Comparative study of CT appearances in renal oncocytoma and chromophobe renal cell carcinoma

Abstract

Background

Renal oncocytoma (RO) and chromophobe renal cell carcinoma (ChRCC) share histologic and some imaging features.

Purpose

To investigate the multidetector computed tomography (MDCT) characteristics of these two tumor types.

Material and Methods

Fifty-six patients with RO and 54 patients with ChRCC were studied retrospectively. MDCT was undertaken to investigate differences in tumor characteristics.

Results

Calcifications were visible in 24 (42.8%) patients with RO and in 11 (20.4%) patients with ChRCC (P = 0.011). 26 patients with RO had stellate scars as did 14 patients with ChRCC (P = 0.025). Spoken-wheel-like enhancement was visible in 41 patients with RO and in 11 with ChRCC (P < 0.001). Thirty-nine (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion (P < 0.001). Two patients with RO had retroperitoneal lymph node enlargement as did 13 patients with ChRCC (P = 0.002). Combined evaluation of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion features were found to have a sensitivity of 99.1% (106 of 107), a speciﬁcity of 100% (3 of 3), a positive predictive value of 100% (106 of 106), and a negative predictive value of 75% (3 of 4). The attenuation of RO tumors was greater than that of ChRCC tumors, normal renal parenchyma on unenhanced CT (P = 0.031). Enhancement was higher with RO than with ChRCC tumors in all phases (P = 0.021, < 0.001, and 0.007, respectively).

Conclusion

CT imaging features such as stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion were more common in RO and they may help in differentiating RO from ChRCC.

Keywords

Renal oncocytoma chromophobe renal cell carcinoma computed tomography (CT)

Introduction

Chromophobe renal cell carcinoma (ChRCC) is postulated to arise from the intercalated cells of the renal cortex. It is interesting that ChRCC and renal oncocytoma (RO) have much in common regarding morphologic, histologic, immunohistochemical, and ultrastructural aspects (1,2). ROs develop from type B intercalated cells of the cortical collecting duct and are indistinguishable from ChRCCs on imaging studies, sharing features such as central scar, and spoke wheel pattern of enhancement (3). Despite these common features, RO and ChRCC exhibit differences in prognosis (4). RO is considered benign in that it does not usually show renal vein invasion, postoperative recurrence, or distant metastases. ChRCC is an RCC subtype with overall favorable prognosis compared with other RCCs, including slower growth and significantly lower rates of metastases, progression, and death (5,6). ChRCC is generally managed via partial nephrectomy when technically possible, and routine radical nephrectomy for this entity would be gross mismanagement, with needlessly exposing the patient to greater risk of reduced renal function and associated higher rate of cardiovascular events from the more extensive surgery (7). The purpose of the present study was to retrospectively characterize multidetector computed tomography (MDCT) characteristics which lead to the histological diagnosis of RO and ChRCC.

Material and Methods

Patients

This study was approved by the Institutional Research Ethics Committee. A search of pathology records and PACS system identified 56 patients with RO and 54 patients with ChRCC who were hospitalized at our hospital between 2001 and 2013. Details of the patient’s gender, surgery or biopsy confirmation, and clinical symptoms were recorded.

CT technique

CT examinations were performed by using a 16 detector row scanner (CT Lightspeed Ultra, GE Healthcare, Milwaukee, WI, USA; [n = 27]), or a 64 detector row scanner CT (Somatom Definition, Siemens Medical Solutions, Forchheim, Germany; [n = 83]). Intravenous contrast material (iopromide, Ultravist 370; Bayer Schering Pharma AG, Berlin, Germany) was administered at a dose of 2 mL per kg of body weight and at a rate of 3 mL/s to a maximum of 150 mL by using a power injector, or at a slower rate (minimum of 2 mL/s) if required when venous access was suboptimal. All patients underwent unenhanced, arterial phase (cortical phase), cortico-medullary phase (medullary phase), and excretory phase (delayed phase). Unenhanced imaging was performed before administering the intravenous contrast agent. The contrast-enhanced CT scans began after 20 s for the arterial phase (cortical phase), 65 s for the cortico-medullary phase (medullary phase), and 300 s for the excretory phase (delayed phase). The scanning parameters varied for the different scanners but were for all phases in the following ranges: tube voltage, 120 kVp (constant for all); tube current, 150–250 mA (depending on patient size); pitch, 0.891–1.6; section thickness, 1.25–5.0 mm; reconstruction interval, 2.5–5 mm, and rotation time, 0.5–0.75 s.

