MDCT Findings of Renal Cell Carcinoma Associated With Xp11.2 Translocation and TFE3 Gene Fusion and Papillary Renal Cell Carcinoma

Transcription

1 Genitourinary Imaging Original Research Woo et al. MDCT in Renal Cell Carcinoma Genitourinary Imaging Original Research Sungmin Woo 1 Sang Youn Kim 1 Myoung Seok Lee 1 Kyung Chul Moon 2 See Hyung Kim 3 Jeong Yeon Cho 1,4 Seung Hyup Kim 1,4 Woo S, Kim SY, Lee MS, et al. Keywords: multiphase renal MDCT, papillary renal cell carcinoma (RCC), RCC associated with Xp11.2 translocation and TFE3 gene fusion DOI: /AJR Received April 1, 2014; accepted after revision June 19, Department of Radiology, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul , Korea. Address correspondence to S. Y. Kim (iwishluv@empas.com). 2 Department of Pathology, Seoul National University Hospital, Seoul, Korea. 3 Department of Radiology, Keimyung University, Dongsan Hospital, Daegu, Korea. 4 Institute of Radiation Medicine and Kidney Research Institute, Seoul National University Medical Research Center, Seoul, Korea. AJR 2015; 204: X/15/ American Roentgen Ray Society MDCT Findings of Renal Cell Carcinoma Associated With Xp11.2 Translocation and TFE3 Gene Fusion and Papillary Renal Cell Carcinoma OBJECTIVE. The purpose of this study was to compare the MDCT features of renal cell carcinoma (RCC) associated with Xp11.2 translocation and TFE3 gene fusion (Xp11 RCC) and papillary RCC. MATERIALS AND METHODS. The study included 19 and 39 patients with histologically proven Xp11 RCC and papillary RCC, respectively, who underwent multiphase renal MDCT before nephrectomy. CT findings were compared between Xp11 RCC and papillary RCC using the Student t test and chi-square test. Subgroup analyses of small (< 4 cm) renal masses for these features were performed. RESULTS. Patients with Xp11 RCC were younger (p < 0.001), and it was more prevalent in women (p = 0.007). Tumor size was greater in Xp11 RCC (p = 0.004) and more common in cystic change (p < 0.001). Calcification and unenhanced high-attenuating areas were more frequent in Xp11 RCC (p = and 0.026, respectively). Xp11 RCCs were more prevalent in lymph node and distant metastasis (p < and p = 0.031, respectively). Xp11 RCC and papillary RCC showed no significant difference in epicenter, margin, and venous and collecting duct invasion (p = ). Although Xp11 RCC and papillary RCC had lower attenuation than the renal cortex on corticomedullary and early excretory phases (p < 0.001), only Xp11 RCCs were hyperattenuating to the cortex on the unenhanced phase (p < 0.001). Xp11 RCCs had significantly higher attenuation compared with papillary RCCs on all phases (p 0.02). Regarding small masses, cystic change, calcification, and lymph node metastasis were still more frequent in Xp11 RCCs (p 0.016). CONCLUSION. Greater size, more cystic change, calcification, high-attenuating areas on unenhanced imaging, and lymph node and distant metastasis were helpful for differentiating Xp11 RCC from papillary RCC. R enal cell carcinoma (RCC) associated with Xp11.2 translocation and TFE3 gene fusion (Xp11 RCC) has recently been identified as a separate entity according to the 2004 World Health Organization classification of kidney tumors [1]. Xp11 RCC most distinctively presents as papillary architecture composed of clear cells, nested architecture with granular eosinophilic cytoplasm [2]. As a result, histopathologic diagnosis and differentiation between Xp11 RCC and papillary RCC may be difficult when only Hematoxylin and eosin staining is performed. Unlike papillary RCC or other subtypes of RCC, Xp11 RCCs overexpress the TFE3 protein from translocations involving TFE3 gene fusions, rendering nuclear labeling for TFE3 specific to the diagnosis of Xp11 RCC [3]. Without such testing, Xp11 RCCs may be misdiagnosed as papillary RCC and could mislead patient management and prediction of prognosis. Papillary RCCs usually have a favorable prognosis compared with the more conventional clear cell RCC [4]. On the other hand, Xp11 RCCs, especially in the adult population, are associated with a poor prognosis, presenting at an advanced stage and more frequently with lymph node metastasis [1, 5, 6]. Given the possible confusion in diagnosis between the two entities and their different prognoses, routine performance of immunohistochemical staining for TFE3 protein could be advocated; however, routine implementation would lead to an unnecessary increase in the health care costs because Xp11 RCCs are truly rare and not of negligible cost. If present, imaging findings suggestive of Xp11 RCC, which are different from papillary RCC, may aid in performing immuno- 542 AJR:204, March 2015

