Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Evidence for oligoclonality and tumor spread by intraluminal seeding in multifocal urothelial carcinomas of the upper and lower urinary tract


Multifocality and recurrence of urothelial carcinoma may result from either the field effect of carcinogens leading to oligoclonal tumors or monoclonal tumor spread. Previous molecular studies, favoring the monoclonality hypothesis, are mostly limited to the urinary bladder. We investigated genetic alterations in a total of 94 synchronous or metachronous multifocal tumors from 19 patients with at least one tumor both in the upper and lower urinary tract. Loss of heterozygosity (LOH) was determined using eight markers on chromosome 9 and one marker on 17p13 (p53). Microsatellite instability was investigated at six loci and protein expression of MSH2 and MLH1 was evaluated by immunohistochemistry. In addition, exons 5–9 of the p53 gene were sequenced. Deletions at chromosome 9 were found in 73% of tumors and at 17p13 in 18% of tumors. There was no significant difference in the frequency of LOH in the upper and lower urinary tract. Deletions at 9p21 were significantly correlated with invasive tumor growth. The pattern of deletion revealed monoclonality of all tumors in nine patients. In five patients there were at least two tumor clones with different genetic alterations. In four of these patients the different clones occured in the bladder and subsequently in the ureter and renal pelvis. All four patients with p53 mutations revealed identical mutations in all tumors. Thus, multifocal urothelial carcinomas are frequently monoclonal, whereas others show oligoclonality, providing molecular evidence for field cancerization. Intraluminal tumor cell seeding appears to be an important mechanism of multifocal occurence and recurrence of urothelial carcinomas.


Synchronous or metachronous multifocal occurrence is a characteristic feature of urothelial carcinoma (Bender and Jones, 1998), but the clonal origin of these multiple tumors remains elusive. One hypothesis proposes that all tumors arise from a single transformed cell and are therefore monoclonal (Sidransky et al., 1992), spreading by either intraepithelial migration or intraluminal seeding (Harris and Neal, 1992). The other hypothesis proposes a field effect of urothelial carcinogens that results in the independent malignant transformation of spatially distinct urothelial cells (Heney et al., 1978). Field cancerization is also presumed to be the cause for multifocal tumor development in the oral cavity and upper aerodigestive tract, although there is molecular evidence for a widespread expansion of a single malignant clone populating extensive areas of the mucosa (Franklin et al., 1997; Partridge et al., 1997). The clonality of urothelial carcinomas is receiving more and more clinical attention, as detection of genetic alterations in urine samples is increasingly utilized for both diagnosis and follow-up of patients with urothelial cancer (Steiner et al., 1997).

Most molecular studies favor the monoclonal origin of urothelial carcinoma with intraepithelial spread or intraluminal seeding (Sidransky et al., 1992; Bender and Jones, 1998; Takahashi et al., 1998; Simon et al., 2001). Some studies, however, provide evidence for the existence of more than one tumor clone, particularly in early stage bladder carcinoma (Paiss et al., 1997; Hartmann et al., 2000). Studies investigating the clonality of urothelial carcinoma are mostly limited to the urinary bladder. Only one recent study investigated simultaneous or consecutive urothelial carcinomas of the bladder and the upper urinary tract, finding a higher frequency of multiple tumor clones (Takahashi et al., 2001).

The aim of this study was to investigate genetic alterations and clonality of combined synchronous or metachronous multifocal tumors of the upper and lower urinary tract in a large number of patients. Overgrowth of one tumor clone by another tumor as well as tumor expansion by intraepithelial migration are unlikely when examining the clonality of upper and lower urinary tract carcinomas because of the large spatial distance of the multiple tumors.

Tumor tissue was carefully microdissected from paraffin-embedded tissue to obtain samples containing at least 80% tumor cells (Dietmaier et al., 1999). Loss of heterozygosity (LOH) analysis at chromosome 9, a genetic alteration generally accepted as an early event in urothelial carcinogenesis (Simoneau et al., 1999), was used as a marker for clonality, since a strong correlation between LOH analysis of chromosome 9q and X-chromosome inactivation analysis has been reported (Sidransky et al., 1992). The great advantage of LOH analysis is the applicability to male patients who make up the majority of patients with urothelial carcinoma. For each arm of chromosome 9, four polymorphic microsatellite markers (Figure 1) were used and one marker was tested for the p53 tumor suppressor gene locus at 17p13. Exons 5–9 of the p53 gene were directly sequenced, and microsatellite instability (MSI) was analysed at six loci using the extended NIH consensus primer panel for colorectal cancer (see Table 1; Boland et al., 1998) as well as immunohistochemical staining for the mismatch repair proteins hMLH1 and hMSH2.

