ajr%2e182%2e3%2e1820551

8/17/2019 ajr%2E182%2E3%2E1820551

1/7

AJ R:182, March 2004 551

Small Renal Cell Carcinomas:Correlation of Size with Tumor Stage,Nuclear Grade, and Histologic Subtype

OBJECTIVE.

Our goal was to correlate the size of renal cell carcinoma with tumor stage,

nuclear grade, and histologic subtype in patients treated using partial or radical nephrectomy.

MATERIALS AND METHODS.

We retrospectively reviewed 213 consecutive renal

cell carcinomas resected at our institution from 1995 through 1999. Three groups of lesions

stratified by size (

≤

3 cm, > 3–5 cm, > 5 cm) were compared with regard to pathologic find-

ings. Statistical significance was assessed using Fisher’s exact test.

RESULTS.

Of 50 lesions 3 cm or smaller, 19 (38%) had extension outside the renal cap-sule (T3 or T4) and 14 (28%) were a high nuclear grade (Fuhrman grade 3 or 4). Lesions 3 cm

or smaller and those greater than 3 cm to 5 cm did not differ statistically with regard to T stage

or nuclear grade. Lesions larger than 5 cm showed a statistically higher T stage (

p

< 0.001)

and nuclear grade (

p

= 0.001) than the other smaller lesions. More non–clear cell tumors were

found in the two groups of smaller lesions (

p

= 0.105) but without statistical significance. The

majority (58%) of the tumors were asymptomatic and had been detected incidentally on

cross-sectional imaging. Lesions larger than 5 cm were significantly more likely to be symp-

tomatic (

p

< 0.001). Seventy-nine percent of the tumors 3 cm or smaller were incidental, and

these lesions did not differ significantly from the symptomatic lesions with regard to stage,

grade, or histology.

CONCLUSION.

In our study population, renal cell carcinomas up to 3 cm, including

asymptomatic lesions, showed a significant incidence of high nuclear grade and tumor exten-

sion beyond the renal capsule; these findings support aggressive management of small

lesions. Symptomatic status was not an adequate discriminator to guide management. A lon-gitudinal study is necessary to further evaluate the efficacy of current patterns of therapy.

enal cell carcinomas comprise up

to 85% of solid renal masses, with

a steadily increasing incidence of

detection [1, 2] and smaller average size at di-

agnosis [3, 4] due, at least partially, to the in-

creased use of cross-sectional imaging. In

several large series, the majority of renal cell

carcinomas [3, 5, 7, 8] were asymptomatic and

found incidentally on noninvasive imaging

performed for unrelated reasons or for screen-

ing. In fact, the classic Grawitz triad of flank pain, gross hematuria, and palpable mass is

now uncommon [3, 4] and is indicative of ad-

vanced disease.

The study of small renal cell carcinomas

potentially involves several areas of current

controversy, including the appropriate man-

agement of incidental small lesions and the

usefulness of screening CT. A key consider-

ation in the management of solid renal masses

detected by serendipity is the nature of these

lesions: Can we anticipate that a significant

percentage of the small renal cell carcinomas

we detect are aggressive tumors that will grow,

invade adjacent structures, and produce me-

tastases? To contribute to the study of this is-

sue, we examined a series of renal carcinomas

resected during a recent 5-year period to deter-

mine the relationship between tumor diameter

and tumor subtype, nuclear grade, and T

stage of each lesion.

Materials and Methods

Retrospective review of the department of pathol-

ogy database at Johns Hopkins Hospital was per-

formed. Patients with renal cell carcinomas

documented at nephrectomy or partial nephrectomy

at Johns Hopkins Hospital from January 1995

through December 1999 were included in this study.

We recorded tumor size; mode of presentation; his-

Raymond M. Hsu

1,2

David Y. Chan

3

Stanley S. Siegelman

1

Received J uly 8, 2003; accepted after revisionSeptember 9, 2003.

