ASPIRIN FOR COLORECTAL CANCER

SAN FRANCISCO — Evidence has long suggested that aspirin might help prevent colorectal adenomas and cancer and might reduce mortality related to colorectal cancer. Some data have also suggested that aspirin can function as "targeted therapy" for the subtype of patients with PIK3CA-mutated tumors.

However, a new study failed to find an association between aspirin and a survival benefit in colorectal cancer patients withPIK3CA mutations.

The results were presented here at the 2014 Gastrointestinal Cancers Symposium.

"Unlike previously published data, we did not confirm a relation between aspirin use and overall survival across our cohort," said study author Nishi Kothari, MD, from the H. Lee Moffitt Cancer Center & Research Institute in Tampa, Florida.

"Aspirin users likely had more comorbidities, such as heart disease and stroke, which could contribute to overall survival. We attempted to account for that by using cancer-specific survival as an end point," explained Dr. Kothari during her presentation.

However, even with this end point, the researchers found that aspirin use did not provide any improvement in survival at any cancer stage. They were also unable to replicate previous data showing an improvement in recurrence-free survival.

Stage IV Benefit Not Sustained

Dr. Kothari and colleagues conducted a retrospective chart review of patients at the Moffitt Cancer Center and at the Royal Melbourne Hospital in Australia. From 1996 to 2009, 1019 patients were diagnosed with colorectal cancer.

Of the 185 patients who harbored a PIK3CA mutation, 26% were aspirin users.

The median age of aspirin users was slightly older, suggesting that aspirin was being regularly prescribed to reduce the incidence of cardiovascular morbidities.

Dr. Kothari explained that because of the difference between their work and published literature, the researchers examined the validity of their study population.

Patients with the PIK3CA mutation — both aspirin users and nonusers — had a higher degree of advanced cancer than patients without the mutation; only 4% of those patients had stage I disease.

Not surprisingly, "increasing age was associated with poorer overall survival, and advancing stage was associated with worse overall and cancer-specific survival," Dr. Kothari reported.

Because the distribution of disease stage was different in the study population than in previously published work, the researchers stratified their results by stage. However, when they looked at the risk for recurrence in patients with stage II or stage III disease, they found that regular aspirin use did not improve overall, cancer-specific, or recurrence-free survival.

When they looked exclusively at the 47 patients with metastatic disease, there was a trend toward an improvement of overall survival (P = .06); however, this was not sustained on multivariate analysis.

Reasons for Differences?

"Of note, we did find significantly worse survival with right-sided cancers in the stage IV population," Dr. Kothari said. "The improved survival seen initially in our stage IV univariate analysis might have been due to a decreased incidence of right-sided cancers," she explained.

It could also be that aspirin use was confounded by the decreased incidence of right-sided cancer in the stage IV patients.

"The make-up of our population was distinct from previously published cohorts, which could have led to our different findings," said Dr. Kothari. "Our patients were unselected and largely from community centers. They also tended to have more advanced-stage disease, which could indicate different tumor biology."

Because this study was retrospective, there was no information on duration of aspirin use or compliance, and follow-up was shorter than in previous studies.

Need for Prospective Study

So where do these data fit with the other literature?

The fact that there was no clear survival benefit with aspirin use seemingly refutes the results of previous studies, even though there was a trend toward a benefit in advanced-stage disease, noted study discussant Neal Meropol, MD, professor of medicine at the Case Western Reserve University School of Medicine in Cleveland.

"However, all of the datasets we have seen so far have challenges," he noted.

"There is no randomization between aspirin and no aspirin, there is reliance on patient reporting of aspirin use, and there is variation in aspirin dosing between patients," Dr. Meropol pointed out. "There is also a mix of disease stages and small sample sizes limit interpretation of all of these studies."

Finally, a problem in all retrospective reporting is that there are limits to the information available, and there are potential unmeasured variables that could affect the relation between aspirin and mutation status, such as the other treatments patients have been receiving.

"A prospective evaluation of aspirin and COX-2 inhibition in the PIK3CA-mutant population is needed," he concluded. "Ongoing planned and randomized trials of aspirin and celecoxib [Celecoxib] in the adjuvant setting will certainly provide additional meaningful data."

The authors have disclosed no relevant financial relationships.

2014 Gastrointestinal Cancers Symposium (GICS): Abstract 386. Presented January 18, 2014.