Evaluation of chromosomal aberrations in patients with benign conditions and reactive changes in urinary cytology

Authors


  • We thank Dr. A. Bachmann (University Hospital, Basel), Dr. Tobias Zellweger (Claraspital, Basel), Dr. R. Chiffelle (Delémont), Dr. T. Eichenberger, and Dr. G.F. Mattarelli (Liestal) for providing patient specimens and related clinical information.

Abstract

BACKGROUND:

Fluorescence in situ hybridization (FISH) is routinely used to help clarify atypical urinary cytology. However, to the authors' knowledge, little is known regarding the frequency of chromosomal aberrations in non-neoplastic conditions that could potentially lead to false-positive FISH results. The objective of the current study was to evaluate the frequency of chromosomal aberrations in benign cells of the urinary tract using the UroVysion FISH test.

METHODS:

The authors analyzed 77 Papanicolaou-stained benign urine cytology specimens with reactive epithelial atypia using a FISH assay detecting the chromosomes 3, 7, and 17 and the gene locus 9p21. A positive test result was defined as an increased copy number of at least 2 chromosomes in ≥ 4 of 25 cells, or > 10 cells with a tetraploid or octaploid pattern, or homozygous or heterozygous deletion of 9p21 (≥ 12cells).

RESULTS:

FISH was positive in 27 of 77 bladder washings (35.1%) including 25 of 65 bladder washings (40.5%) and 2 of 15 voided urines (13.5%) from patients with irritative bladder (15 of 36 patients), a history of radiotherapy (7 of 12 patients), nonspecific cystitis (3 of 11 patients), hematuria (3 of 8 patients), and lithiasis (1 of 4 patients) . In 7 of 27 FISH-positive urothelial specimens, the positivity was solely due a polyploid pattern (tetraploid/octaploid pattern) in > 10 of the cells.

CONCLUSIONS:

Chromosomal aberrations can occur in reactive urothelial cells, with a tetraploid pattern being the most common. Even an aneuploid pattern of FISH signals does not always prove malignancy because it rarely occurs in reactive urothelial cells. Correlation of FISH results with cytomorphology and patient history is crucial to avoid false-positive diagnoses. Cancer (Cancer Cytopathol) 2011;. © 2011 American Cancer Society.

The cytological diagnosis of urinary cytology can be challenging because pronounced and worrisome epithelial atypia can occur in benign conditions.1 The rate of such benign bladder washings with pronounced reactive changes is within the range of 5% to 10% in our laboratory. Chromosomal aberrations in benign cells are generally rare and less complex compared with neoplastic cells.2 Fluorescence in situ hybridization (FISH) is a robust method with which to observe chromosomal aberrations.3 Therefore, this technique has become a helpful tool in cytology to better distinguish reactive from neoplastic lesions.4 The multitarget FISH assay (UroVysion; Abbott Molecular Inc, Des Plaines, Ill) was developed to detect polysomies of the chromosomes 3, 7, and 17, and deletion of the gene locus 9p21.5 This assay has been used successfully to improve the diagnostic yield of urinary cytology.6-8 We found previously that a tetraploid chromosomal pattern occurs in a fraction of benign cells in voided urine specimens and bladder washings from patients with benign prostatic hyperplasia.9-11 Therefore, the presence of benign tetraploid cells needs to be taken into account when defining the cutoff value for a positive FISH result to avoid false-positive diagnoses. Applying stringent cutoff criteria by excluding rare cells with a tetraploid pattern, the FISH assay provides high sensitivity (80%) and specificity (95%) for the detection of malignant urothelial cells.5, 9 Despite this high specificity, occasional chromosomal alterations in benign cells can result in a false-positive diagnosis if based solely on a positive FISH result without consideration of the cytomorphology. The objective of the current study was to determine the frequency of chromosomal aberrations using FISH in selected benign conditions with reactive atypia.

