Real-time Sonoelastography for the Evaluation of Testicular Lesions 1

Transcription

1 Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at Original Research n Ultrasonography Friedrich Aigner, MD Tobias De Zordo, MD Leo Pallwein-Prettner, MD Daniel Junker, MD Georg Schäfer, MD Renate Pichler, MD Nicolai Leonhartsberger, MD Germar Pinggera, MD Vikram S. Dogra, MD Ferdinand Frauscher, MD Real-time Sonoelastography for the Evaluation of Testicular Lesions 1 Purpose: Materials and Methods: To evaluate the feasibility of using real-time sonoelastography (RTE) for the differentiation and characterization of testicular lesions. Institutional review board approval was obtained for this retrospective study, and the requirement to obtain informed consent was waived. Fifty patients (mean age, 42 years; age range, years) with testicular lesions detected with gray-scale ultrasonography (US) and color and/or power Doppler US were evaluated with RTE between December 2004 and August 2010 to assess tissue elasticity of the testes. Stiff or hard lesions were suspected of being malignant. Testicular lesions with normal or decreased tissue stiffness ( soft lesions) were considered benign. Findings from surgery and histopathologic examination were used as the reference standard in 34 cases, and findings from clinical and US follow-up were used as the reference standard in 16 cases. Sensitivity, specificity, negative predictive value, positive predictive value, and diagnostic accuracy were calculated. Results: Conclusion: Thirty-four of the 50 lesions (68%) were testicular tumors and 16 (32%) were of nontumorous origin. RTE showed the presence of hard lesions in all cases of testicular tumors and three cases of nontumorous lesions. Four lesions with an uncertain diagnosis when tested with gray-scale US and color and/or power Doppler US alone were soft at RTE and showed nontumorous character at follow-up. RTE showed a sensitivity of 100%, a specificity of 81%, a negative predictive value of 100%, a positive predictive value of 92%, and an accuracy of 94% in the diagnosis of testicular tumors. RTE demonstrated all testicular tumors as lesions with increased tissue stiffness. Because of its higher specificity, RTE can provide additional information in cases with indeterminate US findings. 1 From the Departments of Radiology (F.A., T.D.Z., L.P., D.J., F.F.), Pathology (G.S.), and Urology (R.P., N.L., G.P), Medical University Innsbruck, Anichstr. 35, 6020 Innsbruck, Austria; Department of Radiology, AKH Linz, Linz, Austria (L.P.P.); and Department of Radiology, University of Rochester, School of Medicine and Dentistry, Rochester, NY (V.S.D.). Received August 16, 2011; revision requested October 4; revision received November 16; accepted December 5; final version accepted December 14. Address correspondence to T.D.Z. ( Tobias@De-Zordo.net). q RSNA, 2012 q RSNA, radiology.rsna.org n Radiology: Volume 263: Number 2 May 2012

2 One percent of all cancer in men is testicular cancer. It is the most common malignancy in male subjects between the ages of 15 and 35 years and accounts for 60% of all cancers in these 2 decades (1). Testicular germ cell tumor is the most common type of testicular cancer, with a frequency of 95% (1,2). The most common symptoms are painless mass or vague discomfort in the scrotum. Ten percent of patients have fever and pain, 10% refer to having had scrotal trauma, and 10% have metastases at diagnosis. Ultrasonography (US) is the imaging modality of choice for the evaluation of scrotal abnormalities (2,3). Owing to a higher vessel density, almost all cancers show increased perfusion at color and/or power Doppler US. A sensitivity of nearly 100% was reported when using the combination of gray-scale and color Doppler US (4,5). Nevertheless, these techniques do not definitely allow differentiation of testicular cancer from benign scrotal masses such as focal orchitis, partial infarction, granuloma, Leydig cell hyperplasia, adrenal rest, hematoma, or lipoma. All of those can mimic testicular cancer, thus decreasing the Advances in Knowledge nn The addition of real-time sonoelastography (RTE) to conventional US may add value in the characterization of testicular lesions: All lesions classified as soft at RTE were shown to be benign (13 of 50 lesions), and all malignant lesions were classified as hard at RTE (34 of 50 lesions); however, hard lesions might be