Tag Archive for: Prostate cancer

You are here: / Prostate cancer

Posts

#RudeFood: Foodporn for a purpose

/1 Comment/by

The Internet is full of weird and wonderful things. Of course, we all know what is most frequently viewed and shared online. That’s right – food! Nonetheless, when celebrity chef Manu Fieldel posted a photo of his latest creation, it certainly made people look long and hard!

View this post on Instagram

Name this dish – I like to call it the ‘Cockembouche’ Does it remind you of anything in par-dick-ular? 🤣 It's bursting with flavour and with one bite, will melt in your mouth. I’m not just posting this for the laugh – I've created this #rudefood dish in support of @anzuptrials who work tirelessly conducting clinical trials to improve the treatment and outcomes of penile, testicular, bladder, kidney & prostate cancer these cancers are the ones that don’t make the headlines and people don’t want to talk about but we need to change that. Share your #rudefood – food porn with a purpose and use the hashtag #rudefood tagging @anzuptrials And let’s raise awareness for these cancers. #foodpornwithapurpose #cockembouche #anzup

A post shared by Manu Feildel (@manufeildelofficial) on

Soon it became clear that this naughty creation had a noble purpose – supporting a campaign to raise awareness of the so-called #BelowTheBelt cancers. While most people may have heard of prostate and bladder cancers, being relatively common, other #BelowTheBelt cancers such as penile and testicular cancers are rarer and relatively unknown. To make matters worse, these cancers affect men either exclusively or predominantly – and we all know how reluctant men can be to go to the doctors.

Hence, the #RudeFood campaign was developed by the Australian and New Zealand Urogenital and Prostate (ANZUP) Cancer Trials Group. ANZUP is the peak co-operative trials group for #BelowTheBelt cancers in Australia and New Zealand. ANZUP has and continues to develop and run many significant clinical trials, including the Enzamet and Enzarad trials for prostate cancer, the Phase III accelerated BEP trial for germ-cell tumours, the sequential BCG-mitomycin trial for bladder cancer and the Eversun and Unison trials in kidney cancer.

The week started with things heating up at ANZUP as they brought #RudeFood to the unsuspecting world!

Manu’s phallic creation was also matched by Ainsley Harriot, Sonia Meffadi and Monty Kulodrovic.

To counterpoint the raunch, there were also poignant personal connections from Simon Leong and Scott Gooding who both described family members who had suffered from prostate cancer.

Over the week, #RudeFood has certainly drawn some attention, including from media outlets such as Mamamia, news.com.au and GOAT. 

A poetic contribution on #RudeFood caught the eye of @UroPoet across the seas. Let us hope this campaign will also lead to greater awareness of #BelowTheBelt cancers and improved outcomes for those affected by them.

Shomik Sengupta is Professor of Surgery at the EHCS of Monash University and visiting urologist & Uro-Oncology lead at Eastern Health. Shomik has particular interests in prostate cancer, including open and robotic prostatectomy, as well as bladder cancer, including cystectomy with neobladder diversion. Shomik is the current leader of the UroOncology SAG within USANZ, and the past chair of Victorian urology training.  Shomik is a Board member and scientific advisory member of the ANZUP Cancer trials group and is heavily involved in numerous clinical trials in GU oncology.

Twitter: @shomik_s 

Article of the week: Management of patients with advanced prostate cancer in the Asia Pacific region: ‘real‐world’ consideration of results from the Advanced Prostate Cancer Consensus Conference 2017

/by

Every week, the Editor-in-Chief selects an Article of the Week from the current issue of BJUI. The abstract is reproduced below and you can click on the button to read the full article, which is freely available to all readers for at least 30 days from the time of this post.

In addition to the article itself, there is an editorial written by a prominent member of the urological community. These are intended to provoke comment and discussion and we invite you to use the comment tools at the bottom of each post to join the conversation. There is also a video produced by the authors describing the ‘real-world’ findings.

If you only have time to read one article this week, it should be this one.

Management of patients with advanced prostate cancer in the Asia Pacific region: ‘real‐world’ consideration of results from the Advanced Prostate Cancer Consensus Conference 2017

Edmund Chionga, Declan G. Murphyb,c, Hideyuki Akazad, Nicholas C. Buchane,f, Byung Ha Chungg, Ravindran Kanesvaranh, Makarand Khochikari, Jason Letranj, Bannakij Lojanapiwatk, Chi-fai Ngl, Teng Ongm, Yeong-Shiau Pun, Marniza Saado, Kathryn Schubachq, Levent Türkeris, Rainy Umbast, Vu Le Chuyenu, Scott Williamsv,r, Ding-Wei Yew, ANZUP Cancer Trials Groupx and Ian D. Davisy,z,r

 

aDepartment of Urology, National University Hospital, National University Health System Singapore, hDivision of Medical Oncology, National Cancer Centre Singapore, Singapore City, Singapore, bDivision of Cancer Surgery, vDivision of Radiation Oncology, Peter MacCallum Cancer Centre Melbourne, yMonash University, zEastern Health, Melbourne, cSir Peter MacCallum Department of Oncology, University Melbourne, Parkville, qAustralian New Zealand Urology Nurses (ANZUNS), Melbourne, VIC, Australia, rANZUP Cancer Trials Group, xLifehouse, Camperdown, Sydney, NSW, Australia, dStrategic Investigation on Comprehensive Cancer Network, The University of Tokyo, Tokyo, Japan, eCanterbury Urology Research Trust, fCanterbury District Health Board, Christchurch, New Zealand, gDepartment of Urology, Yonsei University College of Medicine, Seoul, Korea, iSiddhi Vinayak Ganapati Cancer Hospital, Miraj, India, jSection of Urology, Department of Surgery, University of Santo Tomas, Manila, Philippines, kDivision of Urology, Department of Surgery, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand, lDepartment of Surgery, SH Ho Urology Centre, The Chinese University of Hong Kong, Hong Kong, wDepartment of Urology, Fudan University Shanghai Cancer Center, Shanghai, China, mDivision of Urology, Department of Surgery, oDepartment of Clinical Oncology, Faculty of Medicine, University of Malaya, Kuala Lumpur,Malaysia, nDepartment of Urology, National Taiwan University Hospital, Taipei, Taiwan, sDepartment of Urology, Acibadem University, Istanbul, Turkey, tDepartment of Urology, University of Indonesia, Jakarta, Indonesia, and uDepartment of Urology, Binh dan Hospital, Ho Chi Minh City, Vietnam

