EAU 2015 – Madrid highlights for NMIBC

Rik Bryan.   Cancer Research Birmingham Univesity Head shots day 2.As we have come to expect in the field of bladder cancer research, EAU Madrid did not present any revolutionary or paradigm-shifting data or papers; however, there were some very interesting updates and validations of previously presented novel data that may shift paradigms in the not too distant future. With a number of bladder cancer sessions overlapping I couldn’t be everywhere, but these are my selected highlights, with an emphasis on my interests of NMIBC and translational research. Further scientific content can be found at the EAU Resource Centre: http://rcmadrid2015.uroweb.org

Non-invasive diagnosis
Genomic approaches to the analysis of urine for non-invasive disease diagnosis appear to be approaching the sensitivities and specificities required to be viable alternatives to flexible cystoscopy. Kim van Kessel (from Ellen Zwarthoff’s group in Rotterdam) presented data on their approach (#837): utilising a combination of FGFR3, TERT and PIK3CA mutations with a number of methylation probes for OTX1 and ONECUT2, they were able to detect bladder cancer cases in haematuria patients (n=258, 175 BCs + 83 non-BCs) with a sensitivity of 93% and specificity of 89%. Although this case mix is very different to that which you would see in a UK haematuria clinic population (BC yield c.10%) and requires validation, their assays require very little DNA (only 50ng, such that over 95% of urine samples would yield adequate DNA) and are very cheap (total cost in the research laboratory of c.€20). Similarly, Maria Ribal (Barcelona) presented her data on a gene expression test for the non-invasive diagnosis of BC (#830): refining a 12-gene expression signature down to a 2-gene signature (IGF2, MAGEA3), they achieved an optimal sensitivity of 81% and specificity of 92% for diagnosing 216 BCs out of a total sample of 525. This case mix is more in line with a typical NMIBC surveillance cohort and the assay is cheap, although it relies upon more fragile RNA such that the yields from urine may not be as reliable as for DNA.

However, both studies show significant promise: based upon data from PDD and NBI studies, I often quote a sensitivity and specificity for conventional white light flexible cystoscopy of c.92% for the diagnosis of BC, and so these non-invasive urine-based biomarkers are getting close to clinical utility. They certainly appear more promising than the current generation of commercially available non-invasive urine tests, which, according to WHO/ICUD consensus, do not yet have any value for clinical decision-making (despite what the NICE Guidelines say…). Large multi-centre prospective studies in both the haematuria and surveillance settings are now needed to validate accuracy and to assess reproducibility and commercial viability. Added to this mix, Chris Probert (Liverpool) captivated the NMIBC session (Thematic Session 6) with his electronic nose (‘Odoreader’) approach to non-invasive diagnosis, with preliminary data from controls and predominantly low-grade NMIBCs showing sensitivities and specificities for diagnosis of over 95%. Once again, larger studies in the appropriate populations are needed and this ‘sniffing cancer’ work is ongoing.

Highlighting the urgent need for cheap and accurate non-invasive diagnostics for BC, Sara Hall (Derby) presented local data on 2-week-wait referrals before and after the National Be Clear on Cancer ‘blood in your pee’ campaign commissioned by the UK Department of Health (#833): the number of 2ww referrals increased by 37% from 723 before the campaign to 988 after the campaign, yet significantly fewer patients referred after the campaign were diagnosed with a urological malignancy (13% versus 22%, p<0.001). The campaign has therefore placed a large demand on local urology and radiology services without benefit. Jo Cresswell (Middlesborough) commented from the audience that she and her colleagues had similar findings from the South Tees region. These data are obviously disappointing for the UK Department of Health and the UK urological community.