Imaging analysis

Two radiologists with more than 10 years of experience each, blinded to the final diagnosis, reviewed the CT images in consensus at a picture archiving and communication system (PACS) workstation. The parameters evaluated included tumor calcification and attenuation on unenhanced CT scans. The degree of enhancement on the different phases of the enhanced CT scans were thereafter assessed.

The attenuation of the tumor, normal renal cortex, and medulla were measured during all enhanced phases. The measured tumor area (the region of interest [ROI]) was at the center of the mass to avoid partial volume effects, however, intratumoral calcification and cystic components were excluded from the ROI if central necrosis or calcifications were considered. Normal renal cortex and medulla were measured using uninvolved homolateral renal cortex and medulla. Each 10 mm ROI was measured three times for each phase and the mean value was used.

The enhancement pattern of the tumor was classiﬁed as homogeneous or heterogeneous. The degree of enhancement was based on dynamic CT imaging using Hounsfield unit (HU) attenuation of the tumor, renal cortex, and medulla.

Statistical analysis

Statistical analysis was undertaken using SPSS version 17.0 statistical software (SPSS Inc, Chicago, IL, USA). Numeric data were expressed as means and standard deviations (±SD) and categorical data were expressed as percentages. Evaluated characteristics were compared between the two tumor types using chi-square test, analysis of variance (ANOVA); post-hoc test (Tukey) was performed after ANOVA. Values of P < 0.05 were considered statistically significant.

To assess the diagnostic performance of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion in terms of differentiating ROs from ChRCCs, the sensitivity, speciﬁcity, and positive and negative predictive value of these signs were calculated.

Results

The study included 56 patients (26 women, 30 men) with RO and 54 patients (28 men, 26 women) with ChRCC. The presenting symptoms of RO and ChRCC included flank pain (25 and 28 patients), hematuria (34 and 26 patients), palpable mass (12 and 13 patients), and fever (22 and 34 patients) in each group, respectively.

There was evidence of calcifications in 24 (42.8%) of the RO tumors whereas 11 (20.4%) patients with ChRCC had evidence of calcifications (P = 0.011, Table 1). In 26 (46.4%) patients with RO (Figs. 1 and 2) and in 14 (25.9%) patients with ChRCC the tumors exhibited stellate scars (P = 0.025). There was evidence of spoken-wheel-like enhancement in 41 (73.2%) of the RO tumors compared to 11 (20.4%) patients with ChRCC (P < 0.001). Thirty-nine (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion during the cortico-medullary phase and excretory phase, which signiﬁcantly differentiated ROs from ChRCCs (P < 0.001, Table 1). For differentiating RO from ChRCC, combined evaluation of stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion features was found to have a sensitivity of 99.1% (106 of 107), a speciﬁcity of 100% (3 of 3), a positive predictive value of 100% (106 of 106), and a negative predictive value of 75% (3 of 4). Thirty-six (64.3%) of the 56 RO tumors showed homogeneous enhancement whereas 21 (38.9%) of the 54 ChRCC tumors showed homogeneous enhancement (P = 0.008). Retroperitoneal lymph node enlargement were noted in only two patients with RO whereas 13 patients with ChRCC had lymph nodes metastasis (P = 0.002).

Table 1.

Comparative study of CT appearances in RO and ChRCC.

Main CT findings	RO (n = 56)	ChRCC (n = 54)	P value
Calcification	24 (42.8%)	11 (20.4%)	0.011
Stellate scar	26 (46.4%)	14 (25.9%)	0.025
Homogeneous enhancement	36 (64.3%)	21 (38.9%)	0.008
Spoken-wheel-like enhancement	41 (73.2%)	11 (20.4%)	<0.001
Segmental enhancement inversion	39 (69.6%)	9 (16.7%)	<0.001
Lymph node enlargement/metastasis	2 (3.6%)	13 (24.1%)	0.002

Fig. 1.

RO of left kidney in a 39-year-old-woman. (a) Unenhanced CT scan showed a mild hyperdense mass together with cystic component (arrow). The attenuation of the mass was 47 HU. (b) Contrast-enhanced CT scan showed spoken-wheel-like enhancement and stellate scar (black arrow) on the cortical phase. (c, d) Contrast-enhanced CT scan showed segmental enhancement inversion on corticomedullary and delayed phases (white and black arrows).

Fig. 2.

