Though still in trials, early results show tests may be more accurate than traditional clinical laboratory tests for detecting prostate cancer
Within weeks of each other, different research teams in the US and UK published findings of their respective efforts to develop a better, more accurate clinical laboratory prostate cancer test. With cancer being a leading cause of death among men—second only to heart disease according to the Centers for Disease Control and Prevention (CDC)—new diagnostics to identify prostate cancer would be a boon to precision medicine treatments for the deadly disease and could save many lives.
Thus, these are two different pathways toward the goal of achieving earlier, more accurate diagnosis of prostate cancer, the holy grail of prostate cancer diagnosis.
“There is currently no single test for prostate cancer, but PSA blood tests are among the most used, alongside physical examinations, MRI scans, and biopsies,” said Dmitry Pshezhetskiy, PhD (above), Professorial Research Fellow at University of East Anglia and one of the authors of the UEA study. “However, PSA blood tests are not routinely used to screen for prostate cancer, as results can be unreliable. Only about a quarter of people who have a prostate biopsy due to an elevated PSA level are found to have prostate cancer. There has therefore been a drive to create a new blood test with greater accuracy.” With the completion of the US and UK studies, clinical laboratories may soon have a new diagnostic test for prostate cancer. (Photo copyright: University of East Anglia.)
East Anglia’s Research into a More Accurate Blood Test
Scientists at the University of East Anglia (UEA) worked with researchers from Imperial College in London, Imperial College NHS Trust, and Oxford BioDynamics to develop a new precision medicine blood test that can detect prostate cancer with greater accuracy than current methods.
The researchers evaluated their test in a pilot study involving 147 patients. They found their testing method had a 94% accuracy rate, which is higher than that of PSA testing alone. They discovered their test significantly improved the overall detection of prostate cancer in men who are at risk for the disease.
“When tested in the context of screening a population at risk, the PSE test yields a rapid and minimally invasive prostate cancer diagnosis with impressive performance,” Dmitry Pshezhetskiy, PhD, Professorial Research Fellow at UEA and one of the authors of the study told Science Daily. “This suggests a real benefit for both diagnostic and screening purposes.”
The UK scientists hope their test can eventually be used in everyday clinical practice as there is a need for a highly accurate method for prostate cancer screening that does not subject patients to unnecessary, costly, invasive procedures.
Cedars-Sinai’s Research into Nanotechnology Cancer Testing
Researchers from Cedars-Sinai Cancer took a different approach to diagnosing prostate cancer by developing a nanotechnology-based liquid biopsy test that detects the disease even in microscopic amounts.
Their test isolates and identifies extracellular vesicles (EVs) from blood samples. EVs are microscopic non-reproducing protein and genetic material shed by all cells. Cedars-Sinai’s EV Digital Scoring Assay accurately extracts EVs from blood and analyzes them faster than similar currently available tests.
“This research will revolutionize the liquid biopsy in prostate cancer,” said oncologist Edwin Posadas, MD, Medical Director of the Urologic Oncology Program and co-director of the Experimental Therapeutics Program in Cedars-Sinai Cancer in a press release. “The test is fast, minimally invasive and cost-effective, and opens up a new suite of tools that will help us optimize treatment and quality of life for prostate cancer patients.”
The researchers tested blood samples from 40 patients with prostate cancer. They found that their EV test could distinguish between cancer localized to the prostate and cancer that has spread to other parts of the body.
Microscopic cancer deposits, called micrometastases, are not always detectable, even with advanced imaging methods. When these deposits spread outside the prostate area, focused radiation cannot prevent further progression of the disease. Thus, the ability to identify cancer by locale within the body could lead to new precision medicine treatments for the illness.
“[The EV Digital Scoring Assay] would allow many patients to avoid the potential harms of radiation that isn’t targeting their disease, and instead receive systemic therapy that could slow disease progression,” Posadas explained.
Other Clinical Laboratory Tests for Prostate Cancer Under Development
According to the American Cancer Society, the number of prostate cancer cases is increasing. One out of eight men will be diagnosed with the illness during his lifetime. Thus, developers have been working on clinical laboratory tests to accurately detect the disease and save lives for some time.
In “University of East Anglia Researchers Develop Non-Invasive Prostate Cancer Urine Test,” Dark Daily reported on a urine test also developed by scientists at the University of East Anglia that clinical laboratories can use to not only accurately diagnose prostate cancer but also determine whether it is an aggressive form of the disease.
