Urine cytology in the detection of renal adenocarcinoma

Urine Cytology in the Detection of Renal Adenocarcinoma

FRANCESCO PISCIOLI, MD.’ TERESA PUSIOL, CT,’ PIERANTONIO SCAPPINI, MD,t AND LUClO LUCIANI, MDt

Whereas urine cytology has proved to be of considerable diagnostic value for nonpapillary urothelial carcinoma, carcinoma in situ, and high-gnde urothelial tumors, controversy has arisen over the accuracy of cytology in the diagnosis of renal neoplasms. To establish the reliability of urine cytology as a detection technique, 436 urine specimens from 59 patients with histologically proven renal adenocarcinoma were examined. Malignant cells in urinary sediment were found in 121 (27.75%) specimens from 31 (52.54%) patients. Seven of 20 (35%) patients with Stage I tumor showed neoplastic cells in 16 (16.5%) of 97 urinary samples. Positive cytologic features were found in 36.9% of 84 urinary specimens from 7 (50%) of 14 patients with tumors smaller than 5 cm. Using only imaging methods, the renal neoplasm was diagnosed in 58 (983%) cases. Urine cytology is, therefore, of little value in the diagnostic evaluation of known renal masses and in the detection of early-stage disease. In the current series there was not a single case with positive cytologic findings in which radiology did not reveal the tumor. The cytologic examination of the urinary sediment is unreliable in the diagnosis of radiologically unresolved cases of renal neoplasms. In 15 patients (44%) who had cancer cells in the urine the neoplasm had not invaded the renal pelvis. In 36% of patients with negative urine cytologic findings the renal pelvis was involved by renal adenocarcinoma. In the current study the desquamation of neoplastic cells in the urinary stream did not depend on tumor invasion of the renal pelvis. Based on the assumption that the primary objective of a screening procedure is to achieve an adequate sensitivity for early detection of disease, the results led the authors to conclude that the use of urine cytology as a possible screening test of renal adenocarcinoma is futile. Factors in the current study that diminished the practical value of cytologic examination of urinary sediment included the large number of specimens without malignant cells and the insensitivity of the procedure in diagnosing early renal adenoarcinoma of limited extent. Although this is the largest series reported, the number of cases considered was limited and further detailed studies are mandatory to definitively clarify the value of urine cytology in detecting renal adenocarcinoma.

Cancer 56:2251-2255. 1985.

ECAUSE OF ITS INCREASING FREQUENCY, its unpre- B dictable clinical behavior, and its unresponsiveness to nonsurgical approaches, renal adenocarcinoma (RA) provides a stimulating clinical challenge to urologists and oncologists in terms of detection, prognosis, and management. Epidemiologic data are highly indicative of a real increase in the incidence and mortality of RA in most countries. The reported frequency of RA is underesti- mated because since 67% of RAs are clinically unrecognized tumors.’

When RA is first diagnosed, disseminated malignancy is present in 25% to 57%*” of the patients, and metastases will be evident within 1 year in 94% of patient^.^

Therefore, RA is generally detected as a metastatic malignancy, when the available therapeutic procedures are

The research was supported in part by a “Lega per la Lotta contro i

From the *Institute of Anatomic Pathology and ?Division of Urology,

Address for reprints: Lucio Luciani, MD, Head, Division of Urology,

Accepted for publication January 21, 1985.

Tumori” (Trento) grant.

S. Chiara Hospital, Trento, Italy.

S. Chiara Hospital, 1-38100 Trento, Italy.

of little effect. Surgery is currently the only curative treat- ment, and it is completely effective only in patients with early-stage disease of localized extent. Hence, the best chance for cure of RA lies in early diagnosis.

The principal factor that impedes detection of early lesions is the impossibility of establishing individual risk factors for neoplasm since the etiology is unknown. The use of screening procedures in asymptomatic populations is certainly the best current means of reducing mortality from RA. As an innocuous, noninvasive diagnostic method, cytology might be proposed as a screening technique for renal neoplasms.

In the current article, we report a cytologic study of 436 urine sediments from 59 patients with RA, compare the cytologic findings with tumor stage, and evaluate the reliability of urine cytology in detection of the neoplasm.

Materials and Methods

The current study includes 436 urinary sediments from 59 patients with histologically proven RA. There were 38 men and 21 women, ranging in age from 39 to 82 years.