 

Read the full article

Abstract

Objective

The Asia Pacific Advanced Prostate Cancer Consensus Conference (APAC APCCC 2018) brought together 20 experts from 15 APAC countries to discuss the real‐world application of consensus statements from the second APCCC held in St Gallen in 2017 (APCCC 2017).

Findings

Differences in genetics, environment, lifestyle, diet and culture are all likely to influence the management of advanced prostate cancer in the APAC region when compared with the rest of the world. When considering the strong APCCC 2017 recommendation for the use of upfront docetaxel in metastatic castration‐naïve prostate cancer, the panel noted possible increased toxicity in Asian men receiving docetaxel, which would affect this recommendation in the APAC region. Although androgen receptor‐targeting agents appear to be well tolerated in Asian men with metastatic castration‐resistant prostate cancer, access to these drugs is very limited for financial reasons across the region. The meeting highlighted that cost and access to contemporary treatments and technologies are key factors influencing therapeutic decision‐making in the APAC region. Whilst lower cost/older treatments and technologies may be an option, issues of culture and patient or physician preference mean, these may not always be acceptable. Although generic products can reduce cost in some countries, costs may still be prohibitive for lower‐income patients or communities. The panellists noted the opportunity for a coordinated approach across the APAC region to address issues of access and cost. Developments in technologies and treatments are presenting new opportunities for the diagnosis and treatment of advanced prostate cancer. Differences in genetics and epidemiology affect the side‐effect profiles of some drugs and influence prescribing.

Box 1: Management of advanced prostate cancer in the APAC region: real‐world challenges in implementing the St Gallen APCCC recommendations.

  1. Differences in toxicity: safety data for docetaxel are not fully established in Asian men and concerns about the toxicity profile and risk of neutropaenia may influence prescribing.
  2. Disparities in access to imaging technology: variable access to imaging technology may limit prescribing according to precise definitions.
  3. Disparities in access and cost of treatment: availability and cost of treatments are the most significant factor influencing prescribing decisions in the region; lower‐cost alternatives are not always culturally acceptable, and informed choice is important.
  4. Variability in MDT approaches: the importance of multidisciplinary input to treatment recommendations is understood but MDTs are a challenge in some APAC countries; virtual MDT participation should be encouraged.
  5. Variability in demographics: genetics and epidemiology in Asian men with prostate cancer may result in different treatment responses; collaborative registry studies and trials in APAC populations are likely to be valuable.

Conclusions

As the field continues to evolve, collaboration across the APAC region will be important to facilitate relevant research and collection and appraisal of data relevant to APAC populations. In the meantime, the APAC APCCC 2018 meeting highlighted the critical importance of a multidisciplinary team‐based approach to treatment planning and care, delivery of best‐practice care by clinicians with appropriate expertise, and the importance of patient information and support for informed patient choice.

Read more Articles of the week

 

Article of the week: RS‐RARP vs standard RARP: it’s time for critical appraisal

/by

Every week, the Editor-in-Chief selects an Article of the Week from the current issue of BJUI. The abstract is reproduced below and you can click on the button to read the full article, which is freely available to all readers for at least 30 days from the time of this post.

In addition to the article itself, there are two accompanying editorials written by prominent members of the urological community. These are intended to provoke comment and discussion and we invite you to use the comment tools at the bottom of each post to join the conversation. There is also a podcast by one of our Resident Podcasters describing the article.

If you only have time to read one article this week, it should be this one.

Retzius‐sparing robot‐assisted radical prostatectomy (RS‐RARP) vs standard RARP: it’s time for critical appraisal

Thomas Stonier*, Nick Simson*, John Davis† and Ben Challacombe‡

 

*Department of Urology, Princess Alexandra Hospital, Harlow, ‡Urology Centre, Guy ’s Hospital, London, UK and †Department of Urology, MD Anderson Cancer Center, Houston, TX, USA

 

Read the full article

Since robot‐assisted radical prostatectomy (RARP) started to be regularly performed in 2001, the procedure has typically followed the original retropubic approach, with incremental technical improvements in an attempt to improve outcomes. These include the running Van‐Velthoven anastomosis, posterior reconstruction or ‘Rocco stitch’, and cold ligation of the Santorini plexus/dorsal vein to maximise urethral length. In 2010, Bocciardi’s team in Milan proposed a novel posterior or ‘Retzius‐sparing’ RARP (RS‐RARP), mirroring the classic open perineal approach. This allows avoidance of supporting structures, such as the puboprostatic ligaments, endopelvic fascia, and Santorini plexus, preserving the normal anatomy as much as possible and limiting damage that may contribute to improved postoperative continence and erectile function. There has been much heralding of the excellent functional outcomes in both the medical and the lay press, but as yet no focus or real mention of any potential downsides of this new technique.