MATERIALS AND METHODS

Patients and Cytological Specimens

We selected 77 cytological specimens diagnosed as benign with mostly pronounced reactive changes from the University Hospital of Basel (n = 76) or the Cantonal Hospital St. Gallen (n = 1) in Switzerland. Sixty-two patients (18 women and 44 men; mean age, 60.9 years) had bladder washings and 15 patients (6 women and 9 men; mean age, 47.4 years) underwent voided urine analysis. None of the patients had clinical evidence or a history of previous urinary tract malignancies. We prepared Papanicolaou-stained cytospin preparations as previously described.11 The clinical diagnoses were irritative bladder (36 patients), history of radiotherapy (for prostate cancer in 11 patients and for cervical cancer in 1 patient), cystitis (11 patients), hematuria (8 patients), and urolithiasis (4 patients). Six of the 15 voided urine specimens harbored numerous polyomavirus-infected cells (“decoy cells”) after kidney transplantation. Follow-up data were available for 51 patients, with a mean follow-up of 49.4months (range, 1 month-115 months). Cytology was prospectively evaluated by 2 cytopathologists (L.B. and S.S.). This study was approved by the ethical committee of the University Hospital of Basel (EK 234/10).

FISH and DNA Cytometry

We used the multitarget FISH probe UroVysion as previously described.11 The scoring of 25 cells was performed by cytotechnicians experienced in FISH analysis (A.B, B.G., or M.H). All FISH results were reviewed by 1 cytopathologist (L.B.). Chromosomal aberration in a cell was defined as either multiple gains of chromosomal signals or homozygous or heterozygous loss of 9p21. A relative deletion was diagnosed when the number of the gene locus 9p21 signals was lower than the number of signals of 1 of the 3 chromosomes. A tetraploid (4N) pattern was defined as 4 signals per cell for each FISH probe. Cells with a deviation of the signal number by 1 (ie, 3 or 5 signals) of ≤ 2 of the FISH probes were still considered to be within the analytical tolerance of a tetraploid pattern (eg, 4-3-4-3 or 4-4-5-4). An octaploid (8N) pattern was defined as 8 signals per cell for each FISH probe with a deviation form this pattern by ≤ 2 signals of ≤ 2 of the FISH probes (eg, 8-7-8-9 or 6-8-8-6). The rare cells with a 16N or near 16N pattern were also considered to be polyploid. We defined a FISH result as positive if 1) at least 4 cells had an increased copy number of at least 2 of the chromosomes 3, 7, and 17, or 2) there was a homozygous or heterozygous deletion of 9p21 in at least 12 cells, or 3) if > 10 cells demonstrated a balanced polyploid (ie, 4N or 8N) pattern. Based on the results of our previous studies, we considered a polyploid pattern (ie, 4N or 8N pattern in ≤ 10 cells) to be a negative FISH result, because such alterations are common in reactive conditions.10, 11 This definition of a positive FISH result differs from the manufacturer's recommendation, which defines a positive result either by a gain of multiple chromosome probes in ≥ 4 cells or by a homozygous loss of 9p21 in ≥ 12 cells. DNA image cytometry was performed as previously described.12

RESULTS

Applying the manufacturer's criteria, 49 (63.6%) of the 77 specimens were FISH positive. Using our modified criteria, the number of positive specimens was reduced to 27 (35%). The clinical diagnoses of the 27 FISH-positive specimens included irritative bladder (13 specimens), a history of radiotherapy (7 specimens), cystitis (3 specimens), hematuria (3 specimens), and urolithiasis (1 specimen). An excess of polyploid cells explained the positive FISH result in 9 of the 27 positive urinary specimens and included the following diagnostic categories: irritative bladder (4 patients), a history of radiotherapy (2 patients), cystitis (2 patients), and hematuria (1 patient). Another 7 of 27 positive urinary specimens were positive because of polyploidy excess in combination with aneuploidy in ≥ 4 cells, and in 11 of 27 specimens, a positive FISH result was because of aneuploidy in ≥ 4 cells. We observed occasional umbrella cells with an 8N or even 16N pattern in 14 specimens and 3 specimens, respectively, of the 77 urinary specimens (Fig. 1). None of the specimens was FISH positive because of the 9p21 deletion (deletion in ≥ 12 cells). The results of the 27 FISH-positive specimens are summarized in Table 1. All 6 specimens with decoy cells demonstrated either a diploid or tetraploid pattern but were FISH negative. In contrast, 7 of 12 specimens from patients with a history of radiotherapy to a pelvic organ were FISH positive.