2 MDCT in Renal Cell Carcinoma histochemistry to make the final diagnosis of Xp11 RCC. Although there have been a few reports that have described radiologic findings of Xp11 RCCs [7 11], to our knowledge, no joint assessment of the MDCT imaging findings of Xp11 RCC and papillary RCC, which could be expected to overlap considering their similar pathologic characteristic of papillary architecture, has been performed. Therefore, the purpose of our study was to retrospectively evaluate the MDCT features of Xp11 RCC and papillary RCC. Materials and Methods Patient Selection Institutional review board approval was obtained for this retrospective study, with a waiver of informed consent. A computerized search of the radiologic and pathologic databases at Seoul National University Hospital was performed to search for patients with histopathologically proven Xp11 RCC and papillary RCC from January 2008 to December An additional five patients surgically and pathologically diagnosed with Xp11 RCC at Keimyung University Dongsan Hospital were included. Patients who had not undergone preoperative multiphase renal MDCT (consisting of unenhanced, corticomedullary, and early excretory phases) were excluded. Among 1397 patients who were diagnosed with RCC after surgery, the patients excluded because of other abnormalities had clear cell (n = 1098), clear cell-papillary (n = 19), chromophobe (n = 110), collecting duct (n = 4), undifferentiated (n = 28), multilocular cystic (n = 3), and acquired cystic disease associated RCC (n = 5), leaving 28 and 102 patients with histopathologically proven Xp11 RCC and papillary RCC, respectively. Next, 72 patients were excluded because of a lack of preoperative multiphase renal MDCT. Finally, we identified 58 patients with a mean age of 51.0 ± 16.5 years (range, years) including A B C D E Fig year-old woman with pathologically proven Xp11.2 translocation and TFE3 gene fusion. A C, Axial unenhanced (A), corticomedullary phase (B), and early excretory phase (C) MDCT images show 5.2-cm-sized well-circumscribed mass (arrow) with large necrotic portion (asterisk, B) in right kidney upper pole. Posterior solid component (arrowhead, A) is relatively hyperattenuating to contralateral renal cortex on unenhanced phase (A). Generally, degree of enhancement of solid portion of renal mass is less than that of renal parenchyma. Attenuation values of mass were 51, 98, and 93 HU on unenhanced phase, corticomedullary phase, and early excretory phase, respectively, and they were measured to be 34, 185, and 195 HU in renal cortex on unenhanced phase, corticomedullary phase, and early excretory phase, respectively. D, Coronal plane image shows focal area of calcification (arrow) along superior aspect of renal mass. E, Axial corticomedullary phase image reveals enlarged retrocaval lymph node (arrow) that was confirmed as metastatic node after surgery. 19 patients who had Xp11 RCC (mean age, 40.5 ± 16.4 years; range, years) and 39 (13 and 26 patients with type 1 and 2 papillary RCC, respectively) with papillary RCC (mean age, 56.1 ± 14.1 years; range, 28 82). CT Technique CT was performed using one of several MDCT scanners (Ingenuity, Brilliance 64, and Mx8000, Philips Healthcare; Somatom Definition AS, Sensation 64, and Sensation 16, Siemens Healthcare; Discovery CT750 HD, LightSpeed Ultra, and LightSpeed VCT, GE Healthcare; and Aquilion One, Toshiba Medical Systems). Multiphase renal MDCT is generally acquired with the following protocol at our institution: 120 kvp; mas with tube current modulation; section thickness and reconstruction interval of 5 mm, and unenhanced, corticomedullary phase, and early excretory phase performed by injecting IV contrast media (iopromide, Ultravist 370, Schering) at 3.0 ml/s, with a dose of approximately 2 ml/kg (maximum of 150 ml). Corticomedullary phase and early excretory phase images are acquired approximately and seconds after injection of contrast media [12]. Image Analysis Two radiologists with 3 and 9 years of experience in genitourinary imaging reviewed the im- AJR:204, March

