Good morning, ladies and gentlemen, and welcome to the Guardant Health ECLIPSE Study Update Call. At this time, all lines are in a listen-only mode. If at any time during this call you require immediate assistance, please press star zero for the operator. This call is being recorded on Tuesday, May 9th, 2023. I would now like to turn the conference over to Dr. Craig Eagle, Chief Medical Officer of Guardant Health. Please go ahead.
Great. Thank you very much and welcome everybody. Joining me today, I've got Dr. Chung, gastroenterologist at Massachusetts General Hospital, Professor of Medicine at Harvard Medical School and investigator on the ECLIPSE study and the presenting author of the data this morning. Welcome, Dr. Chung. I also have Dr. Grady, gastroenterologist and Medical Director of the Gastrointestinal Cancer Prevention Program and professor in the Clinical Research Division and professor in the Public Health Sciences Division at Fred Hutch, and professor in the Division of Gastroenterology at the University of Washington, and senior investigator for the ECLIPSE study. Welcome, Dr. Grady. Also joining us is Dr. Lichtenfeld, medical oncologist and Chief Medical Officer at Jasper Health and former Deputy and Interim Chief Medical Officer at the American Cancer Society. Welcome, Dr. Len.
Finally, Dr. Nomitzka, professor in Health System Science Department at the Kaiser Permanente School of Medicine and former scientific director of the United States Preventive Services Task Force. Welcome, Dr. Metzger. Thank you all for being here today. Dr. Chung, perhaps I can start with you. For those who were unable to join your presentation this morning, can you please highlight the major takeaways?
Yes, I'd be happy to. Despite the widespread availability of many colon cancer screening options, there remain persistent barriers to screening, leading to only 59% of eligible individuals being adherent to screening, which is well below the target of 80% set by many leading health organizations. A blood-based colon cancer screening test that's completed as part of a routine healthcare encounter presents an opportunity to increase screening adherence. This morning, we reported on the results of the ECLIPSE study, which was designed to evaluate the performance of a cell-free DNA blood-based colon cancer screening test in an average-risk population that's undergoing a screening colonoscopy. We enrolled individuals at average risk for the development of colon cancer between the ages of 45 and 84 who had not had recent colon cancer screening.
The eligible individuals provided blood sample prior to any preparation for the colonoscopy, and the blood test was evaluated and compared to the colonoscopy. The co-primary endpoints were sensitivity of the test for the detection of colon cancer and specificity for the absence of advanced neoplasia, which was defined as colorectal cancer or an advanced precancerous lesion. In this valuable study population of 7,861 individuals, which included more than 20% of participants who identified as non-white, specifically 12% who self-identified as Black African American, and 13% who identified as Hispanic or Latino, the sensitivity of the assay was 83% and the specificity was 90%, both meeting the study endpoints and demonstrating performance that was on par with currently available stool-based colon cancer screening option.
The sensitivity in stage I through III colon cancers was 81%, and the sensitivity in localized disease was 72%. As a secondary endpoint, ECLIPSE evaluated the sensitivity for the detection of advanced precancerous lesions, which was 13%. The study also showed that the assay had colon cancer sensitivity that was comparable to currently available stool-based screening option. However, it's important to remember that screening programs require consideration of test performance under real-world conditions that integrates patient adherence rates in order to assess clinical effectiveness. In real-world clinical use, this assay has had a patient adherence rate greater than 85%, which far exceeds that of stool-based testing, which ranges anywhere between 43% and 71%, and for colonoscopy, which ranges between 28% and 59%.
With more people being adherent, it means more people get screened, and it demonstrates the potential for this assay to really be highly effective screening tool in the population. In summary, our study shows that the cell-free DNA assay is the first blood-based test with performance comparable to current guideline recommended non-invasive colon cancer screening option. Combined with improved adherence with a blood-based diagnostic, this assay is poised to have a significant population impact on colon cancer screening.
Thank you, Dr. Chung. Just sort of tapping onto your expertise, again, from your perspective as a practicing gastroenterologist, what are the most practical clinical applications of these results?
Colon cancer screening has been shown to positively impact colon cancer-related death, which is why it's so important for people to participate. Despite the multiple screening options available, there remains still significant barriers to completing this, and these include factors like time, access, and comfort with the current options. The fact that this is now a blood-based test that can demonstrate comparable performance in the intended use population is exciting and means now that we have a new option that's available to add to our armamentarium. Given what we know about patient pre-preference for blood-based testing, this option may be viewed as a more acceptable option for those who've either delayed or declined colon cancer screening. Ultimately, the more options we have available, the more likely are people gonna act on it.
The fact that we have options that include colonoscopy, stool, and now blood is a very exciting leap forward to ensure that more people have access to this very valuable preventive intervention.
