Real-time ultrasound elastography a noninvasive diagnostic procedure for evaluating dominant thyroid nodules

Transcription

1 DOI /s ORIGINAL ARTICLE Real-time ultrasound elastography a noninvasive diagnostic procedure for evaluating dominant thyroid nodules Christian Vorländer & Jan Wolff & Said Saalabian & Robert H. Lienenlüke & Robert A. Wahl Received: 9 June 2010 /Accepted: 1 July 2010 # Springer-Verlag 2010 Abstract Purpose Ultrasound elastography (USE) is a newly developed technique for the evaluation of tissue stiffness. It is known that malignancies often show a low-strain value. So far, only limited data for thyroid nodules is available. Methods This study included 309 prospective evaluated patients with dominant, nontoxic thyroid nodules. All patients were referred to surgery. USE was performed preoperatively. Three measuring groups were formed: hard (< 0.15), intermediate ( ), and soft (> 0.31). The measurements were correlated to the final histological findings. Results The strain rated from 0.01 to 0.84 (mean 0.26±0.13). A total of 50 thyroid malignancies (35 papillara carcinoma, 9 medullary carcinoma, and 6 follicular carcinoma) were observed. Patients (81) were within the hard group, 35 of them (43.2%) had thyroid cancer (TC) in final histology. Out of 132 patients in the intermediate group, 15 patients had TC (11.4%). All 96 patients from the soft group showed benign histological results (NPV 100%). Seventy percent of patients with TC were within the hard group (PPV 42%). These results were highly significant (p<0.001). Coarse calcifications and cystic nodules were not connected with reliable measurements and therefore are not suitable for USE. Conclusion USE is a useful adjunctive tool in the workup of thyroid nodules. A low strain value needs surgical intervention, whereas a high strain value predicts a benign histology. It might substitute fine-needle aspiration cytology in the future. C. Vorländer (*) : J. Wolff : S. Saalabian : R. H. Lienenlüke : R. A. Wahl Departement of Endocrine Surgery, Bürgerhospital Frankfurt am Main e.v., Nibelungenallee 37-41, Frankfurt, Germany c.vorlaender@buergerhospital-ffm.de Keywords Thyroid nodules. Elastography. Thyroid ultrasound. Thyroid surgery. Thyroid cancer Introduction Thyroid nodules occur frequently in areas deficient in iodine, such as Germany. It is estimated that up to 25% of the population is affected by thyroid nodules [1]. More than 110,000 thyroid surgeries are performed in Germany each year due to various indications. The prevalence of thyroid carcinoma is 7 8 per 100,000, with 6,000 new cases of thyroid carcinoma diagnosed each year [2, 3]. To date, the most reliable diagnostic imaging method for examining thyroid nodules is B-mode sonography with colour Doppler gain. The significance regarding the expected malignancy varies sharply from study to study [4 6]. Ultrasound elastography (USE) is a new, noninvasive method that measures tissue elasticity using an ultrasound probe. The procedure leverages the fact that soft tissue structures can be compressed more than the hard ones, and that malignant tumours generally consist of harder tissue than do benign lesions. With USE, elasticity can be overlayed directly on the B image using a colour scale from red (soft) to yellow and green (medium elasticity) to blue (hard), as seen in Figs. 1 and 2. To date, several studies applying USE to diagnose thyroid nodules have been published. The indicated specificity and sensitivity of the method varies sharply in the studies, and so far, the number of patients in all investigations has been<100 patients. The examination results in the previous studies were evaluated based on an examiner-dependent division of the elastogram into individual colour patterns [7 10]. Second-generation USE devices can generate a corresponding measured value [strain value (SV)] in addition to the colour display.

