ADULT UROLOGY

IMAGING CHARACTERISTICS OF MINIMAL FAT RENAL ANGIOMYOLIPOMA WITH HISTOLOGIC CORRELATIONS JASON HAFRON, JAMES D. FOGARTY, DAVID M. HOENIG, MAOMI LI, ROBERT BERKENBLIT, AND REZA GHAVAMIAN

ABSTRACT Objectives. To review our experience with minimal fat angiomyolipomas (AMLs) and correlate the confirmed pathologic diagnosis with preoperative radiologic features to evaluate the feasibility of an accurate diagnosis preoperatively. On rare occasions, renal AMLs contain minimal amounts of fat that are not identified on cross-sectional imaging. Methods. From November 1998 to August 2003, 6 patients (seven lesions) underwent renal surgery with the preoperative diagnosis of renal cell carcinoma as determined by preoperative imaging studies, and the finding of AML was unexpected. A single pathologist and radiologist reviewed all pathologic specimens and radiologic examinations. The characteristic findings were recorded and compared with those in published reports. Results. Pathologic review of the specimens demonstrated microscopic amounts of mature adipocytes, abnormally thickened blood vessels, and smooth muscle cells in all seven lesions. The mean estimated fat content was 4.1% (range 3% to 10%). Hounsfield unit measurement of the lesions on unenhanced computed tomography images revealed all lesions to be hyperdense relative to the normal kidney parenchyma and enhanced by at least 90 Hounsfield units or more with the administration of intravenous contrast. Enhancement was homogenous throughout each of the seven masses. Conclusions. Minimal fat renal AML tumors are typically hyperdense relative to the normal kidney parenchyma and demonstrate homogenous enhancement with the administration of intravenous contrast on computed tomography. These results may have important implications when planning partial versus radical nephrectomy by open or minimally invasive techniques. UROLOGY 66: 1155–1159, 2005. © 2005 Elsevier Inc.

R

enal angiomyolipomas (AMLs) have a unique histologic composition of adipose tissue, thickwalled blood vessels, and smooth muscle. The presence of fat enables most of these masses to be diagnosed with great accuracy using current imaging techniques.1 However, this mixed mesenchymal lesion in the extreme form may only contain one element, smooth muscle or adipose tissue, or may only contain minimal fat that is not clearly identified on cross-sectional imaging.2 The rare occasions in which AMLs have no detectable fat on cross-sectional imaging and miniFrom the Departments of Urology, Pathology, and Radiology, Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, New York Reprint requests: Jason Hafron, Department of Urology, Montefiore Medical Center, 3400 Bainbridge Avenue, Bronx, NY 10461. E-mail: [email protected] Submitted: December 12, 2004, accepted (with revisions): June 24, 2005 © 2005 ELSEVIER INC. ALL RIGHTS RESERVED

mal fat on histologic examination have been sporadically reported.3–7 These lesions typically enhance after intravenous contrast administration and are often confused with renal cell carcinoma (RCC). This, in turn, may lead to unnecessary surgical excision. In contrast, the preoperative diagnosis of AML allows for either a watchful waiting or surgical approach (total nephrectomy versus partial nephrectomy), depending on tumor size and/or symptoms. At our institution, 6 patients have had seven renal AML tumors that were removed on the preoperative presumption of RCC. All were found to have a minimal amount of fat on final histologic analysis. We retrospectively reviewed these cases to analyze their distinguishing radiologic and pathologic features. MATERIAL AND METHODS From November 1998 to August 2003, 11 AMLs were removed at our institution for various reasons, including hem0090-4295/05/$30.00 doi:10.1016/j.urology.2005.06.119 1155

FIGURE 1. Photomicrograph showing minimal fat renal AML.

