Article Text

This article has a correction. Please see:

PDF

Suspicious urinary cytology with negative evaluation for malignancy in the diagnostic investigation of haematuria: how to follow up?
  1. G Nabi1,
  2. D Greene2,
  3. M O Donnel3
  1. 1Academic Urology Unit, Department of Surgery, University of Aberdeen, Aberdeen AB25 2ZD, UK
  2. 2Department of Urology, Royal Sunderland Hospital, Sunderland SR4 7TP, UK
  3. 3Department of Pathology, Royal Sunderland Hospital
  1. Correspondence to:
 Mr G Nabi
 Department of Surgery, Medical School, Polwarth Building, University of Aberdeen, Aberdeen AB25 2ZD, UK; g.nabiabdn.ac.uk

Abstract

Aims: To define the natural history of patients with suspicious urinary cytology and negative initial evaluation for malignancy in the investigation of haematuria.

Patients and methods: Data from the hospital information support system on urinary cytology examinations carried out at one centre were audited over a period of 24 months. There were 102 patients who had suspicious urinary cytology for malignant cells with negative initial evaluation. Follow up investigations, treatment, and final outcome were noted.

Results: There were 102 patients with suspicious urinary cytology and negative initial evaluation for malignancy in 24 months, with a mean follow up of 15.7 months. Seventy patients had no obvious pathology on initial investigations. Forty one patients were found to have urological malignancies (29 bladder, eight ureteric, and four prostate) on follow up. All patients diagnosed as having urothelial malignancies on follow up had either persistent suspicious cytology (29) or recurrent haematuria (eight). The mean duration for appearance of lesions was 5.6 months (range, 3–12 months). Three patients had suspicious digital rectal examination and biopsies confirmed adenocarcinoma of the prostate. One patient had urinary retention and transurethral resection of prostate showed prostatic adenocarcinoma. The presence of suspicious cells on repeat urine analysis was the only significant factor in predicting the presence of urothelial tumours (p  =  0.002).

Conclusion: Patients with persistent suspicious/positive cytology or recurrent haematuria need further evaluation and follow up. Asymptomatic patients or patients with obvious benign pathology do not require repeat evaluation. Careful urological evaluation, including prostate, should be carried out in these patients.

  • urinary cytology
  • urological follow up
  • suspicious cells in urine

Statistics from Altmetric.com

The widely used Papanicolaou and Marshal1 classification of urinary cytological examination considers class I–II as negative and class IV–V as positive for the presence of malignant cells. The class III or suspicious group remains controversial. Suspicious urinary cytology in clinical practice necessitates thorough investigations to rule out the presence of urothelial malignancy. Most clinicians treat this as malignant, leading to repeated evaluations and follow up, which in most of these patients remain inconclusive. Lack of guidelines in the literature regarding the management of these patients, variations in urine specimen processing, subjective interpretation of cytological analysis, and lack of a standard method of reporting urinary cytology are some of the difficulties in decision making on this issue. Clearly, there remains a group of patients who are over investigated. We analysed our results in the present study to answer the question of how to follow up suspicious urinary cytology in a patient with a history of haematuria and normal initial investigations? We defined suspicious urinary cytology as (a) not clearly indicative of malignancy or (b) cannot clearly exclude the presence of malignancy. The aims of our study were:

  1. To define the natural history of patients with suspicious urine cytology and negative evaluation for malignancy.

  2. To determine a subgroup of these patients with increased risk of developing malignancy.

  3. To develop guidelines for the follow up of these patients.

PATIENTS AND METHODS

Data from the hospital information support system on urinary cytology examinations carried out at one centre were audited over a period of 24 months. Age, sex, history of persistent microscopic or gross haematuria, history of smoking, occupational exposure, follow up investigations, treatment, and final outcome were noted from the hospital records. A record of clinical profile and investigations generated by suspicious urinary cytology was made. Patients with obvious malignancy on initial investigations were excluded from our study. A group of patients who had suspicious urinary cytology (not clearly malignant or cannot exclude presence of malignancy) with negative initial evaluation was defined. The initial investigations usually comprised flexible cystoscopy, intravenous urogram, and abdomenal ultrasound, followed by rigid cystoscopy with bilateral ureteric washings and retrograde studies, wherever necessary. If washings or retrograde studies showed an abnormality, ureteroscopy (with or without biopsies) was carried out. Patients with negative initial investigations were followed up in clinic after six to eight weeks and repeat urinary cytology was requested. Persistence of suspicious urinary cytology or symptoms required repeat evaluation. Patients in whom repeat urinary cytology at six to eight weeks showed no evidence of malignant cells were followed at three monthly intervals, both clinically and with repeat voided urinary cytological analysis.