2 Fig. 1 A 15 mm papillara carcinoma (PTC). The tumour [A] shows a low elasticity (blue). Normal thyroid tissue [B] shows elastic tissue (red/yellow) with a high strain. The SV of the tumour was The quality of the SV measurement was good (grade 4 [little scale]) The objective of this study was to examine the application of the latest generation of elastography devices to determine their significance in evaluating thyroid nodules. The data were collected prospectively during an indication consultation under routine conditions. Material and method From February 2008 to January 2010, 309 patients were prospectively examined during the indication consultation at the Clinic for Endocrine Surgery at Bürgerhospital in Fig. 2 An 8 mm medullary carcinoma (MTC) shows low elasticity (SV 0.06). Strap muscles demonstrate good elasticity [red, yellow], while carotic artery shows artefacts [black] like all liquid formations. The quality of the SV measurement was grade 3

3 Frankfurt am Main. The inclusion criteria were solid, dominant nodules on one or both thyroid lobes. Patients with an autonomous adenoma, hyperthyreosis, or recurrent goitre were excluded. The examination was performed with the EUB 7500 (Hitachi Medical, Tokyo, Japan) ultrasound system using a linear transducer (EUP-L54M) with a variable frequency from 6 to 13 MHz. USE was performed as part of B mode sonography. During elastography, the elastogram was recorded through slight pressure and release from the ultrasound probe positioned on the neck. The elastogram was recorded within the longitudinal axis of the thyroid lobe. To achieve good examination quality, the device shows a numeric quality scale from 1 to 5. A value of 3 was required to enable a good evaluation. In the elastogram of the affected nodules obtained in this manner, the SV was determined by marking the area to be measured on the screen (Figs. 1 and 2). The SV was calculated from the elastogram as an absolute value using the auto-correlation method (Hitachi Medical). Values of 1.0 indicate maximum elasticity, while values of 0.01 indicate a complete lack of elasticity. Other ultrasound criteria, such as the echogenicity and perfusion pattern of individual nodules, are purposely not taken into account in order to evaluate the value of USE. All patients underwent surgery. The final diagnosis was based on the result of the histopathological examination. The respective elasticity coefficients (SV) measured transcutaneously prior to surgery were assigned to the histological results. Three groups (hard, medium, and soft) of SV were formed; they included the measured ranges 0.15, , and 0.31, respectively. All patient data were entered prospectively in an Access database and statistical evaluation was performed using the BiAS biomathematics software (Epsilon-Verlag, Germany). The sensitivity and specificity, as well as the positive predictive value (PPV) and negative predictive value (NPV), were determined. The significance was determined with the Chi² -test. The significance level was set at p<0.05 to determine the statistical variation. Results In our own 2008 pilot study with N=60 patients, the ability to successfully perform USE on the thyroid was investigated. Because of the extremely heterogeneous surrounding tissue (vessels, muscles, cervical spine, and trachea), it was necessary to examine the influence of these tissue types on the measurement results. For this reason, USE was first performed transcutaneously (in vivo). The measured values were then checked against the values measured from the explanted thyroid gland lobes (ex vivo) of the same patient, under standard conditions on a gel cushion in a saline bath. The results showed a satisfactory correlation between the in vivo and ex vivo values, whereby the elasticity ex vivo was measured to be somewhat softer than in vivo graphic 1 on average, 0.22±0.10 in vivo and 0.23±0.11 ex vivo. Furthermore, this pilot study was followed during an indication consultation. USE was then performed transcutaneously in clinical routine on a total of 309 patients [207 female, 102 male, average age 47.2±13 years (11 85)]. The examined dominant nodules had a size of 28.2± 14 mm (4 83). The measured SV for all patients averaged 0.26±0.13 ( ). All patients underwent surgery in accordance with the Guidelines of the German Society for General and Visceral Surgery. All specimens were sent for histopathological examination. The histological results presented a thyroid carcinoma in 50 cases [35 papillara carcinoma (PTC), 6 follicular carcinoma (FTC), 9 medullary carcinoma (MTC)]; tumour diameter 21±13 mm. The corresponding elasticity coefficients averaged 0.13, 0.18, and 0.15 (Table 1). A follicular adenoma (FA) was diagnosed in 110 patients (SV 0.29). A colloid nodular goitre (SV 0.27) was found in 146 patients. In three patients, two thyroid cysts and one intrathyroidal parathyroid (SV 0.84) were determined. Divided into three groups (soft, medium, and hard), the soft group had 96 patients, all of which had a benign histology. The intermediate group had 132 patients (117 benign, 15 malignant histology). The hard group included 81 patients (35 malignant and 46 benign histologies; Table 2). The differences between the individual groups were highly significant (p<0.001). Patients with a conclusive diagnosis of a malignant tumour were also analysed with respect to other ultrasound criteria. Here, 43/50 patients (86%) with a thyroid carcinoma presented a hypoechoic nodule. A number of 91 patients with an hypoechoic nodule had a benign histology (PPV of echogenicity 32%). In colour Doppler, increased perfusion could be determined in 40/50 (80%) of the carcinoma patients, 196 patients with increased perfusion had a benign histology (PPV of increased perfusion 17%). For 35/50 (70%) patients with a malignant finding, the elasticity coefficient was In the hard group, 46 patients had a benign finding (PPV of elastography 43%) (Table 3). Based on the experience gained during this study, elastography is unsuited for roughly calcified or predominantly cystic nodules. Both cases are characterised by elastograms filled with artifacts, which makes reliable measurements impossible.