orrhage, increasing size, and suspicion of malignancy. In 6 patients, the preoperative diagnosis was RCC as determined by the preoperative imaging studies, and the finding of AML was unexpected. Five patients underwent partial nephrectomy and the sixth, with multifocal lesions, underwent laparoscopic radical nephrectomy. The pathologic and radiologic studies of these patients constituted our study group and were reexamined. All specimens were reviewed by one pathologist (M.L.). The fat content was estimated as the percentage of area occupied by fat vacuoles. Any areas 5 mm or larger that were occupied predominantly by fat-containing cells were also recorded. Positive immunoreactivity for HMB-45, a monoclonal antibody against melanoma-associated antigen, typical of AML, was evaluated in all specimens. All patients underwent preoperative imaging that included ultrasonography, computed tomography (CT), and/or magnetic resonance imaging (MRI). The CT protocol (with or without intravenous contrast and the thickness of axial slices) varied, depending on the clinical indication for the scan. A single radiologist (R.B.) centrally reviewed all images. On ultrasonography, echogenicity and the presence of tumor shadowing were recorded. The attenuation of the lesion on contrast-enhanced and unenhanced CT images, lesion size, location, homogeneity, and pattern of enhancement were noted on the CT scans. The mean attenuation value of the lesion was obtained with multiple measurements within the mass. The pattern of attenuation was characterized by visual inspection. On MRI, the signal intensity of each lesion was evaluated on T1-weighted and T2-weighted images. Gadolinium-enhanced T1-weighted images were used to assess the size, location, and pattern of enhancement. Fat saturation techniques were used to assess the presence of fat.

RESULTS HISTOPATHOLOGIC FINDINGS All tumors consisted of smooth muscle, thickwalled blood vessels, and fat-containing cells with characteristic bundles of smooth muscle emanating from the vessel walls (Fig. 1). The diagnosis of AML was confirmed by positive immunohistochemical staining for HMB-45 in the tumor cells. All the tumors contained 10% or less fat scattered throughout and had no single focus of fat greater 1156

FIGURE 2. (A) Non-contrast-enhanced CT scan showing hyperdense minimal fat AML. (B) Contrast-enhanced CT scan depicting homogenous enhancement.

than 5 mm. The mean estimated fat content was 4.1% (range 3% to 10%). RADIOLOGIC FINDINGS Macroscopic fat was not clearly seen in any of the lesions on cross-sectional imaging. The mean lesion diameter on cross-sectional imaging was 3.2 cm (range 1 to 7.5). Ultrasonography was performed in 3 of the 6 patients, and all lesions demonstrated well-circumscribed masses without tumor shadowing. Two of the masses were hyperechoic and one was isoechoic. On CT, all lesions were well-demarcated, peripheral lesions that induced a contour deformity in the shape of the kidney. Intratumoral calcifications or perinephric changes were not seen in any of the cases. Hounsfield unit measurement of the lesions on unenhanced CT (available for 4 patients) revealed all lesions to be hyperdense (mean 44 Hounsfield units, range 42 to 48) relative to the normal kidney parenchyma (Fig. 2A). All these masses enhanced by at least 90 Hounsfield units or more with administration of intravenous contrast (Fig. 2B). Homogenous enhancement was UROLOGY 66 (6), 2005

FIGURE 3. Contrast-enhanced CT scan showing multicentric minimal fat AML masses with typical AML lesion. (A) Arrow indicates one minimal fat AML tumor. (B) White arrow indicates typical AML lesion, with hypodense areas representing fat. Black arrow indicates second minimal fat AML lesion in same kidney.

noted throughout all seven masses. One male patient had multiple tumors in one kidney. Two masses were suspicious for RCC with homogenous enhancement and no evidence of fat. However, the predominant lesion was a 7.5 ⫻ 5.3-cm fat-containing lesion in the right lower pole of the kidney (Fig. 3). He underwent right laparoscopic radical nephrectomy, which revealed two AMLs with minimal fat and a large classic AML tumor. Two patients underwent MRI. No areas of signal loss consistent with fat were identified in the lesions on the fat saturation sequences. COMMENT The demonstration of fat in a renal mass on radiologic imaging is a diagnostic tool to exclude RCC and is indicative of AML.8,9 Case reports have been published of malignant renal tumors that UROLOGY 66 (6), 2005

