Do Incidental Hyperechoic Renal Lesions Measuring Up to 1 cm Warrant Further Imaging? Outcomes of 161 Lesions

Genitourinary Imaging Original Research Genitourinary Imaging Original Research Ankur M. Doshi 1 Abimbola Ayoola Andrew B. Rosenkrantz Doshi AM, Ayoola A, Rosenkrantz AB Keywords: angiomyolipoma, hyperechoic renal lesion DOI:10.2214/AJR.16.17490 Received October 15, 2016; accepted after revision December 24, 2016. Based on a presentation at the Radiological Society of North America 2015 annual meeting, Chicago, IL. 1 All authors: Department of Radiology, Center for Biomedical Imaging, New York University School of Medicine, Langone Medical Center, 660 First Ave, 3rd Fl, New York, NY 10016. Address correspondence to A. M. Doshi (ankur.doshi@nyumc.org). AJR 2017; 209:346 350 0361 803X/17/2092 346 American Roentgen Ray Society Do Incidental Hyperechoic Renal Lesions Measuring Up to 1 cm Warrant Further Imaging? Outcomes of 161 Lesions OBJECTIVE. The purpose of this study was to determine the outcomes of hyperechoic renal lesions measuring 1 cm or less at ultrasound examination. MATERIALS AND METHODS. This retrospective study included 161 hyperechoic renal lesions measuring 1 cm or less at ultrasound that were evaluated with follow-up ultrasound, CT, or MRI. Follow-up imaging examinations were reviewed to assess for definitive lesion characterization or size stability. RESULTS. Follow-up included 11 unenhanced CT, 39 contrast-enhanced CT, 52 unenhanced and contrast-enhanced CT, two unenhanced MRI, 50 unenhanced and contrast-enhanced MRI, and 87 ultrasound examinations. At CT or MRI 58.4% of lesions were confirmed to be angiomyolipomas. At CT, one lesion represented a stone, and one a hyperdense cyst. At CT or MRI 11.8% of the lesions had no correlate; 3.1% were not visualized at follow-up ultrasound. An additional 23.6% were stable at 2-year follow-up imaging or beyond. Two lesions were evaluated with only contrast-enhanced CT less than 1 month after ultrasound, and the CT images did not show macroscopic fat or calcification or meet the criteria for a simple cyst. These lesions were considered indeterminate. One lesion in a 65-year-old man was imaged with unenhanced and contrast-enhanced CT 23 months after ultrasound, and the CT showed an increase in size, solid enhancement, and no macroscopic fat. This lesion was presumed to represent renal cell carcinoma. Overall, the one lesion presumed malignant and the two indeterminate lesions constituted 1.9% of the cohort. The other 98.1% of lesions were considered clinically insignificant. CONCLUSION. Most hyperechoic renal lesions measuring 1 cm or smaller were clinically insignificant, suggesting that such lesions may not require additional imaging. Patient demographics, symptoms and risk factors for malignancy may help inform the decision to forgo follow-up imaging of such lesions. R enal angiomyolipomas (AMLs) are benign tumors containing varying degrees of fat, dysmorphic blood vessels, and smooth muscle [1]. They are the most common benign renal neoplasm, being identified in as many as 2.1% of cases at autopsy [2]. When a small, solitary, asymptomatic renal lesion is definitely diagnosed as AML, no further follow-up imaging is generally warranted. Intervention or follow-up imaging may be required for AMLs larger than 4 cm or containing aneurysms larger than 5 mm, given increased risk of bleeding in such lesions [3], and if multiple AMLs or a genetic syndrome, such as tuberous sclerosis, are present, in which more rapid growth of AMLs has been observed [4, 5]. AMLs are typically markedly hyperechoic at ultrasound [1]. Forman et al. [6], howev- er, reported that 32% of renal cell carcinomas (RCCs) measuring up to 3 cm were hyperechoic and mimicked AMLs [6]. Given these observations, it is commonly considered that an incidental hyperechoic renal lesion warrants further evaluation with CT or MRI to confirm the presence of macroscopic fat and definitively diagnose the lesion as AML. In addition, authors of a systematic literature review [7] concluded that all noncalcified echogenic lesions detected at ultrasound, regardless of size, be further evaluated with CT. Those authors noted, however, that evidence on the topic was limited. Data to guide the management of hyperechoic lesions measuring 1 cm or less is particularly lacking, and the optimal approach in this group is controversial [2]. For such very small lesions, strategies vary among practices and include performing CT or MRI for further characterization, sur- 346 AJR:209, August 2017