Figure 1.

An example of a reactive umbrella cell with a polyploid chromosome pattern is shown. (A) Papanicolaou-stained, multinucleated urothelial cells with greatly enlarged but otherwise unsuspicious nuclei are shown (× 630). (B) The same urothelial cell as shown in Panel A demonstrated a highly increased number of all fluorescence in situ hybridization (FISH) signals (“crazy chromosomes”). The signals of the individual FISH probes tended to cluster, indicating chromosomal territories within the nucleus. Adjacent smaller cells indicated a normal signal number.

Table 1. Patient Characteristics and Detailed Results of FISH Analysis of the 27 Patients With FISH-Positive Samples
      No. of Cells With Aberrant Chromosomal Patterna 
Patient ID No.Age, YearsGenderFollow-Up, MonthsSpecimen TypeDiagnosisTetraploidOctaploidAneuploid9p21 DeletionResult Categoryb
  • Abbreviations: BW, bladder washing; FISH, fluorescent in situ hybridization; ID, identification; U, voided urine.

  • a

    Twenty-five cells were scored in each specimen.

  • b

    A indicates a positive FISH result due only to polyploidy excess (>10 cells with tetraploidy and/or octaploidy). B indicates a positive FISH result due to polyploidy excess and aneuploidy in ≥4 cells. C indicates a positive FISH result due to aneuploidy in ≥4 cells with no concurrent ploidy excess.

1869Man5BWCystitis18011A
2359Man81UIrritative22300A
2466ManNoneBWIrritative11022A
3768Man68BWPrevious radiation17010A
3861Man43BWPrevious radiation11020A
5274Woman2BWCystitis12121A
6854ManNoneBWIrritative10310A
7133Man35BWIrritative14010A
7380WomanNoneBWHematuria25000A
2837Woman5BWIrritative6552B
2960Man37BWIrritative12073B
3056Man80BWIrritative12381B
4066Man75BWPrevious radiation18340B
4260Man18BWPrevious radiation18250B
4642Woman3BWLithiasis111131B
7551ManNoneBWIrritative11481B
154Man110BWIrritative2074C
2670Man8BWIrritative3072C
2743Man101BWIrritative9050C
3154Man65BWIrritative32170C
3274Man74BWPrevious radiation4142C
3973ManNoneBWPrevious radiation5160C
4170Woman69BWPrevious radiation6170C
5556Man61BWCystitis70142C
6265WomanNoneBWHematuria2041C
6768Man44UHematuria0073C
6980Woman64BWIrritative63153C

Follow-up data were available for 21 of 27 patients with FISH-positive urinary specimens (mean follow-up, 49.9 months [range, 2 months-110 months]). Seven patients underwent subsequent bladder washings, 6 of which were cytologically negative. A 42-year-old woman (Table 1; Patient 46) with ureterolithiasis had a positive FISH result from her bladder washing. Aneuploidy was confirmed by DNA image cytometry. The subsequent bladder washing performed 3 months later was FISH negative and diploid on DNA image cytometry (Fig. 2). A 68-year-old man (Table 1; Patient 67) with urothelial atypia and a positive FISH result in his voided urine specimen had positive cytology and a persistently positive FISH result 11 months later at a control cystoscopy. This patient finally developed a high-grade, noninvasive urothelial carcinoma (pTa) another 21 months later.

Figure 2.