225 1

2252 CANCER November 1 1985 Vol. 56

TABLE I . Relationship Between Cytologic Results and Tumor Stage

No. of No. of patients with Accuracy Stage patients cytology

1 20 7 35 I1 I I 9 82 Ill 15 10 67 1V 13 5 38

In all cases radical nephrectomy was performed after radiologic study (intravenous pyelography, ultrasonography, computed tomography, and arteriography). The unfixed urine specimen was centrifuged at 2500 rpm for 30 minutes. The smears were made on albumen-coated slides, which were fixed in 95% ethyl alcohol and stained with hematoxylin and eosin or by a modified Papanico- laou method.

Urinary sediment of RA showed four different malignant cellular types.’ Multinucleated cells with distinct nucleoli as well as large, clear, vacuolated cells with ec- centric hyperchromatic nuclei repeated the cellular mor- phologic features observed in the parent neoplasm.

Granular eosinophilic cells with pyknotic nuclei and distinct and/or ill-defined cytoplasmic borders were the cancer cells seen most frequently in the urinary sediment and were believed to be the result of degenerative changes caused by the urinary environment.

The cytologic diagnosis was evaluated either as positive or negative. Abnormal, suspicious, or atypical findings were considered as negative. No equivocal category was used. In the histopathologic diagnosis of RA four different categories of cellular types were considered: ( I ) predominantly clear cells; (2) predominantly dark cells; (3) pseudosarcomatous cells, and (4) mixed cells. The size of the tumor and the renal pelvis were evaluated. Three sections of the neoplasm, including one with adjacent kidney, a section of the pelvis, renal artery and vein, ureter, and lymph nodes (if present) were examined histologically.

Patients were staged according to Robson’s groups modification6 of the classification system of Flocks and

Kadesky,’ as follows-Stage I: tumor within the kidney and confined by the renal capsule; Stage 11: capsular invasion and perinephric fat involvement; Stage 111: tumor invading the renal vein or involvement of the regional lymph nodes; Stage IV: distant metastasis or spread to adjacent organs. The accuracy of the technique was cal- culated according to standard method^.^.^

Results

One hundred twenty-one (27.75%) urine specimens from 3 1 (52.54%) patients contained malignant cells.

Table 1 shows tumor stage and percentage of positive cytology cases. Table 2 shows the relationship between the number of specimens and the accuracy of cytology.

Table 3 illustrates the relationship between urinary specimens and the size of the renal tumor. Thirty-one (36.9%) positive samples were found in 7 of 14 patients who had tumors smaller than 5 cm.

In Table 4 we reported the correlation between the cytologic findings and the status of the renal pelvis. Sixteen (64%) of the patients with involvement of the renal pelvis had 60 (26.66%) positive urinary samples of the 225 examined.

In Table 5 the urine cytologic findings are related to the histologic type of the tumor. Positive cytologic results were seen in 68 (20.79%) specimens from 2 1 (47.7%) patients with clear cell carcinoma, in 18 (52.94%) from 4 (57.1%) patients with dark cell carcinoma, and in 35 (55.5%) from 6 (85.7%) patients with mixed cell carcinoma.

The patient with pseudosarcomatous cell carcinoma had 12 negative samples.

The results of imaging methods in detecting renal neoplasms are reported in Table 6. Intravenous pyelography alone disclosed a renal mass in 38% of the patients. U1- trasonography confirmed the pyelographic diagnosis in six patients. Ten patients were studied by the combined use of intravenous pyelography, ultrasonography, and arteriography, and eight by intravenous pyelography and

TABLE 2. Relationship Between Cytologic Results and Number of Urine Specimens

No. of Specimens

1-3 4-10 > 10

Patients with Patients with Patients with Total Positive Total positive Total Positive Total positive Total Positive Total positive

Stage patients cytology specimens specimens patients cytology specimens specimens patients cytology specimens specimens

I 12 4 24 5 6 1 39 5 2 I 36 6 II I I 3 I 8 6 54 31 2 2 41 18

IV 3 2 7 3 4 I 25 2 6 2 93 10 111 8 4 16 4 3 2 21 7 4 4 65 29

No. 9 URINE CYTOLOGY AND RENAL ADENOCA Piscioli ef al. 2253

TABLE 3. Relationship Between Cytology and Tumor Size

Size of the Tumor

I-5cm 6-10cm >10cm

Patients with positive cytology 7 (50%) I 5 (45%) 9 (75%)

Total 14 33 I 2

Patients with negative cytology 7 18 3

arteriography. Six patients were investigated by intravenous pyelography, computed tomography, and arteriography, and three underwent intravenous pyelography, ultrasonography, and arteriography. Three patients underwent intravenous pyelography, ultrasonography, and computed tomography.

Results of intravenous pyelography and arteriography were both negative in one patient, and eight urinary specimens did not contain malignant cells. Exploratory surgery was mandatory because of persistent hematuria. Patho- logic examination of the surgical specimens showed a 1.5- cm papillary dark cell adenocarcinoma.

