Screening, Diagnosis, Treatmen, Monitoring and Support
By reviewing recent innovations in prostate cancer I hope to throw some light on what optimal prostate cancer care could and should look like. Relevant for professionals, but also patients and their loved ones, so that you can raise your expectations and hopefully play your part in accelerating progress in the field.
At Healthskouts, our work is driven by the belief that new technologies have the potential to significantly advance healthcare and improve outcomes for patients. Today. Not some hypothetical time in the future.
We feel a real sense of urgency, even frustration, that things aren’t moving as fast as they should. Scan tech activity targeting any disease or therapeutic area and one quickly feels overwhelmed by the boom in innovation – and subsequent disappointment that such innovation is not more assertively deployed in clinical practice yet. Prostate cancer is a case in point.
Globally, prostate cancer is the second most frequent cancer and the fifth leading cause of cancer death among men. In the US, ⅛ men will develop the disease. As such, the disease will likely impact us all at a personal level, either directly or via people we care about, irrespective of whether we have a professional or academic interest in the disease. So it was for me unfortunately.
Following this patient’s journey from a distance, I have been in (thankful) awe seeing the commitment of the patient’s physicians and the sophistication of the treatment administered. Simultaneously, I have been horrified how wrong things can go too, in ways that really are preventable, for example around PSA-based screening (missed the cancer), treatment guidelines (contradictory, depending on the physician consulted), the monitoring & management of radiotherapy-related complications (utter failure), and the attention to mental health issues (non-existent). All of these problems, and the many more that millions of other men experience today, could be prevented or handled more effectively if the right tools and technologies were properly deployed today.
Using an airline industry analogy, if an ‘optimal’ flight is nearly perfectly safe, on time and comfortable, what would optimal prostate cancer care look like? And let’s look beyond current Standards of Care and Clinical Guidelines, because these look more like preflight checks for physicians than actual standards of care patients ought to expect today. The analogy can be taken further. In the same way airlines compete on factors such as price, punctuality, routes and passenger experience (‘delight’!), how could life sciences companies and providers compete on cost-efficiency, outcomes, treatment options and delightful patient experiences? And how could all of that be made transparent to patients, since in Belgium at least, such transparency is almost non-existent.
In this article, we review a selection of recent innovations that throw light on what such optimal or ideal prostate cancer care could/should look like.
For many men, screening for prostate cancer begins in their 50s, with their GP administering a blood-based PSA test and/or a digital rectal examination (DRE). If PSA or DRE results are concerning, patients are often referred to a urologist for further testing, which could entail a biopsy and possibly an MRI. PSA tests have, however, in recent years become very controversial – even though they are still routinely used in many primary care settings.
While in some cases, a low PSA score may actually predict aggressive prostate cancer (a false negative – as occurred in the case mentioned above), the larger problem is the lack of specificity toward clinically significant disease. PSA levels can be raised for many non-cancer reasons such as infection and even exercise, leading to unnecessary biopsies and associated complications. Moreover, PSA screening has led to the overdiagnosis of clinically insignificant disease (low-risk, localized disease that can be managed appropriately using ‘active surveillance’) and overtreatment, in turn associated with unnecessary biopsy complications and treatment complications such as sexual and urinary difficulties. Research findings that PSA screening saves very few lives at great cost has led the US Preventive Services Task Force in 2012 to issue a grade D recommendation against PSA-based prostate cancer screening. Also in Belgium, PSA screening is not recommended nor reimbursed anymore due to the problems of overdiagnosis and overtreatment. Since the US decision, screening rates of men in the US have declined, but unfortunately this appears to be leading to increases in aggressive high risk prostate cancers in the US.
More and more experts and guidelines therefore argue for a ‘smarter’ screening approach, using better tests or test combinations to improve sensitivity and specificity, or whereby the goal of testing is to identify men who are more likely to have clinically significant disease and would benefit from early diagnosis and treatment, such as the use of anti-androgen therapy.