Tumour resection
The concept of en bloc tumour resection (‘ERBT’, as opposed to TURBT) has been gaining momentum over the last few years. Kramer et al (Hanover) presented results of a somewhat heterogeneous multicentre study utilising a number of different ERBT approaches (electrical and LASER) in 221 patients (#940). They concluded that ERBT was safe regardless of the energy source used and offered high-quality resections of tumours >1 cm; however, LASER ERBT had a significantly lower ‘conversion’ rate to TURBT and a significantly lower haemoglobin loss than electrical ERBT. This represents a useful proof of concept, but future studies need to be better designed to achieve validation and reproducibility and to demonstrate benefit.

The topic of re-TUR was the subject of a ‘Point-Counterpoint’ session (Thematic Session 19). Marko Babjuk (Prague) stated that re-TUR can reduce recurrence rate and determine the need for further treatment, although excellent initial resection should always be the primary objective – the risk of tumour persistence and understaging is still significant, and re-TUR should not just be triggered by an unsuccessful initial procedure. In support of Marko, I personally consider TURBT to be a fundamental urological procedure that likely determines the patient’s burden, interventions and outcome for the following 10 years; perhaps TURBT should be considered in the same technical terms as radical cystectomy or prostatectomy? Food for thought… Amusingly, Thierry Lebret (Suresnes, France) countered with typical gallic charm in a Swiss Toni fashion (The Fast Show, BBC): ‘Performing a bladder resection is like making love – take your time, control the bleeding, enjoy the procedure and control the nerve reflex…’ And that is why Marko is Chairman of the EAU NMIBC Guidelines Committee!

Intravesical therapies
One of the most important pieces of work, surprisingly nestled amongst the posters rather than as a Plenary or Thematic Session presentation, was Richard Sylvester and the EORTC’s systematic review and individual patient data meta-analysis of randomised trials comparing a single immediate instillation of chemotherapy after TURBT to TURBT alone in patients with NMIBC (#939). 11 studies randomised 2278 (94%) of all known 2432 eligible patients: 1161 to TURBT alone and 1117 to a single immediate postoperative instillation of chemotherapy (SIPIC) of either epirubicin, mitomycin C, pirarubicin or thiotepa after TURBT. A single instillation reduced the relative risk of recurrence by 35%, (p<0.001) and the 5-year recurrence rate from 58.8% to 44.8%. Non-randomised comparisons suggested that an SIPIC was most effective for recurrence if given within 2 hours after TURBT. The instillation did not reduce recurrences in patients with a prior recurrence rate >1 recurrence/year or in patients with an EORTC recurrence score ≥ 5 (ie multiple and large tumours that can be identified at TURBT). An SIPIC did not prolong either the time to progression or time to death due to bladder cancer. Notably, there was a small but statistically significant increase in the risk of death due to any cause with an SIPIC (HR = 1.26, 95% CI: 1.05–1.51, p=0.015, 5-year death rates 12.0% versus 11.2%). Sylvester et al concluded that a single immediate instillation of chemotherapy reduces the risk of recurrence in all patients except those with (1) a prior recurrence rate >1 recurrence/year or (2) an EORTC recurrence score ≥ 5. The small increase in the overall risk of death could not be explained and may have been due to chance. Speaking to Richard at his poster (!!), he suggested that these findings would likely be incorporated into the next update of the EAU Guidelines for NMIBC.

In another ‘Point-Counterpoint’ session (Thematic Session 6), Maximilian Burger (Munich) and Kay Thomas (London) debated whether or not BCG is still the gold standard in bladder cancer treatment. Burger argued that it remains the gold standard because the data supporting BCG are robust and there is no effective alternative (although ‘some important study data are due soon’ – see below). He added that BCG is a true and effective therapy for bladder cancer. Kay Thomas, in turn, noted that response rates for BCG are low when used as a single agent; instead, she made a case for a combined regimen with BCG and mitomycin C (MMC), presenting data showing that BCG works better in combination with MMC and results in a reduced recurrence rate (although there is no difference in disease progression).