Discussion

There is a general consensus that urine cytology is the best detection technique for the screening of high-risk in- dustrial workers" who have been exposed to known bladder carcinogens and for the follow-up of patients who have been treated for cancer of the lower urinary

Numerous investigations have documented in depth the effectiveness of the cytologic urinalysis in the diagnosis of in sifu and invasive nonpapillary urothelial carcinoma."-" In these cases the cytologic detection of malignant cells in the sediment can predate the advent of a clinical Occurrence of the cancer since cystoscopic findings may be normal or indistinguishable from those of inflam- matory or other benign conditions.

Considerable controversy has arisen over the value of urine cytology in detecting renal neoplasms. Urinary cytodiagnosis of RA was first described by Papanicolaou.16 who advocated cytologic examination of urine together

TABLE 4. Relationship Between Cytology and Renal Pelvis Involvement

Patients with

positive Patients CYtoIogY with

Total negative patients No. % cytology

Patients with involvement of

Patients without involvement the renal pelvis 25 16 64 9

of the renal wlvis 34 15 44 19

with other diagnostic means as a helpful procedure in recognizing early or hidden urologic cancer. Deden" first found malignant cells in urine samples of asymptomatic patients with uncertain radiologic diagnosis of limited renal tumor. Meisels" suggested that the use of urine cytology can be of value in the detection of kidney carcinoma, and MelamedIg believes that carcinoma of the renal cortex can be diagnosed correctly by examination of the urinary sediment.

Recently Tannenbaum" has put forward the view that with the combined use of careful physical workups, intravenous urograms, and urine cytologic studies it should be possible to identify many more renal neoplasms at an early stage.

On the other hand, urine cytology has been considered by different authors to be unreliable,2' and of limited,22 little,23 q~estionable,'~ modest:4 or ancillary,25 value in diagnosing renal cancer. Also, the overall accuracy of cytologic studies in detecting R4 varies greatly.

A comparison of the results of various investigators shows accuracy ranging from 0%26 to 100%'8,27 in the cytologic diagnosis of renal malignancy.

Employing oil red 0 stain, Hajdu er ~ 1 . ~ ' and Milsten ef aL2' obtained percentages of positive cases in 82% and 85% of patients, respectively, whereas the positive rates using Papanicolaou's technique were 6% and 7%, respectively.

Nevertheless, Milsten ef d2' have cautioned that a positive histochemical finding is meaningful only in pa-

TABLE 5. Relationship Between Cytology and Histologic Type of Renal Adenocarcinoma

Patients with Patients with positive negative Positive Negative CytOh3Y cytology specimens specimens

Total Total Histologic type patients No. % No. % specimens No. I No. z

Clear cells 44 21 41.7 23 52.21 321 68 20.19 259 19.2 Dark cells 7 4 57.1 3 42.85 34 18 52.94 16 47.05 Pseudo-sarcomatous cells I - I 100 I 2 - 12 100 Mixed cells 7 6 85.1 I 14.28 63 35 55.55 28 44.4

2254 CANCER November I I985 Vol. 56

TABLE 6. Imaging Methods in Detecting 5 8 Renal Neoplasms

Imaging methods

IVP and/or IVP IVP + US US and/or CT

IVP + u s +cT + arteriography

No. of patients 22 6 10 20

IVP: intravenous pyelography: US: ultrasonography; CT: computed tomography.

tients who have a suspected or known renal mass, because there is a high false-positive rate reported in control groups ( 17%).

Finally, Mount et dm showed that urine cytology using fat staining is an unsatisfactory procedure in the identi- fication of RA because of the high incidence of false-positive (42%) and false-negative (48%) findings.

Review of the literature shows that: (1) the reported data and opinions do not include a sufficient number of cases to permit any conclusive statement to be made con- cerning the value and accuracy of urine cytology in detecting RA; (2) since there are no large studies on the use of cytologic urinalysis in the diagnosis of RA, it is not possible to comment on the usefulness of the test as a detection technique; (3) no attempt to estimate the relationship between positive cytology and tumor stage has yet been reported. This relationship is essential in the establishment of the effectiveness of the method in order to make early diagnosis of the disease; (4) as far as we know, no studies have provided quantitative assessment of the relationship between the number of urine specimens and accuracy of urine cytodiagnosis; and (5) in the identifi- cation of malignant cells in urinary sediment, the oil red 0 stain is not reliable enough to improve the accuracy of the conventional Papanicolaou's method.

Based on the largest series of urine cytology from patients with RA, the current report questions the role and value of urine cytology in detecting renal neoplasms.