veillance sonography to assess for stability, or no imaging follow-up given the presumed very low risk in such lesions [2]. Each of these approaches has limitations. For example, although the presence of macroscopic fat in a renal lesion at CT or MRI is essentially diagnostic of AML, assessment for fat may be challenging in lesions 1 cm or smaller. In addition, both AML and RCC can exhibit slow growth, potentially limiting the role of sonographic size surveillance in their differentiation [8]. Furthermore, additional imaging with CT or MRI increases costs, exposes the patient to imaging-related risks, and may also increase patient anxiety [9]. Finally, although small ( 1 cm) hyperechoic lesions are believed to be benign in a large fraction of cases, actual data to support routinely forgoing further imaging in this size range are lacking. The purpose of this study was to evaluate the outcomes of small ( 1 cm) hyperechoic renal lesions detected with ultrasound. Materials and Methods Patients This institutional review board approved HIPAA-compliant retrospective study was conducted at a single academic medical center. The requirement for written informed consent was waived. A departmental database was searched for reports of adult abdominal ultrasound examinations performed from 2005 to 2008 that contained any of the following phrases: echogenic lesion, echogenic focus, echogenic mass, hyperechoic lesion, hyperechoic focus, hyperechoic mass, AML, and angiomyolipoma. The search revealed 2374 results. The time period was selected to allow at least a 5-year window of imaging follow-up to assess for long-term stability of lesions included at the time of the search. Cases were excluded for the following reasons: results not related to a hyperechoic renal lesion (n = 1977), lesion larger than 1 cm (n = 109), and lack of sufficient follow-up imaging (n = 127). Sufficient follow-up imaging was defined as either CT or MRI any time after the ultrasound or repeat ultrasound at least 2 years after the initial ultrasound. These exclusions resulted in a final cohort of 161 patients. Imaging Technique Because examinations were performed across a range of years and imaging systems, acquisition techniques varied within the cohort. In addition, the follow-up CT and MRI examinations were performed for a variety of reasons other than renal mass evaluation and varied in the use of IV contrast material and number of contrast-enhanced phases. The index ultrasound examinations were performed with Antares, Sequoia, or S2000 units (Siemens Healthcare). MDCT examinations were performed with 16-MDCT or greater systems with an approximately 4-mm slice thickness. MRI examinations were performed with 1.5-T systems (Magnetom Avanto, Symphony, or Sonata, Siemens Healthcare) and torso phased-array coils. All MRI examinations included axial 2D gradient-echo T1-weighted in- and opposed-phase sequences and a 3D fat-suppressed T1-weighted gradient-echo sequence. Image Evaluation The size of the hyperechoic lesion at both the initial index ultrasound and follow-up ultrasound, CT, or MRI was retrospectively measured for purposes of this investigation to facilitate reliable comparison of lesion size between the initial and follow-up studies. The lesion characteristics at ultrasound, including degree of echogenicity and Doppler ultrasound appearance, were based on the initial clinical interpretations and not retrospectively evaluated. For patients included because of subsequent CT or MRI results, a fellowship-trained abdominal radiologist with 2 years of experience directly reviewed the follow-up examination for lesion characterization. First, CT and MR imaging findings were evaluated for evidence of fat, which if present was considered to indicate a diagnosis of AML. For CT, the largest circular ROI possible was placed over the portion of the lesion with the lowest attenuation to identify the presence of macroscopic fat, defined as less than 20 HU. This measurement was performed with an unenhanced acquisition when available [10]. For MRI, lesions were considered to contain macroscopic fat when they had either a region of hyperintensity on a non fat-suppressed T1-weighted image that became hypointense on a fat suppressed T1-weighted image, or etching artifact was present at the interface between the lesion and the renal parenchyma on inand opposed-phase T1-weighted images [11]. When the imaging findings did not indicate a specific diagnosis, the lesions were subjected to further retrospective evaluation, including assessment for enhancement when unenhanced and contrast-enhanced imaging was available, measurement of lesion size at follow-up imaging (including ultrasound, CT, and MRI), and evaluation for other alternative benign diagnoses (e.g., a renal stone or hyperdense cyst). For all modalities (ultrasound, CT, and MRI), a change in size of 4 mm or less at 2-year or later follow-up