4 Table 1 SV of different thyroid malignancies and benign histology results PTC FTC MTC FA SCN N Median 0.13± ± ± ± ±0.11 Range PTC Papillara carcinoma, FTC follicular carcinoma, MTC medullary carcinoma, FA follicular adenoma, and SCN struma colloides nodosa (benign nodular goitre) Discussion Table 2 Number of cases within the three groups (hard, intermediate, and soft) compared to malignant and benign final histology Hard ( 0.15) Intermediate ( ) Soft ( 0.31) Malignant Benign TC (%) Percentage of thyroid malignancies TC Thyroid cancer Even though most thyroid nodules are benign, there is always a risk that a malignant tumour will be discovered during clarification. Palpation and conventional ultrasound represent the base examinations in all patients. However, they depend to a great extent on the experience of the examiner and location of the nodules [11, 12]. USE is a new procedure that measures tissue elasticity by distorting the tissue structures using external pressure. At present, this tissue distortion can be displayed directly as a colour-coded overlay on the B image (Fig. 1). To date, the most comprehensive experience with elastography is in examining nodes in the breast [13]. The goal of elastography is to determine the hardness of the examined tissue by displaying the degree of tissue distortion under externally applied pressure, since soft tissue distorts more than hard tissue. These differences are displayed in colour; soft tissue is shown in red/yellow and hard tissue in blue (Fig. 1). Since most malignant tumours indicate significantly reduced tissue elasticity, an estimation of the expected malignancy can be made as part of the diagnosis. Results from larger patient groups are already available, particularly for breast cancer and prostate lesions [13, 14]. USE of the thyroid is a new procedure. Previous pilot studies indicate patient numbers of N<100 patients. In studies with the objective of evaluating malignancy for various elastographic examination regions, only a portion of the patients were histologically controlled. In all previous investigations, the elastograms were divided by the respective examiner into four to six groups based on specific colour patterns (yellow/red, green/green-yellow, green/blue, and blue) [7, 8, 10]. This provided an initial assignment of specific elastograms to histological results, a basis for obtaining information on the malignancy of lesions. This study, however, is the first to test the examination technique both in clinical routine and with a large number of patients. In addition, all patients were histologically controlled. Results were classified using the absolute measured SV, not an examiner-dependent division by colour scales. The procedure was performed on a nonselected group of patients of a consultation session in a department for endocrine surgery in an endemic goitre area. And this is the reason for the much lower percentage of malignant tumours in this study when compared to the previous pilot studies. The percentage of carcinoma in all 309 patients we examined was around 16.2%. The previous studies [7 10, 15] indicated 21 54%. In these examinations, different inclusion criteria were applied (e.g., rapidly growing nodules, suspicious cytology), and with an average of 66 patients, the based number of patient was much lower than in our study. During elastography examination of the thyroid, the extremely heterogeneous tissue environment (cervical spine with no elasticity, jugular vein that can be fully compressed) coupled with the lack of surrounding homogeneous tissue (such as the fat tissue at the breast) produces a situation where influencing factors are possible. Our pilot feasibility and reliability study on USE of the thyroid showed a satisfactory correlation of the measured values between in vivo and ex vivo thyroid nodules. We therefore concluded that USE can be performed on the thyroid despite the heterogeneous tissue environment. This also made possible a comparison of our measured results with previous studies, all of which were performed in vivo [7 10, 15]. Table 3 Ultrasound characteristics of the 50 cases of thyroid cancer (TC) N=50 (TC 21±13mm) Hypoechoic Central blood flow SV 0.15 N 43/50 40/50 35/50 SV strain value