This is very exciting to be able to get more people up to colon cancer screening. Thank you. Just switching to you, Dr. Grady. The ability to execute the ECLIPSE study during the COVID-19 pandemic is certainly a exciting, impressive endeavor. Perhaps you could take a few moments and walk us through the clinical study design and then, you know, specifically in particular to terms of participant enrollment and inclusion.
Yes, it'd be my pleasure to do so. I will note that the COVID-19 pandemic was an especially challenging time for us practicing gastroenterology, and there are numerous restrictions on the use of healthcare resources for preventative services like colonoscopy. The long-term impacts of this interruption on screening colorectal cancer-related deaths are being felt now and will be felt for years to come. We've seen an increase in the stage of people presenting with colon cancer even in the immediate future in the last couple of years. That's going to continue to get worse, I imagine. That's all the more reason why I echoed Dr.
Chung's excitement about the ECLIPSE study results and the ability to have another colorectal cancer screening option that more people are likely to complete, and that will ensure more people are up to date with screening. To answer your question, the ECLIPSE study is an event-driven study. Given that individuals who are enrolled and blood samples were collected prior to colonoscopy, what was needed to be done was the ECLIPSE study had to over-enroll participants in order to make sure that we had enough colorectal cancers. We also had about 7,800 people with a non-colorectal cancer finding on colonoscopy. Excuse me. Original projections of the baseline population prevalence of colorectal cancer rates were used to inform the target enrollment numbers.
However, it's important to know that when we're, you're designing these studies, that the enrollment numbers will be updated based on the real-world experience that occurs during the clinical trial's enrollment. As in ECLIPSE enrolled, the observed pre-prevalence of colorectal cancer was lower than our original projected prevalence. That led us to need to expand the enrollment in order to reach our target number of cancers. Interestingly, I had the opportunity to participate in a colorectal cancer screening study that was led by the Early Detection Research Network, or the EDRN. That has been going on for many years.
Several years ago we had had a similar experience where we had to update our enrollment projections because the observed versus projected prevalence of colorectal cancer and advanced pre-neoplastic lesions differed from what we originally had anticipated. This is normal clinical trial execution and underscores the importance of an experienced trial team in running these large-scale studies. What we found with the ECLIPSE study was that we enrolled between October 2019 and September 2022, and we had approximately 22,000 healthy individuals who are enrolled in the study. Not all of these individuals completed their screening colonoscopy, and not all of these individuals provided a blood sample. Even though we enrolled 22,000 people, that does not mean that 22,000 people were fully evaluable.
Even of these individuals who had cancers, some of them were also ineligible due to reasons such as they did not meet the inclusion or exclusion criteria or did not provide an adequate study blood sample. Our final evaluable cohort included 65 people with colorectal cancer and 7,796 individuals with a non-colorectal cancer finding on colonoscopy. Individuals with non-colorectal cancer were included for our specificity calculation. Based on our statistical analysis plan, we knew we needed approximately 7,000 people who had a colonoscopy that was negative for advanced dysplasia, and that was needed in order to have a well-powered specificity co-primary endpoint. Running the remaining samples from the patients who are enrolled in the study was not needed in order to define the specificity co-primary endpoint.
In order to identify the cohort of individuals with non-colorectal cancer that were to be included in the specificity endpoint, individuals without colorectal cancer were randomly selected such that the final age distribution of the non-colorectal cancer individuals matched the age distribution of the U.S. population based on the 2020 census. The selection of these individuals was random, was performed prior to running any of the samples with the Shield test and blinded to the any clinical details beyond knowing that they were non-colorectal cancer patients and that they were randomly selected by age. This is a standard approach to the evaluation of diagnostics in screening studies. It's important to note that an alternative approach would have been to run all samples for individuals who met inclusion and exclusion criteria and had a complete and valid colonoscopy and had a complete and valid blood sample result.
However, the additional samples run would've had a minimal impact on the specificity results that we reported in the ECLIPSE trial.
Thank you, Dr. Grady. You know, your perspective, given your experience in leading other colorectal cancer screening study, is very helpful, particularly around the execution of ECLIPSE and appreciate the additional context you provided. Just while I've got you, I was wondering if you could comment on your perspective on the first stage performance of the Shield assay that Dr. Chung mentioned.
Yes, it'd be my pleasure to. As Dr. Chung mentioned, the co-primary endpoints of the ECLIPSE study were sensitivity for colorectal cancer and specificity for the absence of advanced neoplasia as compared to the colonoscopy reference standard. We had a total of 65 colorectal cancers that were included in the final analysis. This morning, Dr. Chung shared some really interesting data about the performance in stage I-III cancers, and that was 81%, and the performance in localized cancer, which was 72%. Now, it's important to note that while these are interesting data, I would caution over over-interpretation of the per stage performance given the number of cancers in any single stage group and that that was too few to provide a robust stage-by-stage comparison. It's important to note the colorectal cancer screening rate.