RESULTS

There were 102 patients with suspicious urinary cytology and negative initial evaluation for malignancy among those in whom cytological examinations were carried out over a period of 24 months. The mean follow up of our study was 15.7 months (range, 10–24). There were 76 male and 26 female patients. The mean age was 61.7 (range, 51–72) years. A history of smoking was present in 60 patients. Patients were divided into two groups (figs 1 and 2). Group 1 (32) consisted of patients with benign pathology on evaluation. Group 2 (70) consisted of patients with no obvious pathology on investigation (fig 2). This last group was further subdivided into four subgroups. Subgroup A (37) consisted of patients who persisted with suspicious cytology or symptoms and developed malignancy diagnosed on further evaluation. The mean duration for appearance of a lesion was 5.6 months (range, 3–12). Subgroup B (four) consisted of patients in whom repeat prostate evaluation confirmed adenocarcinoma of the prostate. Subgroup C (25) comprised patients with persistent normal cytological analysis carried out at least six to eight weeks apart and who showed no evidence of malignancy on further evaluation. Subgroup D (four) comprised patients who had persistent suspicious urinary cytology with no evidence of malignancy on repeat investigation. Patients in subgroup A were subsequently found to have urological tumours (eight ureteric and 29 bladder; fig 3). Urothelial malignancy was diagnosed in 29 of these patients on further investigation after initial negative evaluation as a result of persistent positive or suspicious cytology at the time of the second investigation. Eight patients had recurrent episodes of haematuria and re-evaluation showed urothelial tumours. In subgroup B, three patients had suspicious digital rectal examination results, and biopsies confirmed adenocarcinoma of the prostate. All these patients had previous evaluation for suspected carcinoma of the prostate, including biopsies with no evidence of malignancy. One patient had urinary retention and transurethral resection of the prostate showed adenocarcinoma of the prostate. None of the patients with benign pathology (stones, longterm catheters, pelviureteric junction obstruction, previous instrumentation) or normal cytological analysis on repeat evaluation was found to have evidence of malignancy on follow up. The presence of suspicious cells on repeat urine analysis was the only significant factor in predicting the presence of urothelial tumours (p  =  0.002). Age, sex, and history of smoking were not significant.

Figure 1

Algorithm for the management of suspicious urinary cytology for malignant cells in urological examination.

Figure 2

Diagnostic outcome of group 1 patients (32 patients with benign pathology on evaluation) and group 2 patients (70 patients with no obvious pathology on evaluation).

Figure 3

Final outcome of patients with persistent abnormalities on follow up. Thirty seven patients had urothelial tumours, four had prostatic adenocarcinoma, and four had no evidence of malignancy on further evaluation.

DISCUSSION

In 1864, Sanders first described the presence of exfoliated neoplastic cells in the urine.2 In 1945,1 its use in the diagnosis of urothelial malignancy was reported. Since that time, urine cytology has gained widespread acceptance as a useful investigation of individuals with haematuria. However, despite its extensive use in routine practice, urine cytology has several important drawbacks. Several large series have shown that urine cytology has a low sensitivity, varying between 42% and 66% depending on the population studied. Specificity rates are higher, but rarely exceed 97% in routine practice.3–8

Unlike other forms of cytological assessment, such as breast cytology, urine cytology does not target an abnormal area but relies on neoplastic cells being shed in the urine. It is well recognised that high grade urothelial tumours, including carcinoma in situ, are more likely to shed abnormal cells into the urine. These cells are more atypical, permitting easier and more confident recognition by the reporting cytologist. It is not surprising then that sensitivity rates are higher for tumours of higher grade. The corollary is that low grade tumours are less likely to shed cells into the urine. The cells that are shed vary little from normal urothelial cells, apart from a tendency to be shed in clusters rather than as single cells. Interpretation of these subtle changes is very difficult. Additional problems in urine cytology are that stones, infections, urinary tract instrumentation, and chemotherapy/radiotherapy can all cause changes in the urine cytology picture that mimic malignancy, leading to an erroneous diagnosis by the unwary.9–11

Urine cytology specimens reported as being suspicious of malignancy require further clinical evaluation of the patient by the attending urologist. This can cause undue anxiety in the patient who is worried about the prospect of undiagnosed malignancy and who has to face a series of often uncomfortable and unpleasant investigations. The urologist may be faced with the situation where all other initial investigations are negative apart from urine cytology, and it is often unclear how to proceed with follow up in the absence of specific guidelines. Most studies that have evaluated the outcome of patients who have been reported as having atypical urine cytology show that less that 50% of these individuals are found to have malignancy. This suggests that only half of those with atypical urine cytology need careful follow up, and that currently a large number of patients are subjected to unnecessary investigations, with resultant wastage of valuable resources and time. The aim of our study was to try to assess whether there were specific criteria that could determine those patients who need to be subjected to careful follow up and those patients who could safely be discharged.