Results from a 42-year-old female patient with ureterolithiasis and transient aneuploidy of urothelial cells are shown. (A) A cytospin preparation with Papanicolaou-stained reactive urothelial cells is shown (× 630). (B) Fluorescence in situ hybridization (FISH) of a reactive umbrella cell is shown. The nucleus demonstrated multiple signals for the gene locus 9p21 (gold) and (C) for chromosome 3 (red). (D) DNA cytometry of the reactive urothelial cells is shown. The majority of the cells were diploid (red column at 2N) or tetraploid (pink column at 4N), but there were numerous cells with an abnormal DNA content between 2.5 N and beyond 15 N, representing aneuploidy. (E) A normal DNA cytogram obtained 3 months later is shown, indicating a predominant diploid stemline with a small percentage of tetraploid cells in the G2/M-phase of the cell cycle (pink column 4N). N indicates normalized DNA content defined by calibration with the DNA content of normal-appearing epithelial cells used as the internal control.

Follow-up data were available for 30 of 50 patients with a negative FISH result, with a mean follow-up of 49 months (range, 1 month-15 months). None of these patients developed urothelial carcinoma.

DISCUSSION

Multitarget FISH is a well-established method for the detection of urothelial carcinoma in urinary specimens, being particularly useful in nondefinitive equivocal cytology.7, 11, 13 Although the specificity of the UroVysion multitarget FISH assay ranges between a high of 89% and 96%, a value closer to 100% would be desirable.14 The specificity of the FISH test varies depending on the patient population, the type of specimens, and the cutoff values. Specificity is likely to decrease if the patient population is comprised of a high percentage of patients with reactive conditions of the urinary tract. The analysis of bladder washing specimens compared with voided urine specimens might also be associated with a higher possibility of false-positive results because the manufacturer's criteria for a positive test result were originally based on voided urine specimens, which are less cellular than bladder washings. In the current study, we explored the frequency and type of chromosomal aberrations in patients with reactive conditions of the urinary tract that might lead to false-positive interpretations.

We found in our collective that nearly two-thirds of reactive urinary cytologies were FISH positive when the manufacturer's criteria were strictly applied.5 This high percentage was markedly reduced to 35% when using modified criteria considering the limited diagnostic value of a polyploid chromosomal pattern. This confirms the results from our previous studies, in which we could significantly enhance the performance of multiprobe FISH in urinary cytology by considering cases with rare tetraploid cells to be FISH negative.9-11 This view has been adopted by others who also advise that small numbers of urothelial cells with tetrasomy should not be construed as evidence of neoplasia.6 Tetraploidy in benign urothelial cells was first described in studies using DNA cytometry.15, 16 This is not surprising, because tetraploidy is a feature of dividing cells,17 and polyploidization is common in some normal cell types including glial cells of the cerebellum, hepatocytes, and megakaryocytes.18, 19 We have observed a polyploid FISH pattern most frequently in superficial urothelial (umbrella) cells. This is in keeping with a previous study by Wojcik et al, who found abnormal DNA content by DNA image cytometry in approximately one-half of 12 benign bladder washings when only umbrella cells were analyzed.16 As assumed for hepatocytes, it is likely that polyploidization of the umbrella cells is caused by endoreplication.18, 20 Endoreplication refers to an increase of the DNA content by clearly delineated genome doublings. It is believed that endoreplication and the resulting polyploidy contribute to stress response, allowing the cells to increase the cell volume without division.20 Given the limited diagnostic value of polyploid urothelial cells, there is only a low probability of urothelial neoplasia in cases of few (ie, ≤ 10) tetraploid cells among 25 selected cells in an otherwise FISH-negative urine cytology specimen. Conversely, it is important to note that tetraploidy does not exclude neoplasia, because tetraploid urothelial carcinomas do exist.21 It is interesting to note that none of the positive FISH results was based on 9p21 deletion, suggesting that 9p21 deletion is a specific feature of neoplasia that does not occur in reactive urothelial lesions.