imaging was considered to represent stability. On the basis of the assessments of the follow-up images, a final classification was reached for each lesion whether it could be deemed clinically insignificant as opposed to either representing a malignancy or remaining indeterminate. For lesions deemed to represent a malignancy at follow-up imaging, the degree of echogenicity relative to renal cortex at baseline ultrasound was subjectively assessed, as were ancillary ultrasound features, including shadowing, intralesion cysts, and an anechoic rim. Statistical Evaluation Results were summarized by use of descriptive statistics, including percentages of the various outcomes among the hyperechoic lesions. Analysis was performed with Excel for Apple Macintosh software (version 14, Microsoft). Results Hyperechoic renal lesions measuring 1 cm or less were found in 161 patients (22 men [13.7%]; 139 women [86.3%]; mean age, 62.6 ± 12.5 [SD] years; range, 30 88 years). The mean lesion size was 0.72 ± 0.18 cm (range, 0.4 1.0 cm). Available follow-up included only 2 or more years of follow-up ultrasound for 17.4% (28/161) of patients, only CT for 28.6% (46/161), only MRI for 11.8% (19/161), both CT and MRI for 5.6% (9/161), 2 or more years of follow-up ultrasound in addition to CT for 21.7% (35/161), 2 or more years of follow-up ultrasound in addition to MRI for 7.5% (12/161), and 2 or more years of follow-up ultrasound in addition to both CT and MRI for 7.5% (12/161). Of the 102 patients who underwent follow-up CT, 11 had only unenhanced images available, 39 had only contrast-enhanced images, and 52 had both unenhanced and contrast-enhanced images. Of the 52 patients who underwent MRI, only two underwent an unenhanced examination, and the others underwent an unenhanced and contrast-enhanced examination. A total of 21 patients had both CT and MRI available. Eighty-seven patients underwent follow-up ultrasound examinations. Among the 161 lesions, 58.4% (94/161) were confirmed to represent an AML on CT or MR images (Fig. 1). One lesion (0.6%) represented a stone at CT, and one lesion (0.6%) represented a hyperdense cyst at CT. No CT or MRI correlate was found for 11.8% (19/161) of lesions, and 3.1% (5/161) were not visualized at follow-up ultrasound. An additional 23.6% (38/161) of lesions were stable at 2-year or later follow-up imaging (23 ultrasound, 15 CT; mean follow-up period, 63.6 months; range, 24 115 months). The 15 lesions found at CT that were stable for 2 or more years of followup did not exhibit macroscopic fat or calcification or meet the criteria for a simple cyst. Two AJR:209, August 2017 347

lesions (1.2%) were evaluated with only contrast-enhanced CT less than 1 month after the baseline ultrasound, and the CT did not show macroscopic fat or calcification or meet the criteria for a simple cyst. These lesions were considered indeterminate because no further follow-up imaging was available. One lesion (0.6%), in a 65-year-old man, was evaluated with only unenhanced and contrast-enhanced CT 23 months after baseline ultrasound. CT showed no macroscopic lipid, solid enhancement, and an interval size increase of 6 mm from the baseline ultrasound (10 to 16 mm). This lesion was presumed to represent RCC, although the patient was lost to further followup. On retrospective review, the echogenicity of this lesion was similar to that of renal sinus fat, and the lesions did not exhibit shadowing, internal cysts, or an anechoic rim (Fig. 2). Overall, the one presumed malignant lesion and the two indeterminate lesions constituted 1.9% (3/161) of the series, and the Fig. 1 70-year-old woman with ultrasound finding of 8-mm hyperechoic renal lesion in upper pole of right kidney. MRI and CT characteristics indicate presence of macroscopic fat, confirming diagnosis of angiomyolipoma deemed clinically insignificant in this study. A, Ultrasound image shows hyperechoic renal lesion (arrow) diagnosed as angiomyolipoma at subsequent CT and MRI. B, In-phase T1-weighted MR image shows T1- hyperintense lesion (arrow) corresponding to that in A. C, Opposed-phase T1-weighted MR image shows etching artifact (arrow) at internal interface between lesion and renal parenchyma. D, Fat-suppressed T1-weighted MR image shows confluent internal loss of signal intensity (arrow). E, Unenhanced CT image shows lesion attenuation of 45 HU (arrow). A other 98.1% (158/161) of lesions were considered clinically insignificant. Discussion In our study, approximately 98% of hyperechoic renal lesions measuring 1 cm or less were deemed clinically insignificant, and only one lesion was a presumed RCC. Most of the lesions deemed clinically insignificant represented a benign AML. Other common reasons for deeming