Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Spotify | RSS
Molecular Science Liaisons with Caris Life Sciences, Sachein Sharma, MD, MPH, and Gretchen Hubbard, PhD, detail how AI is used in the fight against prostate cancer. They share how AI is making for ever more individualized treatments, leading to better patient outcomes for men with prostate cancer.
Program Notes
- Sachein Sharma, MD, MPH on LinkedIn
- Gretchen Hubbard, PhD on LinkedIn
- Gretchen Hubbard, PhD on Twitter: @DrHubbard_MSL
- Caris Life Sciences website
- Caris Life Sciences on Twitter: @carisls
- Carlis Life Sciences on Facebook: facebook.com/carislifesciences
- NCCN Guidelines for Patients with Advanced Stage Prostate Cancer – Biomarker Testing (pg15)
The Stay in the Game podcast is sponsored by Cancer Health – online at cancerhealth.com.
Cancer Health empowers people living with prostate cancer and other cancers to actively manage and advocate for their care and improve their overall health. Launched in 2017, cancerhealth.com provides accessible information about treatment and quality of life for people with cancer and their loved ones, along with information about cancer prevention and health policy.
Episode Transcript
Announcer: Welcome to Stay in the Game, conversations about prostate cancer with Ed Randall. Here we’ll chat with doctors, researchers, medical professionals, survivors, and others to share and connect. This show was produced and shared by Fans For the Cure, a nonprofit dedicated to serving men on their journeys through prostate cancer.
The Stay in the Game podcast is sponsored by Cancer Health — online at cancerhealth.com. Cancer Health empowers people living with prostate cancer and other cancers to actively manage and advocate for their care and improve their overall health. Launched in 2017, cancerhealth.com provides accessible information about treatment and quality of life for people with cancer and their loved ones, along with information about cancer prevention and health policy.
Ed Randall: Hi, everybody. I’m Ed Randall, the founder and chief advocacy officer for Fans For the Cure. Welcome back to our Stay in the Game podcast. My two guests on today’s podcast are from Caris Life Sciences, a company that is fighting the good fight against several forms of cancer, including prostate cancer.
In a world where treatment options are becoming increasingly individualized, it is more important than ever to understand the molecular features of a person’s specific tumors and disease in charting an optimal treatment path. An advanced molecular profiling and searching for biomarkers that help dictate the best treatment options is what Caris does best.
Joining us today is Gretchen Hubbard, a PhD from the University of Baltimore, Maryland County, a postdoctoral fellow at Johns Hopkins, and currently a Genitourinary molecular science liaison lead for Caris. We also welcome Dr. Sachein Sharma, an MD and MPH from University of Texas Health Science Center at Houston. Dr. Sharma, who has ten years of research experience in Multi-omics at three different academic institutions is Caris’ molecular science liaison for New York City. And I’m Ed Randall, recipient of a gentleman’s C minus in biology my sophomore year with Brother VC at All Hallows High School in the Bronx, which also spelled the end of my formal training in the sciences.
So I invite you both Dr. Hubbard and Dr. Sharma to jump in and either correct me or clarify an issue should things be veering even slightly off the rails. And this disclaimer before we begin, although Dr. Hubbard and Dr. Sharma are employees of Caris Life Sciences, any opinions given during this podcast are those of the individual and are not endorsed by Caris Life Sciences. Dr. Sharma, we’ll start with you.
Dr. Sachein Sharma: Thank you very much for having us on the podcast today. So AI was first described in the 1950s. Alan Turing was among the first to actually conceive of the idea and John McCarthy coined the term artificial intelligence. Several limitations, though, in early models prevented its widespread acceptance and application into medicine.
And it wasn’t until the early 2000s that these were overcome by a few things like machine learning, deep learning, or natural language processing. And while those are particularly difficult terms perhaps to understand, one well-known example of AI is Watson, and this was initially developed as a computer system to answer questions on the TV show Jeopardy! and then to compete with humans, including one of its most celebrated contestants, Ken Jennings. Since then, though, AI has been applied to clinical practice through risk assessment models, improving diagnostic accuracy and workflow efficiency as well. At Caris, we use artificial intelligence for prognostic and predictive purposes to help us identify the right treatment for the right patient at the right time.