also of benign origin (three of 50 lesions). nn Additional evaluation of testicular lesions with RTE may increase the specificity of US as a diagnostic tool, as found in four patients in our study with indeterminate findings at gray-scale and color Doppler US and soft lesions at RTE that were benign at follow-up. specificity of testicular US (2). Therefore, an imaging technique that can increase the specificity of gray-scale and color and/or power Doppler US in the assessment of testicular lesions would be desirable. In cases of benign lesions, the findings could potentially help avoid unnecessary testicular surgery. Real-time sonoelastography (RTE) is a US technique that enables assessment of tissue elasticity and has shown a clear potential for characterizing various cancers, such as those of the breast or prostate (6,7). In general, most cancers show increased stiffness because they have higher cell and vessel density compared with the surrounding normal tissue (8). Therefore, we performed this study to evaluate the feasibility of using RTE for the differentiation and characterization of testicular lesions. Materials and Methods Patients Approval for this retrospective study was obtained from the local ethical review board, and the requirement to obtain informed consent was waived. Between December 2004 and August 2010, 62 consecutive patients with clinical suspicion of testicular tumor underwent US. Testicular US was performed with gray-scale and color and/or power Doppler US. In addition, the testes were investigated with RTE in the same session. Patient data, histopathology reports, and clinical and US data were retrospectively analyzed. Complete US follow-up and/or histologic findings of the testicular lesions were available for Implication for Patient Care nn Our preliminary data suggest that soft testicular lesions at RTE are more likely of nonmalignant character, whereas hard lesions are most likely malignant; benign lesions, including partial infarction, scarring, and cysts, may behave as hard lesions at RTE. 50 patients (mean age, 42 years; age range, years). Twelve patients (mean age, 44 years; age range, years) with incomplete follow-up or missing histologic data were excluded. Gray-Scale and Color and/or Power Doppler US Gray-scale and color and/or power Doppler US were performed by using a Sequoia 512 unit (Siemens Medical Systems, Mountain View, Calif) with a 14-MHz linear probe (15L8w, Siemens) in all patients to assess the scrotum. Color Doppler settings were as follows: frequency of 10 MHz, pulse repetition frequency of 16 cm/sec, Doppler gain of 50%, frame rate of 12 pictures per second, and low wall filter. Power Doppler settings were as follows: frequency of 10 MHz, pulse repetition frequency of 10 cm/sec, Doppler gain of 50%, frame rate of nine pictures per second, and low wall filter. After careful investigation, lesion size, gray-scale US characteristics, and vascularity were documented. Standardized protocols with axial and transverse scans of the lesions were performed according to Dogra et al (3). All US examinations were performed by one of three experienced uroradiologists (F.F., F.A., or L.P., all with.5 years of experience). All images were transferred to our local picture archiving and communication system. Published online before print /radiol Content codes: Radiology 2012; 263: Abbreviation: RTE = real-time sonoelastography Author contributions: Guarantors of integrity of entire study, F.A., T.D.Z., L.P., F.F.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, F.A., T.D.Z., L.P., D.J., G.S., G.P., F.F.; clinical studies, F.A., L.P., D.J., G.S., R.P., N.L., G.P., F.F.; statistical analysis, T.D.Z., L.P., D.J., G.P.; and manuscript editing, T.D.Z., L.P., D.J., G.S., N.L., V.S.D., F.F. Potential conflicts of interest are listed at the end of this article. Radiology: Volume 263: Number 2 May 2012 n radiology.rsna.org 585

