NEW YORK – New data presented at the American Society of Clinical Oncology Gastrointestinal Cancers conference on Saturday suggests that drugs that have failed in studies of all-comers can show efficacy in patients with minimal or molecular residual disease (MRD) as detected by circulating tumor DNA testing.
The data, presented in an oral session by Jonathan Nowak, a molecular and gastrointestinal pathologist at Brigham and Women's Hospital and the Dana-Farber Cancer Institute, comes from a retrospective reanalysis of samples from a study called CALGB/SWOG 80702 that tested the addition of celecoxib, an NSAID drug, to adjuvant chemotherapy versus chemotherapy alone in patients with stage III colon cancer.
In that initial trial, adding celecoxib did not significantly improve disease-free survival, but the investigators hypothesized that the presence of ctDNA might be able to isolate a subset of patients with a significant benefit.
The researchers used Natera's Signatera testing, which uses upfront tumor tissue sequencing to design a personalized panel of up to 16 targets to test a patient's blood for the presence of tumor-derived DNA fragments.
Samples from a little over 1,000 patients who participated in the original trial could be tested, 189 of whom were ctDNA-positive. Among these, the addition of the celecoxib significantly improved both disease-free and overall survival compared to chemotherapy alone, which it did not in ctDNA-negative patients. The hazard ratios the team calculated indicated that about a fifth of MRD-positive patients will see a benefit from treatment with celecoxib.
Natera Chief Medical Officer Alexey Aleshin said in an interview that although some oncologists have prescribed celecoxib to their adjuvant patients despite its failure in the CALGB/SWOG trial, it is more the exception than the rule.
"There hasn't been an ideal, or promising adjuvant readout in CRC for over 20 years. So, I think the field has really been looking for what could be next, and we think these results are a step in the right direction," he said. "This was one of the largest randomized biobanks [of its kind], and it's a well-designed study. I think the signal really speaks for itself … and we do hope that the clinical community will consider these results as they update their treatment recommendations."
Another heartening aspect of the results is that celecoxib adds little toxicity compared to chemo-escalation strategies that have also been explored in adjuvant colorectal cancer treatment.
"Traditionally, all we've been able to do is add more and more and more chemotherapy. So, this is really one of the first times we're adding something that does not necessarily markedly increase the toxicity for the patient," Aleshin said.
As a company, Natera hopes that this study can serve as a template for both analysis of new emerging treatments in the adjuvant setting and for additional retrospective reviews of past failed trials.
"If we can find such a strong signal with celecoxib, it stands [to reason] that we could go back and analyze other past studies that didn't show a signal in all-comers and maybe replicate that," he said, adding that the results also suggest that MRD testing should perhaps be integrated into adjuvant studies as a rule moving forward.
"This kind of gives clinicians and patients a second shot on goal, where if a study may not show a signal on all-comers, it could still show a very strong signal in ctDNA-positive patients," Aleshin said.
There are likely a dozen or more studies that are prospectively enrolling patients using ctDNA test results to randomize them to different arms, he said. A highly anticipated example is the IMVIGOR 011 trial, which aims to prove that patients who are MRD-positive benefit from added immunotherapy after surgical removal of their bladder cancer.
The prior IMVIGOR 010 study failed to show a benefit from immunotherapy in a cohort that was not defined by MRD status.
The same or even a greater number of retrospective studies, like Nowak and colleagues' reanalysis of CALGB/SWOG 80702, are also in the works, Aleshin said.
The ASCO GI meeting was awash in data on MRD testing, likely because colorectal cancer has been the first foothold for these tests, both in research and in clinical and commercial implementation.
Natera and others shared new data at the meeting from studies validating their respective technologies as prognosticators and predictors of chemotherapy benefit in various tumor stages and settings.
Guardant Health, for example, shared data on the efficacy of its tumor-agnostic test in detecting early-stage rectal cancer patients who need more aggressive surgery based on lingering ctDNA after a more conservative resection.
Also, researchers working with Exact Sciences' MRD assay presented data on the prognostic value of Exact's test in stage III colorectal cancer, showing that the Oncodetect assay, when used longitudinally, was predictive of recurrence across time points, with up to 91 percent sensitivity and 94 percent specificity during surveillance monitoring. Exact's platform, like Natera's, is tumor-informed but can build much larger ctDNA panels with up to 200 targets.
Natera also provided updated data on its own prospective study, BESPOKE-CRC, which showed that a single postsurgical test could predict inferior outcomes in stage II and III CRC patients. For MRD-negative patients, about 92 percent were disease-free after 24 months compared to only about 41 percent of MRD-positive subjects. Signatera-positive patients also benefited from adjuvant treatment, while Signatera-negative patients did not.
Advocates of MRD testing are still debating the ideal, most efficient, and most economic approach for clinical implementation of such tests. Questions involve the relative sensitivities of tumor-informed versus tumor-naïve tests, as well as single time point tests versus longitudinal monitoring.
Although Natera has long stressed the sensitivity benefits of its tumor-informed approach, the company recently announced that it has also developed a tissue-free assay that, akin to others in the field like Guardant Health's test, relies on epigenetic patterns rather than predetermined mutations to discern cancer-associated DNA molecules in the blood.
"We really want to be able to offer emergency testing for all patients and all physicians. And I know some physicians may prefer tissue-free sometimes because tissue can be inadequate, or not available, or take a long time to receive," Aleshin said. "This is something that opens up testing to more patients."
In data shared at the ASCO GI meeting, investigators showed that the new blood-only test was highly concordant with Signatera results, and that it achieved high specificity, something Aleshin said has been a challenge for some other assays.
"We were very excited to see that the main kind of prognostic and predictive claims replicated — that MRD-positive patients did markedly worse than those that are negative, and that if somebody is MRD-positive, they appear to benefit from adjuvant chemotherapy compared to those who are negative," he said.