Subsequently, John Kelly (London) presented eagerly awaited data from the HYMN trial (a randomised, controlled, phase 3 trial comparing hyperthermia plus mitomycin to a second course of BCG or standard therapy in patients with recurrence of non-muscle invasive bladder cancer following induction or maintenance BCG therapy) and Arends et al (Njimegen) presented their data comparing radiofrequency-induced chemohyperthermia with mitomycin C versus BCG for adjuvant treatment of patients with intermediate- and high-risk NMIBC. These data will be published in due course; however, chemohyperthermia appears safe and effective for intermediate- and high-risk papillary NMIBC.

Henao et al (Barcelona) carried out a retrospective study of 275 patients diagnosed with primary G1pTa NMIBC, 155 of which were described as low-risk (G1pTa, solitary and <3 cm) (#950). The cumulative recurrence rates at 3, 6, 9 and 12 months in the low-risk group were 1.9% (3/155), 2.58% (4/155), 5.16% (8/155) and 9% (14/155), respectively; in the non-low-risk group the cumulative recurrence rates were 1.66% (2/120), 5% (6/120), 13.3% (16/120) and 22.5% (27/120). Comparing recurrence rates between low- and non-low-risk G1pTa NMIBCs, they did not find a statistical difference until the 9-month recurrence rate. They concluded that the first 3-month cystoscopy could be avoided in low-risk bladder tumours, and could even be deferred until 12 months after TUR without compromising oncological control (obviously patients must be warned to consult a urologist if they experience symptoms such as haematuria so that cystoscopy can be expedited).

There then followed an interesting discussion between Vijay Ramani (Manchester) in the audience and Marko Babjuk (Session Chair & Chair of the EAU NMIBC Guidelines Committee). Vijay outlined the recent NICE guidelines recommendation of discharging low-risk NMIBC patients to primary care if they have not had a recurrence within 12 months (recommendation 1.4.5), and requested Marko’s opinion. Marko appeared somewhat quizzical about such a suggestion, and stated that these patients should remain on longer term annual surveillance… As an aside, our own data from a large West Midlands bladder cancer cohort (n=1478) with a minimum of 18 years follow-up (and accepted for publication in European Urology Focus) show that 12% of deaths in patients originally diagnosed with G1pTa disease are due to bladder cancer (25% due to other cancers and 63% due to other causes).

Molecular markers
Although I wasn’t present due to a clash of sessions, Seth Lerner (Houston) presented an update on The Cancer Genome Atlas (TCGA) bladder cancer project (Plenary Session 1), utilising multiple genomic approaches to comprehensively characterise bladder tumours. Now with 412 samples in the cohort and mean follow-up of 1.6 years, they have demonstrated that bladder cancer has a high mutation rate (one of the highest genomic instabilities amongst all cancers), a high percentage of potentially targetable pathways, and a high frequency of mutations in chromatin-modifying genes. Further details of the session can be found here: http://urotoday.com/index.php?Itemid=813&option=com_content&view=article&catid=1160&id=79216&utm_source=newsletter_2532&utm_medium=email&utm_campaign=eau-2015-gu-cancer-exclusives-032315

Marta Sanchez-Carbayo (University of The Basque Country) presented her group’s work on prognostic urinary markers (Thematic Session 11). Utilising multiple research platforms, they are awaiting 5-year follow-up in order to identify prognostic urinary genomic and proteomic signatures. However, she described a urinary DNA methylation profile that appeared to be predictive for BCG response.

Subsequently, Bart Kiemeney (Nijmegen) described his genome-wide association study (GWAS) approaches for the identification of prognostic single nucleotide polymorphisms (SNPs) in germline (blood) DNA (Thematic Session 18). Various groups of SNPs appear to be highly predictive of NMIBC recurrence, and also responses to BCG in high-risk NMIBC. It is expected that these data will be published in the next 12 months.

Richard T Bryan
Senior Research Fellow, School of Cancer Sciences, University of Birmingham

Web design and marketing agency Leamington Spa