IS IT TIME TO RE-DESIGN THE HAEMATURIA CLINIC?

Authors


Hugh Mostafid, Department of Urology, North Hampshire Hospital, Basingstoke, Hampshire, RG24 9NA, UK. e-mail: mostafid@ymail.com

Abbreviations
UC

urine cytology

CIS

carcinoma in situ

eGFR

glomerular filtration rate

NMP22

nuclear matrix protein 22

HTA

Health Technology Assessment.

INTRODUCTION

It has been estimated that up to 20% of referrals to urologists are for haematuria [1] but the investigation of haematuria has remained largely unchanged for the last 50 years. The primary goal is the detection of bladder cancer using cystoscopy. Radiological imaging is used to detect upper tract pathology. Urine cytology (UC) may be used to detect carcinoma in situ (CIS) and high-grade TCC. Over the last decade, combining these investigations into a ‘one-stop’ haematuria clinic has become popular.

Recent studies appear to confirm that the haematuria clinic is ‘fit for purpose’ as 19% of patients with macroscopic haematuria investigated will have bladder cancer [2,3]. However, a significant number of patients have a non-malignant diagnosis that nevertheless would also be worth detecting (Table 1) [2,3]. Although nearly half of the patients referred have microscopic haematuria, the incidence of bladder cancer in this group is 5%, less than UTI (13%), nephrological disease (10%) and calculi (8%).

Table 1.  Incidence of findings in a haematuria clinic [2,3]
FindingIncidence, %
Bladder cancer (macroscopic haematuria)19
UTI13
Nephrological disease10
Urinary calculi 8.4
Bladder cancer (microscopic haematuria) 5
RCC 1.5
Upper tract TCC 0.3
Primary carcinoma in situ 0.3

The cost of investigating patients with haematuria is enormous. It has been estimated that in the UK the total annual cost of investigating patients who are found not to have bladder cancer is £33.5 million, one third of the annual £100 million cost of managing patients with non-muscle-invasive bladder cancer [4] but in its current form many of the resources of the haematuria clinic are focused on excluding urological malignancies that are very rare: only 0.3% of patients will have primary CIS and 0.3% will have upper tract TCC. Given these findings, it seems appropriate to ask whether it is time to re-design the haematuria clinic.

Nephrological investigations in the haematuria clinic

Nephrological disease is the third commonest pathology found in the haematuria clinic [3] yet despite this, the importance of screening for nephrological disease in the haematuria clinic has been largely overlooked by urologists. However, 7% of patients attending a haematuria clinic have chronic kidney disease as defined by an estimated glomerular filtration rate (eGFR) of <60 mL/min and at 5 years follow-up 3.6% of patients discharged from a haematuria clinic with no diagnosis were found to be under the care of nephrologists with a reduced eGFR or on dialysis (Derek Fawcett, personal communication). A recent combined consensus statement by the British Association of Urological Surgeons and the Renal Association [5], whilst aimed primarily at primary care, emphasized the importance of a basic nephrological screen in all patients with microscopic (non-visible) haematuria, consisting of a urine dipstick for proteinuria and measurement of blood pressure and eGFR, all of which could be checked in the haematuria clinic.

The role of UC and urinary tumour markers in the haematuria clinic

The traditional role of UC as an additional test in the haematuria clinic has been to detect primary CIS of the bladder or upper tract TCC, which may be missed by cystoscopy and upper tract imaging. However, the value of UC in the haematuria clinic has been questioned [6–8]. In practice, it is often difficult to obtain a satisfactory urine specimen for UC and the result is not immediately available, which means that patients cannot be discharged immediately. A urine-based point-of-care test (nuclear matrix protein 22 [NMP22] BladderChek® test, Kyowa Kirin, Slough, UK) for bladder cancer is now available that is cheaper than UC (£20 vs £38, respectively, in the UK NHS). Although initial reports of the use of NMP22 in the setting of follow-up for bladder cancer showed a relatively low specificity that could limit its usefulness, when used specifically in the haematuria clinic, NMP22 appears to have a high specificity (93%) similar to UC [9,10].

The detection of primary CIS in the haematuria clinic

One rationale for using UC in the haematuria clinic is to detect CIS. However, most CIS occurs in association with papillary TCC, which is detected by cystoscopy. However, the incidence of primary CIS (which is non-papillary and could therefore be missed by cystoscopy) is 0.3%[2]. Although the value of performing a test to detect a disease with such a low incidence could be questioned, the high risk of progression in primary CIS suggests that early detection would be beneficial. The value of UC or NMP22 to detect primary CIS has not been directly studied. However, when used to investigate haematuria in a combined total of 1527 patients [11,12], NMP22 appears to have a similar detection rate for CIS to cystoscopy and UC, with a negative predictive value of 99.9% (Table 2) [11,12].

