Yamashita, MD #{149}Mutsumasa Takahashi, MD #{149}Okimitsu Watanabe, Syunji Yoshimatsu, MD #{149}Sukeyoshi Ueno, MD #{149}Seiji Ishimaru, MD Masayasu Kan, MD #{149}Shinichi Takano, MD #{149}Nobumasa Ninomiya, MD

Yasuyuki

Small Renal and Radiolo

Cell

The authors correlated the radiologic and pathologic findings of 36 patients with small renal cell carcinoma (3 cm in diameter). Tumors were discovered incidentally or by means of mass surveys with ultrasound (US). Computed tomography (CT) and angiography were performed before surgery. Of the 36 tumors, 24 were of

solid (alveolar) architecture, five were of papillary architecture, three were of tubular architecture, and four were of multilocular cystic architecture. Cell arrangement was closely correlated with radiologic appearance, especially in regard to tumor vascularity and echogenicity. Histologically homogeneous tumors of solid architecture were hypoechoic at US and hypervascular at angiography and contrast medium-enhanced CT. Tumors of papillary, tubular, and multilocular cystic architecture were hyperechoic at US and hypovascular at angiography. There was no correlation of cell differentiation and cell architecture with echogenicity or tumor vascularity. Tumors with hemorrhage showed marked hyperattenuation at CT. A tumor capsule was observed in 25 patients (69%); this was demonstrated as a rim at US or angiography. Index

terms:

Cancer

screening,

81.32

Kidney neoplasms, angiography, 81.124 #{149} Kidney neoplasms, CT, 81.1211 #{149}Kidney neoplasms, diagnosis, 81.32 . Kidney neoplasms, US, 81.1298 Radiology

I

1992;

From

MT.,

the

OW.,

mamoto

184:493-498

Departments

S.Y., S.U.)

University

Honjo

Kumamoto

ments

of Radiology

Pathology

(N.N.),

tat, Kumamoto,

#{149}

of Radiology

and

School 860, Japan;

(M.K.), Kumamoto

Japan.

(Y.Y.,

Pathology

(SI.),

of Medicine, and

the

Urology Red

From the

Ku-

1-1-1 Depart-

(ST.), Cross

1991

and Hospi-

RSNA

sri-

entific assembly. Received August 8, 1991; revision requested October 8; fInal revision received March 13, 1992; accepted March 28. Address reprint requests to Y.Y. 0 RSNA, 1992

Carcinoma: Correlation’

R

MD

Pathologic

cell carcinomas (RCCs) are usually large, averaging 7-8 cm in diameter at diagnosis (1). With the

CT/T 8800, CT 9800 (GE Medical Systems, Milwaukee), or Contex (Yokogawa) units with a section thickness of 5-10 mm. CT

recent development of improved imaging techniques, RCCs smaller than 3 cm in diameter have been discovered frequently (2-6). Recently, mass surveys with ultrasound (US) have been performed of healthy people in Japan, and small RCCs have been discovered. It is reported that RCC is one

scans

ENAL

of the most

common

abdominal

ma-

lignancies and is found in 0.06% of the population (7). The various radiologic characteristics of RCC are well documented, but studies of small

RCC have related findings dentally

been

radioiogic of small or with

MATERIALS

limited

(2,3).

We cor-

and pathologic RCC discovered US survey.

AND

mci-

METHODS

We performed mass surveys of 283,667 patients from January 1986 to December 1990. RCCs were found at surgery in 70 of these patients (0.025%). Twenty-seven of the 70 RCCs (38%) were smaller than 3 cm in diameter. In addition, we discovered nine small RCCs incidentally, either with US

or computed

tomography

(CT),

during

the same period. We correlated the radiologic and pathologic findings in these 36 patients with small RCC. Of the 36 patients, 27 were men and nine were women. Ages ranged from 33 to 81 years (average, 54.3 years). All patients were asymptomatic, and no patient had a known history of chronic renal failure or von Hippel-Lindau disease. All patients had unilateral neoplasms; 22 were in the right kidney and 14 were in the left. The diameters of the tumors, obtained from gross pathologic examination, ranged from 15 to 30 mm (average, 22.7 mm). Metastasis was not seen in any of the patients at initial examination or follow-up. Thirtyfour patients underwent total nephrectomy, and two underwent partial nephrectomy. US