Figure 1

Localization of the markers on chromosome 9. Tumor tissue and normal tissue (renal parenchyma) were microdissected from methylen-blue stained sections with a needle under an inverted microscope to obtain at least 80% tumor cells. After proteinase K digestion overnight, DNA was isolated using the Quiagen tissue kit (Quiagen, Germany). LOH analysis was performed as described previously (Hartmann et al., 2000). All LOH were confirmed in a second independent PCR reaction. Primer sequences and PCR conditions can be obtained from the authors or at

Table 1 Cinical data and results of the microsatellite analysis for five patients with occurence of oligoclonal tumors

Overall, 19 patients with 94 synchronous or metachronous multifocal urothelial carcinomas were studied. Every patient had at least one tumor in the upper (renal pelvis, ureter) and one tumor in the lower (bladder, urethra) urinary tract. The average number of tumors per patient was 4.9, and tumor stage was distributed as follows: 39 (42%) pTa, 33 (35%) pT1, 21 (22%) pT2–4 and 1 (1%) CIS. The study included multifocal upper urinary tract tumors in nine of 19 patients, upper urinary tract tumors with subsequent bladder tumors (n=10), simultaneous multifocal tumors (n=2), and also tumors of the ureter and renal pelvis developing after the first bladder cancer (n=7).

Deletions at chromosome 9 and at the p53 locus on chromosome 17p13 were detected in 69 (73%) tumors and 15 (18%) tumors, respectively. Forty-three (46%) tumors revealed LOH at 9p and 55 (59%) tumors had a deletion at 9q. Deletions at both 9p and 9q occured in 29 (31%) tumors. Deletion at 9p alone were observed in 14 (15%) tumors and isolated 9q deletions occurred in 26 (28%) tumors. Interestingly, 13 of 14 (93%) tumors with an isolated 9p deletion were invasive (pT1), compared to only 12 of 26 (46%) tumors with an isolated 9q alteration (P<0.01). The correlation of invasiveness and 9p deletion was independent of simultaneous deletions of the p53 gene which has been shown to be associated with urothelial tumor invasion (Bender and Jones, 1998). There were no significant differences in the rate of genetic alterations at chromosome 9 of tumors of the upper and lower urinary tract, although upper urinary tract tumors showed a deletion frequency 14% lower at 9p and 7% lower at 9q, respectively.

Five patients with a total of 18 tumors showed neither chromosome 9 nor chromosome 17p13 deletions. The deletion pattern of nine patients was compatible with a monoclonal origin of their tumors from a single progenitor cell. Three of these nine patients harboured identical deletions at all investigated chromosomal regions, whereas six patients had carcinomas that exhibited one common deletion as well as diverse alterations at other loci, compatible with development of subclones.

Monoclonality was confirmed by sequencing the p53 tumor suppressor gene. Exons 5–9 were directly sequenced in one upper and one lower urinary tract tumor from each patient. In patients showing either LOH at 17p13 or a p53 mutation, sequencing of the p53 gene was performed in all tumors. Only four (21%) patients with a total of 26 tumors had a p53 mutation. All patients with a p53 mutation showed the identical mutation (Figure 2) in all tumors and revealed corresponding monoclonal results in the LOH analysis. Of 27 tumors with a p53 deletion and/or a p53 mutation in exons 5–9, only 11 (41%) carried both a deletion and a mutation.

Figure 2

Sequencing results of the tumors in four patients with mutations in the p53 gene. Exons 5–9 were amplified separately utilizing a nested PCR approach. After precipitation with polyethylenglycol, 1–4 μg of the PCR product was directly sequenced on an ABI 373. Primer sequences and PCR conditions can be obtained from the authors or at