Read more Articles of the week

 

Editorial: Retzius‐sparing robot‐assisted radical prostatectomy

/by

In their commentary in the current issue of BJUI, Stonier et al. [1] examine the potential technical pitfalls and published results of the Retzius‐sparing technique of robotic radical prostatectomy. The authors reviewed three studies from three different groups [2,3], including a study by our group [4], and raised three specific concerns: the oncological efficacy of the procedure; the long learning curve; and the generalizability of the technique to challenging surgical scenarios. We offer a few clarifications and comments.

The first study on Retzius‐sparing robot‐assisted radical prostatectomy came from the Bocciardi group [2]. This was a prospective, single‐arm study of 200 patients. The authors reported a 14‐day continence rate of 90–92%, a 1‐year potency rate of 71–81% (in preoperatively potent patients undergoing bilateral intrafascial nerve‐sparing) and a positive surgical margin rate of 25.5%. The positive surgical margin rate improved in patients with pT2 disease, from 22% to 9% (P = 0.04) over the course of the study (initial 100 vs subsequent 100 patients), while in patients with pT3 disease, it remained stable at ~45%. Lim et al. [3] also noted an improvement in their overall positive surgical margin rate from 20% to 8% when comparing the initial 25 patients with the subsequent 25 patients. In that study, a standard robot‐assisted radical prostatectomy comparator arm was included and there were no differences in overall positive surgical margin rates (14% in both arms), while continence was better with the Retzius‐sparing approach.

Recognizing the potentially technically challenging nature of the Bocciardi approach, we performed a randomized controlled trial to objectively evaluate the technique. Randomized controlled trials are typically designed to answer a single question. Our trial was designed to determine whether there were differences in the rate of return of urinary continence, the primary benefit that previous non‐controlled studies had reported. This our study clearly showed [4].

Once the trial was completed, post hoc analysis of secondary outcomes was performed [5]. One of these outcomes was the positive surgical margin rate. In our trial, we noted an overall positive surgical margin rate of 25% in the Retzius‐sparing arm vs 13% in the control arm, a difference that did not achieve statistical significance (P = 0.11). Stonier et al. [1] suggested that if the sample size of our trial were doubled, then the positive surgical margin rate in each group would be doubled as well, leading to significance. This conclusion is problematic. The likelihood that doubling the sample size would result in the exact doubling of numbers in all four cells of a 2 × 2 contingency table is estimated at <5% using Fisher’s exact test (this calculation is different from the P value). Furthermore, the surgical margins depend as much on the pathological stage as on surgical approach. In our trial, patients were matched preoperatively for risk in the best manner possible for a pragmatic randomized trial. However, it is impossible to predict and control for the final pathological characteristics. Pathological analysis showed that patients undergoing Retzius‐sparing surgery did have significantly more aggressive disease: ≥pT3 disease in 45% vs 23.3% of patients (P = 0.04) [4, 5]. This, by itself, could account for a substantial difference in surgical margin rates.

In writing our paper, we made no judgements as to whether the Bocciardi or posterior technique is fundamentally superior to an anterior or Menon approach, whether it is easier to perform, how generalizable it is [6], or what the learning curve may be. That is best left to the individual surgeon’s training and judgement. We do suggest, however, that surgical margins be interpreted as a function of pathological variables, and not in isolation, and that it is simplistic to assume that identical results will be obtained by doubling sample size. We suggest that such conclusions are hypothesis‐generating, and should best be explored through a separate, purpose‐designed randomized trial.

Authors: Akshay Sood, Firas Abdollah and Mani Menon

References

  1. Stonier T, Simson N, Davis J, Challacombe B. Retzius‐sparing robot‐assisted radical prostatectomy (RS‐RARP) vs standard RARP: it’s time for critical appraisal. BJU Int 2019; 123: 5–10
  2. Galfano A, Di Trapani D, Sozzi F et al. Beyond the learning curve of the Retzius‐sparing approach for robot‐assisted laparoscopic radical prostatectomy: oncologic and functional results of the first 200 patients with >/= 1 year of follow‐up. Eur Urol 2013; 64: 974–80
  3. Lim SK, Kim KH, Shin TY et al. Retzius‐sparing robot‐assisted laparoscopic radical prostatectomy: combining the best of retropubic and perineal approaches. BJU Int 2014; 114: 236–44
  4. Dalela D, Jeong W, Prasad MA et al. A pragmatic randomized controlled trial examining the impact of the Retzius‐sparing approach on early urinary continence recovery after robot‐assisted radical prostatectomy. Eur Urol 2017; 72: 677–85
  5. Menon M, Dalela D, Jamil M et al. Functional recovery, oncologic outcomes and postoperative complications after robot‐assisted radical prostatectomy: an evidence‐based analysis comparing the Retzius sparing and standard approaches. J Urol 2018; 199: 1210–7
  6. Galfano A, Secco S, Bocciardi AM. Will Retzius‐sparing prostatectomy be the future of prostate cancer surgery? Eur Urol 2017; 72: 686–8

 

Editorial: Reply: RS-RARP vs standard RARP

/by

Since the introduction of robotic surgery in the treatment of patients with prostate cancer (PCa), different surgical innovations have been implemented in order to preserve postoperative functional outcomes while maintaining oncological safety. Sparing the Retzius space during robot‐assisted radical prostatectomy (RARP) was introduced early this decade by Galfano et al [1]. Interestingly, 90% and 96% of patients treated with Retzius‐sparing RARP (RS‐RARP) were continent (no pad/safety pad) at 1 week and 1 year, respectively. Similarly, our group reported a 70% continence rate (no pad) at 1 month after RS‐RARP [2].