a lesion insignificant included nonvisualization of the lesion at follow-up imaging and the finding of long-term stability. In rare instances, follow-up imaging showed the lesion to represent an alternative benign diagnosis (e.g., a stone or hyperdense cyst). Only one lesion was presumed to represent RCC; two lesions were not classified as insignificant because of insufficient follow-up imaging. The overall findings support a favorable outcome for hyperechoic renal lesions measuring 1 cm or less. B D Controversy regarding the management of hyperechoic renal lesions reflects an understanding that AML and RCC may share this ultrasound feature. Previous studies [12, 13] have shown that both RCC and AML can have echogenicity equal to or greater than that of renal sinus fat. Siegel et al. [13] found that AMLs are associated with shadowing, whereas Yamashita et al. [12] found that RCCs may have an anechoic rim and contain intratumoral cysts. Besides having echogenicity similar to that of renal sinus fat, the single presumed RCC in our series did not have any of these ancillary features to allow differentiation from AML. However, this lesion occurred in a 65-year-old man, illustrating a potential role for patient demographic characteristics in influencing risk assessment for such lesions: sporadic AMLs are approximately 4 times as common in women and occur during middle age, whereas RCC occurs nearly twice as often in men and at a mean age in the sixth de- C E 348 AJR:209, August 2017

A Fig. 2 65-year-old man with hyperechoic renal lesion corresponding to presumed renal cell carcinoma. A, Ultrasound image shows 1-cm hyperechoic lesion (arrow) in midpole of right kidney. B and C, CT image obtained 23 months after ultrasound shows lesion (arrow) has increased in size to 1.6 cm and exhibits solid enhancement (increase in attenuation from 24 HU unenhanced [B] to 241 HU contrast-enhanced [C]). Lesion also does not contain macroscopic fat. Though patient was lost to further follow-up, lesion was presumed to represent renal cell carcinoma. cade [14 17]. Investigators attempting to differentiate AML from RCC using ultrasound observed RCC to be more frequent in men [13]. Thus, female sex and younger age further support a suspected diagnosis of AML for an incidental hyperechoic renal lesion. That there was only a single presumed RCC in our cohort may seem a low rate of occurrence. This frequency, however, can be expected to be lower in a cohort of only lesions 1 cm or smaller, as in our study, compared with earlier cohorts that included lesions of all sizes. In addition, in earlier studies of the ultrasound evaluation of renal lesions, investigators may have overestimated the incidence of hyperechoic RCC. With the older equipment used in such studies, sonographers may have underestimated isoechoic or hypoechoic RCC, which may have been more difficult to detect. With ongoing technical improvements, ultrasound has better spatial resolution and tissue contrast with fewer artifacts, potentially allowing more reliable detection and characterization of small solid renal lesions [18]. Nonetheless, there is a paucity of recent literature on studies conducted with updated equipment to better inform the clinical significance of small hyperechoic renal lesions. Although the practice of performing CT or MRI for hyperechoic renal lesions 1 cm or larger is generally accepted, the optimal follow-up of lesions 1 cm or smaller is unclear. This lack of clarity reflects the high likelihood of benignity of very small lesions. Our findings suggest that 1-cm or smaller hyperechoic renal lesions may not require followup imaging given the exceedingly high frequency with which such lesions are clinically insignificant. Factors that influence the determination regarding performing follow-up imaging include patient age, sex, symptoms, and personal history of RCC or predisposing conditions. When a decision is made to forgo any follow-up imaging for an incidental hyperechoic renal lesion, it should be ensured that the lesion meets previously described ultrasound criteria for AML. If imaging is performed, then follow-up CT or MRI and surveillance ultrasound may all be reasonable approaches, according to our observations. Though results of earlier studies have suggested that evaluation of 1-cm or smaller hyperechoic renal lesions may be challenging with CT or MRI owing to partial volume-averaging effects, results with these modalities confirmed a diagnosis of AML in most of the cases in our cohort while also occasionally reliably indicating an alternative diagnosis. Our study had several limitations. First, a pathologic reference standard was not available for the lesions classified as AML. However, specific