was

performed

with

either

transducer with RT-2800, (Yokogawa, Tokyo), or SSD-650CL Tokyo) units. CT was performed 5-MHz

a 3.5-

were

obtained

before

and

after

con-

trast enhancement with 50-100 mL (usually 90 mL) of iopamidol (Japan Schering, Osaka, Japan) or iohexol (Daichi, Tokyo). Both contrast media contained 300 mg of iodine per milliliter. Aortography and selective renal angiography were performed with a conventional technique. All images were reviewed by three radiologists. A whole section of the tumor specimens was obtained, and hematoxylin-eosin staining was performed. Histologic examination was performed independently by two pathologists, who agreed on the diagnosis of RCC in all cases. Only one case of HUrthle cell tumor was observed during our study, but this was not included in our study group. Tumors were classified with regard to (a) cell characteristics (clear vs granular cytoplasm), (b) cell arrangement or tumor architecture (solid [alveolar], papillary, tubular, and cystic), (c) nuclear grade (well, moderately, and poorly differentiated), and (d) interface between the tumor and surrounding kidney (well circumscribed forming a capsule, well circumscribed without a capsule, and infiltrative) (8-10). If various subtypes were mixed in a tumor, the most predominant findings were used to classify the tumor.

RESULTS Pathologic

Findings

Clear cells were dominant in 26 of the 36 patients (72%), granular cells were dominant in seven (19%), and both cell types were equally dominant in three (8%). Twenty-four of the tumors were solid (66%) (Fig 1), five were papillary (14%) (Fig 2), three were tubular (8%) (Fig 3), and four were cystic (11%) (Fig 4). All cystic tumors were multilocuiated. Thirtytwo tumors (89%) were solid, and tumors with granular cells showed

or

RT-3000

(Aloka, with

Abbreviation:

RCC

=

renal

cell carcinoma.

493

papillary

and

Multilocular composed cells. Degeneration prominent mors, one multilocular atypia was cases. Nuclei in 24 tumors

tubular

cystic of either

architecture. tumors clear

were or granular

or hemorrhage was in 13 cases (11 solid tupapillary tumor, and one cystic tumor). Nuclear seen but was mild in most were well differentiated and moderately differen-

tiated in 12; none were poorly differentiated. An interface between the tumor and normal renal parenchyma was clearly seen in 32 patients (89%), and a pseudocapsule, caused by cornpression of normal renal parenchyma and fibrous tissue, was noted in 25 (69%) (Figs 3c, 4c).

US

b.

C.

d.

Findings

Overall, US demonstrated that 14 lesions (39%) were hypoor isoechoic and 22 (61%) were hyperechoic, when compared with normal renal parenchyma. Most histologically hornogeneous tumors with tubular, papillary, or multilocular cystic architecture were hyperechoic (Figs 2a, 4a), whereas solid tumors tended to be either hypoechoic or isoechoic (Fig la) (Table 1). Tumors with inhomogeneous architecture because of cysts, hemorrhage, or degeneration appeared either hyperechoic or hypoechoic. There was no correlation between echogenicity and cell characteristics or cell differentiation. A hypoechoic rim was observed in 21 tumors (58%). At histologic examination, a pseudocapsule was observed in these tumors (Figs 2, 4).

CT Findings Most tumors were seen as a hypoor isoattenuating mass at unenhanced CT; this was especially seen in tumors of solid architecture (mostly clear cell type). Tumors with papillary or tubular architecture (all granular cell type) were of iso- to slightly hyperattenuation (Figs 2b, 3a; Table 2). Tumors with hemorrhage showed marked hyperattenuation (Fig 4b). Most of the tumors were homogeneous before and after administration of contrast material, but intratumoral areas of low attenuation were observed in 10

(28%), consistent with cavity or necrosis. Contrast enhancement mors was significantly

a large

cystic

in solid tuhigher than in

papillary, tubular, and multilocular cystic tumors (P < .01). After contrast enhancement, the interface between

494

a.