“The injudicious use of urine cytology in individuals who have an obvious cause of haematuria, such as urinary tract stones, is to be avoided because it may lead to suspicious results being generated”

Thirty one percent of our patients with an initial suspicious urinary cytology result and negative initial evaluation were found to have a benign disease that could explain their abnormal result. None of this group had abnormal cytology results on repeat follow up and all of the patients are now asymptomatic. This group highlights the problem of benign pathology causing significant atypia in voided urothelial cells. It also emphasises the need to provide accurate and complete clinical details to the pathology department when submitting the specimen. Urine cytology specimens are high volume specimens. Frequently, the request form arrives with no clinical details, which can lead to misinterpretation of the changes by the reporting cytopathologist. This problem is compounded by increasing workload pressures on pathology laboratories, which prevents follow up of each abnormal urine specimen. The injudicious use of urine cytology in individuals who have an obvious cause of haematuria, such as urinary tract stones, is to be avoided because it may lead to suspicious results being generated. This emphasises the need for urologists to have strict criteria governing which individuals should be subjected to urine cytology to avoid this happening.

Take home messages

  • Patients with suspicious urinary cytology with negative initial evaluation should have repeated urinary cytology analysis carried out at least six to eight weeks later

  • Asymptomatic patients or patients with obvious benign pathology do not require repeat evaluation

  • Those with persistent positive cytology or recurrent haematuria need further evaluation and follow up and careful evaluation of the prostate should be carried out in these patients

Four patients in our study group had persistently atypical cells in their urine despite initial normal evaluation. All of these patients were found to have abnormal prostates on review examination, and were confirmed to have prostate adenocarcinoma on histology. In our present series, all patients showed high grade adenocarcinoma (Gleason score > 7) on histology. This is similar to that reported previously by Rupp et al.12 Several reports have described the cytological features of malignant cells from prostatic carcinoma within voided urine samples. Krishnan and colleagues13 described the cells as having oval nuclei with smooth borders, powdery evenly distributed nuclear chromatin, large prominent nucleoli, and lacking appreciable pleomorphism. However, the recognition of these subtle differences is difficult, and distinguishing these cells from atypical cells of urothelial type may be almost impossible except in rare circumstances. In male patients with persistently atypical cells in their urine, careful assessment of the prostate is recommended, particularly when there is no evidence of malignancy on initial evaluation. In selected cases, there may also be a role for immunohistochemical staining for prostate specific antigen on the urine sample to help identify the profile of the atypical cells.14,15

Forty one (40%) of our patients had persistent haematuria or persistent atypical cells in their urine despite a negative evaluation for malignancy at the initial diagnostic investigation. Thirty seven of these patients were subsequently found to have a urothelial malignancy on repeat evaluation. Eight of these occurred in the ureter and 29 in the bladder. The mean duration for the appearance of these lesions was 5.6 months (range, 3–12). The reasons why the bladder tumours were not readily apparent at initial assessment are unclear. It may be that the tumours were of small size, were in areas of the bladder not readily assessable with a flexible cystoscope, or that the operator was inexperienced or under time pressure because of a high throughput of patients. Our study shows that patients who have persistently abnormal urine cytology should be assessed very carefully, preferably with rigid cystoscopy, and by an experienced urologist before the bladder and upper tracts can be said to be normal.

There still remains a small group of four patients (3%) within our study who showed suspicious cytology despite careful follow up. As in other areas of cytology, there will always be a small false positive rate even in the most expert centres. However, it is probably worthwhile reviewing the specimens in this group and considering seeking an expert opinion. It is difficult to recommend how long these patients should be followed for. Because most of the tumours in our series presented in the first 12 months after an abnormal cytology result, it would seem appropriate to review them regularly for 12–18 months and then to consider discharging them back to primary care

CONCLUSIONS

Patients with suspicious urine cytology with a negative initial evaluation should have a repeat urine cytology examination six to eight weeks later. Asymptomatic patients with negative repeat cytology and patients with an obvious benign pathology that can explain their abnormal result do not require repeat evaluation. Patients with persistent positive cytology or recurrent haematuria need further careful evaluation, including assessment of the prostate. Most patients in this group will subsequently be found to have an underlying malignancy.

REFERENCES

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Correction
    BMJ Publishing Group Ltd and Association of Clinical Pathologists