The fast urinary continence recovery after RS‐RARP is related to several anatomical factors: the anterior Retzius space is kept intact; the urinary bladder is not dropped; the endopelvic fascia and puboprostatic ligaments are preserved; and there is minimal distortion of the supporting urethral tissues. A recent study reported [3] that less bladder neck descent was observed during postoperative cystogram in patients treated with RS‐RARP than in those treated with standard RARP.

In a recent randomized controlled study, the postoperative continence rate at 1 week was 48% in standard RARP compared with 71% in RS‐RARP (P = 0.01), and this difference was maintained at 3 months (86% standard RARP vs 95% RS‐RARP; P = 0.02). At 1 year, however, the effect on urinary continence difference was muted (93.3% standard RARP vs 98.3% RS‐RARP; P = 0.09) [4]. Similarly, Chang et al. [3] found that the higher continence rate at 1 week (73.3% RS‐RARP vs 26.7% standard RARP; P = 0.000) had vanished at 1 year (100% vs 93.3%; P = 0.15). By contrast, a large recent prospective series showed that the superiority of RS‐RARP in terms of higher early urinary continence was maintained at 1 year (97.5% RS‐RARP vs 68.5% standard RARP) [5].

In addition to a higher early continence rate, RS‐RARP has a lower incidence of postoperative inguinal hernia occurrence compared with standard RARP [6]. Theoretically, RS‐RARP may provide several other potential advantages. It may be advantageous if patients require future surgery necessitating access to the Retzius space and dropping of the bladder, such as an artificial urinary sphincter implantation, an inflatable penile prosthesis insertion, or kidney transplantation. In addition, in patients with previous inguinal hernia repair using mesh, it enables the avoidance of anterior adhesions by accessing the prostate directly from the Douglas pouch. Notably, large‐size glands and/or middle‐lobe, advanced/high‐risk PCa, and patients with previous prostatic surgeries can be managed safely with RS‐RARP in experienced hands.

Undoubtedly, oncological safety is our main concern in treating cancer. To determine the effectiveness of new treatment methods, long‐term follow‐up is warranted. Biochemical recurrence (BCR) is widely used as a primary oncological outcome to assess PCa treatment success. To our knowledge, after radical prostatectomy, ~35% of patients are at risk of developing BCR in the next 10 years. Currently, there are insufficient data regarding the oncological outcomes of RS‐RARP. Only four articles have compared early oncological outcomes between RS‐RARP and standard RARP, and there was no significant difference (Table 1).

More recently, we reported on the mid‐term oncological outcomes of 359 patients who underwent RS‐RARP. The median follow‐up was 26 months. Although this period is not long enough to reach a meaningful conclusion on the oncological safety of RS‐RARP, it is the longest follow‐up period reported in literature. Overall, the positive surgical margin (PSM) rate was 30.6% (14.6% in pT2 and 40.8% in pT3a disease) and the BCR rate was 14.8%. In terms of functional outcomes, the urinary continence rate at 1 year was 93.9% [7]. Interestingly, 164 patients (45.7%) of our cohort had high‐risk PCa. In these patients, the PSM rate was 41.2%, the BCR rate was 22%, and the 3‐year BCR‐free survival (BCRFS) rate was 72%. We compared our results with those in patients with high‐risk PCa treated with standard RARP in the literature. In studies that used the D’Amico criteria the median follow‐up ranged from 12.5 to 37.3 months, the PSM rates were 20.5% to 53.3%, the BCR rates were 17.4% to 31% and the 3‐year BCRFS rates were 41.4% to 86%. In studies that used the National Comprehensive Cancer Network criteria, the median follow‐up ranged from 23.6 to 27 months, the PSM rates were 29% to 38%, the BCR rates were 9.4% to 33%, and the 3‐year BCRFS rates were 55% to 66% [7].

In summary, RS‐RARP is a novel surgical approach which is associated with better urinary continence recovery in the first few months compared with standard RARP [2,3,4,5]. This superiority might be maintained [5] or equalized at 1 year [3,4]. A few studies have compared the early oncological results between RS‐RARP and standard RARP and no significant difference was found [2,3,4,5]. Recently, our group reported the mid‐term oncological outcomes of patients with high‐risk PCa treated with RS‐RARP and these were similar to those of large studies of conventional RARP. This confirms effective and safe mid‐term BCR control after RS‐RARP, while the long‐term oncological results are awaited [7]. Currently, >4 000 cases of RS‐RARP are performed worldwide and more centres are beginning to use and converting to Retzius‐sparing surgery. All centres are experiencing faster recovery of continence. Thanks are due to Drs Galfano and Bocciardi for exploring and sharing this surgical frontier.

 

References

  1. Galfano A, Di Trapani D, Sozzi F, et al. Beyond the learning curve of the Retzius‐sparing approach for robotassisted laparoscopic radical prostatectomy: oncologic and functional results of the first 200 patients with ? 1 year of follow‐up. Eur Urol 2013; 64: 974‐80
  2. Lim SK, Kim KH, Shin TY et al. Retzius‐sparing robot‐assisted laparoscopic radical prostatectomy: combining the best of retropubic and perineal approaches. BJU Int 2014; 114: 236–44
  3. Chang LW, Hung SC, Hu JC et al. Retzius‐sparing robotic‐assisted radical prostatectomy associated with less bladder neck descent and better early continence outcome. Anticancer Res 2018; 38: 345–51
  4. Menon M, Dalela D, Jamil M et al. Functional recovery, oncologic outcomes and postoperative complications after robot‐assisted radical prostatectomy: an evidence‐based analysis comparing the Retzius sparing and standard approaches. J Urol 2018; 199: 1210–7
  5. Sayyid RK, Simpson WG, Lu C et al. Retzius sparing robotic assisted laparoscopic radical prostatectomy: a safe surgical technique with superior continence outcomes. J Endourol 2017; 31: 1244–50
  6. Chang KD, Abdel Raheem A, Santok GDR et al. Anatomical Retzius‐space preservation is associated with lower incidence of postoperative inguinal hernia development after robot‐assisted radical prostatectomy. Hernia 2017; 21: 555–61
  7. Abdel Raheem A, Kidon C, Alenzi M et al. Predictors of biochemical recurrence after retzius‐sparing robot‐assisted radical prostatectomy: analysis of 359 cases with a median follow‐up of 26 months. Int J Urol 2018; 25: 1006–14

