Cancer Care

A New FDA Approval Furthers the Role of Genomics in Cancer Care

Posted by

0


A recent drug approval by the Food and Drug Administration (FDA) marks another milestone in the treatment of cancer. The action by FDA expands the growing list of approved uses for the immune checkpoint inhibitor pembrolizumab (Keytruda).

In this case, it is now approved to treat some adults or children with advanced cancer, regardless of the type of cancer they have, if the patient’s tumor has a large number of genetic mutations—also called a tumor mutational burden-high (TMB-H) cancer. 

Because this is what is known as an accelerated approval, this new indication for pembrolizumab must be confirmed in additional studies. There also are a number of factors for oncologists to consider when deciding whether this treatment is a good option for their patients. I’ll return to those issues later. 

But from a broader perspective, this new approval represents a continued march toward the use of genomics to guide cancer treatment, including for children with cancer. 

It reinforces the idea that, as a routine matter, oncologists should be discussing tumor genomic testing (or biomarker testing) with their patients with advanced cancer, in particular those for whom there are no effective standard-of-care treatments. This applies to testing not only for biomarkers like TMB that might predict response to immunotherapies, but also for genetic biomarkers that identify those who might benefit from a specific targeted cancer treatment.  

This type of biomarker testing has helped patients with many types of cancers, even hard-to-treat cancers such as brain and pancreatic, by identifying therapies that might be more effective and less toxic for them. But it’s important to note that biomarker testing has important limitations and, clearly, there are many patients who do not benefit from this testing.  

This new approval for pembrolizumab marks the fourth time that FDA has approved a drug to be used in a “tissue-agnostic” manner—that is, for any cancer type—based solely on the results of biomarker testing. Given that there are now many approved cancer therapies whose use requires genomic biomarker testing (via what are known as “companion diagnostics”), I believe that genomic biomarker testing should be strongly considered for all patients with advanced cancers for which no known effective treatments currently exist.

Given this increasing role of genomic biomarker testing in cancer care, it is important to note that many patients do not benefit from molecularly targeted treatments because they are never offered that testing.  

For example, a recent review in lung cancer, where genomic testing is particularly critical for selecting the most appropriate treatments, found that 10%–40% of patients are not being tested for molecular alterations that predict response to FDA-approved therapies, even though such testing is recommended by clinical guidelines. 

This lack of biomarker testing, which can lead to the use of treatments with the potential to markedly improve how long patients live, is concerning. At NCI, we’re particularly worried about the possibility that less genomic biomarker testing in patients from underserved populations could further exacerbate existing cancer-related health disparities.

Learning about Cancer and Tumor Mutational Burden

Having a TMB-H tumor is not a guarantee that pembrolizumab will be effective, but it is an opportunity, based on a robust evidence base that was developed in part by NCI-supported research and investigators.

It’s also important to note that this is not the first tissue-agnostic FDA approval for pembrolizumab. In May 2017, it was approved to treat any cancer that has specific genetic alterations that affect a cell’s ability to repair damaged DNA, known as microsatellite instability high (MSI-H) tumors. MSI-H and TMB-H are related, with most MSI-H tumors also showing increased TMB. The converse, however, is not true—that is, many TMB-H tumors are not MSI-H. 

Moreover, MSI status does not necessarily need to be established through extensive tumor genomic testing, whereas TMB status generally does. So having a drug approval based on TMB is important, because it is more common in tumors than MSI-H, and thus should promote far greater adoption of tumor genomic testing.

Cancer cells that are MSI-H or TMB-H tend to produce a large number of mutant proteins, which are then often displayed on the cells’ surface as “neoantigens.” Since normal cells do not display such mutant proteins, the production of neoantigens can be a signal that draws the immune system’s attention. That, in turn, can increase the likelihood that treatments like pembrolizumab, which strengthen the anti-tumor immune response, will help immune cells kill those tumors.

In the clinical trial on which FDA based this new approval, called KEYNOTE-158, nearly 30% of patients with TMB-H tumors had a response to treatment—meaning they experienced at least 30% reduction in the size of their tumors—including some whose tumors disappeared completely. In contrast, only 6% of patients who were not TMB-H exhibited a response to pembrolizumab. In half of the participants who responded to the treatment, their tumors did not begin to grow again for at least 2 years!