Radiology

#{149}

Figure 1. scan shows homogeneous pole of the

Small RCC (25 x 25 mm) with solid architecture in a 57-year-old woman. (a) US slightly hypoechoic mass in the left kidney (*) with clear margins and relatively internal architecture. (b) Angiogram shows hypervascular mass in the upper right kidney. Arterial displacement and a translucent rim are seen (arrowheads). (c) Macroscopic section shows welt-demarcated tumor with thin capsule. The internal architecture of the tumor is relatively homogeneous although small cysts (*) and hemorrhage are seen.(Hematoxylin-eosin stain; original magnification, x 1.) (d) Microscopic section shows solid arrangement of clear cells. Large nests of tumor cells are separated by a stroma that is characteristically endowed with prominent sinusoidlike vessels. (Hematoxylin-eosin stain; original magnification, x 100.)

the tumor and normal renal parenchyma was clear in most cases. No patient demonstrated regional lymph node enlargement. Calcification was observed in only one patient, who had

a multilocular

Angiographic

cystic

tumor

(3%).

DISCUSSION

Findings

Angiography demonstrated hypervascular tumors in 19 of the 36 patients (53%) and hypovascular or avascular tumors in 17 (47%). Tumor vascularity was closely correlated with tumor architecture. All hypervascular tumors were solid (Fig 1). Hypovascularity was observed papillary, tubular, or multilocuiar tic tumors and in solid tumors degeneration or fibrosis (Table There was no correlation between tumor vascularity and cell type atypia.

Arteriovenous

not observed in any of the patients. A translucent rim and displacement of the arteries around the tumor were observed in nine of 19 hypervascular tumors with a capsule (Fig ib).

shunting

Prevalence RCC

and

Detection

of Small

or cell

RCC is the most common malignancy of the kidney, accounting for approximately 2% of adult malignancies (11). With the recent development of new imaging modalities, the number of incidentally found RCCs is increasing (5,6). In our study, all tumors were clearly demonstrated with both US and CT, although the literature indicates CT to be superior for the detection of small RCCs (3,4,6,

was

12-14).

in cyswith 3).

August

1992

b.

C.

1 Figure papillary

2.

Small RCC architecture

Angiogram

(not

(20 x 20 mm) in a 42-year-old

shown)

with man.

did not show

any

US scan shows hyperechoic mass in the right kidney with irregular internal architecture (*) surrounded by anechoic halo (arrowheads). (b) Unenhanced CT scan (section thickness, 10 mm) shows slightly hyperattenuating mass (arrowheads) compared with surrounding kidney. (c) Contrast material-enhanced tumor

vessels

or

tumor

stain.

(a)

CT scan (section thickness, 10 mm) shows well-demarcated, hypoattenuating mass (arrowheads). The increase in CT attenuation number

e.

Pathologic

Characteristics

Pathologically, most small RCCs in our study were well encapsulated and well differentiated. Although it is believed that small RCC has a long preinvasive period, metastasis from small RCC has been described by many authors (2,3,5,15,16). It is very important to detect RCC at an early stage. In some small RCCs, pathologic differentiation from adenoma can be difficult. However, in all of our cases, pathologic examination demonstrated unequivocal evidence of renal malignancy. Solid (alveolar) architecture is the most common architecture, especially in large tumors (8,17). In our series of small RCC, tumors of nonsolid architecture (papillary, tubular, or multilocular cystic architecture) were relatively frequent. To our knowledge, the prognostic importance of cell arrangement has not been well clarified (9); however, tumors of papillary architecture are usually associated with a better prognosis (18-20). Large RCCs are usually composed of different subtypes and exhibit various degrees of degeneration, and, thus, may have a variety of patterns Volume

184

#{149} Number

2

at radiologic examination. Conversely, small RCCs are usually homogeneous in architecture and display degeneration less frequently. Therefore, we could precisely compare radiologic and histologic findings.

