Urine cytology after flexible cystoscopy

Authors


Harry W. Herr, 1275 York Avenue, New York, NY 10021, USA.
e-mail: herrh@mskcc.org

Abstract

OBJECTIVE

To correlate urine cytology findings before and after flexible cystoscopy.

PATIENTS AND METHODS

A total of 153 patients undergoing surveillance for bladder tumour provided voided urine for cytology before and immediately after flexible cystoscopy.

RESULTS

Of the 153 patients, 116 had negative urine cytology before and after (96%) a visibly normal cystoscopy and 37 had positive urine cytology before and after cystoscopy that showed recurrent tumour.

CONCLUSIONS

Urine cytology immediately after flexible cystoscopy correlates well with results of urine cytology before cystoscopy.

INTRODUCTION

Urine cytology, in addition to flexible cystoscopy, is considered essential in the diagnosis, evaluation and surveillance of patients with bladder tumours. A recent study by McVey et al.[1] reported that flexible cystoscopy causes morphological changes, particularly papillary aggregation, that might be diagnosed as urothelial malignancy. Although 40% of urine samples collected immediately after flexible cystoscopy resembled low-grade papillary tumours, these changes were transient, and urine samples collected a day or more after cystoscopy were normal. To avoid false-positive diagnosis of malignancy, the authors recommended postponing sampling of urine for ≥ 24 h after flexible cystoscopy.

In our clinic, voided urine is routinely collected for cytological evaluation before cystoscopy. Urine is collected after the procedure if not obtained before, because our patients often face troublesome inconvenience and significant burdens commuting back to the clinic a day or so later simply to give another urine sample. We tested the accuracy of urine cytology in voided urine specimens collected immediately after outpatient flexible cystoscopy in patients undergoing surveillance for recurrent bladder tumours.

PATIENTS AND METHODS

During March 2005, we used flexible cystoscopy to evaluate 153 consecutive patients with a history of bladder tumour. Each patient gave informed consent, and the study was approved by the institutional review board. Povidone-iodine was used to prepare the patients, lidocaine jelly (2%) was instilled into the urethra, and cystoscopy was performed with a 16.2 F Olympus cystovideoscope. Sterile water served as irrigating fluid. Each patient gave a voided urine sample before and immediately after cystoscopy. Both urine samples were collected immediately in Cytolyt solution (Cytyc Corporation, Boxborough, MA, USA) for fixation. The cytology preparations comprised Papanicolaou-stained ThinPrep® slides (Cytyc Corp.). All urine specimens were evaluated by an experienced cytopathologist blinded as to whether urine was collected before or after cystoscopy. Cells were evaluated for cellularity, the number of neoplastic cells, cellular arrangement, cell size and shape, cell borders, and cytoplasmic, nuclear and nucleolar features. Each specimen contained sufficient cellularity for evaluation. Cytology specimens were interpreted as positive (for malignant urothelial cells), suspicious (atypical cells suspicious for urothelial carcinoma), or negative. In our institution, suspicious urine cytology is regarded as positive.

RESULTS

The 153 patients comprised 113 men (74%) and 40 women (26%) aged 38–93 years. Twenty-four patients were undergoing their first follow-up cystoscopy at 3–4 months after induction BCG therapy. Table 1 shows the results of the urine cytology examinations. All 37 patients (24%) with positive urine cytology before cystoscopy also had a positive or suspicious urine cytology after a visible tumour was detected at cystoscopy. This report focuses on the 116 patients (76%) with a negative urine cytology and a visibly normal cystoscopy. Of these, 111 (96%) had negative urine cytology immediately after cystoscopy and only five specimens (4%) were suspicious, including specimens from three patients who had either a bladder neck contracture or urethral stricture dilated or a ureteric stent removed during cystoscopy. Seven patients had negative urine cytology before and after dilatation of a urethral stricture (four), or removal of an indwelling ureteric stent (three), and another 10 had completed 6 weeks of BCG therapy 2 months earlier. None of the specimens after cystoscopy was frankly positive. There was no difference in urine cytology results between male and female patients.

DISCUSSION

The present study shows that voided urine samples collected immediately after flexible cystoscopy provide reliable specimens for cytology examination. Of 116 patients with negative urine cytology before cystoscopy, 96% who had no visible tumour at cystoscopy also had a negative urine cytology after the procedure.

Why is there a discrepancy between the present study and that of McVey et al.[1]? One explanation might be that we used lubricating gel and sterile water, which might have rendered some cytology preparations uninterpretable or falsely negative because of cell lysis. This seems unlikely, as the procedure was the same in all patients and virtually all patients with positive urine cytology before cystoscopy also gave positive samples after cystoscopy (in essence representing a positive control group).

Another possible explanation for false-positive urine cytology is inflammation associated with BPH, catheters or stents, urethral strictures, or even intravesical (notably BCG) therapy. This is also unlikely, as our pathologists were able, for the most part, to distinguish inflammatory cells from malignant urothelial cells.

The likely explanation rests with differences in pathological interpretation of urine cytology specimens. This raises an obvious and important but often overlooked principle guiding bladder tumour management: the urologist must understand his or her pathologist's practice patterns, and consult frequently with them about individual patients. If a pathologist is prone to interpret papillary aggregates as low-grade TCC, the urologist can incorporate this information appropriately with the cystoscopy findings to direct follow-up urine and cystoscopy examinations and treatment. In our centre, positive or suspicious urine cytology means that cancer is present somewhere in the urinary tract, whereas negative urine cytology is considered by urologists to indicate the absence of high-grade cancer. The present study suggests that flexible cystoscopy does not significantly interfere with accurate cytological interpretation. No test is perfect, but our practice of using the results of urine cytology in context with knowledge of a patient's disease, imaging and cystoscopy has proved to be highly reliable, practical, patient-friendly and safe.

Lastly, as urine cytology after cystoscopy is most likely to be misinterpreted as a low-grade papillary tumour in a patient with a negative cystoscopy, we see no harm in waiting another ≈ 3 months until the next scheduled cystoscopy to repeat the urine cytology, especially as expectant management and longer follow-up intervals seem feasible for small papillary tumours [2].

CONFLICT OF INTEREST

None declared.

Ancillary