analyses. Regression coefficients (), diagnostic odds ratios (DOR) with 95% confidence intervals (95% CI) for the size criterion, and the three new assessed MR criteria border, signal intensity (SI) on contrast-enhanced (CE) T1-weighted images and T2-weighted images for both MR observers Size >10 versus 10 mm Border on T2-weighted images spiculated/indistinct versus smooth/lobulated SI on CE T1-weighted images heterogeneous versus homogeneous SI on T2-weighted images heterogeneous versus homogeneous the degree of border irregularity on T2-weighted images, similar feature that can be observed around a primary homogeneity of the signal intensity on contrast-enhanced tumor site. Thirdly, tumor infiltration and presence of T1-weighted images, and the homogeneity of the signal softening or necrosis within lymph nodes usually show on intensity on T2-weighted images of lymph nodes. the pathological examination an irregular and heterogene The results of univariate analyses, with which we ous pattern. This architectural distortion of the nodal compared the diagnostic performance of all four criteria parenchyma results in an irregular signal intensity on T2under study, indicate that all criteria, when used alone, help to weighted images and in a heterogeneous signal intensity on discriminate between lymph nodes with and without metas- the contrast-enhanced T1-weighted images. tasis, but that the discriminative ability was lowest for degree The discrepancies in diagnostic performance between of homogeneity of signal intensity on contrast-enhanced T1- the two observers and low inter-observer agreement weighted images. This finding also holds after restricting the regarding the new morphological criteria could be analysis to lymph nodes smaller than 10 mm. However, explained by the fact that observer 2, experienced in when compared with the analysis involving all lymph nodes, head and neck radiology, was better aware of the DORs were lower, indicating that within small nodes the diagnostic value of border irregularity and homogeneity discriminative ability is poorer. An exception was border of signal intensity than his colleague who had less irregularity, which for observer 2 was still associated with a experience in the assessment of lymph node involvehigh DOR of about 100 (compared with a DOR of 112 in the ment. In this respect, the results also show that the analysis based on all lymph nodes). criterion size for lymph nodes has limited discriminative Another morphological criterion that has been used is the value for the experienced observer 2. For the more presence of central necrosis. However, central necrosis is experienced observer, border irregularity and heterogeoften seen in the larger involved lymph nodes, which based neity of signal intensity on T2-weighted images are on their size only would have been classified as malignant more decisive in the diagnosis. This is reflected by the anyhow [1720]. The majority of lymph nodes in our series results of the univariate analysis with high DORs for the (82%) showed subcentrimetrical short axial diameters. criteria border and signal intensity on T2-weighted Curtin et al. have already described the diagnostic value images (112 and 28.6, respectively), when compared of the appearance of internal abnormalities in lymph nodes with the low DOR for the size criterion (7.96). Also, in on CT and MRI; however, the performance of MRI was not the multivariate logistic regression analysis high DORs changed significantly by the addition of information on are presented for those criteria (66.2 and 22.6, internal abnormalities [18]. respectively). This is in contrast to the DOR of 1.02 Evaluation of the results of multivariate logistic regression for the size criterion. analysis indicates that, if used in combination with the size criterion, for both observers the scores on border irregularity and heterogeneous signal intensity on T2-weighted images Conclusion contribute significantly to the prediction of the presence of metastatic lymph nodes, whereas the criterion heterogeneity Newly assessed morphological criteria like border irregularof signal intensity on contrast-enhanced T1-weighted images ity and heterogeneity of signal intensity on T2-weighted has no additional diagnostic value. images in addition to size significantly improved the Explanations for the MR features in metastatic lymph detection of cervical lymph node metastases on MRI in nodes could be the following. Firstly, changes of a smooth patients with HNSCC. or lobulated border of normal lymph nodes into a spiculated or indistinct border could be due to direct Open Access This article is distributed under the terms of the extra-nodal tumor infiltration into the peri-nodal fatty Creative Commons Attribution Noncommercial License which tissue. Secondly, this phenomenon could be explained by a permits any noncommercial use, distribution, and reproduction in desmoplastic reaction around the affected lymph node, the any medium, provided the original author(s) and source are credited.


This is a preview of a remote PDF: http://link.springer.com/content/pdf/10.1007%2Fs00330-008-1187-3.pdf

R. B. J. de Bondt, P. J. Nelemans, F. Bakers, J. W. Casselman, C. Peutz-Kootstra, B. Kremer, P. A. M. Hofman, R. G. H. Beets-Tan. Morphological MRI criteria improve the detection of lymph node metastases in head and neck squamous cell carcinoma: multivariate logistic regression analysis of MRI features of cervical lymph nodes, European Radiology, 2009, 626-633, DOI: 10.1007/s00330-008-1187-3