CT and MRI features are considered to be diagnostic of AML in clinical practice, and histologic sampling of these lesions is typically not warranted. In addition, follow-up CT was performed for a range of reasons and with varying protocols, such that some did not include an unenhanced scan. Furthermore, the CT images were generally reconstructed at 4-mm slice thickness. The use of an unenhanced phase and of thinner slices may have improved the detection of very small areas of fat in lesions and hence led to a confident diagnosis of AML in lesions that were otherwise deemed insignificant on B the basis of 2 or more years of stability or classified as indeterminate [19, 20]. In addition, the explanation for the observed hyperechoic focus in the baseline examination for lesions not visualized at follow-up imaging remains unknown. Nonetheless, the lack of a lesion on follow-up images supports the clinically insignificant nature of the initial finding. Finally, the patient with the single case of presumed RCC was lost to follow-up, so the suspected malignant pathologic entity was not confirmed. Conclusion The overwhelming majority of incidental hyperechoic renal lesions measuring 1 cm or less were clinically insignificant at follow-up imaging. Such lesions may not require further follow-up imaging in patients at low risk, a decision based in part on the demographic characteristics. References 1. Katabathina VS, Vikram R, Nagar AM, Tamboli P, Menias CO, Prasad SR. Mesenchymal neoplasms of the kidney in adults: imaging spectrum with radiologic-pathologic correlation. RadioGraphics 2010; 30:1525 1540 2. Schieda N, Avruch L, Flood TA. Small (< 1 cm) incidental echogenic renal cortical nodules: chemical shift MRI outperforms CT for confirmatory diagnosis of angiomyolipoma (AML). Insights Imaging 2014; 5:295 299 3. Yamakado K, Tanaka N, Nakagawa T, Kobayashi S, Yanagawa M, Takeda K. Renal angiomyolipoma: relationships between tumor size, aneurysm formation, and rupture. Radiology 2002; 225:78 82 4. Maclean DF, Sultana R, Radwan R, McKnight L, C AJR:209, August 2017 349

Khastgir J. Is the follow-up of small renal angio- paper of the ACR incidental findings committee. and management of renal angiomyolipoma. J Urol myolipomas a necessary precaution? Clin Radiol J Am Coll Radiol 2010; 7:754 773 2002; 168:1315 1325 2014; 69:822 826 10. Zagoria RJ. Imaging of small renal masses: a 16. Prasad SR, Surabhi VR, Menias CO, Raut AA, 5. Halpenny D, Snow A, McNeill G, Torreggiani medical success story. AJR 2000; 175:945 955 Chintapalli KN. Benign renal neoplasms in WC. The radiological diagnosis and treatment of 11. Israel GM, Hindman N, Hecht E, Krinsky G. The adults: cross-sectional imaging findings. AJR renal angiomyolipoma-current status. Clin Radiol 2010; 65:99 108 6. Forman HP, Middleton WD, Melson GL, Mc- Clennan BL. Hyperechoic renal cell carcinomas: increase in detection at US. Radiology 1993; 188:431 434 7. Farrelly C, Delaney H, McDermott R, Malone D. Do all non-calcified echogenic renal lesions found on ultrasound need further evaluation with CT? Abdom Imaging 2008; 33:44 47 8. Jewett MA, Mattar K, Basiuk J, et al. Active surveillance of small renal masses: progression patterns of early stage kidney cancer. Eur Urol 2011; 60:39 44 9. Berland LL, Silverman SG, Gore RM, et al. Managing incidental findings on abdominal CT: white use of opposed-phase chemical shift MRI in the diagnosis of renal angiomyolipomas. AJR 2005; 184:1868 1872 12. Yamashita Y, Ueno S, Makita O, et al. Hyperechoic renal tumors: anechoic rim and intratumoral cysts in US differentiation of renal cell carcinoma from angiomyolipoma. Radiology 1993; 188:179 182 13. Siegel CL, Middleton WD, Teefey SA, McClennan BL. Angiomyolipoma and renal cell carcinoma: US differentiation. Radiology 1996; 198:789 793 14. Shin NY, Kim MJ, Chung JJ, Chung YE, Choi JY, Park YN. The differential imaging features of fatcontaining tumors in the peritoneal cavity and retroperitoneum: the radiologic-pathologic correlation. Korean J Radiol 2010; 11:333 345 15. Nelson CP, Sanda MG. Contemporary diagnosis 2008; 190:158 164 17. Lipworth L, Tarone RE, McLaughlin JK. The epidemiology of renal cell carcinoma. J Urol 2006; 176:2353 2358 18. Hélénon O, Correas JM, Balleyguier C, Ghouadni M, Cornud F. Ultrasound of renal tumors. Eur Radiol 2001; 11:1890 1901 19. Bosniak MA, Megibow AJ, Hulnick DH, Horii S, Raghavendra BN. CT diagnosis of renal angiomyolipoma: the importance of detecting small amounts of fat. AJR 1988; 151:497 501 20. Kurosaki Y, Tanaka Y, Kuramoto K, Itai Y. Improved CT fat detection in small kidney angiomyolipomas using thin sections and single voxel measurements. J Comput Assist Tomogr 1993; 17:745 748 350 AJR:209, August 2017