Five (26%) patients showed evidence of oligoclonality with at least two different tumor clones (Table 1). In four of these five patients the different clones were identified in the upper and lower urinary tract. In all four patients the bladder tumor developed before the upper urinary tract tumor. Two patients (09, 18) showed deletions of different alleles at chromosome 9q. Because chromosome 9 deletions are the earliest genetic alteration in bladder cancer identified so far (Simoneau et al., 1999), this strongly indicates two different tumor cell clones in these patients. One patient (02) developed a bladder carcinoma with no detectable genetic alterations and an ureteral carcinoma with deletions, as well as microsatellite instability (MSI) at two loci on chromosome 9 and MSI at four of six investigated NIH consensus marker loci and negative immunohistochemical staining for the mismatch repair protein hMSH2. In patient 12, the tumor in the ureter also showed MSI and loss of hMSH2 expression, whereas the bladder tumor revealed no genetic alterations. In patient 07, LOH at all investigated chromosomal regions (9p, 9q, 17p) in the bladder cancer and the absence of these changes in the infiltrating ureter tumor supports the development of a second tumor clone. Although it is theoretically possible that the tumors with additional LOH or MSI in patients 02, 07 and 12 are subclones from the tumors without these changes and therefore tumor progression related, we think that the data strongly suggest the existence of different tumor clones (Table 1; Figure 3a, b).

Figure 3

(a) Results of the LOH analysis of patient 18 at D9S1113 (top) and p53 (bottom); normal tissue (1), bladder tumors (2, 3, 5) and renal pelvis tumor (4); the bladder tumors show a loss of the longer allele at D9S1113 and a loss of the shorter allele at p53, whereas the renal pelvis tumor lost the shorter allele at D9S1113 and shows retention of heterozygosity at p53. (b) Clonal development of multifocal tumors in patient 09 (for symbols see Table 1)

In five (26%) patients the clonal origin of their tumors could not be studied, because these patients had tumors that lacked alterations at the investigated loci.

In summary, the clonal origin of tumors from 14 patients could be determined, with nine (64%) having monoclonal tumors and five (36%) possessing at least two different tumor clones. Unfortunately, four of the five patients with oligoclonal tumors were male and the fifth patient was not informative for a polymorphic marker within the human androgene receptor gene (Sidransky et al., 1992), hindering from the use of the HUMARA assay for detection of the X-chromosomal inactivation pattern in this study.

Little is known about genetic alterations in upper urinary tract tumors. Our results show that there is no significant difference in the frequency of chromosome 9 alterations between upper and lower tract tumors. Chromosome 9 alterations are widely accepted as one of the earliest and most common events in urothelial carcinogenesis (Bender and Jones, 1998). In accordance with Takahashi et al. (2001), we believe that there are biologically and also clinically important differences between the urothelium of the upper and lower urinary tract, which may be just due to different degrees of exposure to carcinogens. In Balkan endemic nephropathy the risk of upper urinary tract, but not bladder cancers is increased (Radovanovic et al., 1985). The recent clinical insights in HNPCC-related tumor entities seems to group upper urinary tract tumors in the category of HNPCC-related tumors (Aarnio et al., 1995; Arzimanoglou et al., 1998), while lower urinary tract tumors show no relation to genetic tumor patterns, thus highlighting biological dissimilarities in different urothelial sites. Our experimental findings of frequent microsatellite instability in tumors of the ureter and renal pelvis, but not in tumors of the urinary bladder (unpublished results) support this view. Takahashi et al. (2001), too, found different patterns of microsatellite alterations and a higher degree of genetic instability in patients with multifocal urothelial carcinomas and involvement of the upper urinary tract.

There was a significant correlation between 9p deletions and invasive urothelial carcinoma, whereas isolated 9q deletions were more often observed in non-invasive tumors. This is an interesting finding worth studying further prospectively, since it could potentially indicate a marker for predicting the clinical behavior of papillary urothelial tumors. Our findings indicate an important role of 9q deletions in the initial development of papillary urothelial tumors, as reported by Simoneau et al. (1999). In contrast, other investigators have identified 9p alterations in non-invasive papillary tumors (Orlow et al., 1995; Hartmann et al., 2000).

Clonality studies in multifocal urothelial cancers will help to understand the biology of urothelial tumors, with potential clinical relevance. Detection of microsatellite alterations or p53 mutations in urine samples as genetic fingerprints of a tumor clone could be used for both screening and follow-up of cancer patients (Sidransky et al., 1991; Steiner et al., 1997). Genetic divergence of different clones could potentially lead to false negative results. In our study nine of 14 (64%) informative patients revealed tumors with a monoclonal origin, consistent with Takahashi et al. (1998), who found monoclonal tumors in 20 of 25 patients. Li and Cannizzaro (1999) demonstrated a monoclonal origin of 35 low-grade pTa tumors from 10 patients by determining the X-chromosome inactivation pattern. Other investigators report similar molecular evidence for a monoclonal origin of multiple urothelial carcinomas (Sidransky et al., 1992; Habuchi et al., 1993).