US

Characteristics

US plays an important role in the detection of small renal lesions, but has been invaluable in the characterization of these lesions. The sonographic findings of large RCC range from hypoechoic to hyperechoic (21,22).

To

our

knowledge,

the

before

and

after

contrast

material

administration was 9 HU. (d) Macroscopic section shows well-demarcated tumor with thin capsule (arrowheads) and homogeneous internal architecture of the tumor with high cellularity. (Hemotoxtyin-eosin stain; original magnification, x 1.) (e) Microscopic section shows papillary arrangement of granular cells. Thin, sparsely vascularized connective tissue stalks are lined with neoplastic cells. (Hematoxylin-eosin stain; original magnification, X 100.)

it

mech-

anism of echogenicity of the tumors has not been clarified. In previous articles about papillary carcinoma (19,23), no consistent sonographic pattern was observed; hypoechoic masses were seen more frequently than hyperechoic masses, which is contrary to our results. This is probably because of the high frequency of cystic necrosis and hemorrhage in large papillary carcinomas (18,19,23). Conversely, in small RCCs with histologically homogeneous architecture, US can correctly depict the

internal architecture of the tumor because potentially confusing degenerative changes, which are occasionally seen in large RCCs, are much less of a problem with small RCCs (24). In correlating US and histologic findings, the cell arrangement was found to be the most important factor for echogenicity in small RCCs. The hyperechogenicity of papillary, tubular, or multilocular cystic tumors may be due to the multiple interfaces formed by small spaces between tumor cells, whereas solid tumors have a compact arrangement, which will appear hypoechoic at US. The irregularity of the internal architecture produced by degeneration or hemorrhage also increases echogenicity. Surprisingly, 61% of small RCCs in our study were hyperechoic. Although angiomyolipoma is usually more hyperechoic, RCC should be included as a differential diagnosis of a hyperechoic mass of the kidney (26).

Radiology

495

#{149}

a.

b.

Figure

3.

Small

RCC (28 x 30 mm)

isoattenuating to slightly shows hypoattenuating

terial

hyperattenuating mass with slight

was I 1 HU. (c) Microscopic

capsule

is seen

around

the

section

tumor

C.

of tubular

architecture

mass contrast

shows

in the left enhancement

tubular

(arrowheads).

in a 61-year-old kidney

man.

(arrowheads). (arrowheads).

arrangement

(a) Unenhanced

of granular

(Hematoxylin-eosin

stain;

CT scan

cells.

Large

original

magnification,

woman.

(a) US scan

cysts

b.

a. Figure

4.

Small

RCC (20 x 20 mm)

irregular internal architecture ating, slightly inhomogeneous

occurred

(increase

of cystic

of CT numbers

after

in a 55-year-old

rim kidney

administration

normal renal structure, were observed around the tumor in small RCCs with hypoechoic rims. Presence of a capsule may assist in evaluating the aggressiveness of the tumor.

(arrowheads). (arrowheads).

shows

(b) Unenhanced CT scan On a contrast-enhanced

of contrast

or tumor stain. (c) Microscopic section shows multilocutated interstitium and the wall of the cysts. The thick capsule (Hematoxylin-eosin stain; original magnification, x40.)

The hypoechoic rim was another characteristic of small RCC. At histologic examination, thick pseudocapsules, formed by compression of the

thickness,

5 mm)

shows

C.

architecture

(*) and a hypoechoic mass in the right

(section

CT scan (section thickness, 5 mm) after administration of contrast macontaining serous fluid are observed (*). A x40.)

(b) Contrast-enhanced The increase in CT numbers

material

cystic is composed

was 9 HU).

mass containing of compressed

hyperechoic

(section CT scan

Angiography

hemorrhage normal renal

mass

in the left kidney

thickness, 10 mm) (not shown), slight

did not demonstrate in the cysts. parenchyma

Tumor and

cells fibrous

with

shows hyperattenuenhancement

any tumor

vessels

are seen both in the tissue (arrowheads).