 

Resident’s podcast: Retzius‐sparing robot‐assisted radical prostatectomy

/by

Maria Uloko is a Urology Resident at the University of Minnesota Hospital. In this podcast she discusses the following BJUI Article of the Week:

Retzius‐sparing robot‐assisted radical prostatectomy (RS‐RARP) vs standard RARP: it’s time for critical appraisal

Thomas Stonier*, Nick Simson*, John Davis† and Ben Challacombe‡

 

*Department of Urology, Princess Alexandra Hospital, Harlow, ‡Urology Centre, Guy ’s Hospital, London, UK and †Department of Urology, MD Anderson Cancer Center, Houston, TX, USA

 

Read the full article

Abstract

Since robot‐assisted radical prostatectomy (RARP) started to be regularly performed in 2001, the procedure has typically followed the original retropubic approach, with incremental technical improvements in an attempt to improve outcomes. These include the running Van‐Velthoven anastomosis, posterior reconstruction or ‘Rocco stitch’, and cold ligation of the Santorini plexus/dorsal vein to maximise urethral length. In 2010, Bocciardi’s team in Milan proposed a novel posterior or ‘Retzius‐sparing’ RARP (RS‐RARP), mirroring the classic open perineal approach. This allows avoidance of supporting structures, such as the puboprostatic ligaments, endopelvic fascia, and Santorini plexus, preserving the normal anatomy as much as possible and limiting damage that may contribute to improved postoperative continence and erectile function. There has been much heralding of the excellent functional outcomes in both the medical and the lay press, but as yet no focus or real mention of any potential downsides of this new technique.

Read more Articles of the week

 

BJUI Podcasts now available on iTunes, subscribe here https://itunes.apple.com/gb/podcast/bju-international/id1309570262

 

Article of the Month: Use of machine learning to predict early biochemical recurrence after robot‐assisted prostatectomy

/by

Every month, the Editor-in-Chief selects an Article of the Month from the current issue of BJUI. The abstract is reproduced below and you can click on the button to read the full article, which is freely available to all readers for at least 30 days from the time of this post.

In addition to the article itself, there is an accompanying editorial written by a prominent member of the urological community. This blog is intended to provoke comment and discussion and we invite you to use the comment tools at the bottom of each post to join the conversation.

If you only have time to read one article this week, it should be this one.

Use of machine learning to predict early biochemical recurrence after robot‐assisted prostatectomy

Nathan C. Wong , Cameron Lam, Lisa Patterson and Bobby Shayegan
Division of Urology, Department of Surgery, McMaster University, Hamilton, ON, Canada

Read the full article

Visual abstract created Rebecca Fisher @beckybeckyfish

Abstract

Objectives

To train and compare machine‐learning algorithms with traditional regression analysis for the prediction of early biochemical recurrence after robot‐assisted prostatectomy.

Patients and Methods

A prospectively collected dataset of 338 patients who underwent robot‐assisted prostatectomy for localized prostate cancer was examined. We used three supervised machine‐learning algorithms and 19 different training variables (demographic, clinical, imaging and operative data) in a hypothesis‐free manner to build models that could predict patients with biochemical recurrence at 1 year. We also performed traditional Cox regression analysis for comparison.

= 0.686) and with a univariate regression model (AUC = 0.865).

Results

K‐nearest neighbour, logistic regression and random forest classifier were used as machine‐learning models. Classic Cox regression analysis had an area under the curve (AUC) of 0.865 for the prediction of biochemical recurrence. All three of our machine‐learning models (K‐nearest neighbour (AUC 0.903), random forest tree (AUC 0.924) and logistic regression (AUC 0.940) outperformed the conventional statistical regression model. Accuracy prediction scores for K‐nearest neighbour, random forest tree and logistic regression were 0.976, 0.953 and 0.976, respectively.

Conclusions

Machine‐learning techniques can produce accurate disease predictability better that traditional statistical regression. These tools may prove clinically useful for the automated prediction of patients who develop early biochemical recurrence after robot‐assisted prostatectomy. For these patients, appropriate individualized treatment options can improve outcomes and quality of life.

Read more Articles of the week

Editorial: Can machine‐learning algorithms replace conventional statistics?

/by

Wong et al. [1] evaluate 19 clinical variables (training data) and three supervised machine‐learning algorithms to predict early biochemical recurrence after robot‐assisted prostatectomy. They further compare the areas under the curve (AUCs) resulting from these algorithms with the AUC of a conventional Cox regression model and conclude that the machine‐learning algorithms can produce accurate disease prognosis, perhaps better than a traditional Cox regression model. As the authors state, predictive models have the potential to better individualize care to patients at highest risk of prostate cancer recurrence and progression.