Our screening tests are evaluated based on overall sensitivity and specificity. The fact that the Shield test demonstrated overall performance that's comparable to stool-based testing leads me to believe this will be a very compelling and viable colorectal cancer screening options for patients and providers.
Thank you, Dr. Grady. Just continue to tap into your expertise in this area. Perhaps you can tell me a little bit more about the interpretation of the secondary endpoint with advanced precancerous lesions with a sensitivity of 13%.
It'd be my pleasure. The primary objective of ECLIPSE was to evaluate the performance of the Shield test for the sensitivity of colorectal cancer. It's important to note that the colorectal cancer screening is the early detection of asymptomatic cancers and the prevention of colorectal cancer-related deaths. Colorectal cancer is a unique cancer in that we have the opportunity for cancer prevention by the detection of these advanced precancerous lesions. It's helpful to note the detection rates of the advanced precancerous lesions for any colorectal cancer screening option. From my perspective, the trend towards improved performance with more advanced disease, going from the advanced precancerous lesion to the early stage cancer to late stage cancer, is consistent with the biology of the disease and the shedding of cell-free DNA.
Based on our understanding of cell-free DNA diagnostics, the results from the ECLIPSE trial are consistent with what I would have expected and are very promising for this new form of a diagnostic test.
Right. Thank you. Just continuing on the theme of delving into the ECLIPSE data, just going back to you, Dr. Chung. In the presentation, you shared information on incompletely staged cancers. Can you please walk us through what that means and the impact of the results you presented today?
Sure, I'd be happy to. In the ECLIPSE study, all cancers had to be staged per AJCC criteria as per protocol. What that means is you need to have to confirm that there was a cancer, and you need to then also perform a full lymph node dissection to understand the risk for systemic disease. What's interesting, given that the focus in our study was on individuals presenting for average-risk colon cancer screening, enrolled individuals were presenting to endoscopy centers for their screening colonoscopy. If the individual was subsequently diagnosed with cancer, then they needed to go to a separate cancer center for treatment. For clinical study execution, this requires working with the local trial site to obtain records and information from the subsequent surgical and oncology providers.
It's not uncommon for individuals to be, “lost to follow-up,” which means that despite all our efforts to obtain the additional information, it was just not possible. We saw this in three individuals who had a diagnosis confirmed of colon cancer, but the staging information was not available, and we just were simply not able to include them in the per stage analysis. Interestingly, there were five other individuals that had what we term malignant polyps, and these individuals are often when they're identified in routine clinical practice, the decision to complete full cancer staging, meaning doing that lymph node dissection, vary significantly based on many clinical factors because the cancer is believed to be just confined to that resected polyp. In many cases, full nodal dissection does not happen.
Because this full nodal dissection was not complete in these individuals, they technically could not have their cancer staged by the AJCC criteria. Therefore, per our own study ECLIPSE protocol, they are considered incompletely staged cancers. However, clinically, these individuals do have a confirmed pathologic T1 cancer, and they get clinically managed as stage 1 cancers. For that reason, in the data presented today, we included these five individuals with malignant polyps with the stage I cancer group for our stage-by-stage analysis. This was a more conservative approach to take, but we felt that this would be the approach that would adequately allow for interpretation of this data as it would look like in the real world. Ultimately, these are challenging lesions for any non-colonoscopy screening test to identify.
I appreciate you walking us through these details around the staging information. Perhaps one final question to both you, Dr. Chung, and to Dr. Grady before I shift to the other panel members. In your opinion, will the Shield detection rate for advanced precancerous lesions and localized cancer inhibit or hinder adoption of the test?
As mentioned previously, the goal of colon cancer screening is early-stage asymptomatic cancer detection. Just today, too many people are not getting screened and are dying of cancer. In fact, in one study, 76% of cancer-related deaths occurred in individuals who were not up to date with screening. Having more options for patients will get more people screened. The performance of this particular test, which is on par with stool-based alternatives, demonstrates that it should be included as a screening option for eligible individuals.
I agree with Dr. Chung's comments. It's interesting to note that in an interview about their recently released study, two leaders in colorectal cancer screening, Dr. Sidney Winawer and Dr. Ann Zauber, remind us that a screening-based test is only as good as its uptake, that any screening test is better than none, and the best one is the one that gets done and gets done well. The Shield test demonstrates performance on par with stool-based tests. The adherence numbers that we have seen so in the first year of the Shield test being made available show that people will complete this test. This is an exciting time for colorectal cancer screening, and I do believe that this test will be used.