Only 27.75% of the urine specimens from 3 1 (52.54%) patients contained malignant cells. Radiologic diagnosis of renal neoplasms was made in 58 (98.3%) cases.

Positive cytologic features were found in 16.5% of 97 urine specimens from 7 (35%) of 20 patients with Stage I disease.

Neoplastic cells exfoliating from tumors smaller than 5 cm were present only in 36.9% of 84 urinary sediments in 7 (50%) of 14 patients. Therefore, urine cytology is of little value in the diagnostic evaluation of known renal masses and in the detection of early RA.

In 15 patients (44%) with cancer cells in the urinary sediment the neoplasm had not invaded the renal pelvis. In 36% of patients with negative urine cytologic findings the renal pelvis was involved by RA.

These results are in contrast to the experience of other authors who reported the presence of malignant cells in cytologic specimens only when the tumor reached a sufficient size to ulcerate into the renal pelvis.14,21,22.24,31,32

In our study, increasing the number of urinary samples did not improve the accuracy of urinary cytology.

The accuracy of roentgenographic diagnosis of renal masses is about 97%.33-36 In the current series, the only patients with negative radiologic results had eight urinary sediments without malignant cells.

In conclusion, we strongly believe that urine cytology is unreliable in detecting the 3% of renal neoplasms that radiology has failed to reveal and is not to be recom- mended for the establishment of the benign or malignant nature of radiologically debatable lesions of the kidney.

Since the main objective of a screening procedure is to detect early disease, the cytologic examination of the urinary sediment as a possible screening test for RA is also questionable.

In our experience factors that diminished the practical value of urine cytology were the large number of urine specimens that do not contain malignant cells and the insensitivity of the procedure for diagnosis of early disease.

Although the current series is the largest reported, it is still limited; therefore, further studies are needed to definitively clarify the value of urine cytology in the detection of RA.

REFERENCES

I . Hellsten S, Berge T, Wehlin L. Unrecognized renal cell carcinoma: Clinical and pathological aspects. Scund J Urol Nephrol 1981; 8:273- 278.

2. Skinner DG, Colvin RB, Vermillion CD, Pfister RC, Leadbetter WF. Diagnosis and management of renal cell carcinoma: A clinical and pathological study of 309 cases. Cancer 1971; 28: 1165-1 177.

3. Montie JE. Stewart BH, Staffron RA, Banowsky LHW, Hewitt CB, Montagne DK. The role of adjunctive nephrectomy in patients with metastatic renal cell carcinoma. J Urol 1977; 117:272-275.

4. Lokich JJ, Harrison JH. Renal cell carcinoma: Natural history and chemotherapeutic experience. J Urol 1975; I 14371-374.

5. Piscioli F, Detassis C, Polla E, Pusiol T, Reich A, Luciani L. Cy- tologic presentation of renal adenocarcinoma in urinary sediment. Acln

6. Robson CJ, Churchill BM, Anderson W. The results of radical nephrectomy for renal cell carcinoma. J Urol 1969; 101:297-301.

7. Rocks RH, Kadesky MC. Malignant neoplasms of the kidney: An analysis of 353 patients followed five years or more. J Urol 1958; 79:

8. MacNeils BS, Keeler E, Adelstein SJ. Primer on certain elements of medical decision making. N Engl J Med 1975; 293:2 I 1-2 15.

9. Vecchio TJ. Predictive value of a single diagnostic text in unselected populations. N EnglJMed 1966 2741 171-1 173.

10. Crabbe JGS, Cresdee WC, Scott TS, Williams MHC. The cytological diagnosis of bladder tumors amongst dyestuff workers. Er J Ind Med 1956; 13:270-276.

1 I . Beyer-Boon ME, deVoogt AJ, van der Velde EA, Brussee JAM, Schaberg A. The efficacy of urinary cytology in the detection of urothelial tumors. Urol Res 1978; 6:3-12.

12. Esposti PL, Moberger G, Zajicek J. The cytologic diagnosis of transitional cell tumours of the urinary bladder and its histologic basis. Acla C.vl01 1970 14145-155.

C ~ l o l 1983; 27~383-390.

196-20 I .

No. 9 URINE CYTOLOGY AND RENAL ADENOCA - Piscioli et al. 2255

13. Esposti PL, Edsmyr F, Tribukeit B. The role of exfoliative cytology

14. Koss LG. Diagnostic Cytology and its Histopathologic Bases, ed.

15. Trott PA. Cytological screening for cancer of the bladder. Proc R

16. Papanicolaou GN. Cytology of the urine sediment in neoplasm

17. Deden C. Cancer cells in urinary sediments. Acfu Radio1 1954;

18. Meisels A. Cytology of the carcinoma of the kidney. Acfu Cyfol

19. Melamed MR. Introduction to cytology of the urinary tract. In: Wied GL, Koss LG, Reagan JW, eds. Compendium on Diagnostic Cy- tology, ed. 4. Tutorials of Cytology. Chicago, 1976; 405-412.