Table 2.  Detection of primary CIS in patients with haematuria using urinary tumour markers
 Reference
Grossman et al.[11]Talwar et al.[12]
N 1331196
CIS   5  1
Cystoscopy   4/5
NMP22   4/5  1/1
UC   3/5  0/1

The detection of upper tract TCC in the haematuria clinic

Another rationale for using UC in the haematuria clinic is to detect upper tract TCC missed by radiological imaging. Once again, the incidence of upper tract TCC is only 0.3%. Recent studies have advocated the use of CT urography in patients with macroscopic haematuria to detect upper tract TCC but ureteric TCC can still be missed [13]. However, as the vast majority of patients will not have upper tract TCC, the expense and unnecessary radiation exposure associated with such a policy in all patients even with macroscopic haematuria could be questioned. Nevertheless, as a delay in detecting upper tract TCC can often have a dramatic impact on prognosis it would seem reasonable to develop a safer and more cost-effective approach to detect such tumours early whilst correctly identifying the vast majority of normal patients who could be spared further radiological exposure. As with primary CIS, the value of UC or NMP22 in detecting upper tract TCC has not been directly studied. When used to investigate haematuria in three studies with 2117 patients (Table 3) NMP22 had a negative predictive value of 100% for upper tract TCC, which suggests it may have a more cost-effective role than CT urography in excluding upper tract TCC in the haematuria clinic [9,11,12].

Table 3.  Detection of upper tract TCC in patients with haematuria using urinary tumour markers
 Reference
Bott et al.[9]Grossman et al.[11]Talwar et al.[12]
N590 1331196
Upper tract TCC  1   1  1
Upper tract imaging  0   0  0
NMP22  1/1   1/1  1/1
UC   0/1  0/1

Is cystoscopy the ‘gold standard’ for detecting bladder cancer in the haematuria clinic?

Cystoscopy has until recently been the gold standard for the detection of bladder cancer with apparently near 100% sensitivity. However, studies using fluorescence cystoscopy have shown that white-light cystoscopy can miss tumours [14]. Furthermore, a recent study questioned the accuracy of cystoscopy alone in detecting bladder cancer [15]. This randomized trial measured the detection rate of bladder cancer at cystoscopy when the urologist had prior knowledge of a positive bladder cancer urine test result compared with a control group. The authors reported that awareness of a positive result significantly improved the bladder cancer detection rate by cystoscopy compared with controls (32% compared with 5%, respectively, P < 0.01). The authors suggest that this was because knowing a patient had a positive result would make the urologists perform a more careful cystoscopy.

How could the haematuria clinic be improved?

It is clear that a simplistic ‘one size fits all’ approach to investigating haematuria is outdated. In other areas of urology, e.g. patients with raised PSA levels, sophisticated algorithms and nomograms have been developed that stratify the patient’s risk of prostate cancer. With this information available, some patients avoid a prostate biopsy whilst other urological conditions, e.g. BPH can be identified and managed.

However, despite forming a large part of the workload of urology departments at enormous cost [1,4], little research has been carried out on the investigation of haematuria. A Health Technology Assessment (HTA) review concluded that ‘there are insufficient data currently to derive an evidence-based algorithm of the diagnostic pathway for haematuria’[16]. The results of a HTA review on behalf of the UK National Institute for Health and Clinical Excellence on the subject are therefore eagerly awaited.

In the meantime, urologists should be aware of the limitations of the traditional haematuria clinic. Cystoscopy may not be as good as we thought at detecting bladder cancer [14,15], whilst we may be missing other diseases (e.g. chronic kidney disease), which should be detected. On the other hand, much time and expense is spent on excluding CIS and upper tract TCC that are in reality very rare. For a start, we should stop thinking of the haematuria clinic simply as a ‘bladder cancer detection clinic’ and be aware that important non-malignant conditions present through the haematuria clinic (Table 1). Most of these can still be excluded in the setting of a one-stop haematuria clinic with little or no additional expense. For example, in patients with microscopic haematuria, measurement of blood pressure, urine dipstick for proteinuria and eGFR measurement, can and should become a standard part of the haematuria clinic. These patients make up half of the referrals to a haematuria clinic and have a similar incidence of nephrological disease (7%) as bladder cancer (5%). They could be investigated separately perhaps in a ‘microscopic haematuria’ clinic, perhaps in conjunction with a nephrologist. This would in turn allow a more efficient use of ‘one-stop’ resources to rapidly diagnose or exclude urological malignancy in patients with macroscopic haematuria. In that setting, a point-of-care urine test could, if positive, alert the urologists to perform a careful cystoscopy, whilst if negative could allow safe discharge without the need for further expensive radiological investigations.

In conclusion, the traditional ‘one size fits all’ set-up of the haematuria clinic requires refinement. Given the enormous cost of investigating haematuria, algorithms or nomograms could be used in conjunction with point-of-care tests to provide a safe, comprehensive and cost-effective ‘one-stop’ haematuria clinic fit for the 21st Century.

CONFLICT OF INTEREST

None declared.

Ancillary