3 Figure 1 Figure 1: Images in a 28-year-old man with seminoma. (a) Image from RTE shows a hard lesion (blue). (b) Corresponding gray-scale US scan shows a hypoechoic mass. (c) Color Doppler US scan shows hyperemia. RTE Imaging RTE was performed by using an EUB 8500 US unit (Hitachi Medical Systems, Tokyo, Japan) and a linear transducer operating with a frequency of 6 13 MHz. All examinations were performed by one of the three uroradiologists. RTE technique estimates the displacement in corresponding A lines, which are obtained by the evaluation of the cross-correlation function of the radiofrequency data set (unfiltered, high-frequency US data). The displacement estimates are used to determine the tissue strain and to reconstruct the Young modulus of elasticity (ie, estimation of time shifts between corresponding A lines) (9). Multicompression imaging is used to improve the signal-to-noise ratio of RTE images (10,11). Local strain is computed under slight compression and decompression applied by using a freehand technique. The testes were fixed on a special plastic device (scrotum plate) to allow the optimal position for performing RTE. The pressure applied to the testicles was adjusted according to the visual indicator for compression seen on the video screen, which indicates the average strain in the region of interest between two frames and displays optimal strain on the screen. Calculation of tissue elasticity distribution is in real time (up to 30 frames per second), and stiffness of testicular tissue is displayed with a color overlay on the B-mode image (Fig 1). Tissue stiffness was encoded from red (soft) to blue (hard) according to the standard settings of the US sys tem provider. For each patient, at least two representative images or cine loops were stored on an external hard disk. Image Interpretation US and RTE images were separately reviewed between November and December 2010 by an independent radiologist with 5 years of experience in RTE (T.D.Z.). First, all available RTE images were evaluated while blinded to all other information, then gray-scale and Doppler US images were assessed, and then clinical history and histologic results were evaluated. On US images, lesion size in all three planes (in millimeters), echogenicity (hypo-, iso-, or hyperechoic compared with normal testicular parenchyma), and color and/or power Doppler signal (hyper-, hypo-, or avascular compared with normal testicular parenchyma) were evaluated. Lesion volume (in cubic millimeters) was calculated by using the ellipsoid formula. For evaluation of sensitivity, specificity, and accuracy, the original diagnosis on the US report was used. On RTE images, tissue stiffness of the overall pattern of the lesions was assessed on the basis of our previous experience (12). Soft tissue (colored red-green) was differentiated from hard or stiff tissue (colored blue). Reference Standard All lesions suspected of being malignant underwent surgical exploration and histopathologic analysis. All patients with a diagnosis of benign lesions comprising inflammation and infarction or those with a US appearance of a simple cyst were followed up with US within 24 hours when infarction was suspected and after 2 3 days and then weekly when inflammation was suspected. Follow-up was performed in case of benign lesions after 6 weeks and 3 months as part of regular patient care. The presence of lesion stability, a decrease in lesion size, a decrease of vascularity, or the disappearance of the lesion allowed the lesion to be defined as a nontumorous entity. Statistical Analysis All statistical analyses were performed with software (SPSS, version 17.0; SPSS, Chicago, Ill). Sensitivity, 586 radiology.rsna.org n Radiology: Volume 263: Number 2 May 2012

4 Testicular Lesions: Findings at Gray-Scale US, Color Doppler US, and RTE Findings at Gray-Scale US Findings at Color Doppler US Findings at RTE Lesions Type Hypoechoic Mixed Echogenicity Hyperechoic Avascular Hypovascular Hypervascular Soft Hard Neoplasia (n = 34) 22 (65) 12 (35) (9) 31 (91) 0 34 (100) Cyst (n = 4) 4 (100) (100) (75) 1 (25) Orchitis (n = 6) 5 (83) 1 (17) (50) 3 (50) 6 (100) 0 Partial infarction (n = 5) 5 (100) (80) 1 (20) 0 4 (80) 1 (20) Scar after biopsy (n = 1) (100) 1 (100) (100) Note. Data are numbers of lesions, with percentages in parentheses. specificity, negative predictive value, positive predictive value, and diagnostic accuracy were calculated by using the numbers of lesions as numerators and numbers of testicles as denominators. To compare quantitative data, the two-sided Student t test was performed; P,.05 was considered indicative of a statistically significant difference. Results Histopathologic examination confirmed 34 testicular lesions as testicular tumors: 13 were seminomas, two were embryonal carcinomas, nine were mixed seminomatous and/or nonseminomatous tumors, one was teratoma, four were Leydig cell