REVIEW ARTICLE
Papillary Transitional-Cell Carcinoma of the Upper Urinary Tract: A Cytological Review Vaidehi Kannan,
M.D.
Urinary cytology is a well-accepted diagnostic proceafure f o r bladder carcinomas but it is utilized less frequefiitly f o r diagnosis of upper urinary tract tumors. Accurate diagnosis depends on a suitable specimen as well as knowledge of diagnostic traps. This review article with several case studies emphasizes the various techniques used to obtain optimal samples, correct interpretative criteria, and diagnostic pitfalls. Technological advances f o r objective grading and predicting the biological behavior of tumors are also discussed Diagn Cytopathol 1990;6:204-209. Key Words: Urinary cytology; Upper urinary tract tumors; Selective sampling; Flow cytometry; Cytomorphology
Upper urinary tract tumors constitute approximately 5%-8% of all the urothelial cancers, and the majority of them are papillary transitional-cell carcinomas.'*' These upper urinary tract carcinomas can pose diagnostic problems for clinicians and radiologists. With recent developments, conservative surgical therapy is possible in selected cases of low-grade, low-stage tumors. Therefore, accurate diagnosis and grading by means of cytologic examination can play a major role in therapeutic decision^.'.^^ All transitional-cell carcinomas, regardless of origin, are morphologically similar. For bladder lesions, urine cytology is a well-accepted procedure. The low sensitivity for diagnosis of upper urinary tract carcinomas by urine cytology can be alleviated if representative samples are obtained by appropriate te~hniques.'~'-''
Historical Perspective In 1864, Sanders from the Mayo Clinic was proba.bly the first to describe bladder cells exfoliated in urine.I2 Ferguson in 1892 and Parmenter in 1925 advocated examinaReceived February 5,1990. Accepted March 2, 1990. From the Departments of Pathology, Lankenau Hospital and Jefferson Medical College, Philadelphia, PA. Address reprint requests to Vaidehi Kannan, M.D., Lankenau Hospital, Lancaster and City Line Aves, Philadelphia, PA 19151.
204
Diagnostic Cytopathology. Vol6, N o 3
tion of urinary sediment as the best means of diagnosing bladder cancer.12 Nonetheless, interest in urinary cytology was largely dormant until Papanicolaou established criteria for interpretation of transitional-cell carcinoma (TCC) in voided urine in 1945. Initial enthusiasm was dampened by reports of low sensitivity as well as falsepositive reports. Crabbe, however, emphasized that, in patients with so-called false-positive reports by urine cytology, tumor often became apparent within 5 yr.13 A significant contribution was the use of membrane filter as reported by Solomon in 1958, which increased the cellular yield and, correspondingly, the accuracy rate. Gradually, therefore, urine cytology has become widely accepted as a routine laboratory procedure.
Specimen Collection Cytologic studies on upper urinary tract lesions can be performed on voided or catheterized urine and brush or lavage specimens. Though voided urine samples are less sensitive for the detection of cancer in the upper urinary t r a ~ t , ' ~ ~the , ~ technique ~ ' ~ , ' ~ of transurethral ureteropyeloscopy allows direct visualization and direct sampling from upper urinary tract.6
Urine Collection Random voided samples, preferably after hydration, are recommended for diagnosis of bladder carcinoma. Marked degenerative changes, however, are often noted in the urine from cells exfoliated from renal pelvis and ureter. Thus, voided urine samples give a high false-negative rate. Selective catheterization can obtain samples from the involved area. By this method and with the use of the membrane filter technique, 13 of 16 cases of TCC of renal pelvis were successfully diagnosed by Eriksson and Johan~son.'~ This method emphasizes the importance of selective sampling for accurate diagnosis. 0 1990 WILEY-LISS. INC.
UPPER URINARY TRACT TUMORS
Brush Cytology In 1973, Gill et al. developed a brush technique for direct sampling and used it successfully in more than 40 upper urinary tract lesions? with an accuracy rate of 78%. An open-tipped ureteral catheter is positioned adjacent to the lesion by conventional retrograde urography. Then a small brush with nylon bristles is passed through the ureteral catheter for direct brushing of the lesion. The cells entrapped on the bristles can be easily retrieved. The good preservation of cells by this technique facilitates a very high accuracy rate.''10 Bibbo et al. diagnosed seven of eight cases by this method; they stressed the absence of cell distortion and false-negative reports, as well as the lack of complications using this technique.'
Lavage Cytology As an alternative to brush cytology for the pre-operative diagnosis of TCC, lavage can be performed. Leistenschneider and Nagel described its use on 101 patients with a sensitivity of 80.5%.16 In lavage, after drainage of the contrast medium from retrograde pyelography, 5-7 ml of physiologic salt solution is injected into the ureter or renal pelvis and collected in a test tube. This procedure is repeated three times for an adequate sampling. An additional 10 ml of bladder content is collected separately after each lavage. Smears are made from the centrifuged specimen, which is fixed and stained according to Papanicolaou method. The advantages of lavage include a wide area of sampling, including poorly accessible areas such as terminal calyces and calyceal neck, which facilitates the detection of occult tumors. Proponents of this method also site its ease and cost-effectiveness compared with the brush cytology method. In our experience, brush and lavage are complementary procedures.