RO of left kidney in a 51-year-old-man. (a) Unenhanced CT scan showed a mild hyperdense mass (arrow). The attenuation of the mass was 47 HU. (b) Contrast-enhanced CT scan showed heterogeneous enhancement on the cortical phase. (c, d) Contrast-enhanced CT scan showed segmental enhancement inversion (white arrows) and stellate scar (black arrow) on corticomedullary and delayed phases.

The attenuation of RO tumors was greater than that of ChRCC tumors (Figs. 3 and 4), normal renal cortex or medulla on unenhanced CT (46.8 ± 5.3 vs. 41.2 ± 2.4, 37.3 ± 4.7 or 31.9 ± 4.3, respectively; P = 0.031, Table 2). Enhancement was higher in RO’s than in ChRCC tumors during all phases (P = 0.021, < 0.001, 0.007, respectively; Fig. 5).

Table 2.

The CT values of RO, ChRCC, normal renal cortex, and medulla.

Phase	Renal cortex (n = 110)	Renal medulla (n = 110)	RO (n = 56)	ChRCC (n = 54)	P value
Unenhanced	37.3 ± 4.7	31.9 ± 4.3	46.8 ± 5.3	41.2 ± 2.4	0.031
Cortical	191.6 ± 9.1	71.8 ± 5.8	88.0 ± 8.7	63.2 ± 5.2	0.016
Medullary	203.3 ± 12.1	108.7 ± 8.3	116.8 ± 9.2	89.8 ± 11.1	0.001
Delayed	116.4 ± 11.8	103.6 ± 7.8	76.7 ± 7.3	64.8 ± 7.2	0.021

On unenhanced CT, the attenuation of RO was higher than that of ChRCC, normal renal cortex, and medulla (P = 0.056, 0.047, and 0.036, respectively). Enhancement was higher with RO than with ChRCC tumors during all phases (P = 0.021, <0.001, and 0.007, respectively).

Fig. 3.

ChRCC of right kidney in a 72-year-old-man. (a) Unenhanced CT scan showed an iso-attenuation mass with a clearly defined margin (arrow). The attenuation of the mass was 42 HU. (b) Contrast-enhanced CT scan showed mild enhancement less than the cortex on the cortical phase (59 HU, arrow). (c) The attenuation of the tumor (72 HU) increased (arrow) during the medullary phase and (d) decreased (62 HU) during the delayed phase.

Fig. 4.

ChRCC of right kidney in a 57-year-old-man. (a) Mild, heterogeneous enhancement was noted on the cortical phase. (b, c) Contrast-enhanced CT scan showed segmental enhancement inversion on corticomedullary and delayed phases (black and white arrows).

Fig. 5.

The CT line graph of RO, ChRCC, normal renal cortex, and medulla on unenhanced and enhanced phase.

Follow-up

Fifty-three patients with RO were followed up. Two patients died within 5 years due to complications unrelated to the tumors while the others were still alive.

Fifty-three patients with ChRCC were followed up. Fourteen patients died within 5 years after the initial diagnosis and with the others still alive.

Discussion

Renal oncocytoma (RO) and chromophobe renal renal cell carcinoma (ChRCC) are both uncommon subtypes of renal tumors. RO of the kidney is an uncommon benign epithelial tumor that can present at any age (8). RO constitutes less than 5% of all renal lesions and is postulated to arise from the type B intercalated cells of the cortical collecting duct. ChRCC constitutes less than 4.5% of all renal carcinomas and is postulated to arise from the intercalated cells of the renal cortex (9).

It is interesting that RO and ChRCC share ontogenetic and histologic features and some imaging findings (10). Because renal oncocytoma and ChRCC share a common histologic background, originating from the intercalated cells of the collecting duct, these two entities inevitably are similar regarding morphologic, histologic, immunohistochemical, and ultrastructural background (10). As a result, despite the low proportion of ChRCC among the subtypes of RCC, ChRCC has been known to cause confusion with RO in the preoperative diagnosis (11).

Our results show that RO are a heterogeneous hyperdense solid tumor together with calcifications or stellate scars on unenhanced CT. However, ChRCC is commonly an isodense or mild hyperdense mass together with cystic components. Other authors have reported that the pathologic basis of hyperdensity of a tumor on unenhanced CT is the presence of minimal intratumoral hemorrhage (hemosiderin deposition) or fibrous septa formation (12,13). Gakis (8) reported that calcifications in the tumors often suggest a benign tumor. In our study, there was evidence of calcifications in 24 of the RO tumors whereas 11 patients with ChRCC had evidence of calcifications (P = 0.011).