5 Fig. 3 Benign thyroid nodule. The elastogram shows the lesion within the same colours than the normal thyroid tissue. The SV was calculated as 0.4, and final histology revealed benign goitre This study shows elastography to be satisfactorily accurate in predicting the malignancy of a cold thyroid nodule (> 70%with an SV 0.15). The predictive power of elastography, however, increases to 100% if SV 0.31 and acceptance of a benign diagnostic finding are included (Fig. 3). In this respect, the result differs from other previous studies, all of which primary objective was the detection of malignant findings. In this study, the 35 papillary carcinomas, on average, showed the lowest elasticity values (SV 0.13%), followed by medullary (N=9) carcinomas at 0.15%. The six FTC were the softest at 0.18%. The variation was relatively large for all three tumour entities (Table 1) and extended, in some cases, significantly into the range of benign findings. In total, however, prior to surgery, 70% of the carcinomas showed an elasticity coefficient 0.15% (Table 2). There were limitations when processing the FTC (N=6); they proved to be softer than the other thyroid malignancies, overlapping the measured values for FA (Table 1). A comparison with other studies is difficult here because FTC occurred only in individual cases [8, 10]. Previously, fine-needle aspiration biopsy (FNAB) was the clear recommendation for clarifying cold nodules due to the accuracy of this method, estimated 80 90% [16, 17]. Even though the risk profile for FNAB (bleeding/infection) is low, it is an invasive and time-intensive procedure (sending to and evaluation of the sample by the pathologist). In contrast, the results of USE are available immediately and can be obtained noninvasively as part of a consultation. The method we used, determining the SV instead of dividing into colour patterns, has the advantages of indepen- Table 4 Available literature for thyroid elastography Rago et al. [7] Hong et al. [10] Rubaltelli et al. [9] Asteria et al. [8] Bürgerhospital Ffm N PPV (%) NPV (%) Colour patterns vs. SV CP CP CP CP SV TC 31 [34%] 49 [54%] 11 [21%] 16 [24%] 50 [16%] Final histology in all cases PPV Positive predictive value [thyroid malignancy], NPV negative predictive value [benign histology], CP colour patterns, SV strain value, TC thyroid cancer

6 dence from the examiner and reproducibility of findings. The subjective impression of the focal finding is therefore not influenced by the surrounding colour patterns, and the decision regarding division into specific groups is eliminated by the absolute measured value of the SV. The division into three groups (hard, medium, and soft) that we used was arbitrarily selected by us and was based on experiences from our internal pilot study (in vivo vs. ex vivo). In the course of the study, however, the measurement limit of more than 0.3% (SV) with respect to a benign finding remained. As part of this study, the elastographic result was analysed as an individual variable. This was done to determine its value, because previous investigations on the value of ultrasound of the thyroid have shown that only the simultaneous existence of multiple ultrasound characteristics (e.g., hypoechogenic, central blood flow, and microcalcification) enable a significant predictive value regarding malignancy [18 20]. Most studies on elastography of the thyroid use a division of results into specific colour groups (colour patterns). Depending on the examiner, four to six groups are created. This division is always based on the subjective evaluation of the examiner. Threshold findings cannot be assigned uniquely. The absolute value we used, the SV, offers the advantage of clear reproducibility and assignment. In most studies, including our investigation, the elastographic image is created by a slight raising and lowering movement with the transducer, and the elastogram is generated in real time. Other authors use the pulsation of a carotid artery as the compression source [21] or generate the elastogram offline [15]. Both are associated with high costs in terms of time and technology. For this reason, realtime elastography has proven to be far more advantageous in clinical application. The internal results for USE indicate a smaller PPV for malignant findings than in comparable studies (Table 4). However, the patient group was much larger and unlike other studies, such as Rago et al. [7], the group was not limited to patients with known, suspected cytology or fastgrowing nodules. This also explains the lower percentage of malignant tumours in the patient group as a whole. The objective of all the studies is to predict suspicious findings. When analysing the results, it can be seen, however, that in all examinations, there is a good PPV for malignancy (55 100%). The predictive probability for a benign finding when there are elastographically soft values (NPV) is, on average, significantly higher (93 98%). In our patient group, NPV was 100% for patients with an SV above 0.30%. This consequently impacts the subsequent clinical procedure to be derived: for elastographically hard measured nodules, additional processing via fine-needle aspiration cytology and/or surgery is clearly indicated. In soft results, the suspicion of malignancy is low. Clarification of malignancy, in such cases, should not be the only indication for surgery. The limiting factors of the method are as follows: The evaluation of roughly calcified nodules (measured as very hard) and predominantly cystic nodules Fig. 4 The measurement of a cystic nodule is dominated by artefacts (black). The solid part of the lesion shows a low SV of This part moves without distortion within the liquid (pseudo-strain)