1

Russell H. Morgan Department of Radiology, J ohnsHopkins Hospital, Baltimore, MD, and Department of

Radiology, J ohns Hopkins University, 601 N Caroline St.,Rm. 4214, Baltimore, MD 21287-0801. Addresscorrespondence to S. S. Siegelman.

2

Department of Radiology, Stanford University MedicalCenter, 300 Pasteur Dr., H1307, Stanford, CA 94304-5105.

3

J ames Buchanan Brady Urological Institute, J ohnsHopkins Hospital, Baltimore, MD.

AJ R

2004;182:551–557

0361–803X/04/1823–551

© American Roentgen Ray Society

R

8/17/2019 ajr%2E182%2E3%2E1820551

2/7

552

AJ R:182, March 2004

Hsu et al.

tologic subtype; tumor stage, according to the 1997

TNM classification system [9, 10]; and nuclear

grade, according to the Fuhrman classification sys-

tem [11].

In several cases for which clinicians used an

older tumor staging scheme, stage was reassigned

using the 1997 TNM criteria. All cases of reassign-

ment involved tumors between 3 and 7 cm withoutcapsule transgression; these tumors were down-

graded from stage II to stage I. In cases with inter-

mediate nuclear grade (i.e., lesions having

characteristics of two nuclear grades), lesions were

assigned the higher of the two grades. Direct review

of pathology samples was performed in cases in

which histologic subtype, tumor stage, or nuclear

grade was ambiguous in the original report.

For data analysis, lesions were stratified by size

into three groups: 3 cm or smaller, group A;

greater than 3 cm to 5 cm, group B; and greater

than 5 cm, group C. Fisher’s exact test was used to

test the independence of the data acquired for the

three groups with regard to tumor stage, nuclear

grade, histologic subtype, and mode of presenta-

tion. Symptomatic and asymptomatic lesions were

likewise compared.

Patients with known or suspected renal masses

were examined using a Somatom 4 helical CT

scanner (Siemens, Malvern, NJ). The examination

began with unenhanced consecutive 3-mm scans

to encompass the kidneys. Next, 115 mL of io-

hexol (Omnipaque 350, Nycomed, Princeton, NJ)

was injected at a rate of 3.5 mL/sec. Patients

weighing less than 115 lb (< 52 kg) were injectedwith 1 mL/lb (2.2 mL/kg) of iohexol. Three-milli-

meter-thick contiguous scans were obtained start-

ing 25 sec after the beginning of the injection. The

examination concluded with a series of 3-mm

scans after a 4-min delay (Figs. 1–3).

All lesions that enhanced more than 15 H were

diagnosed as renal neoplasms. Compliance with

advice for excision was high, with fewer than 1%

of patients refusing surgery.

Results

Our results are presented in three tables. Table 1

summarizes the relationship between lesion size,

Fuhrman nuclear grade, and TNM tumor stage.

Table 2 provides the distribution of histologic tu-

mor subtypes as a function of tumor size. Table 3

compares the Fuhrman nuclear grade and TNM

tumor stage for the symptomatic versus the inci-

dentally detected lesions.