tumors, three were sertoli cell tumors, one was metastasis from adrenal tumor, and one was epidermoid cyst. Sixteen testicular lesions were of nontumorous origin: six were epididymorchitis or orchitis, five were partial infarction, four were cysts, and one was scarring after biopsy. The mean size of the tumors was 21 mm (range, 3 66 mm) 3 17 mm (range, 4 44 mm) 3 19 mm (range, 3 57 mm). The mean size of nontumorous lesions was 9 mm (range, 2 30 mm) 3 5 mm (range, 1 15 mm) 3 5 mm (range, 2 10 mm). The mean volume of tumors was 4564 mm (standard deviation), and the mean volume of nontumorous lesions was 33 mm (P =.019). On gray-scale US scans, tumor lesions appeared as hypoechoic (n = 22) or as mixed hypo- and hyperechoic (n = 12) masses (Table). Fourteen of the nontumorous lesions were hypoechoic, one lesion (focal orchitis) was of mixed echogenicity, and one lesion was hyperechoic. On color and/or power Doppler US scans, tumor lesions were hypervascular in 31 patients and hypovascular in three (one patient each had Leydig cell tumor, seminoma, and epidermoid cyst). Of the 16 nontumorous lesions, three showed a hypervascular pattern all of which were all identified as focal areas of orchitis. All other nontumorous lesions (n = 13) showed a hypovascular (n = 4) or an avascular (n = 9) pattern. On RTE images, all 34 tumorous lesions were found to be tissues with increased stiffness (hard lesions) and were color-coded blue (Fig 1). Only three nontumorous lesions were classified as hard lesions at RTE: One lesion represented scarred tissue, a result of previous biopsy (Fig 2), one lesion was a cyst, and one lesion represented partial infarction. However, histologic examination was not performed in these three lesions because medical history (biopsy, infarction) and pathognomonic US findings (cyst) pointed to a lesion with a nontumorous origin. Overall, 13 lesions were classified with soft tissue elasticity at RTE: Six lesions were orchitis, four were partial infarction, and three were testicular cysts. In particular, RTE helped differentiate between a tumorous and nontumorous entity in four patients: In two patients, small focal areas of hypovascular orchitis (lesion diameter,8 mm) were demonstrated as soft lesions at RTE (Fig 3) and showed no growth at the follow-up examination. In the other two patients, RTE demonstrated small areas of partial infarction (lesion diameter,10 mm) as soft lesions, suggesting a nontumorous origin. In all 13 patients with soft lesions, results of follow-up confirmed a nontumorous origin. Statistical analyses showed a sensitivity of 100% (34 of 34 lesions), a specificity of 81% (13 of 16 lesions), a negative predictive value of 100% (13 of 13 lesions), a positive predictive value of 92% (34 of 37 lesions), and an accuracy of 94% (47 of 50 lesions) for RTE in testicular tumor diagnosis. For conventional US (gray-scale and color and/ or power Doppler US), we calculated a sensitivity of 100% (34 of 34 lesions), a specificity of 75% (12 of 16 lesions), a negative predictive value of 100% (12 of 12 lesions), a positive predictive value of 89% (34 of 38 lesions), and an accuracy of 92% (46 of 50 lesions). Discussion In a preliminary study, Pallwein et al (13) evaluated 20 testicular masses with RTE, showing that all tumors were behaving as hard tissue at RTE (n = 15). This criterion has already showed promising results for the detection of breast cancer and prostate cancer (6,7). Although we know that gray-scale US is able to depict testicular tumors in most cases, we also know that benign entities can mimic tumor at gray-scale and color and/or power Doppler US (2). The differentiation of benign tissue from tumor is sometimes difficult with conventional US techniques. We have demonstrated the additional value Radiology: Volume 263: Number 2 May 2012 n radiology.rsna.org 587