3 Woo et al. ages in consensus at a PACS system (Maroview, Marosis). Both reviewers were aware that all patients had undergone nephrectomy for Xp11 RCC or papillary RCC but were blinded to the clinical, surgical, and histologic findings. The reviewers evaluated each renal mass for the following: size as the greatest diameter on the axial plane; epicenter (endophytic vs exophytic) according to the location of the lesion center with regard to the outline of the kidney [13]; solid or cystic character, categorized into solid (no radiologic evidence of cysts or necrosis), predominantly solid (< 25% of cystic portion or necrosis), mixed solid and cystic (25 75%), and predominantly cystic (> 75%); presence of calcification; presence of high attenuation other than calcification; margin of the renal mass (well-defined vs infiltrative); vascular or collecting system invasion; and lymph node or distant metastasis. Furthermore, the attenuation values of the renal mass and adjacent renal cortex were measured by placing ROIs by a third radiologist with 7 years experience. Care was taken to avoid partial volume effects from calcification and cystic change or necrosis. To decrease the effect of other intrinsic factors on the attenuation values, we also calculated a corrected attenuation value (CAV) according to the following equation [13, 14]: CAV p = L p (ca p / A p ), where L is the lesion of interest; A, the aorta; p, the scanning phase; CAV p, the CAV of the lesion for each scanning phase; and A p, the measured attenuation value of the aorta for each phase; ca p is a set constant value (50, 200, and 120 HU for the unenhanced phase, corticomedullary phase, and early excretory phase, respectively) for each phase. Pathologic Analysis All pathologic analyses were performed by a uropathologist with 17 years of experience. Nephrectomy specimens were fixed in 10% formalin and embedded in paraffin, and 4-µm-thick sections were stained using H and E and assessed for any possible features of Xp11 RCC. For the final diagnosis of Xp11 RCC, overnight immunohistochemical staining for TFE was performed [3]. In addition, other immunohistochemical staining for vimentin, CD10, epidermal growth factor receptor, e-cadherin, melanosome (HMB45), CD117, Ki-67, and cytokeratin 7 were also routinely performed at our institution for differential diagnosis between subtypes of RCC. TABLE 1: Patient and Tumor Characteristics of 58 Patients With Xp11 Renal Cell Carcinoma (RCC) and Papillary RCC Characteristic Xp11 RCC (n = 19) Statistical Analysis Continuous data and categoric data are expressed as mean ± SD and percentages, respectively. The evaluated variables were compared between Xp11 RCC and papillary RCC using the unpaired Student t test or chi-square test. In addition, subgroup analyses of small renal masses (size < 4 cm) for imaging findings were performed. Statistical analyses were performed with software (MedCalc, version , MedCalc Software and PASW, version 18, SPSS). A p value less than 0.05 was considered to indicate a significant difference. Results Patient and Tumor Characteristics Patient and tumor characteristics are summarized in Table 1. The mean age was significantly lower in patients with Xp11 RCC (p < 0.001). Of note, there were three Xp11 RCC patients who were 20 years old or under, whereas all patients diagnosed with papillary RCC were older than 20 years. Xp11 RCC was significantly more prevalent in women compared with papillary RCC (p = 0.007). The mean tumor size was significantly greater in patients with Xp11 RCC (p = 0.004). The distribution of solid and cystic RCCs was significantly different between the two entities (p = 0.001). Xp11 RCCs were solid, predominantly solid, mixed solid and cystic, and predominantly cystic in one (5.3%), 10 (52.6%), five (26.3%), and three (15.8%) patients, respectively. Papillary RCCs were solid, predominantly solid, mixed solid and cystic, and predominantly cystic in 24 (61.5%), seven (17.9%), five (12.8%), and Papillary RCC (n = 39) p Age (y) 40.5 ± 16.4 (11 73) 56.1 ± 14.4 (28 82) < Sex (M:F) 42.1:57.9 (8:11) 79.5:20.5 (31:8) Size (cm) 6.9 ± 4.5 ( ) 3.4 ± 2.0 ( ) Exophytic epicenter 36.8 (7/19) 51.3 (20/39) Solid