This is where new technologies come into play. New blood tests such as 4Kscore® and prostate health index (PHI) that combine different prostate kallikreins, including Intact PSA and hk2, have been shown to be more effective in detecting aggressive prostate cancers. CE-IVD-labeled Proclarix is another such blood test that can be combined with a PSA test showing promise. Some urine biomarkers (e.g. PCA3, HOXC6, in tests developed by companies like MDxHealth and ExoMed) are also proving more accurate than PSA screening.Blood tests that combine individual tests, including genetic tests (e.g. Stockholm-3 model – STHLM3), also show great promise.
Genetic testing is particularly promising because while most prostate cancers don’t appear to have a hereditary predisposition, some aggressive and metastatic cancers clearly do. In fact, mutations in genes related to DNA repair (BRCA1, BRCA2, CHECK2, ATM), development of the prostate gland (HOXB13) and Lynch syndrome – may be responsible for up to 10% of all prostate cancers, especially aggressive cancers. Most of these genes have clear implications for early cancer screening and the targeted treatment of advanced prostate cancer. As a result, there are increasing calls for the use of multigene panel testing, both for diagnosed and as yet undiagnosed patients (who, for example, have a family history of aggressive prostate cancer). Stratify Genomics, for example, has developed a prostate cancer genetic test (Prompt PGS) that anyone can order online. More on the horizon are liquid biopsy tests based on immunogenomics platforms. Immunis.ai is one such company that is using RNA sequencing technology and machine learning for the development of prostate cancer-specific diagnostic algorithms.
These and more such innovations in the biomarker field, also through cross cancer learnings, will clearly contribute towards improved screening, diagnosis and treatment of prostate cancer, but adoption in routine clinical practice is slow. One problem with such an explosion in innovation is that clinicians and payers have difficulty evaluating and prioritising the new biomarkers (and combinations of markers). A lot more validation research, including head-to-head comparisons, is needed. Access and affordability are also issues. The danger is that validation researchers and regulators simply cannot keep pace with the acceleration in biomarker discovery. Our team at Healthskouts is particularly interested in novel approaches for accelerating the adoption of new biomarkers (and other innovations in the digital health field) in trials and clinical practice. Let us know if you have insights on this!
The standard diagnostic test for prostate cancer is a prostate biopsy which in most cases involves placing an ultrasound probe in the rectum to visualize the prostate, and a needle is used to sample the prostate in multiple areas. The biopsy is used to make a diagnosis and is also used to ‘stage’ the disease, i.e. determine how aggressive the cancer is to what extent it has spread. The problems with biopsies, besides being frightening and unpleasant for patients, is that they can cause infections and that they can sometimes miss clinically significant cancers.
Fortunately, advances in MRI technology enable clinicians to use this imaging tech to better evaluate a patient’s risk for cancer (before a biopsy, to further prevent unnecessary biopsies) and to guide biopsies, ensuring that they accurately target abnormal areas in the prostate. As a result, EU and US guidelines now recommend MRI before biopsy, leading to large increases in the caseload of radiology practices.
AI technology is now coming to the rescue to automate and improve the analysis of MRI images. Companies such as Ezra, Quantib, Lucida Medical and Bot Image have developed AI systems for automatically detecting prostate cancer from MRI and targeting areas for biopsy. Ezra isn’t just an AI company; it is a radiology practice specialised in MRI scans for cancer screening. The company targets health consumers directly and through employers to book annual full body scans.
If a decision is made to go ahead with a prostate biopsy, the pathology lab comes into play, with the task to not only diagnose the disease but to ‘grade’ (using the Gleason grading system – the higher the Gleason score the more likely the cancer will grow and spread quickly) it as best as possible since this informs the treatment and management strategy.
Also here, AI technology is coming into play. Paige Prostate was the first AI tool approved by the FDA for in-vitro diagnostic use. It identifies areas on the biopsy image with the highest likelihood of having cancer. Israeli startup, Ibex Medical Analytics, and South Korean company, Deep Bio, are similarly developing AI tools that support pathologists in their work. The stated goals of these solutions are to improve pathologists’ efficiency and reduce error rates.