Percutaneous Translumbar Approach With the advent of fine-needle aspiration biopsy, Lang et al. introduced the percutaneous translumbar approach under local anesthesia to selectively brush the renal pelvis and calyceal systems.17An 18-gauge, thin-walled needle, 15-20 cm in length, is used. At the level of L2-3 in the posterior axillary line, with the patient in the prone position, the needle is introduced and the renal pelvis is penetrated under fluoroscopic control. The urine is evacuated and the pelvis is washed with saline. Using air or nitrous oxide, pyelography is performed. A WillsonEskridge nylon bronchial brush mounted on a guidewire deflector system is introduced through the sheath and selectively advanced to brush into the superior, mid, and inferior calyceal regions, renal pelvis, or ureter. After the brush is advanced past the lesion, the tip of the guidewire is deflected further to ensure maximum engagement of the brush when retracting the wire. After each pass,
smears are taken, fixed immediately in 95% alcohol, and stained with Papanicolaou stain. The detached brush is immersed in 10% buffered formalin, and the tissue fragments are teased out for histological preparations. By this method, multiple samplings can be taken from different areas. This invasive procedure, however, has gained little popularity.
Cytologic Findings The cytologic features for evaluation of TCC are similar throughout the genitourinary tract by any of these procedures. They include (1) papillary fragments; (2) an altered nuclear/cytoplasmic (N/C) ratio in favor of the nucleus; (3) nuclear hyperchromatia with coarsely clumped chromatin; (4)nuclear membrane irregularity; (5) macronucleoli; and (6) tumor diathesis. Malignant cells from low-grade tumors may display only minimal morphologic deviations from normal cells and must be correctly identified. As the tumor grade increases, changes become more pronounced. The low-grade papillary tumors shed cells in large papillary fragments, sometimes with a fibrovascular core. These fragments have an architecturally crowded appearance with peripheral pallisading of cells. The cells toward the center overlap, with nuclear crowding. Smaller groups form cell balls with clusters with nuclear overlapping. The individual cells contain moderate amounts of cytoplasm, better appreciated in the peripheral cells with peripheral nuclei (Fig. 1). With increasing grades of the tumor, increasingly numerous small dyshesive groups and single cells are noted. These single cells are usually elongated with moderate amounts of cytoplasm. Nuclear abnormalities include increases in the N / C ratio, nuclear membrane irregularity, coarse chromatin with uneven distribution, and variable numbers of macronucleoli. The relatively clean background of the lowgrade tumors shifts to a tumor diathesis composed of necrotic debris, blood, and degenerating neutrophils. The cytohistologic correlation utilizing these criteria approximates 74%,18 with a somewhat lower sensitivity (30%-40%) for low-grade TCC and a higher sensitivity (90%-95%) for poorly differentiated tumors. With experience, subtle changes can be recognized; in fact, Murphy et al. reported positive cytology in 62% of cases of low-grade TCC.19 Table I compares the cytomorphology of low-grade and high-grade TCCs.
Diagnostic Pitfalls Factors that can confuse the smear pattern include (1) catheter artifact; (2) lithiasis artifact; (3) inflammatory atypia; and (4) degenerative changes. In Table 11, the cellular alterations caused by these conditions are summarized and compared with those of papillary TCC. Cell dislodgement by the catheter tip may form papilDiagnostic Cytopathology,Vol6, No 3
205
VAIDEHI KANNAN Table 11. Comparison of Similar Appearing Entities
Cvtoloaic findinm
Papillary clusters Increased N / C ratio Coarse chromatin Nuclear membrane irregularity
Catheter artifact and lithiasis
Injammation and deaeneration
+
*
-
-
-
-
-
-
Papillary TCC
+ + +
Diathesis, increased nuclear size, and hyperchromasia are usually present in TCC and may be present in other two conditions as well.
cells with dark nuclei and occasional macronucleoli. The neutrophils, blood, and debris may be mistaken for a tumor diathesis. The nuclei, however, appear uniformly dark (smudged). Additionally, the coarsely clumped chromatin with parachromatin clearing, which is associated with carcinoma, is not seen with inflammation.20
Case Examples The following cases illustrate the importance of correlation of cytomorphologic findings and clinical presentation. They are summarized in Table 111.
Case I Fig. 1. Papillary fragment of malignant tumor cells from a low-grade carcinoma of the ureter: brush cytology. Enlarged nuclei with coarse hyperchromasia and slight nuclear membrane irregularity are seen. (Papanicolaou stain, x400).
lary groups, which may be misinterpreted as low-grade papillary carcinoma. The relatively small nuclei, however, are vesicular and display finely granular chromatin (Fig. 2). Lithiasis can also induce papillary clusters and isolated cells. Additionally, there may be some anisocytosis and anisonucleosis. Again, however, hyperchromasia and coarseness of chromatin are usually absent. Atypia and degeneration associated with inflammatory conditions can give rise to isolated and small groups of Table 1. Comparative Cytomorphology of Low-Grade and High-Grade TCC Low-grade
Cells Arrangement Size Cytoplasm N/C ratio Nuclei Size Borders Chromatin Nucleoli
206
High-grade
Papillae; palisades; loose clusters Uniform Homogenous Minimal change
Isolated and loose clusters Pleomorphic Variable Marked change
Enlarged Minimal irregularity Fine; regular Small or absent
Enlarged Irregular Coarse; irregular Variable
Diagnostic Cytopathology, Vol6, No 3
A 64-yr-old woman had renal colic. Voided urine cytology showed cells highly suspicious for TCC. Cystoscopy, however, was negative. Intravenous pyelogram (IVP) demonstrated a filling defect in the right renal pelvis. Selective washings from this area revealed exfoliated malignant cells (Fig. 3). Upon right nephro-ureterectomy, there was a papillary TCC, grade 1-11, confined to the mucosa in right ureteropelvic area (Fig. 4).