Our attention was drawn to highly activated yet cytologically benign umbrella cells with greatly enlarged nuclei and an impressive polyploidy pattern with approximately 8 or rarely even 16 signals per each of the 4 chromosomal signals. We refer to such a polyploid chromosome pattern in apparently benign urothelial cells as the “crazy chromosome phenomenon” (Fig. 1). It is interesting to note that the signals of these cells were not randomly distributed throughout the nucleus but rather were clustered in certain nuclear areas. To date, we have not observed such a degree of clustering of signals in FISH-positive cells in urothelial carcinoma. Clustering of FISH signals might reflect nonrandom territorial organization of the genome within the nucleus that is believed to be critically involved in global regulation of gene expression.22 It has been suggested that disruption of chromosome territories could play a role in tumorigenesis.23-25 Further studies are needed to determine whether the pattern of chromosomal territories could serve as a differentiating feature between benign and malignant urothelial cells with a polyploid FISH pattern.

The current study identified benign conditions of the urinary tract that require special attention, including decoy cells, radiotherapy-induced changes, and urolithiasis. Decoy cells are polyomavirus-infected renal tubular or urothelial cells occurring in voided urine specimens of immunosuppressed patients. Because they can demonstrate remarkable atypia, cytopathologists who are unfamiliar with this cell type can easily misdiagnose them as malignant urothelial cells.26, 27

Intriguingly, decoy cells are found to be aneuploid by DNA cytometry, further corroborating the false-positive diagnosis.13 In contrast, none of the 6 samples with decoy cells was found to be positive in FISH analysis with the human DNA probes. This is in keeping with a previous study revealing only few FISH-positive cases.13 Radiation-induced atypia is another notorious pitfall in cytology. Greater than one-half of cases in the current study with a history of radiotherapy to the pelvic area were FISH positive. Although radiation-induced chromosomal aberrations and cytological atypia can persist for many years, the absolute risk of bladder cancer after radiotherapy for cancer of the prostate or the uterine corpus is only marginally increased.20, 28, 29 Inquiries regarding the patient's medical history are the key in this situation. The current study was limited by missing follow-up data in approximately 30% of patients with a positive FISH result. Nevertheless, we considered the likelihood of urothelial neoplasia in these patients to be very low given the benign cytology and the lack of history or clinical suspicion of urothelial carcinoma.

Some authors claim that aneuploidy never occurs in normal cells.17 However, we observed urothelial cells with an aneuploid FISH pattern in several benign conditions, including a case with reversible aneuploidy in a patient with urolithiasis. In addition, it has been reported that seminal vesicle cells that occasionally can be encountered in voided urine are mostly aneuploid by DNA cytometry.30

The cytological specimens in the current study were highly enriched for unusual or pronounced reactive changes. In routine diagnostics, we ignore such apparently reactive cells for FISH analysis because they do not provoke any suspicion of malignancy. Therefore, the relatively high rate of false-positive FISH results in the current study does call into question the high specificity reported in previous studies using atypical or unselected specimens. However, it draws attention to a possible source of false-positive results. The current study data emphasized that FISH analysis of atypical urinary specimens should always be interpreted by or with the assistance of an experienced cytopathologist to correlate the chromosomal signal patterns with morphology. This approach helps to avoid false-positive diagnoses in morphologically well-defined reactive cells such as reactive umbrella cells, seminal vesicle cells, and decoy cells. In addition, clinical information regarding previous treatment can be critical as in the case of decoy cells (immunosuppression) or radiotherapy-induced atypia.

Conclusions

The high frequency of chromosomal polysomies in reactive urothelial cells is a potential source of false-positive FISH results. Cells with a polyploid chromosomal pattern are of limited diagnostic value and should be interpreted in synopsis with the morphology and patient's medical history because they are common in reactive conditions.

FUNDING SOURCES

No specific funding was disclosed.

CONFLICT OF INTEREST DISCLOSURES

Drs. Bubendorf and Savic have received Speaker's Bureau honoraria and financial research support from Abbott Molecular.

Ancillary