Ed Randall: Dr. Hubbard.
Dr. Gretchen Hubbard: Thank you. Great example of how Caris is using artificial intelligence is to help develop predictive signatures. One signature, which we’ve termed MI FOLFIRSTai, is the first clinically validated predictor of chemotherapy response for patients who have been newly diagnosed with metastatic colorectal cancer. So oftentimes for first-line metastatic colorectal cancer, patients will receive either FOLFOX followed by FOLFIRI or FOLFIRI followed by FOLFOX. But historically, it’s been up to the physician to decide the sequence of those therapies, and there’s really not been a way to predict which sequence is better for patients.
So Caris has this large database of molecular data from profiling patients as well as matched clinical outcomes. And using this database, Caris has been able to identify a signature which predicts response to FOLFOX, which is an Oxaliplatin-based chemotherapy, which is used often with the biologic bevacizumab. So the AI signature that’s been developed is run for every patient with colorectal cancer, and it predicts whether the patient is likely to have an increased or decreased benefit to this regimen. So this AI predictor has been now clinically validated and it has been shown to improve the treatment outcome in terms of overall survival of these patients when the sequence of therapies follows the prediction.
Ed Randall: What are some of the ways that Caris Life Sciences uses artificial intelligence and machine learning to determine a more precise diagnosis, and with it, a more individualized treatment plan? Dr. Sharma.
Dr. Sachein Sharma: So, as Gretchen just mentioned, Caris has a large database with matched molecular and outcomes data. This allows us to interrogate important questions. So in about 3% to 5% of cases, there’s a question of where the tumor originates in the body. And these cases are often termed cancers of unknown primary or cut cases. In addition, there are cases where the patient may present with clinical ambiguity or in an atypical way. Then that makes it harder for a pathologist to determine a definitive diagnosis using immunohistochemistry. And that’s typically been the gold standard method. Caris set out to develop a best-in-class predictor of tumor of origin known as MI GPSai or genomic prevalence score. Gretchen, can you expand on that, please?
Dr. Gretchen Hubbard: Thanks, Sachein. So when a patient’s biopsy is sent to Caris, a comprehensive analysis of the patient’s tumor, which includes both DNA and RNA sequencing, as well as immunity to chemistry, is done. And the goal is to interrogate all parts of the tumor in order to find potential therapies or clinical trials that the patients may qualify for. So when Caris sequences the patient’s tumor, it takes the DNA and RNA information from that patient and compares it to 21 different tumor types that Caris or has developed signatures for.
One of those signatures is for prostate cancer, specifically prostate adenocarcinoma. So the algorithm predicts a percent similarity to the signatures, which provides information as to where the tumor originated. For example, if a patient is diagnosed with prostate cancer and sent to Chris and GPS predicts it’s 99% prostate, then the GPS prediction and diagnosis are concordant with one another. But if the predictor says 99% lung adenocarcinoma, then the Caris pathologists will notify the ordering physician and discuss whether additional tests are reasonable to confirm the original diagnosis.
So it’s important to note that GPS alone is not a diagnosis, but there’s additional factors that are taken into consideration for each case. But we have seen cases where, after a discordance in additional testing, GPS has resulted in a change of diagnosis for the patient, and ultimately this has changed the course of treatment and potentially the outcome for the patient as well.
Ed Randall: Dr. Hubbard, let’s stay with you. Specifically, how does AI represent an improvement in better predicting a patient’s response to a specific treatment?
Dr. Gretchen Hubbard: Traditionally, to identify a patient for a targeted therapy, something like a PARP inhibitor that is often used in prostate cancer, if you’ve heard of olaparib or rucaparib, or to identify patients for immune checkpoint inhibitors, if you’ve heard of like pembrolizumab, atezolizumab, companies like Caris would test for the presence of biomarkers.