Case 2 A 70-yr-old woman presented with cardiac problems. A urine specimen, accidentally sent to the cytology laboratory, revealed markedly atypical cells highly suggestive of carcinoma. Cystoscopy was negative. Due to persistent abnormal cells exfoliated in voided urine samples, the patient had an IVP; this revealed a filling defect in the left lower ureter. Brushings from this area revealed cells consistent with a low-grade papillary TCC. The resected grade 1-11 neoplasm was confined to mucosa.
Case 3 A 70-yr-old woman was investigated for microscopic hematuria. In the voided urine were cells highly suggestive of carcinoma. Although cystoscopy was negative, IVP suggested a right-upper-pole mass, which was not substantiated by ultrasound examination. Based on the cytologic findings, however, urethroscopy and pyeloscopy were performed. At the upper pole infundibulum leading to the posterior calyx, there was a raised red irregular lesion. Brushings, washings, and biopsy from this area were
UPPER URINARY TRACT TUMORS
Fig. 2. Catheter artifact, showing a papillary group of cells with small uniform nuclei, fine chromatin pattern, and regular nuclear membranes (Papanicolaou stain, x250).
consistent with a papillary TCC. The patient underwent right nephro-ureterectomy for TCC, grade 11. Case 4 A 71-yr-old man presented because of chest pain and shortness of breath. Ultrasound, preformed because of upper abdominal discomfort, revealed a right renal pelvic mass. Although voided urine contained only atypical but degenerated cells, cystoretrograde brushing cytology showed papillary TCC, confirmed by biopsy. All four patients had negative cystoscopic findings. Due to persistent abnormal urinary cytology, further investigation with selective cytologic sampling resulted in the detection and treatment of neoplasms of the upper urinary tract.
New Frontiers The development of laser and computer technology has led to the new fields of clinical flow cytometry and nuclear morphometry. These new tools offer a means of objective grading by quantitative cell analysis in urothelial tumors. They a r e also helpful in providing prognostic informati~n.~J--~~ Flow cytometry (FCM) allows high-speed (3,000 cells/ sec) automated analysis and sorting of cells in suspension that pass through a laser light source. Cytophotometry is a microscope-based system to analyze the optical parame-
Fig. 3. Group of exfoliated malignant transitional epithelial cells from renal washings showing nuclear enlargement, coarse chromatin, and slightly irregular nuclear membrane (Papanicolaou, x400).
ters of cells on a glass slide. Nuclear morphometry is part of it. Both methods have advantages. But by FCM, more than 10,000 cells can be studied in less than 5 min, whereas nuclear morphometry is time-consuming and limited to 200-500 cells/slide. The basic principle of FCM that is useful in evaluating urothelial carcinoma is that normal cells are euploid and have a known amount of DNA, whereas malignant cells tend to be aneuploid. Cells are first labeled with probes specific for a variety of cell constituents. These probes are labeled with dyes that bind stoichiometrically to the cells and fluoresce when passed through a laser beam. By electronics, the light energy is converted into digital signals. The data are then analyzed by the computer. Thus far, much work with this technique has utilized bladder washings; however, it can also be applied to
TABLE 111. Case Examples Cytologic diagnosis Case numbers
Tumor site
1 2
Ureteropelvic junction Mid-ureter Calyx, upper pole Pelvis
3
4
Genitourinary symptoms
Voided urine
Renal Colic None Hematuria None
Atypical cells Suspicious Suspicious Atypical cells
Selective catheter urine Carcinoma Unsatisfactory Probable carcinoma -
Brushings -
Carcinoma Carcinoma Carcinoma
In all four cases, histology showed low-grade TCC. In cases 1-3, cystoscopy results were normal; cystoscopy was not performed in case 4. Diagnostic Cytopathology, Vol6. No 3
207
VAIDEHI KANNAN
Conclusions Though nephro-ureterectomy has been recommended for all upper urinary tract cancers,27 local resection for well-differentiated, localized tumors has now become a therapeutic ~ p t i o n . ~ This , ~ , ' optional ~ treatment depends on early, accurate preoperative diagnosis. Highman, in a series of 54 patients with upper urinary tract tumors, used urine cytology successfully for diagnosis'8; however, its value is contr~vestral'~"~ because of degeneration artifacts. The advanced techniques of urological endoscopy has made it possible to get direct brushings and washings from the upper urinary tract, and these samples provide well-preserved cells.4f6However, it behooves us to watch out for the diagnostic pitfalls in interpreting the cells obtained by these techniques. New technological developments in this field include DNA flow cytometry with ploidy analysis and cytomorphometry. Where indicated, these techniques are helpful to diagnose and consistently grade TCCs, and they also help for p r o g n ~ s i s . ~ ' - ~ ~ It is well-known that good clinical correlation with the radiographic impression as well as the necessary information about the technique of sampling form important and necessary parts of the final diagnosis. As has been previously pointed out, the success of cytological diagnosis in this difficult area depends on close cooperation among the clinician, radiologist, and cytopathol~gist.'~.'~
Acknowledgment The author thanks Tilde S. Kline, M.D., for editorial assistance and Carmen Goley for secretarial help. Fig. 4. Corresponding tissue section to that in Fig. 3, showing low-grade papillary TCC (H&E, x 100).
