ASCO25 Recap: CHALLENGE, DESTINY-Breast09, and More

Jun 19, 2025

Ask episode

Chapters

Transcript

Episode notes

Dr. John Sweetenham and Dr. Erika Hamilton highlight key abstracts that were presented at ASCO25, including advances in breast and pancreatic cancers as well as remarkable data from the use of structured exercise programs in cancer care.

Transcript

Dr. Sweetenham: Hello, and welcome to the ASCO Daily News Podcast. I'm your host, Dr. John Sweetenham. Today, we'll be discussing some of the key advances and novel approaches in cancer care that were presented at the 2025 ASCO Annual Meeting.

I'm delighted to be joined again by the chair of the Meeting's Scientific Program, Dr. Erika Hamilton. She is a medical oncologist and director of breast cancer and gynecologic cancer research at the Sarah Cannon Research Institute in Nashville, Tennessee.

Our full disclosures are available in the transcript of this episode.

Dr. Hamilton, congratulations on a fantastic meeting. From the practice-changing science to the world-renowned speakers at this year's Meeting, ASCO25 really reflected the amazing progress we're seeing in oncology today and the enormous opportunities that lie ahead of us. And thanks for coming back on to the podcast today to discuss some of these advances.

Dr. Hamilton: Thanks, Dr. Sweetenham. I'm happy to join you today. It really was an impactful ASCO Annual Meeting. I probably am biased, but some great research was presented this year, and I heard lots of great conversations happening while we were there.

Dr. Sweetenham: Yeah, absolutely. There was a lot of buzz, as well as a lot of media buzz around the meeting this year, and I think that's probably a good place to start. So I'd like to dive into abstract number LBA3510. This was the CHALLENGE trial, which created a lot of buzz at the meeting and subsequently in the media.

This is the study that was led by the NCI Canada Clinical Trials Group, which was the first randomized phase 3 trial in patients with stage III and high-risk stage II colon cancer, which demonstrated that a post-treatment structured exercise program is both feasible and effective in improving disease-free survival in this patient group.

The study was performed over a long period of time and in many respects is quite remarkable. So, I wonder if you could give us your thoughts about this study and whether you think that this means that our futures are going to be full of structured exercise programs for those patients who may benefit.

Dr. Hamilton: It's a fantastic question. I think that this abstract did create a lot of buzz. We were very excited when we read it. It was highlighted in one of the Clinical Science Symposium sessions. But briefly, this was a phase 3 randomized trial. It was conducted at 55 centers, so really a broad experience, and patients that had resected colon cancer who completed adjuvant therapy were allowed to participate. There were essentially 2 groups: a structured exercise program, called 'the exercise group,' or health education materials alone, so that was called just 'the health education group.' And this was a 3-year intervention, so very high quality. The primary end point, as you mentioned, was disease-free survival.

This actually accrued from 2009 to 2024, so quite a lift, and almost 900 patients underwent randomization to the exercise group or the health education group. And at almost 8 years of follow-up, we saw that the disease-free survival was significantly longer in the exercise group than the health education group. This was essentially 80.3% of patients were disease-free in exercise and 73.9% in the health education group. So a difference of over 6 percentage points, which, you know, at least in the breast cancer world, we make decisions about whether to do chemotherapy or not based on these kind of data.

We also looked at overall survival in the exercise group and health education group, and the 8-year overall survival was 90.3% in the exercise group and 83.2% in the health education group. So this was a difference of 7.1%. Still statistically significant.

I think this was really a fantastic effort over more than a decade at over 50 institutions with almost 900 patients, really done in a very systematic, high-intervention way that showed a fantastic result. Absolutely generalizable for patients with colon cancer. We have hints in other cancers that this is beneficial, and frankly, for our patients for other comorbidities, such as cardiovascular, etc., I really think that this is an abstract that deserved the press that it received.

Dr. Sweetenham: Yeah, absolutely, and it is going to be very interesting, I think, over the next 2 or 3 years to see how much impact this particular study might have on programs across the country and across the world actually, in terms of what they do in this kind of adjuvant setting for structured exercise.

Dr. Hamilton: Absolutely.

So let's move on to Abstract 3006. This was an NCI-led effort comparing genomic testing using ctDNA and tissue from patients with less common cancers who were enrolled in but not eligible for a treatment arm of the NCI-MATCH trial. Tell us about your takeaways from this study.

Dr. Sweetenham: Yeah, so I thought this was a really interesting study based, as you said, on NCI-MATCH. And many of the listeners will probably remember that the original NCI-MATCH study screened almost 6,000 patients to assess eligibility for those who had an actionable mutation. And it turned out that about 60% of the patients who went on to the study had less common tumors, which were defined as anything other than colon, rectum, breast, non–small cell lung cancer, or prostate cancer. And most of those patients lacked an eligible mutation of interest and so didn't get onto a trial therapy.