5 Figure 2 Figure 2: Images in a 38-year-old man who had previously undergone biopsy. (a) Image from RTE shows a small hard lesion. (b) Gray-scale US scan shows a hyperechoic area. No signals were seen on Doppler US scan (not shown). Final diagnosis was scarring after biopsy. Figure 3 Figure 3: Images in a 48-year-old man. (a) Image from RTE shows a soft lesion. (b) Gray-scale US scan shows a small lesion (5 mm). No signals were seen on color Doppler US scan (not shown). After treatment with standard antibiotics, lesion had disappeared at follow-up. Final diagnosis was focal orchitis. of RTE for differentiating between tumorous and nontumorous lesions in the current study. In our study population, RTE demonstrated all testicular tumors as hard lesions, resulting in a sensitivity of 100% for the diagnosis of testicular cancer (100% sensitivity was also achieved with gray-scale and color Doppler US). Our data are in line with those from a recently published article by Grasso et al (14), who evaluated 41 testicular lesions and found that RTE confirmed the results of gray-scale and color Doppler US in 38 cases. Some benign intratesticular lesions are associated with increased stiffness at RTE; however, adding RTE to gray-scale and Doppler US increased specificity from 75% to 81% in our series. Thus, our data suggest that adding RTE findings of testicular lesions to findings from grayscale and color and/or power Doppler US may increase the specificity of testicular US. Three benign testicular lesions were associated with increased stiffness. One of these had previously undergone biopsy, one was scarring after prior orchitis, and one was a cyst. Although fibrotic tissue (scar) is likely to appear stiff at RTE, the cystic lesion most probably appeared hard owing to a high internal pressure, as was shown in cysts of the breast (15). In our series, we could exclude neoplasia in the absence of increased tissue stiffness, that is, RTE demonstrated nontumorous lesions such as orchitis, partial infarction, or cysts as soft tissue. Interestingly, four small lesions of unclear origin at gray-scale and color and/ or power Doppler US were classified as soft at RTE. Two lesions were confirmed as focal orchitis and two were confirmed as partial infarction at follow-up. All four cases remained unclear when explored with gray-scale and color and/or power Doppler US alone, and follow-up was performed. We believe that adding RTE findings in cases with indeterminate testicular lesions may provide higher confidence levels in excluding or confirming the presence of cancer. Contrast-enhanced US enables visualization of testicular microvascularization and may thus aid in the preoperative assessment of testicular lesions with hypervascularization as an important feature in the diagnosis of malignancy (16). In addition, magnetic resonance (MR) imaging has been advocated as an adjunct to US examination in cases of uncertainty, but this imaging modality cannot always help distinguish testicular cancers from benign entities (17,18). MR imaging may be helpful for diagnosing testicular hematoma and tubular ectasia of the rete testis but remains an expensive investigation tool and lacks availability (19). RTE, in contrast, is relatively inexpensive and may add important information to that obtained with conventional US imaging. Our study had limitations. Our population size is relatively small (n = 50), although it is comprised of a large variety of tumorous and nontumorous entities. However, larger series with more lesions such as Leydig cell hyperplasia, adrenal rests, hematomas, lipomas, and granulomatous lesions (sarcoid, 588 radiology.rsna.org n Radiology: Volume 263: Number 2 May 2012

6 tuberculosis) should be investigated in the future. US and RTE were performed during the same imaging session; therefore the initial interpretation of the US scans could have been influenced by the RTE findings. Another limitation is the lack of histopathologic analysis in our benign lesions. RTE was applied to patients with known lesions diagnosed with grayscale US and so has not been evaluated in the detection of malignancy. However, we consider RTE to be complementary to conventional US and not a stand-alone imaging modality. In addition, RTE was performed with only one US system. Comparison with different RTE systems would be desirable. Last, RTE, as a freehand investigation tool, is operator dependent like gray-scale US; therefore, inter- and intraoperator and inter- and intraobserver analyses should be evaluated in the future. Furthermore, new developments of RTE, including shear wave technology, strain ratio, and histograms, may further help differentiate between benign and malignant lesions. In conclusion, in this study, RTE has been proved to be capable of depicting testicular cancers as lesions with an increased stiffness. This imaging modality should allow for higher confidence in defining unclear testicular lesions at gray-scale and color and/or