and cystic pattern Solid 5.3 (1/19) 61.5 (24/39) (< 0.001) a Predominantly solid 52.6 (10/19) 17.9 (7/39) Mixed solid and cystic 26.3 (5/19) 12.8 (5/39) Predominantly cystic 15.8 (3/19) 7.7 (3/39) Calcification 68.4 (13/19) 23.1 (9/39) Unenhanced high attenuation 68.4 (13/19) 35.9 (14/39) Infiltrative margin 5.3 (1/19) 5.1 (2/39) Renal vein invasion 5.3 (1/19) 7.7 (3/39) Collecting duct invasion 15.8 (3/19) 10.3 (4/39) Lymph node metastasis 57.9 (11/19) 2.6 (1/39) < Distant metastasis 15.8 (3/19) 0 (0/39) Note Data age and size are mean ± SD with range in parentheses. Other data are percentages with raw numbers in parentheses. a Solid versus others compared using chi-square test. three (7.7%) patients, respectively. When we dichotomized the tumors into those with and without radiologic evidence of cystic change or necrosis, papillary RCCs were significantly more prevalent in the solid pattern (p < 0.001). Both calcification and areas of noncalcified high attenuation on unenhanced images were more commonly found in Xp11 RCCs (p = and p = 0.026, respectively). Lymph node (p < 0.001) and distant metastases (p = 0.031) were more prevalent in Xp11 RCCs than in papillary RCCs. Specifically, metastases to the adrenal gland, lungs, and bone in Xp11 RCC patients were detected and pathologically proven (n = 1) or diagnosed with follow-up CT and bone scans (n = 2). Xp11 RCC and papillary RCC were not statistically different for the following binary imaging features: epicenter (p = 0.403), margin (p = 1.000), venous invasion (p = 1.000), and collecting duct invasion (p = 0.673). Representative cases of Xp11 RCC and papillary RCC are shown in Figures 1 and 2, respectively. Attenuation Values of the Renal Mass and Renal Cortex The measured and corrected attenuation values of Xp11 RCCs and papillary RCCs along with the background renal cortex are 544 AJR:204, March 2015

4 MDCT in Renal Cell Carcinoma TABLE 2: Measured Attenuation Values of Renal Mass and Adjacent Renal Parenchyma in 58 Patients With Xp11 Renal Cell Carcinoma (RCC) and Papillary RCC Xp11 RCC (n = 19) Papillary RCC (n = 39) p Attenuation (HU) summarized in Tables 2 and 3, respectively. The attenuation values of the renal mass were significantly higher in all phases in Xp11 RCC compared with papillary RCC (p 0.02). The mean corticomedullary phase attenuation of the renal cortex was statistically greater in Xp11 RCC patients than in papillary RCC patients (p = 0.002). Regarding the CAV of the renal mass and renal cortex, only the CAVs of the renal mass in the corticomedullary phase and early excretory phase were significantly higher in Xp11 RCCs compared with papillary RCCs (p 0.003). When comparing the attenuation values of the renal mass and the renal cortex, Xp11 RCCs exhibited higher attenuation than the renal cortex on the unenhanced phase but showed lower attenuation values than the cortex on the corticomedullary and early excretory phases for both the measured and corrected attenuation values (p < 0.001). However, in cases of papillary RCC, although there was no significant difference in the unenhanced attenuation values between the renal mass and Renal Mass Renal Cortex Renal Mass Renal Cortex Renal Mass a Renal Cortex b Unenhanced phase 41.9 ± 7.7 (29 55) 32.1 ± 4.5 (21 39) 34.9 ± 11.4 (12 69) 31.4 ± 6.9 (12 48) < c c Corticomedullary phase ± 31.4 (57 174) ± 55.1 ( ) 62.3 ± 23.7 (26 133) ± 42.1 (85 264) < < c < c Early excretory phase 94.8 ± 15.9 (63 120) ± 33.9 ( ) 74.8 ± 21.3 (37 123) ± 37.5 (99 265) < c < c Note Values are mean ± SD with range in parentheses. a From unpaired Student t test to compare attenuation values of Xp11 RCC and papillary RCC. b From unpaired Student t test to compare attenuation values of renal cortex between patients with Xp11 RCC and papillary RCC. c From paired Student t test to compare attenuation values of renal mass and background renal cortex. TABLE 3: Corrected Attenuation Values of Renal Mass and Adjacent Renal Parenchyma in 58 Patients With Xp11 Renal Cell Carcinoma (RCC) and Papillary RCC Attenuation (HU) Xp11 RCC (n = 19) Papillary RCC (n = 39) p Renal Mass Renal Cortex Renal Mass Renal Cortex Renal Mass a Renal Cortex b Unenhanced phase 49.7 ± 11.3 (27 70) 37.6 ± 4.8 (31 47) 43.3 ± 14.8 (12 82) 38.5 ± 9.0 (26 57) < c c Corticomedullary phase 73.1 ± 20.1 (46 113) ± 24.3 ( ) 50.7 ± 26.9 (19 170) ± 48.0 (73 281) < c < c Early excretory phase 87.6 ± 17.7 (57 120) ± 19.8 ( ) 71.9 ± 17.7 (45 110) ± 21.3 ( ) < c < c Note Values are mean ± SD with range in parentheses. a From unpaired Student t test to compare corrected attenuation values of Xp11 RCC and papillary RCC. b From unpaired Student t test to compare corrected attenuation values of renal cortex between patients with Xp11 RCC and papillary RCC. c From unpaired Student t test to compare corrected attenuation values of renal mass and background renal cortex. cortex (p = and for the measured and corrected attenuation values, respectively), the corticomedullary phase and early excretory phase attenuation values of the renal mass were lower than that of the cortex (p < 0.001). On the basis of the attenuation values (measured and corrected) of the renal mass and of the cortex for each of the phases, both Xp11 RCC and papillary RCC were considered to show a gradual enhancement pattern. Subgroup Analysis of Tumors Less Than 4 cm There were seven and 26 patients with Xp11 RCC and papillary RCCs, respectively, that were smaller than 4 cm in size (Table 4). The distribution of solid and cystic small RCCs was different between the two entities (p < 0.001). Xp11 RCCs were solid, predominantly solid, and mixed solid and cystic in one (14.3%), three (42.9%), and three (42.9%) patients, respectively. Papillary RCCs were solid (88.5%, 23/26), predominantly solid (11.5%, 3/26), and mixed solid and cystic (3.8%, 1/26). When the tumors were dichotomized into those with and without radiologic evidence of cystic change or necrosis, small papillary RCCs were significantly (p < 0.001) more prevalent in the solid pattern (Fig. 3). Small Xp11 RCCs were significantly more common in calcification (p = 0.016) and were more prevalent in noncalcified high-attenuating areas with borderline significance (p = 0.068). Despite the small size, Xp11 RCCs frequently manifested with lymph node metastasis (57.1%, 4/7, p = 0.001), whereas none were seen in papillary RCCs. Xp11 RCC and papillary RCC were not statistically different for the following binary imaging features: epicenter (p = 0.398), margin (p = not applicable), venous invasion (p = not applicable), collecting duct invasion (p = not applicable), and distant metastasis (p = 0.212). Although the prevalence of distant metastasis was not significantly different, there was one case of distant metastasis to the lungs and bone from a 2-cm sized Xp11 RCC, whereas none of the papillary RCCs had distant metastasis (Fig. 4). AJR:204, March

5 Woo et al. Discussion In our study, we found that several imaging features on multiphase renal MDCT were useful for differentiating Xp11 RCC from papillary RCC. Xp11 RCCs were significantly greater in size, more commonly contained cystic or necrotic portions, contained calcification and noncalcified high-attenuating areas, and more frequently manifested with lymph node and distant metastasis. Furthermore, patients with Xp11 RCCs had a female predilection and were more prevalent in younger patients (< 20 years) when compared with papillary RCCs. Thus, on the basis of our study results, we suggest that multiphase renal MDCT could be helpful in the differentiation between Xp11 RCCs and papillary RCCs, taking into consideration the patient characteristics. Our study showed that Xp11 RCCs manifested more commonly as a heterogeneous mass with cystic and necrotic portions, high unenhanced attenuation, and calcification when compared with the more solid-appearing papillary RCC with less frequent calcification and unenhanced high attenuation. Our results are in agreement with a few previous reports that have suggested consistent imaging findings of Xp11 RCCs despite the small number of patients [7 11]: heterogeneous mass because of the presence of cystic and necrotic foci [8 10], hyperdense attenuation on unenhanced CT [7, 9, 11], and calcification [9 11]. In the case of papillary RCC, it is known that tumors of type 2 TABLE 4: Tumor Characteristics of 33 Patients With < 4 cm-sized Xp11 Renal Cell Carcinoma (RCC) and Papillary RCC Characteristic Xp11 RCC (n = 7) Papillary RCC (n = 26) p Exophytic epicenter 28.6 (2/7) 53.8 (14/26) Solid and cystic pattern < Solid 14.3 (1/7) 88.5 (23/26) (< 0.001) a Predominantly