contained fat but all of these also had areas of calcification.10 More recently, three reports have described 4 patients with RCC containing fat without calcifications on CT.11–13 However, most would agree the distinctive radiographic characteristic of fat density (Hounsfield units of ⫺90 to ⫺30 CT) without areas of calcification in a renal mass is indicative of AML. Nonetheless, CT may fail to identify fat in a renal AML with a preponderance of mesenchymal elements (blood vessels, muscle) or microscopic amounts of fat. Bosniak et al.14 showed that thin (5-mm) cuts and non-contrastenhanced CT scans offer the greatest probability to demonstrate macroscopic amounts of fat. However, in all the AMLs in our study, thin-slice CT scanning was unable to detect the microscopic amounts of fat within the lesions. On unenhanced CT images in this series, all lesions were homogenous and hyperdense compared with the normal renal parenchyma. Previous series have reported similar findings.4,6,7 Kim et al.,6 in the largest series of minimal fat AML, reported homogenous hyperdense findings on unenhanced scans in only 53% of the minimal fat AML renal lesions. However, in the same study, 22% of the patients with RCC had hyperattenuation on unenhanced CT scans. Hyperdense renal masses on unenhanced CT images is not specific for minimal fat AML and can be a finding of complicated benign renal cysts, renal metastasis, leiomyoma, and even RCC.4 Therefore, hyperdense and homogeneity on unenhanced CT scans are features commonly seen but are clearly not diagnostic of these lesions. Another characteristic feature noted in our series and in previous series was homogenous enhancement.4,6,7 The homogenous enhancement was probably a result of the uniform solid components of these lesions. The predominant component in all tumors was smooth muscle, without any evidence of hemorrhage or necrosis. In contrast, RCCs usually show heterogeneous enhancement and commonly have hemorrhage or necrosis within the tumors at pathologic examination.15 Combining ultrasound findings with CT findings may be helpful. Most AML lesions on ultrasonography are markedly hyperechoic relative to the renal parenchyma secondary to multiple fat/ smooth muscle interfaces.1,16 The finding of a discrete focus with marked echogenicity against the backdrop of hypoechoic renal cortex is a sonographic finding highly suggestive of intratumor fat or AML.2 However, hyperechoic and hypoechoic areas suggestive of fat do not apply to all renal AMLs. Paivansalo et al.7 demonstrated that tumors with a low percentage of fat (5% to 10%) were associated with an echo texture that was isoechoic or hypoechoic relative to the renal parenchyma, 1157

and AMLs with a greater percentage of fat were highly echogenic. Jinzaki et al.4 and Trigaux et al.3 also found that minimal fat AMLs appear homogeneously isoechoic and hypoechoic on ultrasonography. In the current series, the echo texture in the tumors was hyperechoic in 2 patients and isoechoic in 1 patient. Therefore, ultrasonography in our experience did not aid in the diagnosis of minimal fat AML. MRI was unable to detect fat within the lesions in this series. On T1-weighted and T2-weighted MRI using fat saturation techniques, none of the AMLs showed significant intensity that was similar to that of fat. This was not surprising, as macroscopic fat would have to be present for this technique to demonstrate the presence of fat. Opposed phase imaging, also known as chemical shift or out-ofphase imaging, is another valuable technique in MRI to assess for fat. It is commonly used to evaluate adrenal lesions for the presence of fat to help assess whether an adrenal nodule is an adenoma or other adrenal lesion. This technique was not used in either of our patients. However, it is unlikely to be of help in the assessment of minimal fat AMLs for two reasons. First, it detects the presence of microscopic intracellular fat on the basis of the presence of adequate amounts of fat and water being present in the same imaging pixel. If adequate amounts of fat and water are present in the same pixel, a lesion will lose signal on the out-of-phase images compared with the in-phase images. However, if only minimal amounts of fat or water are present in the same pixel, no detectable signal loss will occur. It is well known that a significant number of adrenal adenomas are lipid poor and do not demonstrate this expected signal loss on out-ofphase images, and the same would probably be expected with lipid-poor AMLs.17 In the present series, 1 patient had multicentric AMLs on preoperative imaging without any history of tuberous sclerosis. Two of the lesions had no evidence of fat and were highly suspicious for RCC and one lesion clearly had fat present. The incidence of AMLs synchronous with RCC is 7%,18 suggesting that in the presence of multiple small AMLs and non-fat-containing, enhancing solid renal masses, renal exploration and excision of the suspicious mass is warranted. Even though our sample size was small, hyperdense and homogenous enhancement was demonstrated in all the lesions and should be verified in larger series. One disadvantage of our study was that all the CT scans did not follow the same protocol; ideally these should be standardized. In particular, scanning thickness and timing were different between patients, preventing us from assessing the enhancement patterns during the course of the study. Recently, prolonged en1158