The authors should be commended for their adoption of machine‐learning algorithms to better interpret the vast volumes of clinical data and assess prognosis after robot‐assisted prostatectomy. This should represent another step forward for the management of prostate cancer, where tailored treatment is now largely based on the clinical risk stratification of the disease [2]. Incidentally, we are also in an era where we are seeing aspects of artificial intelligence (machine learning being a subset of it) vastly transform how we view and process data in everyday life. This has been true in medicine as well, particularly for prostate cancer [3].

While our own research group has also evaluated machine‐learning algorithms to process surgeon performance metrics and predict clinical outcomes after robot‐assisted prostatectomy [4], I want to express a word of caution. Utilization of machine learning does not in itself imply automatic superiority over conventional statistics [5] despite literature that has demonstrated so [3]. The success of predictive models in machine learning still relies on the quality of data introduced and careful execution of the analysis. In our experience, it works best when highly experienced clinicians and data scientists are working hand in hand.

Furthermore, I would argue that the results of this present study do not necessarily show that machine learning is superior to conventional statistics, but rather it highlights an inherent advantage of machine learning. While traditional analyses require the a priori selection of a model based on the available data, machine learning has more flexibility for model fitting [6]. Additionally, inclusion of variables in traditional analyses is constrained by the sample size. In contrast, by design, machine learning models thrive on their ability to consider many variables concurrently, and as such, have the potential to detect underlying patterns that may otherwise be undetectable when data are examined effectively in individual silos.

We look forward to the external validation of the methodology described in the present article. Big and diverse data are critical requirements of machine learning. A multi‐institutional, multi‐surgeon cohort is necessary to confirm the findings in this report. A further step from there is the adoption of such prediction models into clinical use. The ultimate question is how improved prognostic data may influence surgeon and patient decisions.

Conflict of Interest

Dr Hung reports personal fees from Ethicon, Inc, outside the submitted work.

References

  1. Wong NC, Lam C, Patterson L, Shayegan B. Use of machine learning to predict early biochemical recurrence following robotic prostatectomy. BJU Int 2019; 123: 51–7
  2. D’Amico AV, Whittington R, Malkowicz SB et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy or interstitial radiation therapy for clinically localized prostate cancer. JAMA 1998; 280: 969–74
  4. Hung AJ, Chen J, Che Z et al. Utilizing machine learning and automated performance metrics to evaluate robot‐assisted radical prostatectomy performance and predict outcomes. J Endourol 2018; 32: 438–445
  5. Kattan MW. Comparison of Cox regression with other methods for determining prediction models and nomograms. J Urol 2003; 170 (6 Pt 2): S6–9
  6. Hung AJ, Chen J, Gill IS. Automated performance metrics and machine learning algorithms to measure surgeon performance and anticipate clinical outcomes in robotic surgery. JAMA Surg 2018; 153: 770–1

RE: National implementation of multi-parametric MRI for prostate cancer detection – recommendations from a UK consensus meeting

/1 Comment/by

Letter to the Editor

National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting [1]

Dear Sir,

Appaya et al report on an expert consensus meeting regarding the implementation of multi-parametric magnetic resonance imaging (mpMRI) for prostate cancer detection [1]. A key item related to ‘who can request an mpMRI’ for patients with suspicion of prostate cancer. The panel unanimously agreed that GPs should not be able to. This is perhaps unsurprising, given the panel was composed entirely of specialists. The authors did consider inviting a GP: however, their assumption was that other than for this question a GP would have little to add. We believe this was a critical omission, as improving access to diagnostic testing in primary care could improve prostate cancer diagnosis,  urology outpatient workloads, patient experiences, and outcomes.

The vast majority of cancer diagnoses, including prostate, occur in symptomatic patients presenting to primary care [2]. GPs already have direct access to diagnostic testing for several other cancer types in the NHS – all endorsed by NICE guidance – including gastroscopy, colonoscopy, flexible sigmoidoscopy, MRI head, ultrasound, and CT abdomen. Each can be requested in primary care, with GPs retaining clinical responsibility for the investigation findings [3].

An oft-raised concern with GP direct access to diagnostic tests is that it will lead to inappropriate referrals. However, a recent systematic review of direct access cancer testing in primary care found no significant difference in pooled cancer conversion rate between GP and specialist requests (except for gastroscopy) and no significant difference in the appropriateness of referrals. Time from referral to testing was shorter with GP direct access testing and patient satisfaction was high [4].

If implementation of pre-biopsy mpMRI truly does reduce the need for biopsy in 27% of men, as suggested from the PROMIS [5] and PRECISION [6] trials, then direct access testing in primary care could significantly reduce referrals for suspected prostate cancer. At a time of limited NHS resources and more ambitious targets for cancer diagnosis and treatment times, this could ease the pressure on Urology departments across the UK.

There are several unanswered questions with regard to the optimal use of pre-biopsy mpMRI: for example, which men to test, how to safely follow-up mpMRI-negative patients, and MRI capacity. Including GPs in these discussions would not just be courteous: it may find better answers.

Dr Samuel Merriel1, Dr Fiona Walter2, Prof Willie Hamilton3

1 Clinical Research Fellow, University of Exeter

2 Principle Researcher in Primary Care Cancer Research, University of Cambridge

3 Professor of Primary Care Diagnostics, University of Exeter

References

  1. Appayya MB, Adshead J, Ahmed HU, Allen C, Bainbridge A, Barrett T, et al. National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting. BJU Int. 2018;122(1):13–25.
  2. Emery JD, Shaw K, Williams B. The role of primary care in early detection and follow-up of cancer. Nat Rev Clin Oncol. 2014;11:38–48.
  3. National Collaborating Centre for Cancer. Suspected cancer [Internet]. NICE. London; 2015.
  4. Smith CF, Tompson AC, Jones N, Brewin J, Spencer EA, Bankhead CR, et al. Direct access cancer testing in primary care: a systematic review of use and clinical outcomes. Br J Gen Pract. 2018;(August):1–10.
  5. Ahmed HU, El-Shater Bosaily A, Brown LC, Gabe R, Kaplan R, Parmar MK, et al. Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study. Lancet [Internet]. 2017;389(10071):815–22.
  6. Kasivisvanathan V, Rannikko AS, Borghi M, Panebianco V, Mynderse LA, Vaarala MH, et al. MRI-Targeted or Standard Biopsy for Prostate-Cancer Diagnosis. N Engl J Med [Internet]. 2018.