References
washings and brushings from throughout the urinary tract. In nuclear morphometry, the glass slides are pirojected on a graphic tablet and the nuclear areas, m,agnified x 100, are measured. After computation of data obtained from several cells, a mean value is derived. Tumors with a mean nuclear area of more than 95 p2 have been categorized as high grade with aggressive behavior.22 There is good correlation between FCM and urine cytology with sensitivities of 74% and 85%, respectively, on bladder washings.25Sensitivity increases to 9'3%with the combination of these two m ~ d a l i t i e s .DNA ~ ~ , ~ploidy ~ also has prognostic significance because of a significantly higher recurrence rate in patients with aneuploid. tumors than in those with diploid carcinomas.22 Thus, ploidy and mean nuclear area analysis of TCCs offer insight into their biologic nature, facilitating therapeutic options. Aneuploid tumors and those with a high mean nuclear area could be considered for aggressive treatment to prevent relapse.22
1. Seldenrijk CH, Verheggen WJHM, Velduhizen RW, et al. Cytomorphometry and cytology in the diagnosis of transitional cell carcinoma of the upper urinary tract. Acta Cytol 1987;31:137-41. 2. Wagle DG, Moore RH, Murphy GP. Primary carcinoma of the renal pelvis. Cancer 1974;33:1642-8. 3. Blute ML, Segura JW, Patterson DE, Benson RC, Jr, Zincke H. Impact of endourology on diagnosis and management of upper urinary tract urothelial cancer. J Urol 1989;14:1298-301. 4. Ooms ECM, Kurver PHJ, Velduizen RW, Alons CL, Boon ME. Morphometric grading of bladder tumors in comparison with histologic grading by pathologists. Human Pathol 1983;14:144-50. 5 . Hoffman JL, Bagley DH, Lyon ES, et al. Endoscopic diagnosis and treatment of upper-tract urothelial tumors. A preliminary report. Cancer 1985;SS:1422-8. 6. Silverberg SG. Principles and practice of surgical pathology. 2nd ed. New York Churchill & Livingstone, 1990:1504-8. 7. Koss LG. Diagnostic cytology & its histopathologic basis. 3rd ed. Philadelphia: Lippincott, 1979:784-6. 8. Bibbo M, Gill WB, Norris MJ, et al. Retrograde brushing as a diagnostic procedure of ureteral, renal pelvic and renal calyceal lesions. A preliminary report. Acta Cytol 1974;18:13741. 9. Gill WB, Lu CT, Thomsen S. Retrograde brushing: a new technique for obtaining histologic and cytologic material from ureteral, renal pelvic and renal calyceal lesions. J Urol 1973;109:573-8. 10. Grill WB, Lu CT, Bibbo M. Retrograde ureteral brushing. Urology 1978;12:279-83.
208
Diagnostic Cytopathology.Vol6, No 3
UPPER URINARY TRACT TUMORS 11. Blute RD, Gittes RR, Gittes RF. Renal brush biopsy: survey of indications, technique and results. J Urol 1981;12:146-9. 12. Lewis RW, Jackson AC Jr, Murphy WM, et al. Cytology in the diagnosis and follow-up of transitional cell carcinoma of the urothelium: a review with a case series. J Urol 1976;116:43-6. 13. Crabbe JGS. Cytological methods of control for bladder tumors of occupational origin. In: Wallace DM, ed. Tumours of the bladder. Vol. 2. Edinburgh: E & S Livingstone, 1959:56-76. 14. Eriksson 0, Johansson S. Urothelial neoplasms of the upper urinary tract: a correlation between cytologic and histologic findings in 43 patients with urothelial neoplasms of the renal pelvis of ureter. Acta Cytol 1976;20:20-5. 15. Zincke H, Aguilo JJ, Farrow GM, et al. Significance of urinary cytology in the early detection of transitional cell cancer of the upper urinary tract. J Urol 1976;116:781-3. 16. Leistenschneider W, Nagel R. Lavage cytology of the renal pelvis and ureter with special reference to tumors. J Urol 1980;124:597-600. 17. Lang EK, Alexander R, Barnett T, et al. Brush biopsy of pyelocalyceal lesions via a percutaneous translumbar approach. Radiology 1978;129:623-7. 18. Highman WJ. Transitional cell carcinoma of the upper urinary tract: a histological and cytopathological study. J Clin Pathol 1986;39: 297-305. 19. Murphy W, Soloway MS, Jukkola AF, Crabtree NM, Ford KS. Urinary cytology and bladder cancer. The cellular features of transitional cell carcinoma. Cancer 1984;53:1555-65.
20. Koshikawa T, Leyh H, Schenck U. Difficulties in evaluating urinary specimens after local mitomycin therapy of bladder cancer. Diagn Cytopathol 1989;5:117-21. 21. Koss LG, Czernicak B, Herz F, Wersto RP. Flow cytometric measurements of DNA and other cell components in human tumors: a critical appraisal. Hum Pathol 1989;20:528-48. 22. Blomjous ECM, Schipper NW, Baak JPA, et al. The value of morphometry and DNA Row cytornetry in addition to classic prognosticators in superficial urinary bladder carcinoma. Am J Clin Pathol 1989;91:243-8. 23. Merkel DE, Dressler LG, McGuire WL. Flow cytometry, cellular DNA content, and prognosis in human malignancy. J Clin Oncol 1987;5:1690-703. 24. Tribukait B, Gustafson H, Esposti P. Ploidy and proliferation in human bladder tumors as measured by flow cytometeric DNAanalysis and its relation to histopathology and cytology. Cancer 1979;43: 1742-5 1. 25. Gander T. Application of flow cytometry. DNA quantitation in tumor pathology. J Clin Irnmunoassay 1989;12:30-5. 26. Klein FA, White FKH. Flow cytometry deoxyribonucleic acid determinations and cytology of bladder washings: practical experience. J Urol 1988;139:275-8. 27. Auld CD, Grigor KM, Fowler JW. Histopathological review of transitional cell carcinoma of the upper urinary tract. Br J Urol 1984;56:485-9.