Thank you, Dr. Chung and Dr. Grady, for walking us through not just the study details, but also providing the clinical perspective and context of this data, as talked about. I just wanna switch gears a little bit and talk about how these tests will be interpreted in terms of the broader colorectal cancer screening landscape. Dr. Len, if I can start with you. How do you interpret these test results in the context of currently available options?
Thank you, Craig. As Doctors Grady and Chung already mentioned, these results are on par with currently available stool-based screening options. Let's talk for a moment about what that really means. In published studies on the performance of stool-based testing, the sensitivity for colorectal cancer screening ranges from 74%-92%, and the specificity ranges from 87%-95%. The ECLIPSE study we're discussing today demonstrated that Shield has a sensitivity of 83% and a specificity of 90%. This performance is on par with stool-based testing with the added benefit of improved adherence with blood-based testing.
Thank you, Dr. Len, for putting that in context for us. Historically, the American Cancer Society has been very proactive with new technologies and modalities for screening that will positively impact the cancer burden in healthy individuals. Based on your experience with American Cancer Society, can you tell us a bit about how American Cancer Society may approach evaluation of Shield as a new colorectal cancer screening option?
Well, reputable organizations such as the American Cancer Society, who develop guidelines for the early detection of cancer, look very carefully at the evidence regarding screening tests, including sensitivity, specificity, and adherence, and how any given test will effectively help reduce the cancer burden. The data presented today shows that Guardant Shield offers sensitivity and specificity in line with other guideline-approved and reimbursed tests for the early detection of colorectal cancer, and there's evidence that adherence is actually better and more consistent than with other colorectal cancer early detection tests. ACS has previously shared in public comments to the Centers for Medicare & Medicaid Services that they view adherence as an important factor in considering colorectal cancer screening options, going so far as to suggest that adherence should be part of coverage decisions for CMS.
For this reason, it's my opinion that a test like Shield with comparable performance to stool-based options will be seriously considered by ACS for guideline inclusion, especially if there are supporting data demonstrating improved adherence to colorectal cancer screening with incorporation of Shield as an option. Ultimately, what everyone wants are screening tests that help reduce the burden of colorectal cancer, Guardant Shield takes us further down the path of accomplishing that goal.
Thank you, Dr. Len. Perhaps now if it's switched to you, Dr. Doubeni. Can you walk us through what factors into USPSTF decisions around recommendations of screening options?
Sure. Yes. Thank you, Dr. Eagle. The U.S. Preventive Services Task Force has really published extensively about its methods on how it evaluates preventive services and screening tests. Some of the factors that they consider are, do the studies have appropriate research design to answer the questions being asked? What are the studies results, but are they the studies of good quality? Are the results generalizable to the primary care population? Again, the general population that this test is being evaluated in. They also look at how large the studies are, how many studies there are, and are there consistent study results across multiple studies. The USPSTF does not look at just necessarily just one study, but many studies.
They prioritize clinical research data studies and real-world data as their primary source of information to assess preventive care and screening. ECLIPSE has an adequate study design, observational study design that's adequately powered to the intended screening population, which is a general average-risk population. As Dr. Len mentioned, there are additional studies ongoing that the USPSTF will likely look at having to do with real-world uptake and adherence, and those studies will also inform USPSTF recommendations. I do want to note that the ECLIPSE study has a real strength in that it basically was able to recruit large number of diverse people from diverse age groups and racial ethnic backgrounds from many different healthcare settings, and I think that's a huge plus for the ECLIPSE study.
I also wanted to mention that modeling is used by the USPSTF as supplemental data that can help inform the decisions about, you know, about screening, when to screen, what age to start and stop when there are gaps in the real world data. That modeling is seen as secondary to real world data from real clinical studies. I think the other thing that I wanted to note is that the U.S. Preventive Services Task Force is really focused on understanding the impact of these tests at a large population level. As such, you can see that when they do modeling the test, the different modalities, whether it's colonoscopy or stool-based tests or a blood test, the tests are evaluated as compared to not getting any kind of screening at all. That's how they do the modeling.
The USPSTF usually state that there's a number of recommended tests that are acceptable. They usually do not rank any one of the recommended tests over another because ultimately it's up to the patients and their providers to decide which test is the right test for them. I mean, I think that this is really an exciting time that there is another... Thanks for letting me give my opinion.
Yeah. Thank you, Dr. Dubeni. Really appreciate your perspective. Just wanna wrap things up a little bit just to acknowledge and thank you to all the panelists for providing a unique perspective from the ECLIPSE data through to how it impacts the clinic through to then guidelines and USPSTF. I just wanna thank those that listened, and have a great day. Thank you.
Thank you, ladies and gentlemen.