Companies such as PathAI and Nucleai take AI a step further by using it to identify pathology-based biomarkers that predict response to therapy. These companies are therefore working in partnership with life sciences companies (e.g. PathAI with Roche, Nuclei with Merck KGaA) to identify new biomarkers and develop these as companion diagnostics. (Note that these, and many more AI Software as Medical Device products, are in our curated solution libraries).
Despite advances in the treatment of prostate cancer, the therapy options are still surprisingly limited, especially for localised tumors. Patients with low risk localised cancers usually are not treated but placed on ‘active surveillance’. A key challenge here is developing indicators for determining which men should go on active surveillance and which require treatment. The liquid biopsy, MRI and pathology based techniques discussed in the previous sections play an obvious role here.
Another major issue is how to monitor patients over time. Repeated biopsies carry risks and are obviously burdensome for patients, leading to adherence problems over time. Hence there is a need to develop more accurate, non-invasive and personalised monitoring approaches. Risk calculators such as the Canary Prostate Active Surveillance Study and PRIAS model have been developed to personalise the timing of biopsies and decide when and if to switch to active treatment. Such models are, however, still based on PSA tests, DREs and prostate biopsies, so it will be interesting to see new calculators being developed that include – or are based solely on – more advanced liquid biopsy tests and MRI.
If a decision is made to pursue active treatment, the best possible treatment options need to be selected. Options here are surgery (e.g. removal of the prostate), radiotherapy, chemotherapy, hormonal therapy (androgen-directed therapies – ADT) and immunotherapy. Making the right choices is challenging because there is substantial heterogeneity in histology, genetics and clinical outcome. Also, many new types of increasingly complex and precise treatments are being developed. While clinical decision making is still mostly based on PSA levels, tumor staging and pathologic Gleason scores, the opportunities for Precision Medicine are rapidly expanding. Epic Sciences, for example, has developed a blood test that identifies patients who are likely to be resistant to ADTs (hormone therapy). Also many of the drugs for late stage prostate cancer currently in development are precision medicine drugs and will require companion diagnostic tests.
Surgery is moving forward too, with advances in less invasive ablation techniques (focal therapy). For example, Francis Medical is developing a vapor ablation therapy to treat prostate cancer. Avenda Health has developed AI-powered therapy using minimally invasive laser ablation, and Profound Medical has got 510(k) clearance for its Tulsa-Pro device, an ablation system that uses real-time MRI alongside robotically-driven transurethral ultrasound to ablate prostate tissue without incisions or radiation.
Entirely new approaches based on the microbiome also show some promise. Research has shown that men with hormone-resistant prostate cancer have a different gut bacteria pattern than patients who respond well to hormone therapy (ADT). Possible outcomes of this research are the development of a microbiome test that could predict the effectiveness of ADT therapy, and further along the development of a probiotic or fecal transplant aimed at preventing resistance to ADT treatment.
One important consequence of this innovation in prostate cancer treatment and the associated shift to precision medicine is that medical decision making in prostate cancer has become increasingly complex, with no single physician having the brain power to process and store the avalanche of information about new diagnostics tests and therapies. This is where Clinical Decision Support Systems come into play. To address these issues, Siemens Healthineers has developed a CDSS system called AI Pathway Companion (not specific to prostate cancer) that integrates data from medical records, patient engagement & ePROs, clinical guidelines and information about diagnostics tests and therapeutic options, to generate tailored clinical dashboards and decision support for physicians. Navify from Roche offers a similar platform.