During the 5 years encompassed by this re-

view, 213 renal cell carcinomas were resected

in 211 patients. Of these lesions, 162 were

conventional clear cell carcinoma; 33, papil-

lary; 10, chromophobe; zero, tubular; two,

mixed; and two, unknown. The remaining four

were categorized as “other,” which included

lesions with sarcomatoid change and with on-

cocytic components. With regard to nuclear

grade, 12 were F1; 105, F2; 79, F3; 16, F4; and

one, unknown. TNM staging revealed 81 T1,

18 T2, 105 T3, and nine T4 lesions.

Group A consisted of 50 lesions that were 3

cm or smaller. Of these lesions, 31% were not

pure clear cell. Seventy-two percent were a

low nuclear grade (grade 1 or 2), and 62%

were confined to the kidney. The mode of

presentation was documented in 43 of the le-

sions in this group, with 34 presenting inci-

A B

C

Fig. 1.—54-year-old man with incidentally detected 2-cm enhancing left renalmass (arrow 1).A, Unenhanced CT scan shows lesion with attenuation of 46 H and normal adja-cent renal parenchyma (arrow 2 ), which measured 42 H.B, Early contrast-enhanced CT scan shows that lesion appears exophytic but wellcircumscribed. Attenuation of lesion is 67 H (enhancement =21 H). Normal adja-cent renal parenchyma (arrow 2 ) was 127 H.C, Delayed contrast-enhanced CT scan shows lesion with attenuation of 95 H (en-hancement =49 H). After patient underwent partial nephrectomy, pathologic ex-amination revealed papillary renal cell carcinoma, nuclear grade 2, and tumorpenetration of renal capsule (T3).

8/17/2019 ajr%2E182%2E3%2E1820551

3/7

CT of Small Renal Cell Carcinomas


553

dentally and nine symptomatically. Of the

incidental lesions in this group, 74% were a

low nuclear grade, and 53% were confined to

the kidney.

Group B consisted of 59 lesions that were

larger than 3 cm but less than or equal to 5

cm. In this group, 26% were not pure clear

cell. Sixty-seven percent were nuclear grade

1 or 2, and 63% were confined to the kidney.

The mode of presentation was documented

in 55 of the lesions in this size stratification,

with 39 presenting incidentally and 16 pre-

senting symptomatically.

Group C was composed of 104 lesionsthat were larger than 5 cm. Of these lesions,

18% were not pure clear cell. Forty percent

were nuclear grade 1 or 2, and 30% were

confined to the kidney. The mode of presen-

tation was documented in 86 of the lesions in

this size stratification, with 34 presenting in-

cidentally and 52 that were symptomatic.

Analysis using Fisher’s exact test showed

no statistically significant difference between

groups A and B with regard to histologic

subtype (

p

= 0.93), nuclear grade (

p

=

0.881), or tumor stage (

p

= 0.731). Further

analysis showed no significant difference in

the percentage of lesions that were a high nu-

clear grade (F3 or F4, p

= 0.992) or a high tu-

mor stage (T3 or T4, p

= 0.731).

Comparison of groups B and C showed a

statistically significant difference in nuclear

grade (

p

= 0.001) or tumor stage (

p ≤

0.001).

Accordingly, there was a significant increase

in the percentage of lesions with nuclear

grade high enough (F3 or F4) to have prog-

nostic significance (

p

= 0.001) and tumorstage high enough (T3 or T4) to have prog-

nostic significance (

p

< 0.001) in group C. A

trend of higher probability of clear cell his-

tology was also documented (

p

= 0.105).

Comparison of the frequency of incidental

lesions showed no statistically significant

difference between groups A and B (

p

=

0.247), but a significant decrease was found

in incidental lesions in group C (

p

< 0.001)

when compared with group A or B. A trend

of higher tumor stage (

p

= 0.05) and higher

nuclear grade (

p

= 0.202) was noted for

symptomatic lesions; however, given our

sample size, we did not show this trend to be

independent of tumor size except in group B,

which was composed of symptomatic lesions

that were significantly more likely to be a

higher stage (

p

= 0.05).

Discussion

The 1969 Robson [12] and 1978 TNM [13]

staging schemes did not incorporate specificsize measurements. The 1987 TNM criteria

[14] defined lesions that were confined to the

kidney as T1 if 2.5 cm, and the generally ac-

cepted 1997 TNM criteria [12, 13] raised this

threshold to 7.0 cm. This substantial increase is

a point of debate, and based on survival data,

proposals include lowering the T2 threshold to

approximately 4–5 cm or discriminating be-

tween T1a and T1b subclassifications [15, 16].

