Arvind Dasari, MD, MD Anderson Cancer Center, Houston, Texas, discusses study results which assessed the clinical utility of adding ctDNA monitoring to standard colorectal cancer treatment to better identify patients eligible for curative-intent metastasis-directed therapy. Results highlighted the value of this addition to the management of this patient population. 

These data were presented at ESMO Gastrointestinal Cancers Congress in Barcelona, Spain. 

Transcript: 

Hello, I am Arvind Dasari, I'm a professor in the department of GI medical oncology at MD Anderson Cancer Center. Today I'll go over some data that were presented by our group at ESMO GI earlier this year. 

This was a study looking at the clinical utility of including circulating tumor DNA, or ctDNA for short, monitoring in standard-of-care colorectal cancer surveillance. As background, the current standard-of-care surveillance for colorectal cancer includes serial visits with history and physical, [carcinoembryonic antigen] CEA, colonoscopy, and cross-sectional imaging for patients who have had a curative-intent treatment. However, multiple studies and meta-analyses have shown mixed results for such an approach in improving survival and also have questioned the utility of such surveillance. ctDNA testing in contrast, has shown superior prognostic and predictive power in detecting recurrences compared to CEA. 

In the current study, the objectives were during surveillance, in ctDNA-positive patients, [to] identify the patterns of recurrences in these patients and compare a curative-intent metastasis-directed treatment in patients who are ctDNA-positive versus in those who have an elevated CEA. The hypothesis here with this objective was that metastasis-directed treatment was a surrogate for identifying patients with oligometastatic disease, that is patients who in spite of a recurrence, could still be potentially cured of their disease. The second objective was during surveillance, in ctDNA-negative patients, to quantify conditional risk, that is the risk of recurrence preconditioned on the fact that patient is ctDNA-negative after a certain period of time and also look at the temporal patterns of recurrence in these patients with ctDNA-negative results. For the dataset, we used data from 2 large prospective observational trials, the Japanese GALAXY trial and the US BESPOKE trial. Both these trials enrolled patients with colorectal cancer who were monitored serially with tissue informed assay Signatera^TM. Cumulatively, we had over 3100 patients and over 15000 time points. The median follow-up in these trials was 24 to 30 months. 

Looking at the key takeaways from the analysis of the data, number one was that in patients who had transient clearance of ctDNA, that is patients who were positive for MRD after surgery and whose ctDNA cleared with adjuvant chemotherapy, but then have a recurrence of their MRD, that is they had detectable ctDNA again after a period of time, it happens very quickly. Over 70% by 1 year and over 90% by 18 months and practically all patients by 24 months, so that's key takeaway number one– in patients who are going to have a recurrence of their MRD, it happens by the 18-to-24-month mark.  

Key takeaway number two, we also noted that ctDNA is a much better predictor for MDT or metastasis-directed therapy and thus by an extension in detecting oligometastatic recurrence as compared to CEA. For instance, looking at the GALAXY trial data and looking at the rates in ctDNA-positive versus -negative patients and looking at the rates in CEA-elevated versus non-elevated patients and breaking these data into stage 2/3 versus stage 4, what we see is that the MDT rates in ctDNA-positive versus -negative patients in stage 2/3 was 32% versus 2%, and in stage 4 was at 38% versus 9%. However, in contrast, when you look at the similar rates according to CEA, it was 12% versus 5% in stage 2/3 patients, and in fact, there was no difference in stage 4 patients at approximately 20% each in either group. We confirmed these observations using the GALAXY data and found that ctDNA has very significantly higher sensitivity and specificity for MDT in patients as compared to CEA, and this difference was statistically significant both for stage 2/3 as well as stage 4 patients. 

So far we've talked about patients who are ctDNA-positive. Now pivoting to patients who are ctDNA-negative, the next takeaway is that the rates of clinical recurrence in patients who are persistently ctDNA-negative are very low. For instance, patients who are ctDNA-negative after surgery and continue to be ctDNA-negative are MRD-negative by the 18-to-24-month mark [and] their risk of recurrence is minimal to none, telling us that the conditional risk of recurrence if patients are persistently ctDNA-negative by the 18-to-24-month mark is very low and perhaps in these patients we could de-escalate or even stop surveillance safely. This is a very important observation because what we know already from historical data is that patients who undergo surgery with curative intent, in fact, we know that 50% of them, if they're stage 3, for instance, will never have a risk of recurrence but we still do adjuvant therapy and we still do surveillance on all of those patients because we cannot risk stratify these patients in terms of their risk of recurrence and ctDNA really provides that tool, so hopefully we can build on this further. 

To conclude and to recap what we've seen, our data indicates that adding ctDNA testing to the current standard of care in surveillance can better identify patients who are candidates for metastasis-directed therapy or who on extension have oligometastatic disease. We've also seen that ctDNA is significantly better at doing this as compared to CEA. On the flip side, we've also seen that patients who are ctDNA-negative, and who are ctDNA-negative at the 18-to-24-month mark, have a very low risk of recurrence. Collectively, all these observations really highlight the need to integrate ctDNA into surveillance and tailor surveillance based on the serial ctDNA results. The hope is that prospective data, both from observational and prospective trials, will confirm and consolidate these observations. Thank you very much.

Source: 

Dasari A, Nakamura Y, Sorscher S, et al. Clinical utility of including circulating tumor DNA (ctDNA) monitoring in standard of care (SoC) colorectal cancer (CRC) surveillance. Presented at ESMO Gastrointestinal Cancers Congress. July 2-5, 2025. Abstract 2O