Although patients on KEYNOTE-158 were not randomly assigned to an alternative therapy or placebo, these types of sustained treatment responses among patients with solid tumors are extremely rare with standard chemotherapies. 

In addition, most participants in KEYNOTE-158 had received numerous other therapies and had no remaining treatment options known to be effective against their cancers. That’s an important point, because FDA’s new approval applies only to people with advanced cancer without any other standard treatment options. 

In general, tumors in children and young adults with cancer do not have a large number of mutations and are therefore less likely to be TMB-H. But some tumors that arise in children are TMB-H, and this approval now means that such young patients have a new, potentially meaningful treatment option available to them as well.

Contribution of Genomics Research

The role of NCI-supported research in helping to identify TMB as a biomarker of immunotherapy response is particularly gratifying. 

For example, although TMB did not even exist as a concept at the time that The Cancer Genome Atlas (TCGA) was launched, researchers used TCGA data to perform wide-scale analyses of the relationship between tumor mutation numbers and immunotherapy response in the first place. The finding is another example of how building large, high-quality, and diverse sets of research data and making them publicly available can yield unexpected insights that lead to meaningful advances in treatment.

In addition, NCI has had a sustained and successful partnership with Foundation Medicine, the company that developed FoundationOne CDx, the test used to identify patients with TMB-H tumors for this approval. NCI-supported researchers made important contributions to the development of the technologies used by this test. 

As should be expected, though, this is an extremely complicated area of research. Even with the best tests, measuring TMB and deciding what that measurement means for individual patients—in particular, estimating their likelihood of response to immunotherapy—is far from straightforward.

So NCI has been working on other fronts to make further advancements that build on the use of TMB, MSI, and other biomarkers to help guide treatment decisions. For example, NCI has partnered with Friends of Cancer Research and several other organizations on the TMB Harmonization ProjectExit Disclaimer, which is working to standardize how mutational burden is measured in tumors using different technologies and approaches. This project will help to ensure that TMB is used in the most effective manner possible to guide decisions about patient care.

NCI is also supporting research through the Cancer MoonshotSM on the impact of TMB and other biomarkers on how patients respond to immunotherapy. That effort is being carried out by researchers who are part of the Moonshot-funded Cancer Immune Monitoring and Analysis CentersExit Disclaimer and through a large public–private partnership called the PACT initiativeExit Disclaimer.

Good News, But More To Learn

This new approval is exciting, and it will have an immediate impact on patient care. However, it’s important to stress that TMB status is just one piece of information about a tumor. Its relevance to treatment may depend on a number of factors, including a patient’s physical ability and willingness to tolerate further treatment. 

In KEYNOTE-158, most patients with TMB-H tumors didn’t benefit from pembrolizumab, and a small percentage of patients whose tumors had a low TMB did respond to treatment. This suggests we still have more to learn in order to predict who will respond to immunotherapy drugs.

There is evidence, for example, that the extent of TMB required to improve the likelihood of responding to immunotherapy may vary from cancer to cancer. Information such as whether cancer type influences treatment response is important, because no cancer treatment comes without side effects, including financial toxicity. 

Larger studies of pembrolizumab, which are required by FDA as part of an accelerated approval, and other research being supported by NCI can hopefully provide answers to these questions. Those answers can then help patients and their doctors have informed discussions and engage in a true process of shared decision making. For patients with advanced cancer and their families, such discussions are vital.

Especially during these hectic times, one can become cynical. But as I’ve said previously, NCI is committed to addressing the challenges of the COVID-19 pandemic while also supporting the research required to address the most pressing needs for those with cancer and heralding important progress. The expanding role and value of immunotherapy as a cancer treatment is good news for patients, and that’s something I think is worth celebrating.

Artifical Intelligence Changing Cancer Care

Posted by

0


Artificial Intelligence (AI) is beginning to change the practice of medicine and will increasingly play a major role in the management of health and cancer. AI is the development of computer systems to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between language. The computer systems are able to learn and adapt without following explicit instructions, by using algorithms and statistical models to analyze and draw inferences from patterns in data.

