Cystoscopic surveillance of patients with non-muscle-invasive bladder cancer revisited

Several urinary biomarkers to detect patients with recurrence of non-muscle-invasive bladder cancer (NMIBC) are presently tested. To achieve a statistically sound result the recurrence rate has to be prognosticated and from that the number of patients needed to include in the study can be calculated. In our experience we have been surprised about the decreasing recurrence rate making it more difficult to complete this kind of study.

The decreasing trend is obvious from the registry studies. In a population-based study of 538 patients with newly detected bladder cancers, diagnosed 1995–1996 and followed during 5 years, recurrent disease was present in 62% of all patients with NMIBC [1]. In a more recent Swedish nationwide study, a decade later the corresponding rate had decreased to 50% [2]. Data from our national quality registry shows a recurrence rate of around 30% for patients diagnosed 2012–2013 (data on file). Reasons for this positive trend could be improved imaging, more intravesical adjuvant therapy and better learning during urology training.

However the surveillance schedules have been mostly unchanged since they were introduced thus the number of negative cystoscopies have more than doubled. Furthermore, the recommended schedules were based on expert opinions due to lack of high-level evidence. An exception is a Danish randomised study that evaluated the consequences of prolonging the follow-up intervals from 3 to 6 months for patients with Ta tumors with none of the following risk factors: concomitant urothelial dysplasia, tumor grade >2, early recurrence [3]. There was no difference with regard to recurrence, progression and tumor-related death between the two groups but the study was underpowered. Still it indicated the possibility of a less intensive schedule.

The objective of surveillance is to detect recurrence at a time when it is easily cured and secondly to diagnose a change to a more malignant potential usually in the form of progression. Different prognostic scores have been developed to make a more individualised follow-up regimen possible. This has later evolved into a division into low, intermediate and high-risk groups but there is no consensus on the definitions of the specific risk groups. The EAU recommendations based on expert advice are:

• Patients with low-risk Ta tumors should undergo cystoscopy at three months. If negative, subsequent cystoscopy is advised nine months later, and then yearly for five years.

• Patients with high-risk tumors should undergo cystoscopy and urinary cytology at three months. If negative, subsequent cystoscopy and cytology should be repeated every three months for a period of two years, and every six months thereafter until five years, and then yearly.

• Patients with intermediate-risk Ta tumors should have an in-between (individualised) follow-up scheme using cystoscopy.

The aim of this commentary is to try to assess if a less stringent control schedule could suffice for these three categories by answering related questions.

Do we need to follow-up the patients with low-risk Ta tumors more than a year?

First we have to consider the definition of low-risk Ta tumors. In the WHO 1973 classification the differences between papilloma and cancer was described and the former was found to represent 3% in clinical materials. Before that, a much broader definition was used and that was evident from our early registry series starting in the 1960s in which the proportion of papillomas was reduced from 19% to less than 2% after the change [4].

For many urologists it was unsatisfying to give a cancer diagnosis to a group of mostly younger patients with a ‘benign’ tumor. That was one reason for the introduction of the entity papillary urothelial neoplasm of low malignant potential (PUN-LMP) in 1998. In a regional Swedish study with central pathology assessment they accounted for 26% of all Ta tumors [5]. With a follow-up of at least 5 years their recurrence rate was half of the rest of the Ta group and none progressed The authors consequently concluded that this sub classification seemed to add valuable prognostic information. Unfortunately, the introduction of PUN-LMP has not been much adopted and in the most recent Swedish register <2% were classified in that way. After that time, they are not separately registered because of their scarcity. A similar decrease has been found in an international study and based on their results with the similar prognosis for PUN-LMP and Ta-LG carcinomas they propose to stop using PUN-LMP as a separate grade category [6].