Diagnostic Cytopathology.Vo16, No 3
209
Papillary Transitional-Cell Carcinoma of the Upper Urinary Tract: A Cytological Review Vaidehi Kannan,
M.D.
Urinary cytology is a well-accepted diagnostic proceafure f o r bladder carcinomas but it is utilized less frequefiitly f o r diagnosis of upper urinary tract tumors. Accurate diagnosis depends on a suitable specimen as well as knowledge of diagnostic traps. This review article with several case studies emphasizes the various techniques used to obtain optimal samples, correct interpretative criteria, and diagnostic pitfalls. Technological advances f o r objective grading and predicting the biological behavior of tumors are also discussed Diagn Cytopathol 1990;6:204-209. Key Words: Urinary cytology; Upper urinary tract tumors; Selective sampling; Flow cytometry; Cytomorphology
Upper urinary tract tumors constitute approximately 5%-8% of all the urothelial cancers, and the majority of them are papillary transitional-cell carcinomas.'*' These upper urinary tract carcinomas can pose diagnostic problems for clinicians and radiologists. With recent developments, conservative surgical therapy is possible in selected cases of low-grade, low-stage tumors. Therefore, accurate diagnosis and grading by means of cytologic examination can play a major role in therapeutic decision^.'.^^ All transitional-cell carcinomas, regardless of origin, are morphologically similar. For bladder lesions, urine cytology is a well-accepted procedure. The low sensitivity for diagnosis of upper urinary tract carcinomas by urine cytology can be alleviated if representative samples are obtained by appropriate te~hniques.'~'-''
Historical Perspective In 1864, Sanders from the Mayo Clinic was proba.bly the first to describe bladder cells exfoliated in urine.I2 Ferguson in 1892 and Parmenter in 1925 advocated examinaReceived February 5,1990. Accepted March 2, 1990. From the Departments of Pathology, Lankenau Hospital and Jefferson Medical College, Philadelphia, PA. Address reprint requests to Vaidehi Kannan, M.D., Lankenau Hospital, Lancaster and City Line Aves, Philadelphia, PA 19151.
204
Diagnostic Cytopathology. Vol6, N o 3
tion of urinary sediment as the best means of diagnosing bladder cancer.12 Nonetheless, interest in urinary cytology was largely dormant until Papanicolaou established criteria for interpretation of transitional-cell carcinoma (TCC) in voided urine in 1945. Initial enthusiasm was dampened by reports of low sensitivity as well as falsepositive reports. Crabbe, however, emphasized that, in patients with so-called false-positive reports by urine cytology, tumor often became apparent within 5 yr.13 A significant contribution was the use of membrane filter as reported by Solomon in 1958, which increased the cellular yield and, correspondingly, the accuracy rate. Gradually, therefore, urine cytology has become widely accepted as a routine laboratory procedure.
Specimen Collection Cytologic studies on upper urinary tract lesions can be performed on voided or catheterized urine and brush or lavage specimens. Though voided urine samples are less sensitive for the detection of cancer in the upper urinary t r a ~ t , ' ~ ~the , ~ technique ~ ' ~ , ' ~ of transurethral ureteropyeloscopy allows direct visualization and direct sampling from upper urinary tract.6
Urine Collection Random voided samples, preferably after hydration, are recommended for diagnosis of bladder carcinoma. Marked degenerative changes, however, are often noted in the urine from cells exfoliated from renal pelvis and ureter. Thus, voided urine samples give a high false-negative rate. Selective catheterization can obtain samples from the involved area. By this method and with the use of the membrane filter technique, 13 of 16 cases of TCC of renal pelvis were successfully diagnosed by Eriksson and Johan~son.'~ This method emphasizes the importance of selective sampling for accurate diagnosis. 0 1990 WILEY-LISS. INC.
UPPER URINARY TRACT TUMORS
Brush Cytology In 1973, Gill et al. developed a brush technique for direct sampling and used it successfully in more than 40 upper urinary tract lesions? with an accuracy rate of 78%. An open-tipped ureteral catheter is positioned adjacent to the lesion by conventional retrograde urography. Then a small brush with nylon bristles is passed through the ureteral catheter for direct brushing of the lesion. The cells entrapped on the bristles can be easily retrieved. The good preservation of cells by this technique facilitates a very high accuracy rate.''10 Bibbo et al. diagnosed seven of eight cases by this method; they stressed the absence of cell distortion and false-negative reports, as well as the lack of complications using this technique.'
Lavage Cytology As an alternative to brush cytology for the pre-operative diagnosis of TCC, lavage can be performed. Leistenschneider and Nagel described its use on 101 patients with a sensitivity of 80.5%.16 In lavage, after drainage of the contrast medium from retrograde pyelography, 5-7 ml of physiologic salt solution is injected into the ureter or renal pelvis and collected in a test tube. This procedure is repeated three times for an adequate sampling. An additional 10 ml of bladder content is collected separately after each lavage. Smears are made from the centrifuged specimen, which is fixed and stained according to Papanicolaou method. The advantages of lavage include a wide area of sampling, including poorly accessible areas such as terminal calyces and calyceal neck, which facilitates the detection of occult tumors. Proponents of this method also site its ease and cost-effectiveness compared with the brush cytology method. In our experience, brush and lavage are complementary procedures.