But with a great deal of foresight, the study group had actually collected plasma samples from these patients so that they would have the opportunity to look at circulating tumor DNA profiles with the potential being that this might be another way for testing for clinically relevant mutations in some of these less common cancer types.

So initially, they tested more than 2,000 patients, and to make a somewhat complicated story short, there was a subset of five histologies with a larger representation in terms of sample size. And these were cholangiocarcinoma, small cell lung cancer, esophageal cancer, pancreatic, and salivary gland cancer.

And in those particular tumors, when they compared the ctDNA sequencing with the original tumor, there was a concordance there of around 84%, 85%. And in the presentation, the investigators go on to list the specific mutated genes that were identified in each of those tumors.

But I think that the other compelling part of this study from my perspective was not just that concordance, which suggests that there's an opportunity there for the use of ctDNA instead of tumor biopsies in some of these situations, but what was also interesting was the fact that there were several clinically relevant mutations which were detected only in the circulating tumor DNA. And a couple of examples of those included IDH1 for cholangiocarcinoma, BRAF and p53 in several histologies, and microsatellite instability was most prevalent in small cell lung cancer in the ctDNA.

So I think that what this demonstrates is that liquid biopsy is certainly a viable screening option for patients who are being assessed for matching for targeted therapies in clinical trials. The fact that some of these mutations were only seen in the ctDNA and not in the primary tumor specimen certainly suggests that there's some tumor heterogeneity. But I think that for me, the most compelling part of this study was the fact that many of these mutations were only picked up in the plasma. And so, as the authors concluded, they believe that a comprehensive gene profiling with circulating tumor DNA probably should be included as a primary screening modality in future trials of targeted therapy of this type.

Dr. Hamilton: Yeah, I think that that's really interesting and mirrors a lot of data that we've been seeing. At least in breast cancer, you know, we still do a biopsy up front to make sure that our markers, we're still treating the right disease that we think we are. But it really speaks to the utility of using ctDNA for serial monitoring and the emergence of mutations.

Dr. Sweetenham: Absolutely.

And you mentioned breast cancer, and so I'd like to dwell on that for a moment here because obviously, there was a huge amount of exciting breast cancer data presented at the meeting this year. And in particular, I'd like to ask you about LBA1008, the DESTINY-Breast09 clinical trial, which I think has the potential to establish a new first-line standard of care for metastatic HER2+ breast cancer. And that's an area where we haven't seen a whole lot of innovation for around a decade now. So can you give us some of the highlights of this trial and what your thinking is, having seen the results?

Dr. Hamilton: Yeah, absolutely. So this was a trial in the first-line metastatic HER2 setting. So this was looking at trastuzumab deruxtecan. We certainly have had no shortage of reports around this drug, initially approved for later lines. DESTINY-Breast03 brought it into our second-line setting for HER2+ disease and we're now looking at DESTINY-Breast09 in first-line.

So this actually was a 3-arm trial where patients were randomized 1:1:1 against standard taxane/trastuzumab/pertuzumab in one arm; trastuzumab deruxtecan with pertuzumab in another arm; and then a third arm, trastuzumab deruxtecan alone. And what we did not see reported was that trastuzumab deruxtecan-alone arm. But we did have reports from the trastuzumab deruxtecan plus pertuzumab versus the chemo/trastuzumab/pertuzumab. And what we saw was a statistically significant improvement in median progression-free survival, 26.9 months up to 40.7, so an improvement of 13.8 months, over a year in PFS. Not to mention that we're now in the 40-month range for PFS in first-line disease. Really, across all subgroups, we really weren't able to pick out a subset of patients that did not benefit. We did see about a 12% ILD rate with trastuzumab deruxtecan. That really is on par with what we've seen in other studies, around 10%-15%.

I think that this is going to become a new standard of care in the first-line. I think it did leave some unanswered questions. We saw some data from the PATINA trial this past San Antonio Breast, looking at the addition of endocrine therapy with or without a CDK4/6 inhibitor, palbociclib, for those patients that also have ER+ disease, after taxane has dropped out in the first-line setting.

So how we're going to kind of merge all this together is, I suspect that there are going to be patients that we or they just don't have the appetite to continue 3 to 4 years of trastuzumab deruxtecan. And so we're probably going to be looking at a maintenance-type strategy for them, maybe integrating the PATINA data there. But how we really put this into practice in the first-line setting and if or when we think about de-escalating down from trastuzumab deruxtecan to antibody therapy are some lingering questions.

Dr. Sweetenham: Okay, so certainly is going to influence practice, but watch this space for a little bit longer, it sounds as though that's what you're saying.

Dr. Hamilton: Absolutely.