Read the article

Reply by the authors

I thank Dr Samuel Merriel and Dr Fiona Walter for writing in about the role of the GP in requesting prostate MRI studies. Indeed, we reported that our consensus panel unanimously agreed that GPs should not be requesting prostate MRI studies [1]. Their letter in follow-up of this offers an opportunity to explain further this recommendation.

We adopted a UCL-RAND based methodology to conduct the consensus meeting. Whilst we would have liked to have broader representation on the panel, one of the key metrics necessitates that a minimum proportion of participants be able to answer a particular question for that question to be valid. Further broadening of the panel risked delivery of this.

Nonetheless, we agree that GP representation in designing and implementing healthcare is invaluable and should not be ignored and could have all of the benefits for prostate cancer diagnostics that are pointed out by Dr Merriel and Dr Walter.

To clarify the consensus panels discussion on this topic; the panel felt that within the current climate there remained many areas that needed standardisation and improvement if we are to realise the benefits highlighted in PROMIS [2] and PRECISION [3] e.g. diagnostic quality of scans, training of radiologists to report scans all the way through to a consensus from urologists of how to manage patients with a specific scan result.  The panel did not believe that GPs were incapable of managing direct referral services for prostate MRI, only that this needed to be introduced in a controlled fashion if we were to be successful in implementing multi-parametric MRI whilst maintaining its performance and value. The panel felt that direct GP referral should therefore be re-discussed once mechanisms to maintain scan quality and standards of reporting were realised across the UK.

Indeed, GPs should be included in the discussion of which men to test, how to follow-up mp-MRI negative patients and those related to MRI capacity – all of these assume that one is looking at a test that is correctly set-up and reported to a specific standard. Perhaps a consensus on management of patients amongst urologists and primary care physicians is a warranted next step.

Shonit Punwani1

1Centre for Medical Imaging, University College London, UK.

References

  1. Appayya MB, Adshead J, Ahmed HU et al. National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting. BJU Int 2018; 122:13-25.
  2. Ahmed HU, El-Shater Bosaily A, Brown LC, Gabe R, Kaplan R, Parmar MK, et al. Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study. Lancet [Internet]. 2017;389(10071):815–22.
  3. Kasivisvanathan V, Rannikko AS, Borghi M, Panebianco V, Mynderse LA, Vaarala MH, et al. MRI-Targeted or Standard Biopsy for Prostate-Cancer Diagnosis. N Engl J Med [Internet]. 2018.

 

 

RE: National implementation of multi-parametric MRI for prostate cancer detection – recommendations from a UK consensus meeting

/by

Letter to the Editor

National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting [1]

Dear Sir,

We congratulate the authors for their efforts in standardising prostate mpMRI. However, we are concerned that the consensus as reported may place substantial pressure on the diagnostic pathway in a rapidly evolving field. Most departments in the West of Scotland and the UK are striving to offer a routine pre-biopsy MRI service, witnessing doubling of the demand on scanning and radiologist time [2], aggravating the nationwide shortage of radiologists [3].

A standard reporting system is vital to the consistent implementation of diagnostic pathways. PI-RADS version 2 focuses on the standardisation of reading, setting out where and when to use each sequence, and has been widely implemented [4]. This, however, has taken time and its learning curve is ongoing within local hospitals [5]. Its main advantage over a clinico-radiological Likert impression is the reduced flexibility of interpretation of radiological parameters, giving those gaining experience in the field a clear set of definitions to work to. We are concerned that enforcing a second system so quickly will discourage radiologists during their learning curves. The panel is entirely correct in stating that even with the use of Likert system, the reporting will be influenced by PI-RADS criteria. While the urology and uro-radiology communities have widely embraced PI-RAD reporting, official switch to Likert reporting may introduce unnecessary reporting subjectivity before individual radiologists are fully experienced with the use of PI-RADS. Furthermore, with the expectation of PI-RADS v3 in the near future, it may be better to adopt PI-RADS v2 now, and compare PI-RADS v3 and Likert system in due course.

The panel describes dynamic contrast enhancement (DCE) as essential component and we agree that mpMRI incorporating the use of contrast is well established, with particular impact on the distinction between PIRADS 3 and 4 lesions. However, a short 9 minute biparameteric MR protocol was as good as a longer and more elaborate protocol using DCE in detecting clinically significant prostate cancer [6]. In NHS Greater Glasgow and Clyde (NHSGGC), we adopted a pragmatic approach, selecting patients for whom contrast may be particularly informative, i.e. those with hip replacement surgery and those who had previous negative biopsies. A contrast MRI prostate scan takes at least 10 minutes longer, and requires nominated radiologist supervision, reducing the overall capacity by 20%. In a 12 month period (2017-2018), 2,333 diagnostic MR prostate scans (20% with contrast) were performed within NHSGGC. For full adoption of contrast prostate MR, it will be necessary to increase the imaging capacity by ~100 supervised imaging lists per annum. Pharmacokinetic modelled DCE parameters (and not visually inspected start of enhancement) have the greatest potential to detect aggressive disease [7,8]. However this approach is currently not deemed feasible for widespread UK use, due to the lack of suitable standardized software.