7 (measurement with significant artefacts) is not possible using USE. This observation is based on both the literature [8] and our data (Fig. 4). Summary USE represents a noninvasive diagnostic procedure for evaluating dominant thyroid nodules. The malignant histology of hard on elastography nodules can be predicted with a satisfactory level of certainty. In cases of soft lesions, the predictive probability for a benign finding is very high. Further evaluations are necessary to consolidate the method, in particular, for multi-nodular goitres. The feasibility of the method for the thyroid can be evaluated as reliable given the available data. Cystic and roughly calcified lesions and the inability to sufficiently differentiate between FTC and FA are the method s limitingfactors. References 1. Reiners C, Wegscheider K, Schicha H, Theissen P, Vaupel R, Wrbitzky R, Schumm-Draeger PM (2004) Prevalence of thyroid disorders in the working population of Germany: ultrasonography screening in 96, 278 unselected employees. Thyroid 14(11): Dralle H (2007) Thyroid incidentaloma. Overdiagnosis and overtreatment of healthy persons with thyroid illness? Chirurg 78(8): Dralle H, Sekulla C (2005) Thyroid surgery: generalist or specialist? Zentralbl Chir 130(5): Tamsel S, Demirpolat G, Erdogan M, Nart D, Karadeniz M, Uluer H, Ozgen AG (2007) Power Doppler US patterns of vascularity and spectral Doppler US parameters in predicting malignancy in thyroid nodules. Clin Radiol 62(3): Ivanac G, Brkljacic B, Ivanac K, Huzjan R, Skreb F, Cikara I (2007) Vascularisation of benign and malignant thyroid nodules: CD US evaluation. Ultraschall Med 28(5): Frates MC, Benson CB, Charboneau JW, Cibas ES, Clark OH, Coleman BG, et al (2006) Management of thyroid nodules detected at US: Society of Radiologists in Ultrasound consensus conference statement. Ultrasound Q 22(4):231 8; discussion Rago T, Santini F, Scutari M, Pinchera A, Vitti P (2007) Elastography: new developments in ultrasound for predicting malignancy in thyroid nodules. J Clin Endocrinol Metab 92 (8): Asteria C, Giovanardi A, Pizzocaro A, Cozzaglio L, Morabito A, Somalvico F, Zoppo A (2008) US-elastography in the differential diagnosis of benign and malignant thyroid nodules. Thyroid 18 (5): Rubaltelli L, Corradin S, Dorigo A, Stabilito M, Tregnaghi A, Borsato S, Stramare R (2009) Differential diagnosis of benign and malignant thyroid nodules at elastosonography. Ultraschall Med 30(2): Hong Y, Liu X, Li Z, Zhang X, Chen M, Luo Z (2009) Real-time ultrasound elastography in the differential diagnosis of benign and malignant thyroid nodules. J Ultrasound Med 28(7): Ghassi D, Donato A (2009) Evaluation of the thyroid nodule. Postgrad Med J 85(1002): Tan GH, Gharib H, Reading CC (1995) Solitary thyroid nodule. Comparison between palpation and ultrasonography. Arch Intern Med 155(22): Thomas A, Degenhardt F, Farrokh A, Wojcinski S, Slowinski T, Fischer T (2010) Significant differentiation of focal breast lesions calculation of strain ratio in breast sonoelastography. Acad Radiol Miyagawa T, Tsutsumi M, Matsumura T, Kawazoe N, Ishikawa S, Shimokama T (2009) Real-time elastography for the diagnosis of prostate cancer: evaluation of elastographic moving images. Jpn J Clin Oncol 39(6): Lyshchik A, Higashi T, Asato R, Tanaka S, Ito J, Mai JJ et al (2005) Thyroid gland tumor diagnosis at US elastography. Radiology 237(1): Chow LS, Gharib H, Goellner JR, van Heerden JA (2001) Nondiagnostic thyroid fine-needle aspiration cytology: management dilemmas. Thyroid 11(12): Yang J, Schnadig V, Logrono R, Wasserman PG (2007) Fine-needle aspiration of thyroid nodules: a study of 4703 patients with histologic and clinical correlations. Cancer 25;111(5): Rago T, Vitti P, Chiovato L, Mazzeo S, De Liperi A, Miccoli P (1998) Role of conventional ultrasonography and color flow-doppler sonography in predicting malignancy in cold thyroid nodules. Eur J Endocrinol 138(1): Rago T, Vitti P (2008) Role of thyroid ultrasound in the diagnostic evaluation of thyroid nodules. Best Pract Res Clin Endocrinol Metab 22(6): Alexander EK, Heering JP, Benson CB, Frates MC, Doubilet PM, Cibas ES, Marqusee E (2002) Assessment of nondiagnostic ultrasound-guided fine needle aspirations of thyroid nodules. J Clin Endocrinol Metab 87(11): Dighe M, Bae U, Richardson ML, Dubinsky TJ, Minoshima S, Kim Y (2008) Differential diagnosis of thyroid nodules with US elastography using carotid artery pulsation. Radiology 248 (2):