The most recent example gaining attention in the news is ChatGPT – a natural language processing tool driven by AI technology. ChatGPT allows human-like conversations and much more.  The model can answer questions, and assist with tasks such as composing emails, essays, and code.

Machine learning is a subfield of artificial intelligence already making an impact in oncology. Broadly defined, machine learning is the capability of a machine to imitate intelligent human behavior and perform complex tasks in a way that is similar to how humans solve problems.

AI is beginning to play a vital role in cancer care

In November 2022, Google announced an AI algorithm designed to help screen for breast cancer that will now be part of commercial mammograms. Google licensed its AI technology to iCAD, a medical technology company that provides breast cancer detection services to health care facilities around the world.

In a study published in the medical journal Nature Google’s algorithm for mammograms performed better than radiologists. The study involved mammograms from more than 91,000 women in the U.S. and the U.K. In the U.S. Google’s system lowered the false positive rate by 6%, and in the U.K. by 1.2%. The machine learning algorithm also decreased false positives by 9% in the U.S. and nearly 3% in the U.K.

And that’s just the beginning, machine learning has the ability to improve over time – the more data from mammograms that are fed into the algorithm, the better the algorithm gets at detecting the smallest differences that distinguish normal tissue from potentially cancerous tissue. Women receiving mammograms using the AI-based system will have their information fed back into the algorithm

Potential advantages with the Google AI algorithm.

  • The AI-based system could make mammography more widely available around the world.
  • The system could replace radiologists or at least relieve the burden on radiologists, since many nations require two readings of a mammography image.
  • Cost of care could be reduced.
  • Early detection will be continuously improved.

Machine-based learning will begin to recognize patterns that human eyes can’t see which should further reduce the False Negative rates of the algorithm ultimately leading to improved care.

In age of misinformation, cancer care community seeks new ways to improve communication

Posted by

0


An important component of cancer care is communication with patients and caregivers about their disease.

But in a communication age where social media is king, cancer misinformation travels far and fast.

Erqi Liu Pollom, MD, MS
Connecting with patients to combat cancer misinformation may require a multifaceted approach on multiple platforms, according to Erqi Liu Pollom, MD, MS. “I think we can do better in reaching out to patients and filling that vacuum of information so they’re not out there searching on their own,” she said. Source: Stanford Cancer Center.

“What keeps me up at night is thinking about the root causes of the spread of misinformation — the disinformation endeavors and what they may do to public health and people’s emotions and relationships, fostering hate between groups and division in society,” Wen-Ying Sylvia Chou, PhD, MPH, program director in the health communication and informatics research branch of the behavioral research program at NCI, told Healio.

Chou referenced an editorial commentary she wrote a little over a year ago for American Journal of Public Health on the results of an analysis of the “#China-virus” hashtag on Twitter during the COVID-19 pandemic, including how it “emboldened xenophobia and racist sentiments.”

Wen-Ying Sylvia Chou, PhD, MPH

Wen-Ying Sylvia Chou

“Online and offline interactions are intertwined. Powerful people on social media can affect the real world and cause harms in vulnerable people; it’s not just words,” Chou said. “That makes me think that, in a way, a cancer diagnosis is one way for someone to become more vulnerable. Without policies in place to protect the vulnerable and change how social media platforms operate, I worry that things are not going to get better and, if anything, get worse.”

The spread of cancer misinformation online isn’t limited to any specific disease or social media platform. It’s everywhere, according to experts in the field who are working diligently to identify the best ways to combat it.

Healio spoke with oncologists and leading communication experts in science and medicine about the prevalence of cancer misinformation on social media, important research into the issue and possible strategies to improve communication platforms and disseminate trustworthy information more effectively.

A problem ‘across multiple health topics’

Skyler B. Johnson, MD, physician-scientist at Huntsman Cancer Institute and assistant professor of radiation oncology at The University of Utah, discovered the scale and scope of cancer misinformation when his wife was diagnosed with cancer.

Like any other person, he went online to do some quick research. But unlike many people without a medical background, Johnson could identify misinformation.

Skyler B. Johnson, MD

Skyler B. Johnson

“I came away from that experience with quite a bit of empathy for patients,” Johnson said during an interview last year with Healio. “I encounter a lot of patients through my oncology practice who believe misinformation they read online and through social media, and it has led to delays in their diagnosis and treatment. It has led to refusal of proven cancer treatments and, ultimately, their early deaths.”