In summary, we welcome the consensus recommendations for a unified prostate cancer imaging diagnostic strategy, but propose a pragmatic evolving approach that is sympathetic to the constraints of local and regional resources, both in terms of imaging capacity and uro-radiology reporting expertise.

Elizabeth Day1, Amit Patel2, John Morrison2, Thomas Hambrock3, Hing Y Leung1,4

1 Department of Urology, NHS Greater Glasgow and Clyde, Glasgow, G12 0XH

2 Department of Radiology, NHS Greater Glasgow and Clyde, Glasgow, G12 0XH

3 Department of Radiology, The Christie NHS Foundation Trust, Manchester M20 4BX

4 CRUK Beatson Institute for Cancer Research, Glasgow G61 1BD

 

References

  1. Appayya MB, Adshead J, Ahmed HU et al. National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting. BJU Int 2018; 122:13-25.
  2. Day E, Nalagatla S, Shin JS et al. Triaging patients to primary biopsy or prostate MRI based on digital rectal examination improves the detection rate of TRUS biopsy and avoids unnecessary biopsies. JCU 2018.
  3. The Royal College of Radiologists. Scottish patients at risk from radiologist shortages. 2018.
  4. Padhani AR, Weinreb J, Rosenkrantz AB et al. Prostate Imaging-Reporting and Data System Steering Committee: PI-RADS v2 Status Update and Future Directions. Eur Urol 2018.
  5. Hansen NL, Koo BC, Gallagher FA et al. Comparison of initial and tertiary centre second opinion reads of multiparametric magnetic resonance imaging of the prostate prior to repeat biopsy. Eur Radiol 2017; 27:2259-2266.
  6. Kuhl CK, Bruhn R, Krämer N et al. Abbreviated Biparametric Prostate MR Imaging in Men with Elevated Prostate-specific Antigen. Radiology 2017; 285:493-505.
  7. Vos EK, Litjens GJ, Kobus T et al. Assessment of prostate cancer aggressiveness using dynamic contrast-enhanced magnetic resonance imaging at 3 T. Eur Urol 2013; 64:448-55.
  8. Hambrock T, Vos P, Hulsbergen-van de Kaa C et al. Prostate cancer: computer-aided diagnosis with multiparametric 3-T MR imaging–effect on observer performance. Radiology 2013; 266:521-30

Read the article

Reply by the authors

I read with great interest the letter composed by Day et al regarding our article: National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting (Appayya et al [1]).

The authors of the letter highlight several important points that were indeed discussed in detail at the consensus meeting.

The first being the pressure on diagnostic services that is expected to result from the implementation of a strategy adopting multi-parametric MRI prior to biopsy. We concur that there is a real risk that without additional funding and the specific training of specialist radiologists in prostate MRI, our aspirations to implement pre-biopsy MR will meet limited success. The consensus panel did agree that this remains an area where national support and prioritisation will be a key driver of success or failure.

The second point raised by Day et al was in regard to recommendation for the use of Likert verses PI-RADS version 2 reporting systems. This again was a well debated item at the consensus meeting. To clarify the consensus discussion, the panel felt that the PI-RADS v2 system was a good system to use when training to report prostate MRI, specifically as it has a very rigid set of definitions. Indeed, the intention of the panel in recommending Likert reporting was not to disregard the PI-RADS system, but to highlight that when sufficiently experienced we also use other factors in scoring that were not as yet incorporated into PI-RADS v2. Indeed, many experienced radiologists knowingly/unknowingly do not adhere to strict PI-RADS v2 scoring. For example, PI-RADS v2 recommends that clinical details (including PSA) should not influence interpretation yet it is almost ubiquitous that we report in light of the PSA and are now recommending PSA density measures to help us guide practice. Furthermore, many centres in the UK do not perform DCE MRI, yet report score as a PI-RADS v2– although DCE MRI as Day et al point out is a key component of the PI-RADS v2 scoring system. Any reports produced for patients that have had previous treatment cannot be scored by PI-RADS v2 criteria, yet there are ongoing examples of radiologists stating a PI-RADS v2 score within such reports.

In reality, Day et al are correct in pointing out that PI-RADS v3 may indeed resolve many of these issues.  PI-RADS v3 was not as far developed when the consensus meeting took place.

With regard to a bi-parametric verses multi-parametric approach; I would agree that most significant tumours will be detected with a bi-parametric MRI and that it will be a small minority that would benefit from DCE MRI. The difficulty to date has been that DCE MRI applications are very varied across studies and have generally concentrated on the use of pharmacokinetic parameters – which has limited studies looking at the value of DCE-MRI. Indeed, such protocols are difficult to themselves apply. The panel felt that a high resolution short DCE protocol that lasted 3 minutes could be implemented with visual inspection of the early arterial phase image for the detection of early enhancement indicative tumour. They also acknowledged that further research is required to establish the benefits of DCE MRI.

I would like to thank Day et al for their letter, as it is important to highlight these areas from the consensus paper in order (i) to recognise where national support is required, (ii) to help clarify any areas which more detailed description could not be provided within the paper, and (iii) to continue to recognise areas where research and further technique refinement is required.

Shonit Punwani1

1Centre for Medical Imaging, University College London, UK.

References

  1. Appayya MB, Adshead J, Ahmed HU et al. National implementation of multi-parametric magnetic resonance imaging for prostate cancer detection – recommendations from a UK consensus meeting. BJU Int 2018; 122:13-25.

 

 

© 2024 BJU International. All Rights Reserved.