A B

C

Fig. 2.—67-year-old man with incidentally detected 3-cm en-hancing exophytic left renal mass (arrow ).A, Unenhanced CT scan shows lesion with attenuation of 49 H.B, Early contrast-enhanced CT scan shows lesion with attenua-

tion of 90 H (enhancement =41 H).C,Delayed contrast-enhanced CT scan shows lesion with atten-uation of 94 H (enhancement =45 H). After patient underwentpartial nephrectomy, pathologic examination revealed papillaryrenal cell carcinoma, nuclear grade 3, and penetration of renalcapsule (T3).

8/17/2019 ajr%2E182%2E3%2E1820551

4/7

554


Hsu et al.

The debate regarding the appropriate sizecriteria for staging has fueled a great deal of

analysis about how tumor size affects survival.

Not surprisingly, smaller tumors generally por-

tend higher postoperative survival rates [16],

with various statistically significant break-

points reported [15, 16]. Our goal was not toduplicate these efforts, but to correlate renal

cell carcinoma size with other factors that may

independently influence management and out-

come. We correlated tumor size with incidence

of extracapsular spread (T3 or T4 status), Fu-

hrman nuclear grade, and histologic subtype,all of which have been validated as indepen-

dent prognostic factors [15].

Siegelman et al. [18] conducted a study of

100 consecutive patients diagnosed with re-

nal cell carcinoma in the 1960s to evaluate

A B

C

Fig. 3.—56-year-old woman with incidentally detected 3-cm enhancingwell-circumscribed left renal mass (arrow 1).A, Unenhanced CT scan shows lesion with attenuation of 43 H and normaladjacent renal parenchyma (arrow 2 ), which measured 40 H.B, Early contrast-enhanced CT scan shows lesion with attenuation of 73 H

(enhancement =32 H).C,Delayed contrast-enhanced CT scan shows lesion with attenuation of 78H (enhancement =37 H). After patient underwent partial nephrectomy,pathologic examination revealed papillary renal cell carcinoma, nucleargrade 2, with penetration of renal capsule (T3).

TABLE 1 Summary of the Distribution of Lesions by Fuhrman Nuclear Grade and 1997 TNM Tumor Stage

Lesion Size Total No.Fuhrman Nuclear Grade TNM Stage

F1 F2 F3 F4 Unknown T1 T2 T3 T4

≤ 3 cm

No. 50 4 32 13 1 31 0 18 1

% 8 64 26 2 62 0 36 2

>3–5 cm

No. 59 6 33 18 1 1 37 0 19 3

% 10 57 31 2 63 0 32 5

>5 cm

No. 104 2 40 48 14 13 18 68 5

% 2 38 46 13 13 17 65 5

Total no. 213 12 105 79 16 1 81 18 105 9

8/17/2019 ajr%2E182%2E3%2E1820551

5/7



555

Note.—Papillary and chromophobe tumors are associated with a better prognosis than clear cell renal cell carcinomas.

TABLE 2 Distribution of Histologic Subtypes of Lesions

Lesion Size

No. of Lesions Percentage of Lesions

Total Clear Cell Papillary Chromophobe Tubular Mixed Other Unknown Clear CellPapillary and

Chromophobe

≤ 3 cm 50 35 12 2 0 0 1 0 70 28

>3–5 cm 59 43 12 2 0 1 0 1 74 24>5 cm 104 84 9 6 0 1 3 1 82 14

Note.—Mode of presentation was unknown for seven that were ≤ 3 cm, four that were >3–5 cm, and 18 that were >5 cm.

TABLE 3 Number and Percentage of Incidental and Symptomatic Lesions by Size, Fuhrman Nuclear Grade, and TNM Tumor Stage

Lesion Size and Mode

of Presentation

Total

No.