In July, Johnson and colleagues published results of a study in Journal of the National Cancer Institute that showed one-third of the most popular cancer treatment articles on social media included inaccurate information. Perhaps even more troubling: Most of the misinformation they identified had the potential to harm patients with approaches that could have negative effects on treatment quality and survival.

In a separate, mixed-methods experimental study last year, published in Journal of Cancer Education, Chou and colleagues integrated eye-tracking, survey and interviews to examine how participants assessed cancer messages’ believability when viewing simulated Facebook posts. They wanted to see how people discern credible information from noncredible posts, what caught their attention, where they looked and how they made decisions.

“We found that participants spent a lot of time viewing (dwelling on) the nonevidence-based posts,” Chou said. “It may be because inaccurate messages were generally more vivid and shocking or even emotionally provocative. They have stories, such as, ‘This is how my daughter was vaccinated and she was never the same person.’ There’s a lot for people to look through. And that parallels a lot of the misinformation tactics. Narrative is a powerful tool, but it can also be a tool to instill negative emotions such as doubt and fear.”

In 2018, researchers from NYU Langone Health investigated prostate cancer information on YouTube, which they wrote was the most popular of all platforms, with use by 73% of U.S. adults.

They found that 77% of the top 150 videos on prostate cancer had misinformative or biased information. Additionally, many videos contained outdated information regarding screening guidelines and treatment recommendations.

Another finding echoed that of many social media cancer misinformation studies: misinformative videos had more social media engagement (numbers of comments, thumbs up/down interaction, etc.), including more than twice as many average views per day, than videos with factual information.

“I think it’s extremely concerning that a lot of biased and misinformative health content is reaching so many people,” Stacy Loeb, MD, MSc, PhD (Hon), urologist in the department of urology at NYU Langone Medical Center and one of the authors of the study, told Healio.

Stacy Loeb

Stacy Loeb

Loeb and colleagues took on another YouTube study involving the quality of bladder cancer information. They found 67% of overall information was moderate to poor and 21% of the videos, which had more than 1.2 million views, contained a moderate to high amount of misinformation.

“We’re seeing this misinformation problem across multiple health topics,” Loeb said. “We recently published a study on the quality of prostate cancer information on TikTok. … Among the videos that had objective information, there was a significant amount of misinformation.

“The bottom line is this is a cross-cutting issue that goes across many topics and platforms,” Loeb added. “It’s not specific to prostate cancer or to YouTube, but rather a major problem in society, as there is an increasing amount of unfiltered, user-generated content that is posted online and broadly disseminated.”

Hopeful stories with convenient answers

Brian G. Southwell, PhD, has tried to identify the motivations behind cancer misinformation and the reasons for its existence. He published a 320-page book, “Misinformation and Mass Audiences,” in 2018 and has regularly contributed his communication expertise to the health care sphere.

“It’s a really important question that does not get asked enough: Why is (this misinformation) out there?” Southwell told Healio.

Brian G. Southwell, PhD

Brian G. Southwell

Southwell, senior director of science in the public sphere at Research Triangle Institute (RTI) International, used the example of a social media influencer-turned-scammer. Belle Gibson rose to fame as a blogger who suggested she overcame brain cancer because of a diet she adhered to, only to be later exposed as a fraud who didn’t even have cancer.

“You can judge her, of course. But how did she rise to fame?” Southwell asked. “Because she was telling a hopeful story. If some people hear you can prevent breast cancer with cannabis oil or whatever, that might sound great. I think there is this certain degree of storytelling, telling stories people want to hear.”

As Southwell said, emotion plays a vital role, as someone either with cancer or with a loved one with cancer is easily drawn to stories of hope, regardless of their veracity, on social media. Stories of misinformation are also more novel and intriguing, whereas factual studies generally take time and produce results incrementally rather than overnight.

“There is nothing magical about false information, it’s just that the people who are creating it have an advantage relative to carefully peer-reviewed journal articles framed in a mundane way,” Southwell said. “‘Halloween Candy is Going to Cure Cancer.’ If you could write the headline first and just fill in the details … you have the advantage of being as sensational as you want.”