solid 42.9 (3/7) 7.7 (2/26) Mixed solid and cystic 42.9 (3/7) 3.8 (1/26) Predominantly cystic 0 (0/7) 0 (0/26) Calcification 71.4 (5/7) 19.2 (5/26) Unenhanced high attenuation 57.1 (4/7) 19.2 (5/26) Infiltrative margin 0 (0/7) 0 (0/26) NA Renal vein invasion 0 (0/7) 0 (0/26) NA Collecting duct invasion 0 (0/7) 0 (0/26) NA Lymph node metastasis 57.1 (4/7) 0 (0/26) Distant metastasis 14.3 (1/7) 0 (0/26) Note Data are percentages with raw numbers in parentheses. NA = not applicable. a Solid versus others compared using chi-square test. pathology and large size can be associated with areas of necrosis and hemorrhage [15, 16]. Furthermore, papillary RCCs could occasionally appear as cystic masses because of inherent architecture or from secondary cystic degeneration and extensive necrosis [17]. Otherwise, papillary RCCs are known to more commonly appear as a homogeneous and uniform solid tumor when compared with clear cell RCC [15], which is similar to most papillary RCCs seen in our study. Regarding the unenhanced high attenuation in Xp11 RCC, previous investigators have reported this finding to correlate with hemorrhagic or proteinaceous fluid or densely packed cellular components on histopathology [7, 9]. On the other hand, papillary RCCs have been shown to infrequently appear hyperdense on unenhanced phase imaging (8.7%) [13]. With regard to calcification, although it was more common in Xp11 RCCs, it should be noted that it was not a rare finding in papillary RCCs (23.1%). Among the subtypes of RCC, papillary RCC shows A Fig year-old woman with pathologically proven papillary renal cell carcinoma. B C A C, Axial unenhanced (A), corticomedullary phase (B), and early excretory phase (C) MDCT images show well-circumscribed mass (arrow) in left kidney. Mass is solid in nature without definite cystic change or areas of necrosis. No calcification was present. Generally, enhancement degree of mass was poor, with attenuation values measuring 26, 48, and 55 HU in contrast with 36, 174, and 162 HU in renal cortex on unenhanced phase, corticomedullary phase, and early excretory phase, respectively. No lymph nodes were significantly enlarged. There was no evidence of distant metastasis. 546 AJR:204, March 2015

6 MDCT in Renal Cell Carcinoma A C calcification more commonly than does clear cell RCC [18]. Therefore, although calcification has been noted as a common and characteristic finding of Xp11 RCC, its value in differentiating between Xp11 RCC and papillary RCC remains questionable. Both Xp11 RCC and papillary RCC were poorly enhanced compared with the background renal cortex on the basis of measured and corrected attenuation values in our study. This could be expected when considering previous reports in the literature. It is well known that papillary RCCs are hypovascular in nature, showing low degrees of enhancement on imaging studies [15, 18]. Regarding Xp11 RCC, the first report of CT and MRI findings by Kato et al. [7] described Xp11 RCC as hypovascular on both contrast-enhanced CT and MRI. Recent studies have observed that Xp11 RCCs showed a lower degree of enhancement than the normal renal cortex on all enhanced phases, which is in line with our results [19]. However, it should be noted that Xp11 RCCs were characteristically hyperdense to the renal cortex on unenhanced imaging, unlike papillary RCCs. Furthermore, Xp11 RCCs had higher attenuation values (both measured and corrected) than papillary RCCs on the corticomedullary phase and early excretory phase. Therefore, although both tumors seem to be hypovascular, quantitative measurement of attenuation values, provided that they are further validated in the future, may have a role to play in the differentiation between Xp11 RCC and papillary RCC. Overall, Xp11 RCCs showed more aggressive features compared with papillary RCCs. Considering the more heterogeneous nature of the mass in terms of greater necrotic and cystic portion, Xp11 RCCs were greater in mean tumor size and more frequently showed lymph node and distant metastasis. This is congruent with the well-known fact that Xp11 RCCs are clinically aggressive [10]. Because mean tumor size was also greater in Xp11 RCCs than papillary RCCs in our study population, we performed a subgroup analysis of small tumors (< 4 cm) to B Fig year-old man with pathologically