Monitoring and Support
While it may seem obvious that patients with a severe illness such as prostate cancer need close monitoring and continuous support, in practice most urologists and oncologists only have visibility over their patients during the minutes spent in the consultation room. The other 99,9% of the time, patients are often forced to deal with problems such as treatment complications and mental health issues on their own. While remote monitoring using digital patient diaries, ePRO and Patient Engagement platforms is gradually taking hold in many hospitals, it does not appear to be in routine use in many prostate cancer clinics (there is very little data on this). In our view, this is a major gap, not just a gap in the care of individual patients, but also a missed opportunity to gather detailed, comprehensive and continuous outcomes data. Collecting such real-world data and making it available for analysis and research is crucial, especially in the coming era of precision medicine. Fortunately, there really is nothing stopping clinics from implementing a monitoring setup: the content, tools and evidence already exist.
With regard to the content, numerous validated instruments have been developed to collect patient reported outcomes and assess relevant Quality of Life parameters. The ICHOM Standard Set for Advanced Prostate Cancer, for example, is an international recommendation of the outcomes that matter most to patients with Advanced Prostate Cancer. The authors (all clinicians) urge all providers around the world to start measuring these outcomes to better understand how to improve the lives of their patients.
Studies have also shown the feasibility of using digital health applications to report quality of life changes in real-time during prostate radiotherapy. In lung cancer, web-based symptom monitoring has even been shown to be associated with increased survival. A possible explanation is that symptoms suggesting adverse events or recurrence were detected earlier.
The tools for remote monitoring are also available. Numerous oncology-specific patient engagement platforms have been developed, most notably Noona (now part of Varian, in turn part of Siemens Healthineers) and Vine Health in the UK. In Belgium, Awell has developed an automated prostate cancer pathway for monitoring and managing patients according to the ICHOM standard. Monitoring with long boring questionnaires should also be a thing of the past. Besides the fact that the user interface design of modern patient engagement platforms is superb, new tools such as PatchAI (teamed up with Roche to support oncology patients) are using AI-driven virtual assistants to engage patients.
A new generation of digital health tools go beyond monitoring and actually intervene with evidence-based Digital Therapeutics. Oleena, for example, is a Digital Therapeutic in oncology that not only enables the patient to report symptoms but also intervenes by evaluating the symptom severity and providing a personalised recommendation. The patient then self-manages their symptom based on the recommendation, and if the severity is too high, the system connects the patient to the care team.
Toward Integrated Digital Care Pathways
We hope that the prostate cancer innovations reviewed here offer an inspiring glimpse of what prostate cancer care will and ought to look like, sooner rather than later – especially if we succeed in stitching together all these solutions in an integrated, end-to-end care pathway. In complete contrast to the case I witnessed in close proximity, this future comprises a journey where the patient and his GP – through genetic testing earlier in life – were prepared and vigilant about the increased risk of aggressive prostate cancer. Where regular liquid biopsy tests caught the disease at an early stage, and where accurate MRI-based staging and minimally invasive ablation surgery stopped the disease in its tracks. Where the patient’s physicians agreed on the best possible course of action, informed by the best and latest clinical guidelines and evidence coming from large-scale RWE datasets and trials. Meaning that if the disease had advanced, the precision therapies with best predicted response would have been deployed, without having to resort to trial & error. Treatment-related complications would have been spotted and addressed immediately. And finally, the data generated by this case would feed into international cohorts and RWE databases, driving drug discovery and improved clinical guidelines.
As is clear, however, the key obstacles to achieving such levels of care are not technical but are more to do with business models, systems, culture and management. Science and technological progress marches on. But implementing such innovations in clinical practice is another matter. The key challenge – and opportunity – in our opinion is the ‘stitching together’ of innovations in end-to-end, cradle-to-grave digital care paths. Such paths will traverse different levels of care and involve multiple stakeholders, including providers, life sciences companies, government and of course the patient – who ultimately is empowered to make informed, educated choices.
If you think we missed something in this story about prostate cancer innovations, or if you have a relevant solution that we should be aware of and add to our curated database of digital health solutions, please don’t hesitate to reach out! If you think we can help you improve prostate cancer and bring delight in the patient journey, then we’d love to hear from you. Finally, a shout out to the most excellent Movember movement, for their efforts in raising awareness about men’s health and funding important research in prostate cancer, especially around biomarkers.