Fuhrman Nuclear Grade 1997 TNM Tumor Stage

F1 F2 F3 F4 Unknown T1 T2 T3 T4

≤ 3 cm

No. 50 4 32 13 1 31 0 18 1

% 8 64 26 2 62 0 36 2

Incidental

No. 34 2 23 6 3 17 1 16 0

% 6 68 18 9 50 3 47 0

Symptomatic

No. 9 0 5 4 0 7 0 2 0

% 0 56 44 0 78 0 22 0

>3–5 cm

No. 59 6 33 18 1 1 37 0 19 3

% 10 57 31 2 63 0 32 5

Incidental

No. 39 3 25 9 1 1 27 1 11 0

% 8 66 24 3 69 3 28 0

Symptomatic

No. 16 1 10 5 0 7 0 9 0

% 6 63 31 0 44 0 56 0

>5 cm

No. 104 2 40 48 14 13 18 68 5

% 2 38 46 13 13 17 65 5

Incidental

No. 34 1 11 17 5 6 4 23 1

% 3 32 50 15 18 12 68 3

Symptomatic

No. 52 0 25 19 8 5 11 35 1

% 0 48 37 15 10 21 67 2

Total

No. 213 12 105 79 16 1 81 18 105 9% 6 50 37 8 38 8 49 4

Incidental

No. 107 6 59 32 9 1 50 6 50 1

% 6 55 30 8 46 6 47 1

Symptomatic

No. 77 1 40 28 8 19 11 46 1

% 1 52 36 10 25 14 60 1

8/17/2019 ajr%2E182%2E3%2E1820551

6/7

556


Hsu et al.

the relationship between the manner of clini-

cal presentation and the eventual outcome af-

ter nephrectomy [17]. There were 16 patients

in whom a tumor was detected by serendipity

during excretory urography. The indication

for the studies was hypertension, prostatic en-

largement, or another condition considered

unrelated to a suspicion of renal neoplasm.

The mean size of the 16 tumors was 6.2 cm,

which is considerably smaller than renal car-

cinomas detected because of hematuria

(mean size, 10 cm), flank pain (mean size,

11.7 cm), or palpation of a mass (mean size,

12.1 cm). Fifteen of these 16 patients were

long-term survivors free of recurrent tumor

[18]. From that study, the authors concluded

that, as one would intuit, the ideal time to de-

tect renal cell carcinoma is when the lesions

are small and asymptomatic. These research-

ers asserted that every patient who had an im-

aging study that included the kidneys should

be screened for early renal cell cancer. The

subsequent introduction of CT, sonography,

and MRI greatly increased the number of

subjects available for evaluation. Serendipity

has become the most frequent means by

which renal cell carcinomas is detected.

Radiologists scrutinizing each kidney on

imaging studies performed for other purposes

are performing a screening study. They should

be ready to deal with the criticisms that have

been directed at such efforts [19]. A key issue

is whether length-time bias is a serious consid-

eration in current practice. This concept holds

that if there is great variability in the biologicbehavior of renal cell carcinomas, the less ag-

gressive, slower-growing neoplasms are more

likely to be detected at screening because they

are present in the preclinical state for a longer

period. Are we making surgical candidates of

patients with renal neoplasms that might re-

main dormant for decades and never become

clinically evident during their lifetimes? It is

not possible to provide a definitive answer to

this challenging question. The issue of screen-

ing bias is particularly pertinent because

screening of asymptomatic adults for renal

cancer using sonography has been used in Ja-

pan [20] and Germany [21] for several years.This study offers support to justify an aggres-

sive approach to the management of small re-

nal cell carcinomas detected by serendipity.

In our study population, small renal tumors

up to 3 cm in diameter were found to exhibit

several statistical trends of interest. Of clini-

cal prognostic significance, 28% were nu-

clear grade 3 or higher, and 38% extended

beyond the renal capsule. Even for asymp-

tomatic lesions, these figures were 26% and

47%, respectively. These percentages were

similar to those for lesions in the next group

(i.e., > 3–5 cm) but were significantly less

than those in the group composed of lesions

larger than 5 cm.

Nuclear grading according to the Fuhrman

classification system [11] is widely accepted

and has been shown to confer prognostic sig-

nificance. A sharp increase in metastases and a

decrease in survival have been noted for pa-

tients with lesions that are nuclear grade 3 or 4

[22]. TNM tumor staging is even more predic-

tive than the Fuhrman classification system.