Percutaneous Translumbar Approach With the advent of fine-needle aspiration biopsy, Lang et al. introduced the percutaneous translumbar approach under local anesthesia to selectively brush the renal pelvis and calyceal systems.17An 18-gauge, thin-walled needle, 15-20 cm in length, is used. At the level of L2-3 in the posterior axillary line, with the patient in the prone position, the needle is introduced and the renal pelvis is penetrated under fluoroscopic control. The urine is evacuated and the pelvis is washed with saline. Using air or nitrous oxide, pyelography is performed. A WillsonEskridge nylon bronchial brush mounted on a guidewire deflector system is introduced through the sheath and selectively advanced to brush into the superior, mid, and inferior calyceal regions, renal pelvis, or ureter. After the brush is advanced past the lesion, the tip of the guidewire is deflected further to ensure maximum engagement of the brush when retracting the wire. After each pass,
smears are taken, fixed immediately in 95% alcohol, and stained with Papanicolaou stain. The detached brush is immersed in 10% buffered formalin, and the tissue fragments are teased out for histological preparations. By this method, multiple samplings can be taken from different areas. This invasive procedure, however, has gained little popularity.
Cytologic Findings The cytologic features for evaluation of TCC are similar throughout the genitourinary tract by any of these procedures. They include (1) papillary fragments; (2) an altered nuclear/cytoplasmic (N/C) ratio in favor of the nucleus; (3) nuclear hyperchromatia with coarsely clumped chromatin; (4)nuclear membrane irregularity; (5) macronucleoli; and (6) tumor diathesis. Malignant cells from low-grade tumors may display only minimal morphologic deviations from normal cells and must be correctly identified. As the tumor grade increases, changes become more pronounced. The low-grade papillary tumors shed cells in large papillary fragments, sometimes with a fibrovascular core. These fragments have an architecturally crowded appearance with peripheral pallisading of cells. The cells toward the center overlap, with nuclear crowding. Smaller groups form cell balls with clusters with nuclear overlapping. The individual cells contain moderate amounts of cytoplasm, better appreciated in the peripheral cells with peripheral nuclei (Fig. 1). With increasing grades of the tumor, increasingly numerous small dyshesive groups and single cells are noted. These single cells are usually elongated with moderate amounts of cytoplasm. Nuclear abnormalities include increases in the N / C ratio, nuclear membrane irregularity, coarse chromatin with uneven distribution, and variable numbers of macronucleoli. The relatively clean background of the lowgrade tumors shifts to a tumor diathesis composed of necrotic debris, blood, and degenerating neutrophils. The cytohistologic correlation utilizing these criteria approximates 74%,18 with a somewhat lower sensitivity (30%-40%) for low-grade TCC and a higher sensitivity (90%-95%) for poorly differentiated tumors. With experience, subtle changes can be recognized; in fact, Murphy et al. reported positive cytology in 62% of cases of low-grade TCC.19 Table I compares the cytomorphology of low-grade and high-grade TCCs.
Diagnostic Pitfalls Factors that can confuse the smear pattern include (1) catheter artifact; (2) lithiasis artifact; (3) inflammatory atypia; and (4) degenerative changes. In Table 11, the cellular alterations caused by these conditions are summarized and compared with those of papillary TCC. Cell dislodgement by the catheter tip may form papilDiagnostic Cytopathology,Vol6, No 3
205
VAIDEHI KANNAN Table 11. Comparison of Similar Appearing Entities
Cvtoloaic findinm
Papillary clusters Increased N / C ratio Coarse chromatin Nuclear membrane irregularity
Catheter artifact and lithiasis
Injammation and deaeneration
+
*
-
-
-
-
-
-
Papillary TCC
+ + +
Diathesis, increased nuclear size, and hyperchromasia are usually present in TCC and may be present in other two conditions as well.
cells with dark nuclei and occasional macronucleoli. The neutrophils, blood, and debris may be mistaken for a tumor diathesis. The nuclei, however, appear uniformly dark (smudged). Additionally, the coarsely clumped chromatin with parachromatin clearing, which is associated with carcinoma, is not seen with inflammation.20
Case Examples The following cases illustrate the importance of correlation of cytomorphologic findings and clinical presentation. They are summarized in Table 111.
Case I Fig. 1. Papillary fragment of malignant tumor cells from a low-grade carcinoma of the ureter: brush cytology. Enlarged nuclei with coarse hyperchromasia and slight nuclear membrane irregularity are seen. (Papanicolaou stain, x400).
lary groups, which may be misinterpreted as low-grade papillary carcinoma. The relatively small nuclei, however, are vesicular and display finely granular chromatin (Fig. 2). Lithiasis can also induce papillary clusters and isolated cells. Additionally, there may be some anisocytosis and anisonucleosis. Again, however, hyperchromasia and coarseness of chromatin are usually absent. Atypia and degeneration associated with inflammatory conditions can give rise to isolated and small groups of Table 1. Comparative Cytomorphology of Low-Grade and High-Grade TCC Low-grade
Cells Arrangement Size Cytoplasm N/C ratio Nuclei Size Borders Chromatin Nucleoli
206
High-grade
Papillae; palisades; loose clusters Uniform Homogenous Minimal change
Isolated and loose clusters Pleomorphic Variable Marked change
Enlarged Minimal irregularity Fine; regular Small or absent
Enlarged Irregular Coarse; irregular Variable
Diagnostic Cytopathology, Vol6, No 3
A 64-yr-old woman had renal colic. Voided urine cytology showed cells highly suspicious for TCC. Cystoscopy, however, was negative. Intravenous pyelogram (IVP) demonstrated a filling defect in the right renal pelvis. Selective washings from this area revealed exfoliated malignant cells (Fig. 3). Upon right nephro-ureterectomy, there was a papillary TCC, grade 1-11, confined to the mucosa in right ureteropelvic area (Fig. 4).