So let's move on to GI cancer. Abstract 4006 reported preliminary results from the randomized phase 2 study of elraglusib in combination with gemcitabine/nab-paclitaxel versus the chemo gemcitabine/nab-paclitaxel alone in patients with previously untreated metastatic pancreatic cancer. Can you tell us more about this study?

Dr. Sweetenham: Yeah, absolutely. As you mentioned, elraglusib is actually a first-in-class inhibitor of GSK3-beta, which has multiple potential actions in pancreatic cancer. But the drug itself may be involved in mediating drug resistance as well as in some tumor immune response modulation. Some of that's not clearly understood, I believe, right now. But certainly, preclinical data suggests that the drug may be effective in preclinical models and may also be effective in combination with chemotherapy and potentially with immune-modulating agents as well.

So this particular study, as you said, was an open-label, randomized phase 2 study in which patients with pancreatic cancer were randomized 2:1 in favor of the elraglusib plus GMP—gemcitabine and nab-paclitaxel—versus the chemotherapy alone. And upon completion of the study, which is not right now, median overall survival was the primary end point, but there are a number of other end points which I'll talk about in just a moment. But the sample size was planned to be around 207 patients. The primary analysis included 155 patients in the combination arm versus 78 patients in the gemcitabine/nab-paclitaxel arm.

Overall, the 1-year overall survival rate was 44.1% for the patients in the elraglusib-containing arm versus 23.0% in the patients receiving gemcitabine/nab-paclitaxel only. When they look at the median overall survival, it was 9.3 months for the experimental arm versus 7.2 months for chemotherapy alone. So put another way, there's around a 37% reduction in the risk of death with the use of this combination arm.

The treatment was overall well-tolerated. There were some issues with grade 1 to 2 transient visual impairment in a large proportion of the patients. The most common treatment-related adverse effects with the elraglusib/GMP combination was transient visual impairment, which affected around 60% of the patients. Most of the more serious treatment-related adverse events included neutropenia, anemia, and fatigue in 50%, 25%, and 16% of the patients, respectively.

So the early results from this study show a significant benefit for 1-year overall survival and for median overall survival with, as I mentioned above, a significant reduction in the risk of death. The authors went on to mention that the median overall survival for the control arm in this study is somewhat lower than in other comparable trials, but they think that this may be related to a more advanced disease burden in this particular study.

Of interest to me was that right now: there is no apparent difference in progression-free survival between the 2 arms of this study. The authors described this as potentially indicating that this may be related in some way to immune modulation and immune effects on the tumor, which, if I'm completely honest, I don't totally understand. And so, the improvement in overall survival, as far as I can see at the moment, is not matched by an improvement in progression-free survival.

So I think we probably need to wait for more time to elapse to see what happens with the study. And so, I think it certainly is an interesting study, and the results are intriguing, but I think it's probably a little early for it to actually shift the treatment paradigm in this disease.

Dr. Hamilton: Fantastic. I think we've been waiting for advances in pancreatic cancer for a long time, but this, not unlike others, we learn more and then learn more we don't realize, so.

Dr. Sweetenham: Right. Let's shift gears at this point and talk about a couple of other abstracts in kind of a very different space. Let's start out with symptom management for older adults with cancer. We know that undertreated symptoms are common among the older patient population, and Abstract 11002 reported on a randomized trial that demonstrated the effects of remote monitoring for older patients with cancer in terms of kind of symptoms and so on. Can you tell us a little bit about this study and whether you think this approach will potentially improve care for older patients?

Dr. Hamilton: Yeah, I really liked this abstract. It was conducted through the Veterans Affairs, and it was based in California, which I'm telling you that because it's going to have a little bit of an implication later on. But essentially, adults that were 75 years or older who were Medicare Advantage beneficiaries were eligible to participate. Forty-three clinics in Southern California and Arizona, and patients were randomized either into a control group of usual clinic care alone, or an intervention group, which was usual care plus a lay health worker-led proactive telephone-based weekly symptom assessment, and this was for 12 months using the validated Edmonton Symptom Assessment System.

So, there was a planned enrollment of at least 200 patients in each group. They successfully met that. And this lay health worker reviewed assessments with a physician assistant, who conducted follow-up for symptoms that changed by 2 points from a prior assessment or were rated 4 or greater. So almost a triage system to figure out who needed to be reached out to and to kind of work on symptoms.

What I thought was fantastic about this was it was very representative of where it enrolled. There were actually about 50% of patients enrolled here that were Hispanic or Latinos. So some of our underserved populations and really across a wide variety of tumor types.

They found that the intervention group had 53% lower odds of emergency room use, 68% lower odds of hospital use than the control group. And when they translated this to actual total cost of care, this was a savings of about $12,000 U.S. per participant and 75% lower odds of a death in an acute care facility.