Capsular transgression defines a T3 lesion,

with a significant increase in metastases and a

decrease in survival for patients with T3 or T4

lesions [23].

The significant number of small lesions, in-

cluding asymptomatic lesions, found to have

high nuclear grade and high TNM tumor stage

suggests that many of these lesions would have

progressed to regional distant metastases and

potentially would have become a source of

morbidity and mortality. This finding clearly

supports resection of small tumors, whether in-

cidental or symptomatic.

Some may argue that the similar incidence

of high nuclear grade and tumor stage for le-

sions in groups A (

≤

3 cm) and B (> 3–5 cm)

suggests that a group A lesion may be safely

followed up until it is in the 3- to 5-cm range

and may subsequently be resected if found to

grow or to grow rapidly. However, this propo-

sition is suspect, unless the incidence of inter-val metastases is known to be very low. The

risk–benefit ratio of such an approach might

deserve examination in patients less tolerant of

surgery or with shorter life expectancy.

As for histologic classification, renal cell

carcinoma includes several subtypes. Clear

cell (“conventional”) histology accounts for

70–80%, followed by papillary (10–15%) and

chromophobe (5%), both of which have been

shown to be associated with a higher rate of

survival and lower rate of metastasis than the

clear cell subtype [15]. The tubular subtype

and sarcomatoid change are less common and

confer a worse prognosis. These subtypes can-not be reliably differentiated on imaging.

We found the largest lesions (> 5 cm) to

include more clear cell tumors and fewer

papillary and chromophobe tumors than the

smaller tumors. The distribution of histo-

logic subtypes was similar between the two

groups of smaller lesions. Of the three most

common histologic subtypes, we did not

find a significantly different distribution for

incidental and symptomatic lesions when

corrected for size.

We acknowledge several inherent limita-

tions of this retrospective study. By searching

the pathology database, we defined our study

population as patients with renal cell carci-

noma treated by resection. This criterion ex-

cludes most patients with known metastases,

patients unwilling to undergo or unsuitable for

surgery, and those treated with percutaneous

ablation. However, few patients were treated

by percutaneous ablation during the time

frame of our study, and selection for this op-

tion was primarily based on small size (usually

< 4 cm) or the need to spare nephrons and not

on other independent predictors of outcome.

Our conclusions therefore apply to a close ap-

proximation of the defined study population—

namely, those with renal cell carcinoma who

are candidates for resection.

In fact, despite exclusion of most patients

with known metastases with presumably [24]

higher average nuclear grade and tumor stage,

we found a significant percentage of small le-

sions to have high nuclear grade and tumor

stage. Exclusion of metastatic disease obvi-

ously does however preclude correlation of

metastatic incidence with tumor characteristics

at presentation. Also, follow-up would be re-

quired to determine survival after resection

and metastasis-free survival.

Our study also excludes masses radiologi-

cally presumed to be renal cell carcinoma but

later found to be other malignant or benign en-

tities at tissue diagnosis. Exclusion of these le-sions is mitigated by the fact that most

nonspecific solid enhancing renal masses are

renal cell carcinoma or collecting system tran-

sitional cell carcinoma, which also requires re-

section for cure. Therefore, the argument for

treatment of small presumed renal cell carci-

noma remains strong.

References

1. Ries LAG, Kosary CL, Hankey BF, et al. SEER

cancer statistics review, 1973–1995

. Bethesda,

MD: National Cancer Institute, 1998

2. Chow WH, Devesa SS, Warren JL, Fraumeni JF

Jr. Rising incidence of renal cell cancer in theUnited States. JAMA

2000;281:1628–1631

3. Jayson M, Sanders H. Increased incidence of ser-

endipitously discovered renal cell carcinoma.

Urology

1998;51:203–205

4. Sene AP, Hunt L, McMahon RFT, et al. Renal

carcinoma in patients undergoing nephrectomy:

analysis of survival and prognostic factors.