Table 1 Relationship

between

Tumor

Echogenicity Homogeneous

at US and Cell Arrangement Architecture

(n

=

23)

Inhomogeneous Multilocular

Echogenicity

Solid

/isoechoic Hyperechoic

11 2

Hypo-

Note.-Irthomogeneous hemorrhage,

small

architecture cysts,

or necrosis,

Papillary 0 4

indicates regardless

Tubular 1 2

histologically of the type

Architecture

Cystic

(n

0 3

inhomogeneous

13)

=

2 11

architecture

due to

of cell arrangement.

CT Characteristics Large RCCs usually contain lowattenuation or cystic areas suggestive of intratumoral necrosis at CT (26). The attenuation varies from somewhat less to nearly equal that of normal renal parenchyma; less commonly,

496

Radiology

#{149}

tumors have high attenuation cornpared with the normal renal parenchyma at unenhanced CT (24,27). In our series of small RCC, visualization of a slightly hyperattenuating mass

was relatively common. Tumors of solid architecture usually had low attenuation, whereas tumors of papillary or tubular architecture were isoattenuating to slightly hyperatten-

August

1992

degree of cystic change, degeneration, or fibrosis within the tumor. Another characteristic feature seen at angiography in hypervascular small RCC is the presence of a translucent rim and displacement of the tumor vessels around the tumor. These findings were an indication of the existence of a capsule. At histologic examination, many displaced feeding arteries are seen inside the capsule. The vascuiarity of the capsule is sparse and there appears to be a translucent rim at angiography.

Table 2 Relationship

between

Tumor

at CT and

Appearance

Tumor

Tumor

Tumor

Appearance

at CT Unenhanced

Solid

Papifiary

Multilocular Cystic

Tubular

CT

Hyperattenuation

2

3

Isoattenuation Hypoattenuation

7

2

15

Contrast-enhanced Marked

Slight

ArChiteCtUre

Architecture

0

1 3 0

2 0 2

20 4

1 4

1 2

1 2

8

0

1

1

CT*

enhancement

enhancement

Intratumoral

hypoattenua-

tion

* Marked enhancement crease in CT number ofless

=

increase in CT number than 20 HU.

of more

than

21) HU, slight

enhancement

in-

=

CONCLUSION Cell arrangement or tumor architecture was closely correlated with radiologic characteristics, especially in regard to echogenicity, attenuation at CT, and vascularity of the tumor. With the increasing use of noninvasive imaging modalities, the number of small RCCs discovered wiil increase. However, the imaging characteristics of small RCC are variable and cover a wide spectrum. It would be

Table 3 Relationship

between

Tumor

at Angiography

Vascularity

Tumor

and

Architecture

Architecture

Multilocular

Vascularity

Solid

Papifiary

19 5

Hypervascular

Hypo-/avascular

Tubular

Cystic

0 3

0 4

0 5

difficuit

uating at unenhanced CT. The reason for the difference in attenuation is not clear; however, the number of cells in the tumor may be a factor. Tumor cells are arranged more densely in papillary or tubular architecture than in solid architecture. Tumors with hemorrhage demonstrated markedly high attenuation. Calcification is not infrequently observed, especially in papillary tumors (19,26,28), but was very rare in our

series

of small

RCC

(3%),

consistent

with findings reported by Zagoria et ai(26). The degree of contrast enhancement may be useful for differentiating small RCCs (especially of solid architecture) from cysts. However, papillary or multilocular cystic tumors, which usually show slight enhancement, must be viewed with caution to differentiate them from hemorrhagic or complicated cysts. It would be difficult to distinguish carcinoma from other solid tumors such as H#{252}rthle cell tumor, metastasis, or fibroma with CT alone (6). The degree of enhancement of a mass will depend not only on the vascuiarity of the tumor, but also on the amount of contrast materiai administered, the speed of the injection, and the phase of the scan. Dynamic study may provide useful clues to the amount of tumor vascularity and the presence or absence of a tumor capsule (29).