Case 2 A 70-yr-old woman presented with cardiac problems. A urine specimen, accidentally sent to the cytology laboratory, revealed markedly atypical cells highly suggestive of carcinoma. Cystoscopy was negative. Due to persistent abnormal cells exfoliated in voided urine samples, the patient had an IVP; this revealed a filling defect in the left lower ureter. Brushings from this area revealed cells consistent with a low-grade papillary TCC. The resected grade 1-11 neoplasm was confined to mucosa.
Case 3 A 70-yr-old woman was investigated for microscopic hematuria. In the voided urine were cells highly suggestive of carcinoma. Although cystoscopy was negative, IVP suggested a right-upper-pole mass, which was not substantiated by ultrasound examination. Based on the cytologic findings, however, urethroscopy and pyeloscopy were performed. At the upper pole infundibulum leading to the posterior calyx, there was a raised red irregular lesion. Brushings, washings, and biopsy from this area were
UPPER URINARY TRACT TUMORS
Fig. 2. Catheter artifact, showing a papillary group of cells with small uniform nuclei, fine chromatin pattern, and regular nuclear membranes (Papanicolaou stain, x250).
consistent with a papillary TCC. The patient underwent right nephro-ureterectomy for TCC, grade 11. Case 4 A 71-yr-old man presented because of chest pain and shortness of breath. Ultrasound, preformed because of upper abdominal discomfort, revealed a right renal pelvic mass. Although voided urine contained only atypical but degenerated cells, cystoretrograde brushing cytology showed papillary TCC, confirmed by biopsy. All four patients had negative cystoscopic findings. Due to persistent abnormal urinary cytology, further investigation with selective cytologic sampling resulted in the detection and treatment of neoplasms of the upper urinary tract.
New Frontiers The development of laser and computer technology has led to the new fields of clinical flow cytometry and nuclear morphometry. These new tools offer a means of objective grading by quantitative cell analysis in urothelial tumors. They a r e also helpful in providing prognostic informati~n.~J--~~ Flow cytometry (FCM) allows high-speed (3,000 cells/ sec) automated analysis and sorting of cells in suspension that pass through a laser light source. Cytophotometry is a microscope-based system to analyze the optical parame-
Fig. 3. Group of exfoliated malignant transitional epithelial cells from renal washings showing nuclear enlargement, coarse chromatin, and slightly irregular nuclear membrane (Papanicolaou, x400).
ters of cells on a glass slide. Nuclear morphometry is part of it. Both methods have advantages. But by FCM, more than 10,000 cells can be studied in less than 5 min, whereas nuclear morphometry is time-consuming and limited to 200-500 cells/slide. The basic principle of FCM that is useful in evaluating urothelial carcinoma is that normal cells are euploid and have a known amount of DNA, whereas malignant cells tend to be aneuploid. Cells are first labeled with probes specific for a variety of cell constituents. These probes are labeled with dyes that bind stoichiometrically to the cells and fluoresce when passed through a laser beam. By electronics, the light energy is converted into digital signals. The data are then analyzed by the computer. Thus far, much work with this technique has utilized bladder washings; however, it can also be applied to
TABLE 111. Case Examples Cytologic diagnosis Case numbers
Tumor site
1 2
Ureteropelvic junction Mid-ureter Calyx, upper pole Pelvis
3
4
Genitourinary symptoms
Voided urine
Renal Colic None Hematuria None
Atypical cells Suspicious Suspicious Atypical cells
Selective catheter urine Carcinoma Unsatisfactory Probable carcinoma -
Brushings -
Carcinoma Carcinoma Carcinoma
In all four cases, histology showed low-grade TCC. In cases 1-3, cystoscopy results were normal; cystoscopy was not performed in case 4. Diagnostic Cytopathology, Vol6. No 3
207
VAIDEHI KANNAN
Conclusions Though nephro-ureterectomy has been recommended for all upper urinary tract cancers,27 local resection for well-differentiated, localized tumors has now become a therapeutic ~ p t i o n . ~ This , ~ , ' optional ~ treatment depends on early, accurate preoperative diagnosis. Highman, in a series of 54 patients with upper urinary tract tumors, used urine cytology successfully for diagnosis'8; however, its value is contr~vestral'~"~ because of degeneration artifacts. The advanced techniques of urological endoscopy has made it possible to get direct brushings and washings from the upper urinary tract, and these samples provide well-preserved cells.4f6However, it behooves us to watch out for the diagnostic pitfalls in interpreting the cells obtained by these techniques. New technological developments in this field include DNA flow cytometry with ploidy analysis and cytomorphometry. Where indicated, these techniques are helpful to diagnose and consistently grade TCCs, and they also help for p r o g n ~ s i s . ~ ' - ~ ~ It is well-known that good clinical correlation with the radiographic impression as well as the necessary information about the technique of sampling form important and necessary parts of the final diagnosis. As has been previously pointed out, the success of cytological diagnosis in this difficult area depends on close cooperation among the clinician, radiologist, and cytopathol~gist.'~.'~
Acknowledgment The author thanks Tilde S. Kline, M.D., for editorial assistance and Carmen Goley for secretarial help. Fig. 4. Corresponding tissue section to that in Fig. 3, showing low-grade papillary TCC (H&E, x 100).