And in “UPMC Researchers Develop Artificial Intelligence Algorithm That Detects Prostate Cancer with ‘Near Perfect Accuracy’ in Effort to Improve How Pathologists Diagnose Cancer ,” we outlined how researchers at the University of Pittsburgh Medical Center (UPMC) working with Ibex Medical Analytics in Israel had developed an artificial intelligence (AI) algorithm for digital pathology that can accurately diagnose prostate cancer. In the initial study, the algorithm—dubbed the Galen Prostate AI platform—accurately detected prostate cancer with 98% sensitivity and 97% specificity.
More research and clinical trials are needed before the new US and UK prostate cancer testing methods will be ready to be used in clinical settings. But it’s clear that ongoing research may soon produce new clinical laboratory tests and diagnostics for prostate cancer that will steer treatment options and allow for better patient outcomes.
As the worldwide demand for histopathology services increases faster than the increase in the number of anatomic pathologist and histopathologists, a DP platform that suggests courses of treatments may be a boon to cancer diagnostics
Europe may become Ground Zero for the widespread adoption of whole-slide imaging (WSI), digital pathology (DP) workflow, and the use of image-analysis algorithms to make primary diagnoses of cancer. Several forward-looking histopathology laboratories in different European countries are moving swiftly to adopt these innovative technologies.
Clinical laboratories and anatomic pathology groups worldwide have watched digital pathology tools evolve into powerful diagnostic aids. And though not yet employed for primary diagnoses, thanks to artificial intelligence (AI) and machine learning many DP platforms are moving closer to daily clinical use and new collaborations with pathologists who utilize the technology to confirm cancer and other chronic diseases.
Now, Swiss company Unilabs, one of the largest laboratory, imaging, and pathology diagnostic developers in Europe, and Israel-based Ibex Medical Analytics, developer of AI-based digital pathology and cancer diagnostics, have teamed together to deploy “Ibex’s multi-tissue AI-powered Galen platform” across 16 European nations, according to a Unilabs press release.
Though not cleared by the federal Food and Drug Administration (FDA) for clinical use in the US, the FDA recently granted Breakthrough Device Designation to Ibex’s Galen platform. This designation is part of the FDA’s Breakthrough Device Program which was created to help expedite the development, assessment, and review of certain medical devices and products that promise to provide for more effective treatment or diagnosis of life-threatening or irreversibly debilitating diseases or conditions.
Benefits of AI-Digital Pathology to Pathologists, Clinical Labs, and Patients
According to Ibex’s website, the Galen DP platform uses AI algorithms to analyze images from breast and prostate tissue biopsies and provide insights that help pathologists and physicians determine the best treatment options for cancer patients.
This will, Ibex says, give pathologists “More time to dedicate to complex cases and research,” and will make reading biopsies “Less tedious, tiring, and stressful.”
Patients, according to Ibex, benefit from “Increased diagnostic accuracy” and “More objective results.”
And pathology laboratories benefit from “Increased efficiency, decreased turnaround time, and improved quality of service,” Ibex claims.
According to the press release, AI-generated insights can include “case prioritization worklists, cancer heatmaps, tumor grading and measurements, streamlined reporting tools and more.”
This more collaborative approach between pathologists and AI is a somewhat different use of digital pathology, which primarily has been used to confirm pathologists’ diagnoses, rather than helping to identify cancer and suggest courses of treatment to pathologists.
“This cutting-edge AI technology will help our teams quickly prioritize urgent cases, speed up diagnosis, and improve quality by adding an extra set of digital eyes,” said Christian Rebhan, MD, PhD (above), Chief Medical and Operations Officer at Unilabs, in the press release. “When it comes to cancer, the earlier you catch it, the better the prognosis—so getting us critical results faster will help save lives.” (Photo copyright: Unilabs.)
AI-based First and Second Reads
The utilization of the Galen platform will first be rolled out nationally in Sweden and then deployed in sixteen other countries. The AI-based DP platform is CE marked in the European Union for breast and prostate cancer detection in multiple workflows.
“The partnership with Ibex underlines Unilabs’ pioneering role in Digital Pathology and represents yet another step in our ambition to become the most digitally-enabled provider of diagnostic services in Europe,” Rebhan stated.