Volume

184

Number

#{149}

2

Angiographic Most RCCs angiographic

neovascularity,

Characteristics

arteriovenous

1.

mors of solid hypovascular,

architecture depending

increase.

RCC

with U

these

Tu-

were also on their

Bonsib

SM.

Renal

parenchymal

tumors.

In: Culp DA, Loening SA, eds. Genitourinary oncology. Philadelphia: Lea & Fe-

shunt-

tiaily higher; in our study, 47% of tumors were hypovascular. Vascularity correlated well with cell arrangement. In tumors with solid architecture (the most common type of RCC), a large nest of tumor is separated by a stroma that is characteristically endowed with prominent sinusoidlike vessels, whereas in tumors with papillary architecture, thin, vascuiarized, connective tissue stalks are lined with neoplastic cells (17). The vascular space in the stroma in solid tumors is greater than that in papiliary tumors (Figs ld vs 2e) (8). In larger tumors, the finding of solid architecture may become more frequent and the percentage of tumors with

may

small

tumors aione.

References

have characteristic findings, including

ing, pooling, and tumor staining. A small percentage of RCCs lack these classical angiographic features. Watson et al (30) and Lang et al (31) found hypovascularity in 22% and 10% of tumors studied, respectively. Among these hypovascular tumors, papillary-tubular adenocarcinoma or necrotic tumors are frequently observed (32,33). In small RCC, the frequency of hypovascuiarity is substan-

hypervascuiarity

to differentiate

from other solid imaging studies

2.

biger, Curry renal

1985. NS, Schabel neoplasms:

pathologic Radiology 3.

5.

Amendola MA, Bree RL, Pollack HM, Small renal cell carcinomas: resolving

7.

cell carcinoma:

Mihara

S.

10.

early

Case finding

of cancer

screening

9.

Radiology

discovery

and

increased detection. Radiology 1989; 170: 699-703. Bosniak MA. The small ( 3.0 cm) renal parenchymal tumor: detection, diagnosis, and controversies. Radiology 1991; 179: 307-317. nosis

8.

dilemma.

et al. a 1988; 166:

637-641. Warshauer DM, McCarthy SM, Street L, et at. Detection of renal masses: sensitivities and specificities of excretory urography/ linear tomography, US, and CT. Radiology 1988; 169:363-365. Smith SJ, Bosniak MA, Megibow AJ, et at.

Renal

6.

WL Jr. Small imaging, clinical course.

features, and 1986; 158:113-117.

diagnostic 4.

SI, Betsill diagnostic

detected

for abdominal

efficacy

and prog-

by ultrasonic

organs.

mass

Jpn J Med

Ultrasonics 1990; 17:267-279. Kashgrrlan M, Rosai J. Urinary tract: kidney, renal pelvis and ureter. In: Rosai J, ed. Ackerman’s surgical pathology. 7th ed. St Louis: Mosby-Year Book, 1989; 819-897. Bennington JL, Beckwith JB. Tumors of the kidney, renal pelvis, and ureter. In: Atlas of tumor pathology. Vol 12. Washington, DC: Armed Forces Institute of Pathotogy, 1975; 128-130. Fuhrman SA, Lasky LC, Limas C. Frognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathot 1982; 6:655-663.

Radiology

#{149} 497

11.

Subramanyam parenchymal

BR, Bosniak and capsular

MA. tumors

adults. In: Taveras JM, Ferrucci diology diagnosis-imaging-intervention. 12.

13.

14.

15. 16.

17.

18.

Renal in JT, eds. Ra-

Vol 4. Philadelphia: Lippincott, 1989:1-15. Weyman PJ, McClennan BL, Stanley RJ, Levitt RG, Saget 55. Comparison of cornputed tomography and angiography in the evaluation of renal cell carcinoma. Radiology 1980; 137:417-424. Karp W, Ekelund L, Otafsson G, Olsson A.