Low-risk Ta tumors are characterized in the EAU guidelines by being primary, solitary, low grade, <3 cm and without CIS. Studies evaluating long-term outcomes in patients with primary Ta-LG tumors showed that 2.0 − 6.3% progressed to muscle invasive bladder cancer, 1.2–2.4% died of bladder cancer, and 5-year cancer-specific survival was approximately 98% [2,7]. This was mainly without postoperative chemotherapy with proven efficacy in this group of patients. In a Japanese study with intravesical chemotherapy instillation multiplicity was the most important risk factor both for recurrence and progression whilst, underscoring that this factor excludes patients from the low risk category [8].

The NICE recommendation from 2015 with a cystoscopy at 3 months and at one year seems sufficient for these tumors with low malignant potential. If negative at the latter the patient is told to come back if visible hematuria is seen. If they develop recurrence during this first year, they will of course be upgraded to a higher risk category.

Can we do better than the frequent cystoscopy and urinary cytology investigations in patients with high-risk tumors?

High-risk tumors are characterized by having any of the following criteria: stage T1, High grade or CIS. Also if all these criteria are fulfilled; multiple, recurrent and large (>3 cm) Ta Low grade tumors.

An alternative to frequent follow-up of this category is early radical cystectomy. Unfortunately, the difference in outcome between bladder sparing or initial cystectomy has not been addressed successfully in randomised studies. Our national data indicate that with the former management coupled with a very intensive active surveillance still almost 20% will succumb to their disease within five years [9]. Thus, extravesical dissemination is often unrecognized and consequently systemic therapy an option. The first Nordic neoadjuvant cystectomy trial included also patients with T1 G3 tumors and the cancer specific 5-year survival rate was 11% more in the neoadjuvant arm [10]. Since only 58 patients were in the T1 subgroup this difference could not be statistically verified. We can better detect extravesical spread with better imaging and possibly with liquid biopsies but the jury is still out on how we can best can improve the management of the high-risk tumors.

Can the surveillance of intermediate-risk tumors be simpler for the patient and the urologist?

The first control at 3 months is crucial after a transurethral resection regardless of initial stage and grade. It is a quality control but also an indicator of the prognosis and thus could form the basis for the continued management including the follow-up schedule. A risk-based approach to follow-up has been proposed by the IBCN network [11]. They subclassified intermediate-risk tumors and described a high risk group within, that should have a surveillance strategy similar to that of true high risk tumors. This group includes patients for whom 3 or more of the following factors are present (multiple tumors, tumor size ≥3 cm, early recurrence <1 year, frequent recurrences 4 per year). Thus, they advocate making the risk classification after initial follow-up.

A simplified schedule for the rest of the intermediate-risk tumors could be to double the interval after each recurrence free control and terminate after 5 years. The timing then will be 3, 9, 21 and 45 months.

Discussion

Why is there a need to change the scheduling of follow-up? In a systematic review of the economics of bladder cancer two important targets for decreasing variation and cost were identified [12]. One of these was the management of NMIBC patients where excessive costs are due to vigilant surveillance strategies. These unnecessary surgical procedures contribute to bladder cancer being the most expensive cancer to treat per patient. Besides the cost considerations a survey found that participants experienced preprocedural anxiety and worry about their disease during surveillance specifically with the term cancer involved [13].

In another discipline, the value of gynecologic cancer follow-up was reviewed [14]. When finding no evidence of a positive effect on survival the reviewers argued that the conception of follow-up among physicians, patients, and their relatives needed revision. Follow-up after treatment should have a clearly defined and evidence-based purpose. Based on the existing literature, this purpose should increase focus on other endpoints rather than early detection of relapse and improved survival. These endpoints could be quality of life, treatment toxicity, and economy in their view. Regarding bladder cancer it is time to adapt to the changing clinical scenario and this article is proposition for changes but not as radical as proposed for gynecologic cancers. In the future, the role of improved imaging such as PDD, NBI and cytology has to be delineated. Finally, appropriate prospective trials similar to the cited Danish trial would be valuable. The increasing interest in urinary biomarkers has already initiated prospective trials of alternative follow-up schedules. Examples of this are the UroFollow-study with multiple markers and ultrasound in low risk tumors and the surveillance trial using the Xpert Bladder Cancer Monitor for the high risk category.