The differential diagnosis of a hyperdense renal solid mass can include clear cell renal cell carcinomas (14) and angiomyolipomas with minimal fat. Clear cell renal cell carcinomas larger than 3 cm usually show evidence of intratumoral necrosis, hemorrhage, cystic components, and calcifications. Papillary renal cell carcinoma is commonly homogeneous when poorly differentiated, but has often necrotic and cystic changes, hence appearing to be heterogeneous. Well-differentiated papillary renal cell carcinoma is usually regular in shape and has homogeneous density, dissimilar to RO.

In our study, the degree of enhancement of RO and ChRCC was lower than that of the normal renal cortex during all phases. Enhancement was higher with RO than with ChRCC tumors in all phases. RO enhanced less than cortical structures but more than medullary structures in all phases except for the delayed phase, where it enhanced less than medullary structures. ChRCC tumor enhancement was lower than that of the normal renal medulla in all phases. This enhancement pattern is atypical of hypervascular tumors such as clear cell renal cell carcinoma, renal medullary carcinoma (15), renal angiomyolipoma, and renal angioma (16). These findings suggest that it may be relatively easy to distinguish between ChRCC and tumors with hypervascularity on the basis of differences in enhancement.

In common with RO and ChRCC, papillary renal cell carcinoma, collecting duct carcinoma and renal lymphoma (17) are hypovascular compared to renal parenchyma on enhanced CT imaging. The majority of papillary renal cell carcinomas show slight enhancement during all enhancement phases. However, a previous study reported that the imaging findings with papillary renal cell carcinoma were similar to ChRCC (17).

Our results showed that the CT imaging features of RO and ChRCC are somewhat different, such as spoke-wheel like enhancement, segmental enhancement inversion, and stellate scar within masses during enhanced phases. Histopathological examinations conﬁrm that spoke-wheel like enhancement are attributable to thick vessels running along the ﬁbrous bands strongly stained by contrast material (18). Originally, the spoke-wheel arterial pattern on angiograms has been recognized as a characteristic ﬁnding of RO. However, it is unclear whether similar features can be demonstrated on enhanced CT scans. Additionally, no previous studies have described a spoke-wheel-like enhancement during dynamic CT scans in clear cell renal cell carcinomas (18). In our study there was evidence of spoken-wheel-like enhancement in 41 (73.2%) of the RO tumors whereas 11 (20.4%) of the patients with ChRCC had evidence of spoken-wheel-like enhancement, which signiﬁcantly differentiate between ROs and ChRCCs (P < 0.001).

Woo (19) has reported that segmental enhancement inversion may be helpful in identifying RO. We defined segmental enhancement inversion as the contrast-enhanced biphasic renal CT finding of two distinct regions of enhancement of a renal mass in the corticomedullary phase in which the degree of enhancement reverses in the excretory (delayed) phase. The more enhanced segment correlated with compactly arranged tumor cells with scarcely intervening stroma, and the less enhanced segments correlated with abundant hyalinized stroma. In our study, 39 (69.6%) patients with RO and nine (16.7%) patients with ChRCC showed segmental inversion during the cortico-medullary and excretory phases, which signiﬁcantly differentiate ROs from ChRCCs (P < 0.001). We infer that differences in stroma content are probably responsible for the observed segmental enhancement differences.

A central, sharply deﬁned stellate scar is one of the characteristic CT findings of RO and present in 25–33% of large RO cases. However, Davidson (20) reported that the central scar found on CT scan is not speciﬁc of RO and can also be found in some renal cell carcinomas, but it is unclear whether ChRCC were included in this analysis. In our study, 26 (46.4%) patients with RO and 14 (25.9%) patients with ChRCC exhibited stellate scars (P = 0.025).

Retroperitoneal lymph node enlargement was found in only two patients with RO compared to 13 patients with ChRCC. Clear cell renal cell carcinoma is commonly associated with lymph node metastasis and renal vein infiltration (21,22). Hence, the differential biological behavior of the tumors may also provide useful diagnostic information.

Information regarding the clinical behavior of these two tumors is limited due to its rare incidence. Although the clinical course of these patients is rather indolent, routine follow-up is still mandatory. The correct distinction of these two tumors can lead to better understanding of their clinicopathologic differences, which should aid in developing individualized treatment plans.