20. Tannenbaum M. Surgical and histopathology of renal tumors. Semin Oncol1983; 10485-489.

21. Rosai J. Ackerman’s Surgical Pathology, ed. 6. St. Louis: CV Mosby, 1981; 729.

22. Umiker W. Accuracy of cytologic diagnosis of cancer of the urinary tract. Acta Cytol 1964; 8:189-193.

23. McDonald JR. Exfoliative cytology in genitourinary and pul- monary diseases. Am J Clin Path01 1954; 24:684-687.

24. Pedio G. Clinical applications of urinary cytology with special reference to the diagnosis of urinary tract neoplasms. In: Lubec G , ed. Non-lnvasive Diagnosis of Kidney Disease. Basel: Karger AG, 1983;

25. Bennington J, Beckwith JB. Tumors of the kidney renal pelvis and ureter. In: Atlas of Tumor Pathology, Series 2, Fascicle 12. Wash- ington DC Armed Forces Institute of Pathology, 1975; 127.

in the management of bladder carcinoma. Urol Res 1970 6: 197-200.

3. Philadelphia, Toronto, J. B. Lippincott Co., 1979; 71 1-817.

Soc Med 1976; 69:496-498.

of the urinary tract. J Urol 1947; 57:375-379.

SUPPI 115~3-75.

1963; 7:239-244.

258-277.

26. Evans JA. Halpern M, Finby N. Diagnosis of kidney cancer: An analysis of 100 consecutive cases. JAMA 1961; 175:201-203.

27. Park CH, Britisch C, Uson AC, Veenema RJ. Reliability of positive exfoliative cytologic study of the urine in urinary tract malignancy. J Urol 1969; 102:91-92.

28. Hajdu SI, Savino A, Hajdu EO, Koss LG. Cytologic diagnosis of renal cell carcinoma with the aid of fat stain. Acfa Cytol 1971; 15:31- 33.

29. Milsten R, Frable WJ, Texter JH Jr, Paxson L. Evaluation oflipid stain in renal neoplasms as adjunct to routine exfoliative cytology. J Uroll973; 110169-171.

30. Mount BM, Curtis M, Marshall K, Husk M. Cytologic diagnosis of renal cell carcinoma: Reassessment of fat-soluble staining. Urology 1973; 2421425.

3 I . Foot NC, Papanicolaou GN, Holmquist ND, Seybolt JF. Exfoli- ative cytology in urinary sediments: Review of 2,829 cases. Cancer 1958; 11:127-137.

32. De Voogt HJ, Rathert P, Bayer-Boon ME. Urinary Cytology. Berlin: Springer-Verlag, 1977; 39.

33. Lang ED. Roentgenographic assessment of asymptomatic renal lesions. Radiology 1973; 109:257-269.

34. Lang EK. Asymptomatic space occupying lesions of the kidney: A programmed sequential approach and its impact on quality and cost of health care. South Med J 1977; 70:277-285.

35. Clayman RV, Williams RD, Fraley EE. Current concepts in cancer: The pursuit of the renal mass. N Engl J Med 1979; 300:72-74.

36. Lang EK. Roentgenologic approach to the diagnosis and management of cystic lesions of the kidney: Is cyst exploration mandatory? Urol Clin North Am 1980; 7:677-688.

American Cancer Society Research Development Program in Clinical Investigation

The accelerated pace of cancer research and the need for constant peer surveillance for expeditious translation into strategies for diagnosis and treat- ment suggests that support of the scientifically prepared physician will be critical for ultimate cancer control. To strengthen programs in clinical investigation at both the basic and

technical transfer or “cancer control” levels, the American Cancer Society has extended its Research Development Program in this direction. For such short- term grants the priorities are as follows: peer review (hopefully within 3 months), merit. relevance, urgency, and a direct budget of SS0,OOO or less for I2 to I8 months. All other standard characteristics of ACS grants apply. Write or call for a brochure that describes this funding mechanism.

Application can be made at any time with the Research Department in care of Dr. S. B. Gusberg, Special Consultant to the Departments of Medical Affairs and Research, or Dr. Frank J. Rauscher Jr, Senior Vice President for Research, American Cancer Society, Inc.. 777 Third Avenue, New York. New York 10017.

Documents

Urine cytology in the detection of renal adenocarcinoma