hancement was also seen more often with minimal fat AMLs and may be a potentially useful diagnostic tool in the future.6 Aside from radiologic imaging, no other methods are available to diagnose these lesions confidently. Some investigators have proposed nonradiologic methods for elucidating the nature of these indeterminate masses, specifically percutaneous biopsy. Biopsy of renal masses is limited secondary to the high proportion of false-negative results.19 The treatment of these lesions should follow the accepted principles for the management of RCC, but the presence of these radiologic characteristics may influence the decision toward a nephron-sparing approach. CONCLUSIONS In the current era of minimally invasive surgery and emerging ablative technologies, we have identified radiologic characteristics associated with an unusual benign renal lesion. Urologists should consider these characteristics when discussing with their patients the substantial chance of a benign lesion and should strongly consider a nephron-sparing approach when treating this type of lesion.20 These results may have important implications when planning partial versus radical nephrectomy by open or minimally invasive techniques. REFERENCES 1. Bosniak MA: Angiomyolipoma (hamartoma) of kidney: a preoperative diagnosis is possible in virtually every case. Urol Radiol 3: 135–142, 1981. 2. Raghavendra BN, Bosniak MA, and Megibow AJ: Small angiomyolipoma of the kidney: sonographic-CT evaluation. AJR Am J Roentgenol 141: 575–578, 1983. 3. Trigaux J, Pails C, and Van Beers B: Atypical renal hamartomas: ultrasonography, computed tomography and angiographic findings. J Clin Ultrasound 21: 41– 43, 1993. 4. Jinzaki M, Tanimot A, Narimatsu Y, et al: Angiomyolipoma: imaging findings in lesions with minimal fat. Radiology 205: 497–502, 1997. 5. Sherman JL, Hartman DS, Friedman AC, et al: Angiomyolipoma: computed tomographic-pathologic correlation of 17 cases. AJR Am J Roentgenol 137: 1221–1226, 1981. 6. Kim JK, Park SY, Shon JH, et al: Angiomyolipoma with minimal fat differentiation from renal cell carcinoma at biphasic helical CT. Radiology 230: 677– 684, 2004. 7. Paivansalo M, Lahde S, Hyvarinen S, et al: Renal angiomyolipoma: ultrasonographic, CT, angiographic and histologic correlation. Acta Radiol 32: 239 –243, 1991. 8. Lemaitre L, Claudon M, Dubrulle F, et al: Imaging of angiomyolipoma. Semin Ultrasound CT MR 18: 100 –114, 1997. 9. Davidson AJ, Hartman DS, Choyke PL, et al: Radiological assessment of renal masses: implications for patient care. Radiology 202: 297–305, 1997. 10. Henderson J, Germany R, Peavy P, et al: Fat density in renal cell carcinoma: demonstration with computerized tomography. J Urol 157: 1347–1348, 1997. 11. Schuster TG, Ferguson MR, Baker DE, et al: Papillary renal cell carcinoma containing fat without calcification mimicking angiomyolipoma on CT. AJR Am J Roentgenol 183: 1402–1404, 2004. UROLOGY 66 (6), 2005

12. D’Angelo PC, Gash JR, Horn AW, et al: Fat in renal cell carcinoma that lacks associated calcifications. AJR Am J Roentgenol 178: 931–932, 2002. 13. Lesavre A, Correas JM, Merran S, et al: CT of papillary renal cell carcinomas with cholesterol necrosis mimicking angiomyolipoma. AJR Am J Roentgenol 181: 143– 145, 2003. 14. Bosniak MA, Megibow AJ, Hulnick DH, et al: CT diagnosis of renal angiomyolipoma: the importance of detecting small amounts of fat. AJR Am J Roentgenol 151: 497–501, 1988. 15. Russo P: Renal cell carcinoma: presentation, staging and surgical treatment. Semin Oncol 27: 160 –176, 2000. 16. Hartman DS, Goldman SM, Friedman AC, et al: Angiomyolipoma: ultrasonic-pathologic correlation. Radiology 139: 451– 458, 1981.

UROLOGY 66 (6), 2005

17. Outwater EK, Siegelman ES, Huang AB, et al: Adrenal masses: correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposedphase sequences. Radiology 200: 749 –752, 1996. 18. Blute ML, Malek RS, and Segura JW: Angiomyolipoma: clinical metamorphosis and concepts for management. J Urol 139: 20 –23, 1988. 19. Dechet C, Zincke H, Sebo T, et al: Prospective analysis of computerized tomography and needle biopsy with permanent sectioning to determine the nature of solid renal masses in adults. J Urol 169: 71–74, 2003. 20. McKiernan J, Yossepowitch O, Kattan MW, et al: Partial nephrectomy for renal cortical tumors: pathologic findings and impact on outcome. Urology 60: 1003–1009, 2002.

1159