Circulating Tumor DNA May Help Identify Early Responses in CRC

Pashtoon Murtaza Kasi, MD, MS

Recent evaluations have suggested that due to its shorter half-life compared with other tumor biomarkers, circulating tumor DNA (ctDNA) can be an effective biomarker that allows for assessments of early response.¹

Pashtoon Murtaza Kasi, MD, MS, and fellow investigators presented data at the 2023 Gastrointestinal Cancers Symposium which demonstrated that patients (N = 185) with advanced colorectal cancer (CRC) who were molecular non-responders had 5.9 times the hazard ratio (HR) for death and twice the HR for a time to next treatment (TTNT) event compared with molecular responders.

“[The data gives] this cohort of analysis immense power to answer some very meaningful questions that may or may not be possible through a trial or may take years for a prospective trial and in terms of time and funding to be able to be completed,” Kasi, director for Colorectal Cancer Research at Weill Cornell Medicine in New York, New York, said in an interview with Oncology Nursing News®. “Across a range of cutoffs, whether it's 10%, 20%, 30%, all the way up to 80%, [a] pattern of benefit in terms of progression free survival or overall survival [OS] was seen [in responders] across the board.”

The median OS was not reached (NR; 95% CI, 26.3-NR; CPH HR, 0.17; 95% CI, 0.08-0.36; P < .005 in the chemotherapy cohort of responders (n = 84)and was also NR (95% CI, 23.7-NR; CPH HR, 0.43; 95% CI, 0.24-0.76; P < .005) in the cohort of responders receiving all regimens (n = 109). Median TTNT was 10.3 months (95% CI, 7.3-NR; CPH HR, 0.51; 95% CI, 0.28-0.92; P = .026) vs 10.1 months (95% CI, 8.2-16.1; CPH HR, 0.48; 95% CI, 0.30-0.76; P < .005), respectively.

“These [data] were even more striking than anticipated in terms of how powerful these tools can be not just for the next treatment, because [one] could argue we’re going find out about that in the next scan in 2 months or 3 months anyways, but sometimes having that early insight into who is responding and who is not responding is something that is of value,” Kasi said.

However, non-responders in the chemotherapy cohort (n = 51) experienced a median OS of 11.8 months (95% CI, 8.7-14.6; CPH HR, 5.92; 95% CI, 2.78-12.63; P < .005) while non-responders in the all regimens cohort (n = 81) experienced a median OS of 17.8 months (95% CI,10.5-23.4; CPH HR, 2.43; 95% CI, 1.31-4.18; P < .005). Median TTNT was 5.8 months (95% CI, 4.0-7.5; CPH HR, 1.97; 95% CI, 1.08-3.58; P = .026) and 6.1 months (95% CI, 4.5-7.6; CPH HR, 2.09; 95% CI, 1.31-3.34; P < .005), respectively.

“ctDNA is here to stay,” Kasi said. “Regardless of the platform, regardless of the question, this is something that is having an impact at all phases of a journey of a patient with cancer. You could argue where you have multiple options, that could one escalate or deescalate therapy based on these findings those might be future directions in which some other trials and NSCLC or other cancers are already doing…The principles or results we find here are applicable to other tumor types as well…While we focus on CRC, this is very much applicable to the next patient as long as we have shedding a baseline.”

Kasi noted that ctDNA has moved from an exploratory marker to an integral marker and that depending on the effective therapy or surgery, the cancer DNA will fall in hours to days, no longer than a few weeks, and the change can be seen over time with benefit being shown if a patient clears their ctDNA or has a decline in levels.

“One key aspect to realize which is both a strength and a limitation to keep in mind in these data analysis, is we're talking about outcomes based on aggregated commercial payer health claims data,” Kasi said.

For their analysis, investigators pulled data from the Guardant INFORM database, which includes the claims and de-identified records of more than 220,000 patients who underwent ctDNA testing with Guardant360 from September 2018 to March 2022. Testing began 15 weeks before treatment and another test was performed 3 to 15 weeks after starting treatment.

Kasi explained that registries can provide clear information on patient mortality but looking at health care claims to evaluate treatments is more guesswork; when seeing how long a patient is on a treatment it can be assumed they’re deriving benefit if they remain on it, but reasons for switching may not be as clear —a new bill may signify a different treatment was started, but reasons why, toxicities, maximum response reached or surgery are some possibilities as well. Because of this, Kasi said “we often look at TTNT or TTD, time to treatment discontinuation, as surrogates for arbitrarily progression-free survival.”

“In the post-operative setting, our group looked at the question of if somebody just had surgery, when should those patients be tested,” Kasi said. “[This is] because the concern is immediate post-op; you may have a lot of background noise [that] may muffle the ctDNA so you're simplistically looking for the needle in the haystack.”

Kasi explained that there is a new setting he has called “adjuvant plus” where a patient has detectable ctDNA after receiving adjuvant therapy. The ctDNA allows patients who were not cured from treatment to be identified. However, patients who are not metastatic yet may not have the cancer appear on the scans.

In the study, 58% of patients were classified as molecular responders and 42% were classified as non-responders; 65% received chemotherapy with or without a VEGF inhibitor, 21% received a regimen with anti-EGFR monoclonal antibodies, and 14% received other therapies. For the responders, 16% cleared ctDNA on treatment and 15% were ctDNA low. The trial defined molecular responders as patients with a 50% decrease or greater in mean VAF between pre-treatment and on-treatment ctDNA or patients who were ctDNA low.

“What we're showing here is [ctDNA] clearance,” Kasi said. “The FDA and other regulatory agencies are accepting this as an end point or a co-end point to consider, which means we could get results faster with these novel trial designs [and] that looking at multiple arms at the same time could be a way to expedite drug development.”

Reference

Kasi PM, Weipert C, Nguyen T, et al. Use of circulating tumor DNA (ctDNA) for early assessment of treatment response in patients with advanced colorectal cancer (aCRC): A real-world (RW) analysis. J Clin Oncol. 2023;41(suppl 4):246. doi:10.1200/JCO.2023.41.3_suppl.246