The Ibex website explains that the Galen platform is divided into two parts—First Read and Second Read:
The First Read “is an AI-based diagnostics application that aims to help pathologists significantly reduce turnaround time and improve diagnostic accuracy. The application uses a highly accurate AI algorithm to analyze slides prior to the pathologist and provides decision support tools that enable focusing on cancerous slides and areas of interest, streamline reporting, improve lab efficiency, and increase diagnostic confidence.”
The Second Read “is an AI-based diagnostics and quality control application that helps pathologists enhance diagnostic accuracy with no impact on routine workflow. The application analyzes slides in parallel with the pathologist and alerts in case of discrepancies with high clinical significance (e.g., a missed cancer), thereby providing a safety net that reduces error rates and enables a more efficient workflow.”
“Ibex is transforming cancer diagnosis with innovative AI solutions across the diagnostic pathway,” said Joseph Mossel, Chief Executive Officer and co-founder of Ibex, in the press release. “We are excited to partner with Unilabs to deploy our AI solutions and empower their pathologists with faster turnaround times and quality diagnosis. This cooperation follows a thorough evaluation of our technology at Unilabs and demonstrates the robustness and utility of our platform for everyday clinical practice.”
Use of AI in Pathology Increases as Number of Actual Pathologists Declines
Developers like Unilabs and Ibex believe that DP platforms driven by AI image analysis algorithms can help pathologists be more productive and can shorten the time it takes for physicians to make diagnoses and issue reports to patients.
This may be coming at a critical time. As nations around the globe face increasing shortages of pathologists and histopathologists, the use of AI in digital pathology could become more critical for disease diagnosis and treatment.
A 2019 Medscape survey stated that “One-third of active pathologists are burned out,” and that many pathologists are on the road to retirement.
And in the same year, Fierce Healthcare noted that in a 2013 study, “researchers found that more than 40% of pathologists were 55 or older. They predicted that retirements would reach their apex in 2021. Consequently, by the end of next decade, the United States will be short more than 5,700 pathologists.”
Dark Daily previously reported on the growing global shortage of pathologists going back to 2011.
Even China is struggling to keep up with demand for anatomic pathologists. In 2017, Dark Daily wrote, “China is currently facing a severe shortage of anatomic pathologists, which blocks patients’ access to quality care. The relatively small number of pathologists are often overworked, even as more patients want access to specialty care for illnesses. Some hospitals in China do not even have pathologists on staff. Thus, they rely on understaffed anatomic pathology departments at other facilities, or they use imaging only for diagnoses.”
Thus, it may be time for an AI-driven digital platform to arrive that can speed up and increase the accuracy of the cancer diagnostics process for pathologists, clinical laboratories, and patients alike.
There are multiple companies rapidly developing AI, machine learning, and image analysis products for diagnosing diseases. Pathologists should expect progress in this field to be ongoing and new capabilities regularly introduced into the market.
Hello primary diagnosis of digital pathology images via artificial intelligence! Goodbye light microscopes!
Digital pathology is poised to take a great leap forward. Within as few as 12 months, image analysis algorithms may gain regulatory clearance in the United States for use in primary diagnosis of whole-slide images (WSIs) for certain types of cancer. Such a development will be a true revolution in surgical pathology and would signal the beginning of the end of the light microscope era.
A harbinger of this new age of digital pathology and automated image analysis is a press release issued last week by Ibex Medical Analytics of Tel Aviv, Israel. The company announced that its Galen artificial intelligence (AI)-powered platform for use in the primary diagnosis of specific cancers will undergo an accelerated review by the Food and Drug Administration (FDA).
FDA’s ‘Breakthrough Device Designation’ for Pathology AI Platform
Ibex stated that “The FDA’s Breakthrough Device Designation is granted to technologies that have the potential to provide more effective treatment or diagnosis of life-threatening diseases, such as cancer. The designation enables close collaboration with, and expedited review by, the FDA, and provides formal acknowledgement of the Galen platform’s utility and potential benefit as well as the robustness of Ibex’s clinical program.”
“All surgical pathologists should recognize that, once the FDA begins to review and clear algorithms capable of using digital pathology images to make an accurate primary diagnosis of cancer, their daily work routines will be forever changed,” stated Robert L. Michel, Editor-in-Chief of Dark Daily and its sister publication The Dark Report. “Essentially, as FDA clearance is for use in clinical care, pathology image analysis algorithms powered by AI will put anatomic pathology on the road to total automation.
“Clinical laboratories have seen the same dynamic, with CBCs (complete blood counts) being a prime example. Through the 1970s, clinical laboratories employed substantial numbers of hematechnologists [hematechs],” he continued. “Hematechs used a light microscope to look at a smear of whole blood that was on a glass slide with a grid. The hematechs would manually count and record the number of red and white blood cells.