Perhaps some of these points — the accessibility of hopeful stories with easy solutions and the clickbait headlines — may help explain why a healthy proportion of cancer misinformation involves the use of cannabis.

Researchers from Stanford University and Kaiser Permanente Interstate Radiation Oncology Center published a study in January of 2019 on the space cannabis holds within the realm of cancer misinformation on social media. The researchers found that:

  • online search volume for cannabis and cancer increased at 10 times the rate of standard therapies (more in states where recreational or medical cannabis was legal);
  • cannabis as a cancer cure represented the largest category of social media content (23%) on alternative cancer treatments;
  • of the 40 high-impact new stories on social media referencing “cannabis cancer,” 32 (80%) were false news that proposed cannabis as a cancer cure and only one (2.5%) was an accurate news story that debunked that claim; and
  • the top false news story claiming cannabis as a cancer cure generated 4.26 million engagements on social media; meanwhile, the top news story debunking this false news only received 0.036 million engagements.

Although she supports medical cannabis for symptom control, Erqi Liu Pollom, MD, MS, said some of the results surprised her.

“It seemed like the false articles, the articles with inaccurate information, were more popular,” Pollom, assistant professor of radiation oncology at Stanford University, told Healio. “Accounts that are propagating this misinformation were more popular than the legitimate sources like the cancer societies, cancer physicians, etc. [Stories with misinformation] had broader reach than some of the more legitimate accounts.”

Pollom and colleagues also analyzed Facebook and Twitter posts from 31 leading cancer organizations during the study period. They found a drastic difference in engagement of these organizations’ posts on cannabis vs. the top 10 false stories on cannabis as a cancer cure, including average Facebook likes, comments and shares (97.7 vs. 452,050.1).

In addition, National Comprehensive Cancer Network member organizations in states where medical or recreational cannabis had not been legalized had zero Facebook posts or tweets on cannabis during the study period.

“If organizations like NCCN and ASCO could dedicate resources to provide educational material on a specific topic for patients, that probably would be helpful,” Pollom said. “Maybe they’re just not aware that it’s a big issue. That was one of the reasons we did our study, to see how big of an issue this was.”

AI and an active Twitter account

Although monitoring social media is a challenge, it’s not one too vast for researchers and medical professionals to try to tackle.

In July, Chou contributed to a paper published in NAM Perspectives on identifying credible sources of health information that recommended a “within reach” plan involving a system of spot checks for quality and integrity “supported by machine learning technology” and aided by expert human evaluation.

“We need more social scientists, ethnographers, anthropologists … I have a background in linguistics,” Chou told Healio. “AI with important, meaningful human input, working together with a computer scientists and engineers, is needed — and right now. We’re not doing enough. We’re kind of saying, ‘Well, these are the algorithms’ (causing misinformation to spread) as if humans have nothing to do with it.”

Loeb has contributed to this area along with University of Michigan researchers, including Rada Mihalcea, PhD, professor of computer science and director of the Michigan AI Lab. Work they presented at ACM International Conference on Multimodal Interaction in 2019 involved utilizing linguistic, acoustic and user engagement features to develop classification models to identify misinformation in prostate cancer videos.

“We reported preliminary findings that an algorithm to identify misinformative videos about prostate cancer had 74% accuracy,” Loeb said. “And we’re continuing to work on potential automated solutions to assist in the identification of misinformative content that could be potentially flagged or manually reviewed or removed, or that could be used to help design a smarter search filter to help health consumers identify content that’s more likely to be accurate.”

Another possible solution that doesn’t require AI but does require time and effort is for oncologists to actively disseminate information.

“For health professionals and scientists, an important action item is to participate in the online discourse,” Loeb said. “There is a quote that circulates: ‘The solution to pollution is dilution.’ For those of us who are experts in health topics, it’s important to be a part of the solution by providing high-quality and accurate information or public dissemination.”

Deanna J. Attai, MD

Deanna J. Attai

Loeb (@loebstacy) and other colleagues hold monthly discussion sessions on Twitter about prostate cancer and the latest research that anyone can join. Similarly, Deanna J. Attai, MD, associate clinical professor of surgery at David Geffen School of Medicine at University of California Los Angeles, serves as a co-moderator (@DrAttai) for #BCSM, a weekly, live community forum for breast cancer.