So an example would be a mutation in BRCA1 or the expression of a protein called PD-L1. And biomarkers would help qualify a patient for a particular therapy, but they don’t predict whether a patient will actually respond. So Caris is going one step further than just identifying a biomarker. And we’re using that large database that we’ve discussed to predict signatures of response. So by predicting whether a patient will actually respond to a particular drug, the outcomes of these patients may be improved. And that’s the ultimate goal of molecular profiling and precision oncology.
Ed Randall: To maximize the incredible power of artificial intelligence and machine learning, won’t the quality of research, testing, and data gathering need to keep pace? Won’t this mean in the world of prostate cancer, the clinical trials will need to include a higher percentage of Black men since Blacks are so disproportionately affected by the disease? Dr. Sharma.
Dr. Sachein Sharma: So the short answer is yes. And Caris has been working to address health disparities through our international research collaborative called the Precision Oncology Alliance. And one of its goals is to increase the number of patients from different races and ethnicities that are profiled so that these important questions can be addressed. It’s also important that patients are aware of molecular profiling and ultimately access to this profiling becomes more widely available regardless of socioeconomic status or location. So ideally, this data can be representative of the nations and the world’s diverse population.
Ed Randall: Another advancement in extending life Johnson explains is immunotherapy. That is, precision oncology. Cancer is no longer about one size fits all treatments and has moved to a specific patient and his or her tumors and disease. Can you elaborate what we mean by precision medicine and precision oncology versus how cancer treatment has historically taken place? Dr. Sharma, back to you.
Dr. Sachein Sharma: Sure. So traditional medicine uses stage or where a cancer is, how extensive it is, and if it’s spread to other parts of the body and grade, or how likely cells are to grow and spread, i.e. its aggressiveness. And both of these things guide treatment and management plans. Precision medicine, though, and precision oncology are both tailored to the patient. And this is similar to the design of custom-grade clothing in that precision medicine takes into account a personalized approach by examining factors such as biomarker status to identify the best course of treatment.
And so by accounting for multiple data points, precision medicine and precision oncology are able to suggest the best therapy, whether chemotherapy, targeted therapy, immunotherapy, or even a combination. And the immune microenvironment or the immune cells present in and around that patient’s tumor, like B or T cells, guide this particular selection.
Ed Randall: Dr. Hubbard, your thoughts?
Dr. Gretchen Hubbard: One of the drawbacks of traditional chemotherapy has been the side effects. Chemotherapy not only attacks the tumor cells but can affect normal healthy cells as well. Targeted therapy was developed to target specific genes and proteins that are involved in the growth and survival of only cancer cells. And immunotherapy uses the body’s own immune system to attack cancer cells. Because the treatments are designed to target cancer cells and spare damage to healthy cells, these treatments may have fewer side effects than conventional chemotherapy.
Ed Randall: What are the relevant biomarkers to test for in men who have metastatic castration-resistant prostate cancer? Dr. Sharma.
Dr. Sachein Sharma: This is a very important question. And to do this, I’ll contextualize that in the following way. So damage to our DNA can occur through a variety of mechanisms, including exposure to UV light from the sun or normal processes inside our cells, such as metabolism or replication. And this damage can cause breaks in the strands of our DNA, which under normal conditions can be repaired through a variety of mechanisms.
One of these mechanisms is the homologous recombination repair pathway or HRR pathway. And that’s a network of approximately 15 genes like BRCA1 and BRCA2 as well as their respective RNA and proteins that are critical for the repair of double-stranded breaks in our DNA. Patients with somatic mutations in these genes may qualify for targeted therapy, such as PARP inhibitors like rucaparib or olaparib. These drugs are examples of targeted therapies. and it’s important to identify mutations in all relevant genes within this pathway. And that’s what Caris does through a technique where we sequence the DNA called whole exome sequencing.
Ed Randall: Dr. Hubbard.