Br J

Urol 1992;70:125–134

5. Luciani LG, Cestari R, Tallarigo C. Incidental re-

nal cell carcinoma: age and stage characterization

8/17/2019 ajr%2E182%2E3%2E1820551

7/7



557

and clinical implications—study of 1092 patients

(1982–1997). Urology

2000;56:58–62

6. Reddan DN, Raj GV, Polascik TJ. Management

of small renal tumors: an overview. Am J Med

2001;110:558–562

7. Lee CT, Katz J, Fearn PA, Russo P. Mode of pre-

sentation of renal cell carcinoma provides prog-

nostic information. Urol Oncol

2002;7:135–140

8. Homma Y, Kawabe K, Kitamura T, et al. In-

creased incidental detection and reduced mortal-

ity in renal cancer: recent retrospective analysis at

eight institutions. Int J Urol 1995;2:77–80

9. Flemming ID, Cooper JS, Henson DE, et al.

AJCC

cancer staging manual,

5th ed. Philadel-

phia, PA: Lippincott-Raven, 1997:231–232

10. Swanson DA, Rothenberg HJ, Boynton AL, et al.

Future prognostic factors for renal cell carci-

noma. Cancer

1997;80:997–998

11. Fuhrman SA, Lasky LC, Limas C. Prognostic sig-

nificance of morphologic parameters in renal cell

carcinoma. Am J Surg Pathol 1982;6:655–663

12. Robson CJ, Churchill BM, Anderson W. The re-

sults of radical nephrectomy for renal cell carci-

noma. J Urol 1969;101:297–301

13. Harmer M. TNM classification of malignant tu-

mors

, 3rd ed. Geneva, Switzerland: International

Union Against Cancer, 1978

14. Hermanek P, Sobin LH. TNM

classification of

malignant tumors, 4th ed. Berlin, Germany:

Springer-Verlag, 1987

15. Lau WK, Cheville JC, Blute ML. Prognostic fea-

tures of pathologic stage T1 renal cell carcinoma af-

ter radical nephrectomy. Urology

2002;59:532–537

16. Gettman MT, Blute ML. Update on pathologic

staging of renal cell carcinoma. Urology

2002;60:

209–217

17. Delahunt B, Kittelson JH, McCredie MRE, et al.

Prognostic importance of tumor size for localized

conventional (clear cell) renal cell carcinoma.

Cancer

2002;94:658–664

18. Siegelman SS, Sprayregen S, Bosniak MA, Freed

SZ. Serendipity in the diagnosis of renal carci-

noma. J Can Assoc Radiol 1972;23:251–256

19. Black CB, Welch HG, Advances in diagnostic im-

aging and overestimation of disease prevalence and

the benefits of therapy. N Engl J Med 1993;328:

1237–1243

20. Mihara S, Kuroda K, Yoshioka R, Koyama W.

Early detection of renal cell carcinoma by ultra-

sonic screening: based on the results of 13 years

of screening in Japan.

Ultrasound Med Biol

1999;25:1033–1039

21. Filipas D, Spix C, Schulz-Lampel D, et al. Ultra-

sound screening of renal cell carcinomas [in Ger-

man]. Radiologe

2002;42:612–616

22. Ficarra V, Righetti R, Marignoni G, et al. Prog-

nostic value of renal cell carcinoma nuclear grad-

ing: multivariate analysis of 333 cases. Urol Int

2001;67:130–134

23. Tsui K, Shavarts O, Smith RB, et al. Prognostic in-

dicators for renal cell carcinoma: a multivariate

analysis of 643 patients using the revised 1997

TNM staging criteria.

J Urol 2000;163:1090–1095

24. Kanamaru H, Mori H, Sasaki M, et al. Histologic

characteristics of renal cell carcinomas with lymph

node metastasis. Int J Urol

1997;4:451–455

ajr%2e182%2e3%2e1820551

Documents