Meeting patients where they are online may require a multifaceted approach on multiple platforms. In a study published in May in Practical Radiation Oncology, Pollom and colleagues found patients more apt to use Instagram and physicians more active on Twitter.

“I think we can do better in reaching out to patients and filling that vacuum of information so they’re not out there searching on their own,” Pollom said.

Other means of communication

Getting every oncologist in the world to join Twitter, Facebook and Instagram and spend a specific amount of time each day or week on the social media platforms likely is not feasible.

“One of the barriers, I think — not as an academic physician or researcher — is that we’re promoted on academic papers … writing for our colleagues rather than peers,” Pollom said. “I don’t think there’s a viable infrastructure yet in terms of how we can advance our academic careers on social media. That’s a limitation of having physicians spending more time educating online.”

More long-term solutions may require much larger investments of time and money by a larger entity: health care systems. Southwell said the current information/misinformation climate could be an ideal time for “reimagining” how communication systems function within a health care system.

“We need a lot of investment in this translational space, which I don’t think we have,” he said. “You start out worried about all of these patients running into misinformation, and you have to continue to think about that since patients are a part of the equation, but part of the story and the solution are our health care institutions. What are they doing, necessarily, to provide information?”

Improved in-person communication between oncologists and patients also can play a role in limiting a patient’s possible engagement with misinformation on social media.

In an October 2020 editorial he co-wrote in American Journal of Public Health, Southwell suggested an open-ended question such as, “What have you already heard or learned about your treatment/condition?” as a useful starting point.

“You’re respecting that the patient is seeking out their own information on their condition and also getting a quick beat on where they are looking,” Southwell said. “You’re also starting from a place where you’re not assuming or dismissing (their view) right away.”

Lending an empathetic ear is important, Southwell said, as is understanding that the source of their information may be someone important to them and not just something they came across on their social media feed.

“Maybe this came from their Aunt Edna, and they listen to everything she says,” Southwell said. “Getting some sense of the personal referral (is important) because now you know what you’re up against and you can tread lightly. ‘Well, my mother told me this’ or ‘my wife told me this.’ That’s probably going to carry some weight.”

Unlike traditional media, such as newspapers or television, social media is a two-way communication channel. Similarly, within a clinic, physicians who are adept at communication value the shared decision-making process between physician and patient.

“This includes a discussion of risk, benefits, alternatives and the patient’s preferences related to a medical decision,” Loeb said. “We published a study [in Journal of Medical Internet Research] using data from HINTS, a national survey, showing that patients who perceived a worse quality of communication with their health care provider were more likely to watch health-related videos.

“Bottom line: it is very important that patients have an opportunity to have their questions answered and that they are well-informed by their health care provider,” Loeb said. “Otherwise, it is natural that people will look to other sources for information.”

Telemedicine fails to reduce, may widen disparities in cancer care, study shows

Posted by

0


Patients newly diagnosed with cancer and considered to be of high socioeconomic status had the highest use of telemedicine in the early days of the COVID-19 pandemic, according to a research letter published in JAMA Oncology.

“Patients with cancer cannot afford delays in their treatment. During the early months of the COVID-19 pandemic, many in-person health care visits were cancelled, and telemedicine was rapidly expanded to provide timely care for patients,” Ronald C. Chen, MD, MPH, FASCO, FASTRO, Joe and Jean Brandmeyer endowed professor and chair of the department of radiation oncology at University of Kansas Cancer Center, told Healio. “Although telemedicine has the potential to reduce disparities by broadly improving access to health care, we undertook this study to examine whether the rollout of telemedicine occurred equitably.”

Telemedicine visit rate.
Data derived from Katz AJ, et al. JAMA Oncol. 2021;doi:10.1001/jamaoncol.2021.5784.

Researchers pooled data from the HealthCore Integrated Research Database on 16,006 patients (53% men; 50.7% aged 18 to 64 years) with newly diagnosed breast, lung, prostate or colorectal cancer between Jan. 1 and Aug. 31, 2020.