So I thought this was really interesting for a variety of reasons. One, certainly health care utilization and cost, but even more so, I think any of our patients would want to prevent hospitalizations and ER visits. Normally, that's not a fantastic experience having to feel poorly enough that you're in the emergency room or the hospital. And really showing in kind of concrete metrics that we were able to decrease this with this intervention.

In terms of sustainability and scalability, I think the question is really the workforce to do this. Obviously, you know, this is going to take dedicated employees to have the ability to reach out to these patients, etc., but I think in value-based care, there's definitely a possibility of having reimbursement and having the funds to institute a program like this. So, definitely thought-provoking, and I hope it leads to more interventions.

Dr. Sweetenham: Yeah, we've seen, over several years now, many of these studies which have looked at remote symptom monitoring and so on in this patient population, and many of them do show benefits for that in kinds of end points, not the least in this study being hospitalization and emergency room avoidance. But I think the scalability and personnel issue is a huge one, and I do wonder at some level whether we may see some AI-based platforms coming along that could actually help with this and provide interactions with these patients outside of actual real people, or at least in combination with real people.

Dr. Hamilton: Yeah, that's a fantastic point.

So let's talk a little bit about clinical trials. So eligibility assessment for oncology clinical trials, or prescreening, really relies on manual review of unstructured clinical notes. It's time-consuming, it's prone to errors, and Abstract 1508 reported on the final analysis of a randomized trial that looked at the effect of human-AI teams prescreening for clinical trial eligibility versus human-only or AI-only prescreening. So give us more good news about AI. What did the study find?

Dr. Sweetenham: Yeah, this is a really, a really interesting study. And of course, any of us who have ever been involved in clinical trials will know that accrual is always a problem. And I think most centers have attempted, and some quite successfully managed to develop prescreening programs so that patients are screened by a health care provider or health care worker prior to being seen in the clinic, and the clinical investigator will then already know whether they're going to be eligible for a trial or not. But as you've already said, it's a slow process. It's typically somewhat inefficient and requires a lot of time on the part of the health care workers to actually do this in a successful way.

And so, this was a study from Emory University where they took three models of ways in which they could assess the accuracy of the prescreening of charts for patients who are going to be considered for clinical trials. One of these was essentially the regular way of having two research coordinators physically abstract the charts. The second one was an AI platform which would extract longitudinal EHR data. And then the third one was a combination of the two. So the AI would be augmented by the research coordinator or the other way around. As a gold standard, they had three independent oncology reviewers who went through all of these charts to provide what they regarded as being the benchmark for accuracy.

In a way, it's not a surprise to me because I think that a number of other systems which have used this combination of human verification of AI-based tools, it actually ultimately concluded that the combination of the two in terms of chart accuracy was for the most part better than either one individually, either the research coordinator or the AI alone. So I'll give you just a few examples of where specifically that mattered. The human plus AI platform was more accurate in terms of tumor staging, in terms of identifying biomarker testing and biomarker results, as well as biomarker interpretation, and was also superior in terms of listing medications.

There are one or two other areas where either the AI alone was somewhat more accurate, but the significant differences were very much in favor of a combination of human + AI screening of these patient charts. So, in full disclosure, this didn't save time, but what the authors reported was that there were definite efficiency gains, and presumably this would actually become even more improved once the research coordinators were somewhat more comfortable and at home with the AI tool. So, I thought it was an interesting way of trying to enhance clinical trial accrual up front by this combination of humans and technology, and I think it's going to be interesting to see if this gets adopted at other centers in the future.

Dr. Hamilton: Yeah, I think it's really fascinating, all the different places that we can be using AI, and I love the takeaway that AI and humans together are better than either individually.

Dr. Sweetenham: Absolutely.

Thanks once again, Dr. Hamilton, for sharing your insights with us today and for all of the incredible work you did to build a robust program. And also, congratulations on what was, I think, a really remarkable ASCO this year, one of the most exciting for some time, I think. So thank you again for that.

Dr. Hamilton: Thanks so much. It was really a pleasure to work on ASCO 2025 this year.

Dr. Sweetenham: And thank you to our listeners for joining us today. You'll find links to all the abstracts we discussed today in the transcript of this episode.

Be sure to catch up on all of our coverage from the Annual Meeting. You can catch up on my daily reports that were published each day of the Annual Meeting, featuring the key science and innovations presented. And we'll have wrap-up episodes publishing in June, covering the full spectrum of malignancies from ASCO25.

If you value the insights you hear on the ASCO Daily News Podcast, please remember to rate, review, and subscribe wherever you get your podcasts.

Disclaimer:

The purpose of this podcast is to educate and to inform. This is not a substitute for professional medical care and is not intended for use in the diagnosis or treatment of individual conditions. Guests on this podcast express their own opinions, experience, and conclusions. Guest statements on the podcast do not express the opinions of ASCO. The mention of any product, service, organization, activity, or therapy should not be construed as an ASCO endorsement.