| Objective：the purpose of this study was to compare with conventionalultrasound and elastography scoring system, to validate the value of strain ratioultrasound elastography (UE) in the differentiation between benign andmalignant enlarged cervical lymph nodes (LNs).Materials and methods: B-mode ultrasound (B-US), color Doppler flowimage (CDFI) and UE were examined in89lymph nodes (malignant, n=52;benign, n=37) of70patients, with histological pathology as the referencestandard obtained by ultrasound-guided core needle biopsy. At B-USexamination, LN characteristics were evaluated as follows: border, shape,short-axis diameter, echogenicy, long-to-short-axis diameter ratio (L/T), andhilum. According to the report of the Wu et al (17), the CDFI of lymph nodewas divided into5types. Type I and V were the criteria as benign lymph node,and type II, III, IV were the criteria as malignant. Two kinds of methods,4scores of elastographic classification and strain ratio (SR), were used toevaluate the ultrasound elastograms. The best cut-off point of SR was obtainedusing receiver-operating characteristic (ROC) curve analysis. Chi-square testwas used for comparing qualitative variables. The comparation of area underthe curve (AUC) was analyzed by Z test. Parameters with P <.05wereconsidered statistically significant and all tests were two tailed. SPSS13.0wasused for statistical analysis.RESULTS: Final pathologic results of all the89lymph nodes as follows:in37benign lymph nodes,16were tuberculosis,5were Histiocytic NecrotizingLymphadenitis,16were hyperplasia; in52malignant lymph nodes,14weremetastatic squamous cell carcinoma,12were metastatic adenocarcinoma,6 were the metastatic small cell carcinoma,5were lymphoma(2non Hodgkin’slymphoma,1Hodgkin’s Lymphoma,1follicular lymphoma,1mantle celllymphoma),3were metastatic thyroid papillary carcinoma,1was metastaticthyroid follicular carcinoma,1was adenosquamors carcinoma,10weremalignant lymph nodes can’t be distinguish the sources.(1)Conventional ultrasound In six B-US evaluation criterions, thehighest sensitivity was obtained from short-axis diameter (94.2%), the lowestfrom border (42.3%); the highest specificity was obtained from border (75.7%),the lowest from hilum (35.1%); the highest accuracy was obtained fromshort-axis diameter (71.9%), the lowest from border and echogenicy (56.2%).the sensitivity, specificity and accuracy of B-US were67.3%,70.3%and68.5%.28of37benign lymph nodes were type I or V in CDFI and only9lymphnodes were other types.35of52malignant lymph nodes were type II, III or IV,17malignant lymph nodes were type I or V. The statistical analysis shows thatthe differences are statistically significant (P=000), the odds ratio is6.41,contingency coefficient is0.417. And the sensitivity, specificity and accuracy ofCDFI in distinguish were67.3%,75.7%and70.8%respectively.(2) UE The results of4scoring method and strain ratio were obtainedfrom elastogram. The cutoff point of stain ratio in cervical lymph nodes was1.78, calculated from ROC curve. SR≥1.78was the cutoff point for malignantLNs, SR＜1.78for benign.13of37benign LNs were scoring as1or2point in4scoring method,24were SR＜1.78.46of52malignant lymph nodes werescoring as3or4point in4scoring method,51were SR≥1.78. The sensitivity,specificity and accuracy of4scoring method were88.4%,35.1%and66.3%respectively. The sensitivity, specificity and accuracy of SR were98.1%，64.9%and84.3%. Both of the methods were valuable in the differentiation betweenbenign and malignant enlarged cervical lymph nodes (LNs), and SR was more valuable than4scoring method.Conclusion:1. Ultrasound elastography is valuable in the differentiation betweenbenign and malignant enlarged cervical lymph nodes.2. SR with the cutoff point of1.78is better than4scoring method andB-US in differential diagnosis of benign and malignant of enlarged cervicallymph nodes.3. As conventional ultrasound, UE is expected to be an importantcriterion in differential diagnosis of benign and malignant of enlarged cervicallymph nodes. |
PDF Full Text Request