Receipt of a telemedicine visit within 30 days of cancer diagnosis, which the researchers established with the presence of one or more claims with Current Procedural Terminology of Healthcare Common Procedure Coding Systems codes for telemedicine services, served as the primary outcome.

According to study results, the rate of telemedicine visits increased from 0.4% in January to a peak of 54% in April.

Researchers noted socioeconomic status differences in the patterns of telemedicine uptake, with approximately 66.9% of patients in the highest socioeconomic status index quartile having a telemedicine visit within 30 days of cancer diagnosis compared with 47.4% to 48.6% of patients in the lower socioeconomic quartiles.

Results of multivariable analysis showed that compared with patients in the lowest socioeconomic status index quartile, those in the highest quartile had 31% higher odds of telemedicine use within 30 days of cancer diagnosis (risk ratio = 1.31; 95% CI, 1.17-1.47).

Ronald C. Chen, MD, MPH, FASCO, FASTRO

Ronald C. Chen

“Unequal utilization of telemedicine among [patients with cancer] across the U.S. may widen cancer disparities. Now that telemedicine is no longer new, further studies can examine whether socioeconomic status differences in its use persist,” Chen said. “Additional studies can examine whether less telemedicine use is associated with delayed cancer care and worse outcomes for certain groups of patients.”

For more information:

Ronald C. Chen, MD, MPH, FASCO, FASTRO, can be reached at University of Kansas Cancer Center, Richard and Annette Bloch Cancer Care Pavilion, 2650 Shawnee Mission Parkway, Westwood, KS 66205.

Oncologists are guardedly optimistic about AI. But will it drive real improvements in cancer care?

Posted by

0


Artificial intelligence 
Over the course of my 25-year career as an oncologist, I’ve witnessed a lot of great ideas that improved the quality of cancer care delivery along with many more that didn’t materialize or were promises unfulfilled. I keep wondering which of those camps artificial intelligence will fall into.

Hardly a day goes by when I don’t read of some new AI-based tool in development to advance the diagnosis or treatment of disease. Will AI be just another flash in the pan or will it drive real improvements in the quality and cost of care? And how are health care providers viewing this technological development in light of previous disappointments?

To get a better handle on the collective “take” on artificial intelligence for cancer care, my colleagues and I at Cardinal Health Specialty Solutions fielded a survey of more than 180 oncologists. The results, published in our June 2019 Oncology Insights report, reveal valuable insights on how oncologists view the potential opportunities to leverage AI in their practices.

Limited familiarity tinged with optimism. Although only 5% of responding oncologists describe themselves as being “very familiar” with the use of artificial intelligence and machine learning in health care, 36% said they believe it will have a significant impact in cancer care over the next few years, with a considerable number of practices likely to adopt artificial intelligence tools.

The survey also suggests a strong sense of optimism about the impact that AI tools may have on the future: 53% of respondents said that such tools are likely or very likely to improve the quality of care in three years or more, 58% said they are likely or very likely to drive operational efficiencies, and 57% said they are likely or very likely to improve clinical outcomes. In addition, 53% described themselves as “excited” to see what role AI will play in supporting care.

An age gap on costs. The oncologists surveyed were somewhat skeptical that AI will help reduce overall health care costs: 47% said it is likely or very likely to lower costs, while 23% said it was unlikely or very unlikely to do so. Younger providers were more optimistic on this issue than their older peers. Fifty-eight percent of those under age 40 indicated that AI was likely to lower costs versus 44% of providers over the age of 60. This may be a reflection of the disappointments that older physicians have experienced with other technologies that promised cost savings but failed to deliver.

Hopes that artificial intelligence will reduce administrative work. At a time when physicians spend nearly half of their practice time on electronic medical records, we were not surprised to see that, when asked about the most valuable benefit that AI could deliver to their practice, the top response (37%) was “automating administrative tasks so I can focus on patients.” This response aligns with research we conducted last year showing that oncologists need extra hours to complete work in the electronic medical record on a weekly basis and the EMR is one of the top factors contributing to stress at work. Clearly there is pent-up demand for tools that can reduce the administrative burdens on providers. If AI can deliver effective solutions, it could be widely embraced.