power Doppler US. Although RTE cannot replace gray-scale and color and/or power Doppler US, our data suggest that adding RTE findings of testicular lesions to findings at gray-scale and color and/ or power Doppler US may increase the specificity of testicular US. Disclosures of Potential Conflicts of Interest: F.A. No potential conflicts of interest to disclose. T.D.Z. No potential conflicts of interest to disclose. L.P. No potential conflicts of interest to disclose. D.J. No potential conflicts of interest to disclose. G.S. No potential conflicts of interest to disclose. R.P. No potential conflicts of interest to disclose. N.L. No potential conflicts of interest to disclose. G.P. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: is a paid consultant for Lilly Icos Consultation; received money for expert testimony from Pfizer and Lilly Icos; received a grant or has a grant pending from Lilly Icos Indianapolis; institution received a grant or has a grant pending from Lilly Icos Indianapolis; receives payment for development of educational presentations from Bayer Schering. Other relationships: none to disclose. V.S.D. Financial activities related to the present article: none to disclose. Financial activities not related to the present article: received money for expert testimony; receives royalties from Elsevier. Other relationships: none to disclose. F.F. No potential conflicts of interest to disclose. References 1. Peng X, Zeng X, Peng S, Deng D, Zhang J. The association risk of male subfertility and testicular cancer: a systematic review. PLoS ONE 2009;4(5):e Lesnik G, Nickl S, Kuschnig P, Sinzig M, Hausegger K, Jeschke K. Sonography of the scrotum [in German]. Rofo 2006;178(2): Dogra VS, Gottlieb RH, Oka M, Rubens DJ. Sonography of the scrotum. Radiology 2003;227(1): Guthrie JA, Fowler RC. Ultrasound diagnosis of testicular tumours presenting as epididymal disease. Clin Radiol 1992;46(6): Rifkin MD, Kurtz AB, Pasto ME, Goldberg BB. Diagnostic capabilities of high-resolution scrotal ultrasonography: prospective evaluation. J Ultrasound Med 1985;4(1): Salomon G, Köllerman J, Thederan I, et al. Evaluation of prostate cancer detection with ultrasound real-time elastography: a comparison with step section pathological analysis after radical prostatectomy. Eur Urol 2008;54(6): Regini E, Bagnera S, Tota D, et al. Role of sonoelastography in characterising breast nodules: preliminary experience with 120 lesions. Radiol Med (Torino) 2010;115(4): Krouskop TA, Wheeler TM, Kallel F, Garra BS, Hall T. Elastic moduli of breast and prostate tissues under compression. Ultrason Imaging 1998;20(4): Erkamp RQ, Emelianov SY, Skovoroda AR, O Donnell M. Nonlinear elasticity imaging: theory and phantom study. IEEE Trans Ultrason Ferroelectr Freq Control 2004;51(5): Konofagou EE, Ophir J, Kallel F, Varghese T. Elastographic dynamic range expansion using variable applied strains. Ultrason Imaging 1997;19(2): Varghese T, Zagzebski JA, Frank G, Madsen EL. Elastographic imaging using a handheld compressor. Ultrason Imaging 2002;24(1): Schurich M, Aigner F, Frauscher F, Pallwein L. The role of ultrasound in assessment of male fertility. Eur J Obstet Gynecol Reprod Biol 2009;144(Suppl 1):S192 S Pallwein L, Steiner H, Akkad T, Bartsch G, Frauscher F. Real-time elastography for evaluation of testicular masses: initial experience. Eur Urol Suppl 2006;5(2): Grasso M, Blanco S, Raber M, Nespoli L. Elasto-sonography of the testis: preliminary experience. Arch Ital Urol Androl 2010;82(3): Chiorean AR, Duma MM, Dudea SM, et al. Typical and unusual sonoelastographic patterns of breast cystic lesions: impact on BI-RADS classification. Ultraschall Med doi: /s Published online October 22, Accessed March 15, Lock G, Schmidt C, Helmich F, Stolle E, Dieckmann KP. Early experience with contrast-enhanced ultrasound in the diagnosis of testicular masses: a feasibility study. Urology 2011;77(5): Choyke PL. Dynamic contrast-enhanced MR imaging of the scrotum: reality check. Radiology 2000;217(1): Watanabe Y, Dohke M, Ohkubo K, et al. Scrotal disorders: evaluation of testicular enhancement patterns at dynamic contrastenhanced subtraction MR imaging. Radiology 2000;217(1): Kim W, Rosen MA, Langer JE, Banner MP, Siegelman ES, Ramchandani PUS. US MR imaging correlation in pathologic conditions of the scrotum. RadioGraphics 2007;27(5): Radiology: Volume 263: Number 2 May 2012 n radiology.rsna.org 589