“That changed when in vitro diagnostics (IVD) manufacturers used the Coulter Principle and the Coulter Counter to automate counting the red and white blood cells in a sample, along with automatically calculating the differentials,” Michel explained. “Today, only clinical lab old-timers remember hematechs. Yet, the automation of CBCs eventually created more employment for medical technologists (MTs). That’s because the automated instruments needed to be operated by someone trained to understand the science and medicine involved in performing the assay.”
Primary Diagnosis of Cancer with an AI-Powered Algorithm
Surgical pathology is poised to go down a similar path. Use of a light microscope to conduct a manual review of glass slides will be supplanted by use of digital pathology images and the coming next generation of image analysis algorithms. Whether these algorithms are called machine learning, computational pathology, or artificial intelligence, the outcome is the same—eventually these algorithms will make an accurate primary diagnosis from a digital image, with comparable quality to a trained anatomic pathologist.
How much of a threat is automated analysis of digital pathology images? Computer scientist/engineer Ajit Singh, PhD, a partner at Artiman Ventures and an authority on digital pathology, believes that artificial intelligence is at the stage where it can be used for primary diagnosis for two types of common cancer: One is prostate cancer, and the other is dermatology.
On June 17, Ajit Singh, PhD (above), Partner at Artiman Ventures, will lead a special webinar and roundtable discussion for all surgical pathologists and their practice administrators on the coming arrival of artificial intelligence-powered algorithms to aid in the primary diagnosis of certain cancers. Regulatory approval for such solutions may happen by the end of this year. Such a development would accelerate the transition from light microscopes to a fully digital pathology workflow. Singh is shown above addressing the 2018 Executive War College. (Photo copyright: The Dark Report.)
“It is now possible to do a secondary read, and even a first read, in prostate cancer with an AI system alone. In cases where there may be uncertainty, a pathologist can review the images. Now, this is specifically for prostate cancer, and I think this is a tremendous positive development for diagnostic pathways,” he added.
Use of Digital Pathology with AI-Algorithms Changes Diagnostics
Pathologists who are wedded to their light microscopes will want to pay attention to the impending arrival of a fully digital pathology system, where glass slides are converted to whole-slide images and then digitized. From that point, the surgical pathologist becomes the coach and quarterback of an individual patient’s case. The pathologist guides the AI-powered image analysis algorithms. Based on the results, the pathologist then orders supplementary tests appropriate to developing a robust diagnosis and guiding therapeutic decisions for that patient’s cancer.
In his interview with The Dark Report, Singh explained that the first effective AI-powered algorithms in digital pathology will be developed for prostate cancer and skin cancer. Both types of cancer are much less complex than, say, breast cancer. Moreover, the AI developers have decades of prostate cancer and melanoma cases where the biopsies, diagnoses, and downstream patient outcomes create a rich data base from which the algorithms can be trained and tuned.
This webinar is organized as a roundtable discussion so participants can interact with the expert panelists. The Chair and Moderator is Ajit Singh, PhD, Adjunct Professor at the Stanford School of Medicine and Partner at Artiman Ventures.
The panelists (above) represent academic pathology, community hospital pathology, and the commercial sector. They are:
Because the arrival of automated analysis of digital pathology images will transform the daily routine of every surgical pathologist, it would be beneficial for all pathology groups to have one or more of their pathologists register and participate in this critical webinar.
The roundtable discussion will help them understand how quickly AI-powered image analysis is expected be cleared for use by the FDA in such diseases as prostate cancer and melanomas. Both types of cancers generate high volumes of case referrals to the nation’s pathologists, so potential for disruption to long-standing client relationships, and the possible loss of revenue for pathology groups that delay their adoption of digital pathology, can be significant.
On the flip side, community pathology groups that jump on the digital pathology bandwagon early and with the right preparation will be positioned to build stronger client relationships, increase subspecialty case referrals, and generate additional streams of revenue that boost partner compensation within their group.
Also, because so many pathologists are working remotely, Dark Daily has arranged special group rates for pathology practices that would like their surgical pathologists to participate in this important webinar and roundtable discussion on AI-powered primary diagnosis of pathology images. Inquire at info@darkreport.com or call 512-264-7103.