Need for decision-support tools. Oncologists have historically been reluctant to relinquish control over patient treatment decisions to tools like clinical pathways that have been developed to improve outcomes and lower costs. Yet, with 63 new cancer drugs launched in the past five years and hundreds more in the pipeline, the complexity surrounding treatment decisions has reached a tipping point. Oncologists are beginning to acknowledge that more point-of-care decision support tools will be needed to deliver the best patient outcomes. This was reflected in our survey, with 26% of respondents saying that artificial intelligence could most improve cancer care by helping determine the best treatment paths for patients.

AI-based tools that enable providers to remain in control of care while also providing better insights may be among the first to be adopted, especially those that can help quickly identify patients at risk of poor outcomes so physicians can intervene sooner. But technology developers will need to be prepared with clinical data demonstrating the effectiveness of these tools — 27% of survey respondents said the lack of clinical evidence is one of their top concerns about AI.

Challenges to adoption. While optimistic about the potential benefits of AI tools, oncologists also acknowledge they don’t fully understand AI yet. Fifty-three percent of those surveyed described themselves as “not very familiar” with the use of AI in health care and, when asked to cite their top concerns, 27% indicated that they don’t know enough to implement it effectively. Provider education and training on AI-based tools will be keys to their successful uptake.

The main take-home lesson for health care technology developers from our survey is to develop and launch artificial intelligence tools thoughtfully after taking steps to understand the needs of health care providers and investing time in their education and training. Without those steps, AI may become just another here today, gone tomorrow health care technology story.

How Can Bioinformatics Help Advance Precision Cancer Care?

Posted by

0


Chris Sander PhD

Chris Sander, PhD, director of the cBio Center at Dana-Farber. Sander is co-founder of the computational biology field and a leader in applying its methods to cancer research.

Scattered amongst the letters of genetic code in a tumor cell are telltale mutations and DNA alterations that spur its malignant activity. But there are billions of letters of code and each patient’s cancer is different, with its own particular genetic changes. These changes may dictate how it behaves, how aggressively the cancer progresses, and it may spell out which molecular weaknesses might be successfully attacked with treatments.

Technology like the OncoPanel platform of Dana-Farber’s Profile research program can rapidly decipher the DNA code of 400 cancer-related genes in an individual’s tumors, detecting mutations, missing or extra copies of genes, and other changes. But it’s a process that generates massive volumes of digital data – data which mean nothing until processed and analyzed. Only then, may researchers and oncologists be able to prescribe a precision treatment to target the specific mutation. This is the goal of precision cancer medicine, but it’s often a needle in a haystack search.

Dana-Farber’s Chris Sander, PhD, founded and directs the new cBio Center at DFCI, along with Ethan Cerami, PhD, who leads the Knowledge Systems Group in the Center, to help researchers mine the genomic data, using a user-friendly web-based tool they developed called cBioPortal for 

The cBio Portal software digests the data and presents it in diverse visual formats that help investigators detect patterns of abnormalities across groups of patients and cancer types.

With cBioPortal, a researcher can tap into the Profile tumor base of nearly 8,000 genomic tumor profiles and rapidly compare the mutation pattern in one patient’s tumor to those of hundreds of other patients. Some cancers with certain mutations might have an approved drug targeting those changes; finding the same mutations in other tumor types may lead to new trials of that drug. “We will be able to use cancer genomics to define groups of patients who might be eligible for new kinds of genomically informed clinical trials,” says Sander.

The cBioPortal analysis tools can also help scientists sift genomic data for clues to why a few patients in a clinical trial had dramatic responses to a drug that had little benefit for the majority of other patients. The cause of this “exceptional response” may be a previously undiscovered mutation that made the tumor vulnerable to the drug – and cBioPortal analysis of genomic profile data may help scientists discover it.

The Center’s Knowledge Systems Group has also devised a clinically oriented tool called MatchMiner. “Clinicians use it to recruit patients for clinical trials,” Cerami explains. “They look in the Profile database for patients whose tumors’ mutations match targeted drugs in available clinical trials, and they can have MatchMiner notify them when appropriate new patients are identified.”

The scientists are now working on tools to enable oncologists and patients to sit down together and use MatchMiner to search for clinical trials of drugs targeted to their cancer’s specific mutation pattern.