proven small (< 4-cm sized) Xp11.2 translocation and TFE3 gene fusion. A C, Axial unenhanced (A), corticomedullary phase (B), and early excretory phase (C) MDCT images show 2.6-cm-sized well-circumscribed mass with necrotic portion (arrowheads, C) in right kidney. Peripheral rimlike calcification (arrow) is seen on unenhanced image (A). Solid components of mass are enhanced poorly compared with renal cortex. Attenuation values of mass were 38, 104, and 120 HU, and those of renal cortex were 33, 213, and 180 HU on unenhanced phase, corticomedullary phase, and early excretory phase, respectively. counterbalance the effect of size on more aggressive features. Even in this subgroup of small tumors, Xp11 RCCs showed a significantly higher proportion with radiologic evidence of necrosis and lymph node metastasis. In addition, although the presence of distant metastasis was not a significant discriminating feature between Xp11 RCC and papillary RCC, it was notable that there was one case of distant metastasis to the lungs and bone from a 2-cm-sized Xp11 RCC. A few limitations of our study should be mentioned. First, the retrospective nature, small sample size, and inclusion of cases with preoperative multiphase renal MDCT may have introduced selection bias. However, Xp11 RCC is a very rare subtype of RCC, and the number of cases of Xp11 RCC in our study population (n = 19) outnumbers those in the radiologic literature (n = 1 9) [7 11]. Furthermore, despite the small number of patients, we were able to identify imaging findings that were helpful in the differentiation between Xp11 RCC and papillary RCC. In addition, AJR:204, March

7 Woo et al. multiphase renal MDCT was essential to compare the attenuation and enhancement characteristics between the two entities. Second, we did not perform radiologic-pathologic correlation for each of the specific imaging features. However, this has already been performed in previous studies. For instance, the cystic and necrotic portion on CT has been shown to correlate with necrosis and hemorrhage on pathology, whereas noncalcified high-attenuating solid portions were shown to represent dense cellular components [9]. Therefore, the main focus of our study was to search for imaging features on multiphase renal MDCT that would help in the differential diagnosis of these tumors, which have overlapping histopathologic findings, rather than correlation between imaging findings and histopathologic specimens. Third, although the radiologists were blinded to the pathologic results, they were aware that the renal mass in question was either Xp11 RCC or papillary RCC. Therefore, the comparative results of MDCT features derived from our study may not be applied to comparison with other subtypes of RCC. However, taking into account the attenuation values and degree of enhancement of Xp11 RCCs in our study and the previous literature, we speculate that differentiating Xp11 RCC from other common types of RCC, such as clear cell or chromophobe types, would be less clinically relevant. Clear cell RCCs typically are known to be heterogeneous and hypervascular, with strong enhancement [18]. In addition, chromophobe RCCs tend to be rather homogeneously enhanced on CT and MRI even when manifesting as large renal masses [20]. In contrast, Xp11 RCCs seem poorly enhanced and heterogeneous, with portions of necrosis and cystic change. In conclusion, features of greater size, more cystic change or necrosis, presence of calcification and high-attenuating areas on unenhanced imaging, and lymph node and distant metastasis were helpful for differentiating Xp11 RCC from papillary RCC. Furthermore, although both were hypovascular on contrast-enhanced phases, Xp11 RCCs had significantly higher attenuation values. If properly validated, radiologists may be able to preoperatively suggest the possibility of Xp11 RCC and help guide pathologists to perform immunohistochemical staining for the final diagnosis of Xp11 RCCs. References 1. Lopez-Beltran A, Scarpelli M, Montironi R, Kirkali Z WHO classification of the renal A C D Fig year-old man with pathologically proven Xp11.2 translocation and TFE3 gene fusion with distant metastasis to bone and lung. A and B, Axial unenhanced (A) and early excretory phase (B) MDCT images show 2-cm-sized wellcircumscribed