Herein we demonstrate the existence of more than one tumor cell clone in five patients with combined upper and lower urinary tract tumors, supporting the hypothesis of field effect cancerization of the urothelium by exogenous carcinogens. Yamamoto et al. (1998) used the experimental bladder carcinogen N-butyl-N-(4-hydroxybutyl)nitrosamine in chimeric C3H/HeN–BALB/c mice and observed multiple tumors of different parental origin in 30% of the mice, suggesting polyclonal tumor development. Paiss et al. (1997) found oligoclonality in 36% of pTa bladder tumors and Hartmann et al. (2000) found oligoclonality in 10% of patients with multifocal early stage bladder tumors (pTa, G1). In contrast, none of the 35 pTa grade 1 and 2 tumors in the study of Li and Cannizzaro (1999) showed oligoclonality. Spruck et al. (1994) reported the existence of different clones in three patients with CIS and simultaneous bladder tumors. It is possible that oligoclonality is more common in early lesions, and that tumor progression to invasive and/or poorly differentiated stages leads to monoclonality by overgrowth of one clone. The only other study investigating multifocal tumors in the upper and lower urinary tract provided also molecular evidence for the existence of different tumor clones within one patient (Takahashi et al., 2001). However, both patients with multifocal upper urinary tract tumors and patients with upper urinary tract tumors developing after the first bladder cancer were not studied. Interestingly, four of five patients with different tumor clones in our study developed tumors in the renal pelvis or ureter after the initial bladder cancer (see Table 1).

We identified a group of four patients with identical p53 gene mutations in the tumors of the upper and lower urinary tract. Habuchi et al. (1993) also reported identical p53 mutations in four patients with metachronous tumors of the renal pelvis/ureter and bladder. Waridel et al. (1997) found identical p53 mutations in the normal epithelium of patients with multifocal tumors of the upper aero-digestive tract in 76% of cases, providing evidence for expansion and spread of a p53-mutant tumor clone. The finding of identical mutations in upper and lower urinary tract tumors suggests that the p53 mutation occurred before tumor spread. The distance between such tumors of the renal pelvis and bladder indicates that clonal expansion occurs by intraluminal seeding of malignant urothelial cells rather than by intraepithelial migration (Harris and Neal, 1992; Habuchi et al., 1993; Xu et al., 1996; Li and Cannizzaro, 1999).

In conclusion, there were no significant differences in the frequency of deletions on chromosome 9 in tumors of the upper and lower urinary tract. Our results provide evidence for possible oligoclonality in a subset of urothelial carcinomas. Multifocal occurrence and recurrence of urothelial cancers appear to be the result of both monoclonal tumor spread by intraluminal seeding and, less frequently, the development of distinct tumor clones as a result of field effect carcinogenesis.


  1. Aarnio M, Mecklin JP, Aaltonen LA, Nystrom-Lahti M, Jarvinen HJ . 1995 Int. J. Cancer 64: 430–433

  2. Arzimanoglou II, Gilbert F, Barber HR . 1998 Cancer 82: 1808–1820

  3. Bender CM, Jones PA . 1998 Carcinoma of the Bladder: Innovations in Management. Petrovich Z, Baert L and Brady LW (ed) Springer: Heidelberg/New York pp. 37–51

  4. Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshleman JR, Burt RW, Meltzer SJ, Rodriguez-Bigas MA, Fodde R, Ranzani GN, Srivastava S . 1998 Cancer Res. 58: 5248–5257

  5. Dietmaier W, Wallinger S, Bocker T, Kullmann F, Fishel R, Rüschoff J . 1997 Cancer Res. 57: 4749–4756

  6. Dietmaier W, Hartmann A, Wallinger S, Heinmöller E, Kerner T, Endl E, Jauch K-W, Hofstädter F, Rüschoff J . 1999 Am. J. Pathol. 154: 83–95

  7. Franklin WA, Gazdar AF, Haney J, Wistuba II, La Rosa FG, Kennedy T, Ritchey DM, Miller YE . 1997 J. Clin. Invest. 100: 2133–2137

