Our study was focussed on the treatment of patients with colon cancer who have regional lymph node involvement. We know that despite optimal surgery and adjuvant chemotherapy, which is standard of care for these patients, about 20-60% will eventually develop recurrent disease. We know that we need strategies beyond standard chemotherapy to reduce the risk of this recurrent disease. Our study was focussed on whether post-resection circulating tumour DNA status could help guide adjuvant treatment decisions for patients.
What was the study design?
The study that we used was from a phase III clinical trial called Alliance or CALGB /SWOG 80702. This was a study of patients with resected adenocarcinoma of the colon without metastatic disease but that had stage 3 disease, so regional lymph node positivity. This was a study across the United States and Canada that enrolled about 2,500 patients and randomised them in a 2x2 fashion to receive either three versus six months of adjuvant FOLFOX chemotherapy and also celecoxib or placebo for three years. Celecoxib, as you may know, is a prostaglandin PTGS2 inhibitor, similar to aspirin but with greater selectivity, and there has been significant evidence in the colon cancer literature that prostaglandin synthesis inhibition can reduce the risk of adenoma formation and reduce the risk of cancer formation in patients that have a familial predisposition to develop adenomas or cancer. We also know in prospective observational cohort studies that treatment with aspirin improves survival for patients. So this was a clinical trial formally testing our hypothesis that we’d previously seen in observational cohort studies.
What were the results to this study?
In total, we were able to analyse ctDNA from about 940 of our 2,500 patients. Remember, this was a retrospective analysis, ctDNA wasn’t baked into the study; the study was designed more than a decade ago. But we found that in that subset of 940 patients in our post-hoc analysis positivity after resection was highly predictive of a worse disease free survival as well as worse overall survival, as we expected from prior studies. That’s no surprise but it confirms that the assay is working appropriately here. But what we also found when we looked at patients stratified by celecoxib use was that patients that had received celecoxib and were ctDNA positive had a significant benefit from celecoxib whereas patients that were ctDNA negative had no benefit from celecoxib versus placebo. So this result was true both for disease free survival with a hazard ratio of 0.55 as well as for overall survival with a hazard ratio of 0.58. Those results then suggest that you might be able to use ctDNA positivity after resection to determine whether patients should receive celecoxib or not as an adjuvant therapy.
What is the clinical significance of these results?
That’s a great question. So there’s a lot of literature now, and it’s really widely agreed, that ctDNA status after colon cancer resection is predictive of survival. But what we’ve had really very little evidence for is whether it is something that can actually inform a therapy that you select for patients. That’s really what we’d like to have – something that you could test that should say patients should get more or less chemotherapy or chemo A versus chemo B or perhaps some other agent entirely.
Our study is notable because it’s one of the first to demonstrate that there is a therapeutically predictive value for ctDNA. In this particular case it’s for the use of celecoxib which is a drug that has a well-understood side effect profile, similar to aspirin although less severe in some ways, that is not only well understood but is also well managed. It’s a relatively inexpensive drug, it’s something that can be taken orally and so it potentially puts a real option on the table for patients in terms of reducing their risk of recurrence after resection and those patients that do recur very frequently do end up dying of their disease. So it’s an exciting opportunity to show that ctDNA testing can actually be used to guide therapeutic decisions for patients with stage 3 colon cancer.
Is there anything else you would like to add?
The last point I would make is that these results are really exciting to us but it’s important to keep in mind that this was a post-hoc analysis of patients that were enrolled in the 80702 trial. ctDNA was not used to stratify patients to receive one therapy or another and as exciting as the results are, they probably will need some validation in an external cohort or in an additional clinical trial before you could become fully comfortable implementing what we found here as standard of care. But it’s definitely thought provoking and suggests directions that we should go for future testing.