Computed

tomography,

ultrasound

in staging

angiography renal

20.

logic and biologic 21.

cell carcinoma. 22.

Gray-scale

hypernephromas. 765. Charboneau

137:757-

RR, Ernst

EC III,

Blei CL, Hartman OS, Friedman AC, Davis CJ. Papillary renal cell carcinoma: ultra-

SL, eds. Robbins pathologic ease. 4th ed. Philadelphia: 1011-1081.

25.

Zappasodi neti C.

26.

Zagoria RJ, Wofrnan CT features of renal

RJ, Blath

sonic,

pathologic

correlation.

10:429-434. Foster WL Jr. Roberts

Radiology

#{149}

30.

32. RA Jr.

Dunnick NR. Sonography of small renal masses with indeterminate density characteristics on computed tomography. Urol Radiot 1988; 10:59-67.

renal

RA.

F, Sanna G, Fiorentini G, FrassiSmall hyperechoic nodules of the

parenchyma.

31.

JCU 1982;

L Jr. Halvorsen

GM, Ketalis

AJR 1984; GW,

PP.

143:291-

Witten

Calcified

DM,

renal

masses. a review of ten years experience at the Mayo Clinic. Radiology 1972; 103:503508. Zeman RK, Cronan JJ, Rosenfietd AT, et at.

33.

tumor vascutarity. Radiology 1988; 167: 393-396. Watson RC, Fleming RJ, EvansJA. Arteriography in the diagnosis of renal cell carcinoma: review of 100 cases. Radiology 1968; 91:888-897. Lang EK,Johnson B, Chance HL, et al. Assessment of avascular renal mass lesions. South Med J 1972; 65:1-10. Blath RA, Mandlla-Jimenez R, Stanley RJ. Clinical comparison between vascular and avascular renal cell carcinoma. J Urot 1976; 115:514-519. Weiss RM, BeckerJA, Davidson AJ, Lytton

B.

Angiographic

appearance

illary-tubutar adenocarcinomas. 1969; 102:661-664.

of renal papJ Urol

JCU 1985; 13:321-324. NT, Karstaedt cell carcinoma

emphasis on relation to tumor Radiot 1990; 25:261-266.

498

of

1980;

mographic dilemma. 294. Daniel WW, Hartman

Renal cell carcinoma: dynamic thin-section CT assessment of vascular invasion and

et

spectrum

Radiology JW, Hattery

CBJr,

Coleman BG, Arger PH, Mintz MC, et al. Hyperdense renal masses: a computed to-

Farrow

Urology

Muthern

sonographic

24.

R, Stanley

28.

29. PH,

Renal cell carcinoma at autopsy. J Urol 1%7; 97:978-982. Talamo TS, Shonnard JW. Small renal adenocarcinoma at autopsy. J Urol 1980; 124:132-134. Cortan RS, Kumar V, Robbins SL. The kidney. In: Cortan RS, Kurnar V. Robbins

Mancilla-Jimenez

AG.

basis of disSaunders, 1989;

27.

et al. Spectrum of sonographic findings in 125 renal masses other than benign simple cyst. AJR 1983; 140:87-94. 23.

SI, Thomas

characteristics.

1982; 1:16-20. Coleman BG, Arger

at.

and

Acta Radiol 1981; 22:625-633. Cronan JJ, Zeman RK, Rosenfleld AT. Comparison of computed tomography, ultrasound and angiography in staging renal cell carcinoma. J Urol 1982; 127:712714. Hajdu

19.

Papillary renal cell carcinoma. Cancer 1976; 38:2469-2480. Press GA, McClennan BL, Melson GL, et at. Papillary renal cell carcinoma: CT and sonographic evaluation. AJR 1984; 143: 1005-1009. Bard RH, Lord B, Fromowitz F. Papillary adenocarcinoma of the kidney. II. Radio-

N, et at. with

size. Invest

August

1992