and predominantly solid mass in right kidney (arrow, A and B). On early excretory phase (B), areas of nonenhancing foci, suggestive of necrosis or cystic change (arrowhead), are suspected despite small size of mass. Solid component of mass (51, 82, and 101 HU on unenhanced phase, corticomedullary phase [not shown], and early excretory phase, respectively) enhanced less than that of renal cortex (39, 168, and 175 HU, respectively). C, Axial image in lung window setting reveals well-defined solid nodule (arrow) in right lower lobe. This nodule showed significant interval growth on follow-up imaging and was confirmed as metastasis (not shown). D, Bone window setting image shows well-defined sclerotic lesion (arrow) in T10 vertebral body, which was confirmed as metastasis on follow-up MRI and bone scans (not shown). tumors of the adults. Eur Urol 2006; 49: Armah HB, Parwani AV. Xp11.2 translocation renal cell carcinoma. Arch Pathol Lab Med 2010; 134: Argani P, Lal P, Hutchinson B, Lui MY, Reuter VE, Ladanyi M. Aberrant nuclear immunoreactivity for TFE3 in neoplasms with TFE3 gene fusions: a sensitive and specific immunohistochemical assay. Am J Surg Pathol 2003; 27: Steffens S, Janssen M, Roos FC, et al. Incidence and long-term prognosis of papillary compared to clear cell renal cell carcinoma: a multicentre study. Eur J Cancer 2012; 48: Moch H. An overview of renal cell cancer: pathology and genetics. Semin Cancer Biol 2013; 23: Ross H, Argani P. Xp11 translocation renal cell carcinoma. Pathology 2010; 42: Kato H, Kanematsu M, Yokoi S, et al. Renal cell carcinoma associated with Xp11.2 translocation/ TFE3 gene fusion: radiological findings mimicking papillary subtype. J Magn Reson Imaging 2011; 33: Dang TT, Ziv E, Weinstein S, Meng MV, Wang Z, Coakley FV. Computed tomography and magnetic resonance imaging of adult renal cell carcinoma associated with Xp11.2 translocation. J Comput Assist Tomogr 2012; 36: Koo HJ, Choi HJ, Kim MH, Cho KS. Radiologic- B 548 AJR:204, March 2015

8 MDCT in Renal Cell Carcinoma pathologic correlation of renal cell carcinoma associated with Xp11.2 translocation. Acta Radiol 2013; 54: Liu K, Xie P, Peng W, Zhou Z. Renal carcinomas associated with Xp11.2 translocations/tfe3 gene fusions: findings on MRI and computed tomography imaging. J Magn Reson Imaging 2014; 40: He J, Huan Y, Qiao Q, Zhang J, Zhang JS. Renal carcinomas associated with Xp11.2 translocations: are CT findings suggestive of the diagnosis? Clin Radiol 2014; 69: Jinzaki M, Tanimoto A, Mukai M, et al. Doublephase helical CT of small renal parenchymal neoplasms: correlation with pathologic findings and tumor angiogenesis. J Comput Assist Tomogr 2000; 24: FOR YOUR INFORMATION Mark your calendar for the following ARRS annual meetings: April 19 24, 2015 Toronto Convention Centre, Toronto, ON, Canada April 17 22, 2016 Los Angeles Convention Center, Los Angeles, CA April 30 May 5, 2017 Hyatt Regency New Orleans, New Orleans, LA April 22 27, 2018 Marriott Wardman Park Hotel, Washington DC 13. Zhang YY, Luo S, Liu Y, Xu RT. Angiomyolipoma with minimal fat: differentiation from papillary renal cell carcinoma by helical CT. Clin Radiol 2013; 68: Ruppert-Kohlmayr AJ, Uggowitzer M, Meissnitzer T, Ruppert G. Differentiation of renal clear cell carcinoma and renal papillary carcinoma using quantitative CT enhancement parameters. AJR 2004; 183: Herts BR, Coll DM, Novick AC, et al. Enhancement characteristics of papillary renal neoplasms revealed on triphasic helical CT of the kidneys. AJR 2002; 178: Yamada T, Endo M, Tsuboi M, et al. Differentiation of pathologic subtypes of papillary renal cell carcinoma on CT. AJR 2008; 191: Vikram R, Ng CS, Tamboli P, et al. Papillary renal cell carcinoma: radiologic-pathologic correlation and spectrum of disease. RadioGraphics 2009; 29: Kim JK, Kim TK, Ahn HJ, Kim CS, Kim KR, Cho KS. Differentiation of subtypes of renal cell carcinoma on helical CT scans. AJR 2002; 178: Zhu QQ, Wang ZQ, Zhu WR, Chen WX, Wu JT. The multislice CT findings of renal carcinoma associated with XP11.2 translocation/tfe gene fusion and collecting duct carcinoma. Acta Radiol 2013; 54: Prasad SR, Humphrey PA, Catena JR, et al. Common and uncommon histologic subtypes of renal cell carcinoma: imaging spectrum with pathologic correlation. RadioGraphics 2006; 26: AJR:204, March