Working from tissue slides similar to those used by surgical pathologists, the algorithm accurately detects prostate cancer with an impressive 98% sensitivity
It could be that a new milestone has been reached on the road to using artificial intelligence (AI) to help anatomic pathologists diagnose cancer and other diseases. A research collaboration between a major American university and an Israeli company recently published a study about the ability of an AI algorithm to correctly diagnose prostate cancer.
The scientists trained the Galen Prostate AI to recognize prostate cancer by having it examine images from over a million parts of stained tissue slides taken from patient biopsies. Expert pathologists labeled each image to teach the algorithm how to distinguish between healthy and abnormal tissue. The AI was then tested on 1,600 different tissue slide images that had been collected from 100 patients seen at UPMC who were suspected of having prostate cancer.
“Humans are good at recognizing anomalies, but they have their own biases or past experience,” said Rajiv Dhir, MD, Chief Pathologist and Vice Chair of Pathology at UPMC Shadyside Hospital, Professor of Biomedical Informatics at University of Pittsburgh, and senior author of the study, in a UPMC news release. “Machines are detached from the whole story. There’s definitely an element of standardizing care.”
The image above is “of prostate cancer (represented by the heatmap) detected by the Ibex Galen Prostate [AI] solution on a biopsy that was previously diagnosed as benign by the pathologist,” stated an Ibex news release announcing the UPMC study. (Photo copyright: Ibex.)
UPMC Algorithm Goes Beyond Cancer Detection, Exceeds Human Pathologists
The researchers also noted that this is the first algorithm to extend beyond cancer detection. It reported high performance for tumor grading, sizing, and invasion of surrounding nerves—clinically important features of pathology reports.
“Algorithms like this are especially useful in lesions that are atypical,” Dhir said. “A nonspecialized person may not be able to make the correct assessment. That’s a major advantage of this kind of system.”
The algorithm also flagged six slides as potentially containing abnormal tissue that were not flagged by human pathologists. However, the researchers pointed out that this difference does not mean the AI is better than humans at detecting prostate cancer. It is probable, for example, that the pathologists simply saw enough evidence of malignancy elsewhere in the patients’ samples to recommend treatment.
Other Studies Where AI Detected Prostate Cancer
The UPMC researchers are not the first to use AI to detect prostate cancer. In February, The Lancet Oncology published a study from researchers at Radboud University Medical Center (RUMC) in the Netherlands who developed a deep learning AI system that could determine the aggressiveness of prostate cancer in certain patients.
For that research, the RUMC scientists collected 6,000 biopsies from more than 1,200 men. They then showed the biopsy images along with the original pathology reports to their AI system. Using deep learning, the AI was able to detect and grade prostate cancer according to the Gleason Grading System (aka, Gleason Score), which is used to rate prostate cancer and choose appropriate treatment options. The Gleason Score ranges from one to five and most cancers obtain a score of three or higher.
“Systems such as ours can be used in different ways. First, it can be used to screen biopsies and to filter out the easy (benign) cases. This could reduce the workload for pathologists,” said Wouter Bulten, a PhD candidate at Radboud who worked on the study, in an interview with HemOnc Today. “Second, the system can be used as a second opinion after the pathologist’s initial read. The system can flag a case if its opinion differs from that of the pathologist. It also can give feedback during the first read, showing the pathologist where to look. In this case, the pathologist needs only to confirm the opinion of the AI system.”
Can Today’s AI Outperform Human Pathologists?
In their research, the Radboud team discovered that their AI system was able to achieve pathologist-level performance and, in some cases, even performed better than human pathologists. However, they do not foresee AI replacing the need for pathologists, but rather emerging as another method to use in cancer detection and treatment.
“We see our system as an additional tool that the pathologist can use. Although our system performs very well, it still makes mistakes,” stated Bulten. “These mistakes are often different from those a human would make. We believe that when you merge the expertise of the pathologist with the second opinion of an AI system, you get the best of both worlds.”
According to the American Cancer Society, prostate cancer is the second most common cancer among men in the US, after skin cancer. The organization estimates there will be approximately 191,930 new cases of prostate cancer diagnosed and about 33,330 deaths from the disease in the US in 2020.
Though the UPMC study focused only on prostate cancer, the scientists believe their algorithm can be trained to detect other types of cancer as well. AI in clinical diagnostics is clearly progressing, however more studies will be required. Nevertheless, if AI can truly become a useful tool for anatomic pathologists to detect cancer earlier, we may see a welcomed reduction in cancer deaths.
