Do Tumor-Naive MRD Assays Have a Role in Resected CRC? Cremolini Weighs in

Findings from an analysis of the GALAXY study indicate that circulating tumor DNA (ctDNA) may have a role to play in helping refine adjuvant therapy in colon cancer management, according to Chiara Cremolini, MD. Cremolini added that this may be particularly useful in tailoring treatment intensity based on minimal residual disease (MRD) status although the technology is not yet ready for primetime use.

At the 2024 ASCO Annual Meeting investigators presented findings from the GALAXY study in CIRCULATE-Japan (NCT04089631), which investigated the role of ctDNA in assessing MRD in patients with colon cancer following radical resection. The study utilized a novel tissue-naive ctDNA assay on a subset of participants.

Initial findings showed that the tissue-naive test had a sensitivity of 61.1% four weeks after surgery, which increased to 83.3% in subsequent longitudinal assessments. Despite the test’s high specificity of 87.9% and 89.5% at these respective timepoints, its lower sensitivity suggests that it is not yet suitable for routine use.

The study also compared this new assay with more traditional tissue-informed assays, which incorporate both genomic and methylation markers and are noted for their higher sensitivity. Cremolini emphasized the need for further clinical trials to fully understand how these tests can be integrated into clinical practice and influence treatment decisions based on ctDNA profiles.

In discussing the broader implications of these findings, Cremolini also noted the challenges posed by low-shedding tumors, which can result in false-negative ctDNA results, underscoring the importance of understanding the biological variability of colon cancer to prevent undertreatment of certain patient subgroups.

Cremolini is a medical oncologist and associate professor in medical oncology at the University of Pisa in Italy.

OncLive: Could you explain the rationale and design of the GALAXY study?

Cremolini: [The GALAXY study in CIRCULATE-Japan] is a very good example of prospective research in the field of MRD in colon cancer. The idea was to explore the detection of MRD by means of analyzing ctDNA [with] liquid biopsy among patients with radically resected colon cancer in the real world. Therefore, in this trial patients undergoing radicalization of colon cancer underwent multiple blood draws for liquid biopsy and were analyzed both immediately after the surgery and longitudinally at different time points. GALAXY provided a very high level of evidence about the prognostic impact of ctDNA in these patients.

What were some of the key findings observed from this analysis?

The primary analysis of GALAXY showed the negative prognostic impact of ctDNA independent and regardless of all the other prognostic variables. [We previously evaluated] the potential predictive impact [of ctDNA] with regard to the efficacy of adjuvant chemotherapy. [To do this we] performed a tissue-based test that analyzes each individual patient tissue sample and then searches a few selected allelic variants with a high allelic frequency. As a result [we] built a kind of tailored test to evaluate the presence of ctDNA in each individual patient based on the specific molecular characteristics of that tumor. 

At this year’s ASCO meeting we found a tissue-naive test for the detection of ctDNA that was applied to a subgroup of patients enrolled in the GALAXY trial. So not the tissue-informed technology that is able to detect ctDNA without the preliminary analysis of the tissue sample. This [tool] was applied on 80 patients who were enrolled among the 1000s of patients enrolled in the GALAXY study.

How can oncologists start incorporating these tests into clinical practice?

This tissue-naive test has a lot of potential points of strength, and the strongest point is the short turnaround time, because you do not need to collect the tissue sample and analyze the tissue sample before assessing ctDNA.

The sensitivity that has been shown in the landmark analysis [with] one liquid biopsy four weeks after surgery, was around 60%, but it was increased with the longitudinal assessment. In the longitudinal evaluation of these patients, the sensitivity of the tests is as high as 83%. [We also showed that] the specificity is very high, both in the landmark analysis and in the longitudinal assessment.

Therefore these tests are very good tools to select patients who need adjuvant chemotherapy and potentially to even intensify the adjuvant therapy in this subgroup with ctDNA since the test has a very high specificity. On the other hand since the sensitivity is not so high, especially four weeks following surgery, excluding a patient from receiving adjuvant therapy due to ctDNA negativity based on this test is not yet ready for primetime.

How might the tumor-informed Tempus Dx assay be used in clinical practice?

Nowadays, the tissue-informed tests seem more informative and more sensitive compared with the tissue-naive ones that include both methylation and genomic markers. Clinical trials will be needed to understand how a treating physician should implement these results in daily clinical practices so [that they better understand] the role of ctDNA in driving treatment choices in radically resected patients.

Another important point is that whichever technology we are using, there is [going to be] a subgroup of patients that has low shedding disease. In this case, we see low levels of ctDNA. This [can cause] false negative cases in clinical trials and clinical experiences [regardless of] all the technologies [that we have] for the detection of ctDNA. It’s not only a matter of sensitivity of the technology, but also of the biology of the disease. We need to understand these biological features of the disease in order not to undertreat a subgroup of patients.

Could you touch on a few key findings from some of the other CRC trials that stood out to you?

The most relevant news in the field of CRC is the TRANSMET trial [NCT02597348], which was a randomized trial in patients with liver-limited disease that were deemed unresectable, both at baseline and following an upfront chemotherapy regimen. In this trial, roughly 80 patients were randomized to chemotherapy alone or a liver transplantation following chemotherapy. The trial showed a really significant advantage in the liver transplantation arm in terms of five-year overall survival, which was the primary end point of the study, even more so in the per-protocol population compared with the intention-to-treat population, where the hazard ratio was 0.37. [This showed] a clear benefit from this procedure that has never been considered in the treatment armamentarium for patients with metastatic CRC.

These data are big news and something that we will need to include in our multidisciplinary evaluations for our patients with liver-limited metastatic CRC.

Looking beyond the presentations at ASCO, what research is needed to further advance CRC care globally?

An area of improvement and a hot field of research is making immunotherapy [more effective] in mismatch repair–proficient [pMMR]/microsatellite stable tumors. In these tumors the classical immunotherapy approaches do not work. We need to figure out how to select the subgroups of patients with pMMR tumors that may derive benefit from immunotherapy approaches or identify new strategies, potentially with the new drugs that are on the horizon to make that happen for our patients with pMMR, metastatic CRC.

Reference

Nakamura Y, Kristiyana K, Lo C, et al. Predicting recurrence using a tumor-uninformed ctDNA assay detecting MRD in patients with resected stage II or III colorectal cancer: subset analysis from the GALAXY study in CIRCULATE-Japan. J Clin Oncol. 2024;42(suppl 3):21. doi:10.1200/JCO.2024.42.3_suppl.21