  8. Habuchi T, Takahashi R, Yamada H, Kakehi Y, Sugiyama T, Yoshida O . 1993 Lancet 342: 1087–1088

  9. Harris AL, Neal DE . 1992 N. Engl. J. Med. 326: 737–740

  10. Hartmann A, Roesner U, Schlake G, Dietmaier W, Zaak D, Hofstaedter F, Knuechel R . 2000 Lab. Invest. 80: 709–718

  11. Heney NM, Daly J, Prout GR, Nieh PT, Heaney JA, Trebeck NE . 1978 J. Urol. 120: 559–560

  12. Hermanek P, Hutter RVP, Sobin LH, Wagner G, Wittekind C (ed) . 1998 TNM-Classification of Malignant Tumors. 4th edition Springer: Berlin, Heidelberg, New York

  13. Li M, Cannizzaro A . 1999 Hum. Pathol. 30: 1197–1200

  14. Orlow I, Lacombe L, Hannon GJ, Serrano M, Pellicer I, Dalbagni G, Reuter VE, Zhang ZF, Beach D, Cordon-Cardo C . 1995 J. Natl. Cancer Inst. 87: 1524–1529

  15. Paiss T, Woehr GM, Vogel W, Kleinschmidt K, Hautmann RE . 1997 J. Urol. 157: 161

  16. Partridge M, Emilion S, Pateromichelakis S, Phillips E, Langdon J . 1997 Oral Oncol. 33: 332–337

  17. Radovanovic Z, Krajinovic S, Jankovic S, Hall PW, Petkovic S . 1985 J. Cancer Res. Clin. Oncol. 110: 181–183

  18. Sidransky D, von Eschenbach A, Tsai YC, Jones P, Summerhayes I, Marshall F, Paul M, Green P, Hamilton SR, Frost P . 1991 Science 252: 706–709

  19. Sidransky D, Frost P, von Eschenbach A, Oyasu R, Preisinger A, Vogelstein B . 1992 N. Engl. J. Med. 326: 737–740

  20. Simon R, Elize E, Schafer KL, Burger H, Semjonow A, Hertle L, Dockhorn-Dwornicz HJ, Bocker W . 2001 Cancer Res. 61: 355–362

  21. Simoneau M, Aboulkassim TO, LaRue H, Rousseau F, Fradet Y . 1999 Oncogene 18: 157–163

  22. Spruck CH, Ohneseit PF, Gonzalez-Zulueta M, Esrid D, Miyao N, Tsai YC, Lerner SP, Schmütte C, Yang AS, Cote R, Dubeau L, Nichols PW, Hermann GG, Steven K, Horn T, Skinner DG, Jones PA . 1994 Cancer Res. 54: 784–788

  23. Steiner G, Schoenberg MP, Linn JF, Mao L, Sidransky D . 1997 Nature Medicine 3: 621–624

  24. Takahashi T, Habuchi T, Kakehi Y, Mitsumori K, Akao T, Terachi T, Yoshida O . 1998 Cancer Res. 58: 5835–5841

  25. Takahashi T, Kakehi Y, Mitsumori K, Akao T, Terachi T, Kato T, Ogawa O, Habuchi T . 2001 J. Urol. 165: 672–677

  26. Waridel F, Estreicher A, Bron L, Flaman JM, Fontolliet C, Monnier P, Frebourg T, Iggo R . 1997 Oncogene 14: 163–169

  27. Xu X, Stower MJ, Reid N, Garner RC, Burns PA . 1996 Clin. Cancer Res. 2: 1795–1800

  28. Yamamoto S, Tatematsu M, Yamamoto M, Fukami H, Fukushima S . 1998 Carcinogenesis 19: 855–860

Download references


We thank Monika Kerscher, Astrid Köhler, Rene Krieg, Anne Pietryga-Krieger, Joachim Rauch, Andrea Schneider, Silvia Seegers and Lisa Weber for excellent technical support. This research was supported by Grant 10-1096-HaI from the Dr Mildred Scheel Foundation of Cancer Research to A Hartmann and R Knuechel.

Author information



Corresponding author

Correspondence to Arndt Hartmann.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hafner, C., Knuechel, R., Zanardo, L. et al. Evidence for oligoclonality and tumor spread by intraluminal seeding in multifocal urothelial carcinomas of the upper and lower urinary tract. Oncogene 20, 4910–4915 (2001).

Download citation


  • urothelial carcinoma
  • upper and lower urinary tract
  • LOH
  • oligoclonality
  • p53 mutation

Further reading


Quick links