Our study has several limitations. First, pathologic specimens were sectioned in a variety of planes, but only axial, sagittal, and coronal planes were available on CT images. Therefore, it was difﬁcult to meticulously correlate the imaging ﬁndings with the histopathologic ﬁgures. Second, few patients were included in the study and finally, the retrospective nature of this study might have introduced some form of patient selection bias.

In conclusion, CT imaging features such as stellate scar, spoken-wheel-like enhancement, and segmental enhancement inversion were more common in RO than in ChRCCs and may help in differentiating RO from ChRCC.

Footnotes

Conflict of interest

None declared.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Skolarus

Serrano

Berger

. The distribution of histological subtypes of renal tumors by decade of life using the 2004 WHO classification. J Urol 2008; 179: 439–443.

Lopez-Beltran

Scarpelli

Montironi

. 2004 WHO classification of the renal tumors of the adults. Eur Urol 2006; 49: 798–805.

Kim

Cho

Moon

. Segmental enhancement inversion at biphasic multidetector CT: characteristic finding of small renal oncocytoma. Radiology 2009; 252: 441–448.

Siegel

. Differentiation of oncocytoma and renal cell carcinoma in small renal masses (<4 cm): the role of 4-phase computerized tomography. J Urol 2012; 188: 1722–1723.

Mues

Landman

. Small renal masses: current concepts regarding the natural history and reflections on the American Urological Association guidelines. Curr Opin Urol 2010; 20: 105–110.

Rosenkrantz

Hindman

Fitzgerald

. MRI features of renal oncocytoma and chromophobe renal cell carcinoma. Am J Roentgenol 2010; 195: W421–427.

Pedrosa

Sun

Spencer

. MR imaging of renal masses: correlation with fndings at surgery and pathologic analysis. Radiographics 2008; 28: 985–1003.

Gakis

Kramer

Schilling

. Small renal oncocytomas: Differentiation with multiphase CT. Eur J Radiol 2011; 80: 274–278.

Barthelemy

Feng

. S100A1: a powerful marker to differentiate chromophobe cell renal carcinoma from renal oncocytoma. Histopathology 2007; 50: 642–647.

10.

Kim

Ahm

. Differentiation of subtypes of chromophobe cell carcinoma. Hinyokika Kiyo 2004; 50: 865–868.

11.

Rohan

Kao

. Gene expression profiling separates chromophobe cell renal carcinoma from oncocytoma and identifies vesicular transport and cell junction protein as differentially expressed genes. Clin Cancer Res 2001; 12: 6937–6945.

12.

Silverman

Mortele

Tuncali

. Hyperattenuating renal masses: etiologies, pathogenesis, and imaging evaluation. Radiographics 2007; 27: 1131–1143.

13.

Kato

Kanematsu

Yokoi

. Renal cell carcinoma associated with Xp11.2 translocation/TFE3 gene fusion: radiological findings mimicking papillary subtype. J Magn Reson Imaging 2011; 33: 217–220.

14.

Sauk

Hsu

Margolis

. Clear cell renal cell carcinoma: Multiphasic multidetector CT imaging features help predict genetic karyotypes. Radiology 2011; 261: 854–862.

15.

Blitman

Berkenblit

Rozenblit

. Renal medullary carcinoma: CT and MRI features. Am J Roentgenol 2005; 185: 268–272.

16.

Abercrombie

Holmes

Ball

. Diagnosis and management of renal angioma. J R Soc Med 1992; 85: 625–627.

17.

El-Sharkawy

Siddiqui

Aleem

. Renal involvement in lymphoma: prevalence and various patterns of involvement on abdominal CT. Int Urol Nephrol 2007; 39: 929–933.

18.

Raman

Johnson

Allaf

. Chromophobe renal cell carcinoma: multiphase MDCT enhancement patterns and morphologic features. Am J Roentgenol 2013; 201: 1268–1276.

19.

Woo

Cho

Kim

. Segmental enhancement inversion of small renal oncocytoma: Differences in prevalence according to tumor size. Am J Roentgenol 2013; 200: 1054–1059.

20.

Davidson

Hayes

Hartman

. Renal oncocytoma and carcinoma: failure of differentiation with CT. Radiology 1993; 186: 693–696.

21.

Yoon

Nam

Rha

. Collecting duct carcinoma of the kidney: CT and pathologic correlation. Eur J Radiol 2006; 57: 453–460.

22.

Tsuda

Kinouchi

Tanikawa

. Imaging characteristics of papillary renal cell carcinoma by computed tomography scan and magnetic resonance imaging. Int J Urol 2005; 12: 795–800.