In the context of clinical trial data and personal experience, experts weigh the value of circulating tumor DNA as a tool in cancer care.
John H. Strickler, MD: Dr Dietrich, what are, in your experience, some of the pros and cons of the use of circulating-tumor DNA [ctDNA] in the clinic?
Martin Dietrich, MD, PhD: Well, it depends on the setting. I think, again, for descriptive characterization, I think there’s no hesitation anymore. I fully agree with Dr Bardia’s sentiments about the basically ubiquitous use in multiple settings, not only to afford baseline characterization but really for tracking developing mutations that may guide therapy in different directions. What challenges we’re facing with the use of MRD [minimal residual disease] monitoring is currently that we’re really lacking confidence in the prospective nature of that data into the interventional predictive nature of the data. I do think that, for example, in the different settings, we could see potential benefits, the hopes would go into avoiding or reducing at least the number of imaging methodologies, reducing the exposure to contrast and radiation, and reducing cost, hopefully, in the end. One of my concerns is—and I think we’ve mentioned this in a similar setting—to see other tumor markers, that technology that may even be more sensitive, although there’s some contradictive data with more concern about creating anxiety without necessarily having the opportunity to intervene earlier. We may create a time bias for intervention when we’re seeing the rise of ctDNA. We’re getting faced with these tests that are now ubiquitously ordered in the community, and then the question is what to do with these tests. When I see a lung cancer turning positive or a breast cancer turning positive in the adjuvant setting, where we typically don’t even recommend proactive monitoring beyond the clinical, local regional assessments, what do I do with this information? The concern is that the disease is going to come back, but I’m not sure that I would be able to intervene earlier or cure this earlier. Then one could say, “Well, if I see a positive ctDNA test, I might be able to increase scan frequency.” Maybe detect the recurrence earlier, but one of the reasons why we’ve never seen proactive monitoring in the guidelines is because we really haven’t shown that we were able to improve outcomes with proactive monitoring, and therefore, we really just tried to avoid it and monitor it clinically. In the metastatic setting, I think the pros and cons would certainly be a little bit different. We have a much more acute setting. I think we would be able to see, certainly, there’s no doubt in my mind that these titers at least and serial time points are helping us decide about disease space. For example, if you have a bone-only breast cancer where imaging is very difficult to interpret, I think they could very much complement, especially the marker-negative breast cancers. This might be a very helpful setting. We’re not necessarily lacking the confidence in the assay here, but like I said, how to integrate this I think is still heavily provider dependent.
I hope that with additional data we’ll be able to simplify monitoring and hopefully by detecting things earlier and intervening more aggressively, although possibly improving outcomes but again, this is a challenge. Another one that I have is the utility for the MRD monitoring is for difficult imaging findings. Whether it is a lung mass that we’re trying to evaluate or if there is a peritoneal distribution or in a diffuse carpeting kind of representation. I think that certainly a high level of sensitivity and the presence of ctDNA could certainly be utilized as a way of supporting a more aggressive approach. Otherwise, you may need to use interventional methodologies to get tissue.
I think there’s plenty of fantasy in that technology where we can use it in multiple different ways and I think it’s being used in multiple different ways. Part of the reason why we have such a heterogeneity in application is because, at this point at least, we’re still lacking clear guidance on how often, how useful and obviously how predictive these tests are. We had a very similar controversy 20 years ago in breast cancer with circulating tumor cells and they turned out to be a reasonably good prognostic marker but unfortunately never really changed the way we were guiding therapy. They didn’t help in therapy selection and so like I said for me the assays mature, the different applications still must prove their worth in their respective settings. Though, like I said, I’m excited about what it can provide.
John H. Strickler, MD: Wonderful, and Dr Bardia do you see ctDNA replacing other monitoring strategies? Or maybe just augmenting them?
Aditya Bardia, MD, MPH: Both. I think in metastatic breast cancer it is something that is being used to augment scans, as was mentioned by Dr Dietrich. Sometimes we have bone only metastases and it’s tough to know whether there’s pseudo-progression vs true progression or if it’s just tumor flare. We can use tumor markers, and we do use tumor markers in this setting, but sometimes tumor markers are normal. There have been a couple of studies that have looked at ctDNA in this situation and very similar to tumor markers if you’re seeing a rise in ctDNA that is suggestive of disease progression. On the other hand, if there’s a decline this could just be a tumor flare. So that’s valuable in that setting and that’s why I like to get circulating DNA at a baseline because if you have a baseline, then you can look at change and that could be valuable. In this setting, it’s more like a tumor marker but just a more sensitive tumor marker. In terms of where it could replace, it could be in the MRD setting. We are not there yet but if we do demonstrate clinical utility of monitoring ctDNA in the adjuvant setting, I think that would be a game changer and there would be early adoption of that.
Transcript edited for clarity.