References
washings and brushings from throughout the urinary tract. In nuclear morphometry, the glass slides are pirojected on a graphic tablet and the nuclear areas, m,agnified x 100, are measured. After computation of data obtained from several cells, a mean value is derived. Tumors with a mean nuclear area of more than 95 p2 have been categorized as high grade with aggressive behavior.22 There is good correlation between FCM and urine cytology with sensitivities of 74% and 85%, respectively, on bladder washings.25Sensitivity increases to 9'3%with the combination of these two m ~ d a l i t i e s .DNA ~ ~ , ~ploidy ~ also has prognostic significance because of a significantly higher recurrence rate in patients with aneuploid. tumors than in those with diploid carcinomas.22 Thus, ploidy and mean nuclear area analysis of TCCs offer insight into their biologic nature, facilitating therapeutic options. Aneuploid tumors and those with a high mean nuclear area could be considered for aggressive treatment to prevent relapse.22
1. Seldenrijk CH, Verheggen WJHM, Velduhizen RW, et al. Cytomorphometry and cytology in the diagnosis of transitional cell carcinoma of the upper urinary tract. Acta Cytol 1987;31:137-41. 2. Wagle DG, Moore RH, Murphy GP. Primary carcinoma of the renal pelvis. Cancer 1974;33:1642-8. 3. Blute ML, Segura JW, Patterson DE, Benson RC, Jr, Zincke H. Impact of endourology on diagnosis and management of upper urinary tract urothelial cancer. J Urol 1989;14:1298-301. 4. Ooms ECM, Kurver PHJ, Velduizen RW, Alons CL, Boon ME. Morphometric grading of bladder tumors in comparison with histologic grading by pathologists. Human Pathol 1983;14:144-50. 5 . Hoffman JL, Bagley DH, Lyon ES, et al. Endoscopic diagnosis and treatment of upper-tract urothelial tumors. A preliminary report. Cancer 1985;SS:1422-8. 6. Silverberg SG. Principles and practice of surgical pathology. 2nd ed. New York Churchill & Livingstone, 1990:1504-8. 7. Koss LG. Diagnostic cytology & its histopathologic basis. 3rd ed. Philadelphia: Lippincott, 1979:784-6. 8. Bibbo M, Gill WB, Norris MJ, et al. Retrograde brushing as a diagnostic procedure of ureteral, renal pelvic and renal calyceal lesions. A preliminary report. Acta Cytol 1974;18:13741. 9. Gill WB, Lu CT, Thomsen S. Retrograde brushing: a new technique for obtaining histologic and cytologic material from ureteral, renal pelvic and renal calyceal lesions. J Urol 1973;109:573-8. 10. Grill WB, Lu CT, Bibbo M. Retrograde ureteral brushing. Urology 1978;12:279-83.
208
Diagnostic Cytopathology.Vol6, No 3
UPPER URINARY TRACT TUMORS 11. Blute RD, Gittes RR, Gittes RF. Renal brush biopsy: survey of indications, technique and results. J Urol 1981;12:146-9. 12. Lewis RW, Jackson AC Jr, Murphy WM, et al. Cytology in the diagnosis and follow-up of transitional cell carcinoma of the urothelium: a review with a case series. J Urol 1976;116:43-6. 13. Crabbe JGS. Cytological methods of control for bladder tumors of occupational origin. In: Wallace DM, ed. Tumours of the bladder. Vol. 2. Edinburgh: E & S Livingstone, 1959:56-76. 14. Eriksson 0, Johansson S. Urothelial neoplasms of the upper urinary tract: a correlation between cytologic and histologic findings in 43 patients with urothelial neoplasms of the renal pelvis of ureter. Acta Cytol 1976;20:20-5. 15. Zincke H, Aguilo JJ, Farrow GM, et al. Significance of urinary cytology in the early detection of transitional cell cancer of the upper urinary tract. J Urol 1976;116:781-3. 16. Leistenschneider W, Nagel R. Lavage cytology of the renal pelvis and ureter with special reference to tumors. J Urol 1980;124:597-600. 17. Lang EK, Alexander R, Barnett T, et al. Brush biopsy of pyelocalyceal lesions via a percutaneous translumbar approach. Radiology 1978;129:623-7. 18. Highman WJ. Transitional cell carcinoma of the upper urinary tract: a histological and cytopathological study. J Clin Pathol 1986;39: 297-305. 19. Murphy W, Soloway MS, Jukkola AF, Crabtree NM, Ford KS. Urinary cytology and bladder cancer. The cellular features of transitional cell carcinoma. Cancer 1984;53:1555-65.
20. Koshikawa T, Leyh H, Schenck U. Difficulties in evaluating urinary specimens after local mitomycin therapy of bladder cancer. Diagn Cytopathol 1989;5:117-21. 21. Koss LG, Czernicak B, Herz F, Wersto RP. Flow cytometric measurements of DNA and other cell components in human tumors: a critical appraisal. Hum Pathol 1989;20:528-48. 22. Blomjous ECM, Schipper NW, Baak JPA, et al. The value of morphometry and DNA Row cytornetry in addition to classic prognosticators in superficial urinary bladder carcinoma. Am J Clin Pathol 1989;91:243-8. 23. Merkel DE, Dressler LG, McGuire WL. Flow cytometry, cellular DNA content, and prognosis in human malignancy. J Clin Oncol 1987;5:1690-703. 24. Tribukait B, Gustafson H, Esposti P. Ploidy and proliferation in human bladder tumors as measured by flow cytometeric DNAanalysis and its relation to histopathology and cytology. Cancer 1979;43: 1742-5 1. 25. Gander T. Application of flow cytometry. DNA quantitation in tumor pathology. J Clin Irnmunoassay 1989;12:30-5. 26. Klein FA, White FKH. Flow cytometry deoxyribonucleic acid determinations and cytology of bladder washings: practical experience. J Urol 1988;139:275-8. 27. Auld CD, Grigor KM, Fowler JW. Histopathological review of transitional cell carcinoma of the upper urinary tract. Br J Urol 1984;56:485-9.
Diagnostic Cytopathology.Vo16, No 3
209