Dr. Gretchen Hubbard: In addition to the biomarkers like BRCA1 and BRCA2 that Sachein mentioned, there are also genomic signatures that are what we call tumor agnostic and important to test for. So tumor mutational burden, microsatellite instability, and genomic loss of heterozygosity are three signatures Caris tests for in every patient regardless of the tumor type. And we do this by DNA sequencing.
Additionally, Caris tests for mismatch repair proteins by a technique called immunohistochemistry. So the goal is to identify all the potential therapies and clinical trials for patients with not only prostate cancer but other cancers as well. And it’s really important that a multifactorial approach is taken. And this is how Caris approaches molecular profiling by interrogating the DNA, RNA, and proteins for each patient.
Ed Randall: Would the presence of these biomarkers in parents, aunts, uncles, and siblings change or expand the definition of family history when a man is evaluating his risk of developing prostate cancer? Dr. Sharma.
Dr. Sachein Sharma: Thanks for the question. So while our platform provides somatic testing for advanced or metastatic prostate cancer patients, family history may often be as informative in the diagnostic odyssey. And a relevant example would be germline mutations such as BRCA1, which can be passed on from parents to their children. Another example could be mismatch repair deficiency, and this may indicate Lynch syndrome, which is an inherited disorder in which affected individuals have a higher than normal chance of developing colorectal cancer and certain other types of cancer, particularly before the age of 50.
Germline testing, though, is recommended for different prostate cancer states. In the metastatic setting, it’s used to determine PARP or platinum candidacy, clinical trial eligibility, and to inform family cancer risks.
Ed Randall: What is the advantage of identifying these specific biomarkers and what effect has Caris molecular profiling had on optimizing treatment plans versus what was able to be offered, say, even ten years ago? Dr. Hubbard.
Dr. Gretchen Hubbard: Ten years ago, or even greater, traditional recommendations were treatments like chemotherapy, hormone therapy, and radiation therapy, either alone or in combination, depending on the particular cancer type. By testing the DNA, RNA, and protein biomarkers, Caris now does comprehensive molecular profiling that can identify targeted therapies like PARP inhibitors and immunotherapies. And this ultimately offers greater access to additional therapies for patients and clinical trial options for patients who have advanced or metastatic disease.
Ed Randall: Our men’s support group has taught us nothing if not that a man and often his caregiver must be their own strongest advocates in finding their way to the right doctors and then to a plan of care that balances the best treatment of the disease with the lowest odds of dealing with life-changing side effects. For men who have been diagnosed with metastatic prostate cancer, how do they go about getting their oncologists to consider the Caris Life Sciences approach to sequencing and treating their disease? Dr. Sharma.
Dr. Sachein Sharma: So for each cancer type, there are clinical practice guidelines developed by an alliance of the world’s leading cancer centers. These are called the National Comprehensive Cancer Network, or NCCN guidelines. And your health care teams may use these guidelines to discuss treatment and management plans with patients and caregivers. Knowing the most up-to-date guidelines for metastatic prostate cancer can be helpful when discussing molecular tests with treating physician.
Ed Randall: Dr. Hubbard.
Dr. Gretchen Hubbard: It is similarly important to become educated about molecular profiling as a whole because not all companies are equal in terms of what they test for. So it’s vital to take a comprehensive approach by examining, as I’ve mentioned, the DNA, RNA, and protein for each patient’s tumor. So one of the resources we welcome the listeners to consider is going to the Caris Life Sciences website at carislifesciences.com to find out more information about comprehensive molecular profiling.
Ed Randall: We wish you the best of success in your continuing lifesaving work. Thank you so much, Dr. Hubbard and Dr. Sharma, for being with us. I’m Ed Randall. Thank you so much for joining us on our Stay in the Game podcast, Fans For the Cure.
Dr. Sachein Sharma: Thank you for your time.
Dr. Gretchen Hubbard: Thank you so much.
Announcer: Thanks for listening to the show. You can find program notes and a full transcript at the charity’s website, fansforthecure.org. Be sure to subscribe to our podcast in iTunes, Spotify, Stitcher, and everywhere. good podcasts are available. And if you like what you heard, a positive review on iTunes will help other people also find our show.
