December 07, 2010
Post Date - 7 - December - 2010
Daily aspirin use appeared to lower the risk of death from cancer in randomized trial participants, a meta-analysis found.
Cancer death rates were 21% lower in study participants assigned to aspirin after five years (95% CI 8% to 32%), and the reduction persisted among those followed for 20 years, according to pooled data on more than 25,000 individuals in eight randomized studies.
The meta-analysis, led by Peter M. Rothwell, FMedSci, of John Radcliffe Hospital in Oxford, England, was published online in The Lancet.
The researchers called the results "the first reliable evidence that aspirin prevents noncolorectal cancer in humans," which had been suggested for a number of solid-tumor cancers on the basis of earlier studies.
Rothwell and colleagues also suggested that the findings should alter calculations of risks and benefits of daily aspirin in individual patients.
"Our analyses show that taking aspirin daily for five to 10 years would reduce all-cause mortality (including any fatal bleeds) during that time by about 10%," the researchers wrote.
But they also cautioned that the findings, by themselves, do not prove that aspirin prevents cancer or even cancer death.
They noted that before either conclusion would be fully justified, additional studies would be needed to:
- Determine what effect, if any, daily aspirin has on cancer incidence
- Analyze the effect on gynecologic cancers, which were not adequately represented in the meta-analysis
- Examine aspirin effects with follow-up longer than 20 years, in case there is a late rebound effect
- Look for effects occurring in the first five years that might have escaped detection in the current analysis
The American Cancer Society agreed that it would be premature to put everyone on daily aspirin.
"Even low-dose aspirin can substantially increase the risk of serious gastrointestinal bleeding," warned Eric Jacobs, PhD, an ACS official, in a statement.
"It will take time for expert guidelines committees to carefully consider these new results together with the existing evidence and consider for whom the benefits of aspirin use are likely to outweigh the risks."
On the other hand, several physicians contacted by MedPage Today and ABC News indicated that daily aspirin could be considered for at least some patients as chemoprevention.
Stefan Glück, MD, PhD, associate chief of hematology/oncology at the University of Miami said in an e-mail that he had already been recommending daily aspirin to patients.
He said the meta-analysis findings "only confirm what has been known for a while -- the nature of the study just makes the statement stronger."
Randy Wexler, MD, a family physician at Ohio State University in Columbus, commented that a general recommendation would not be warranted yet, "but in patients who do have risk factors (such as family history) this would be something to discuss and consider."
On the other hand, an oncologist at Memorial Sloan-Kettering Cancer Center in New York City toldMedPage Today and ABC News that the studies included in the meta-analysis weren't designed to address cancer prevention, and hence the findings need confirmation.
"What is badly needed is a prospective, disease-specific study that shows with placebo-control that either incidence, recurrence, or death are significantly reduced," said Clifford Hudis, MD, chief of breast medicine at MSKCC, in an e-mail.
He noted that the selection of participants for these trials may not be relevant since they were selected based on cardiovascular risk and "the primary goal of the included studies was to reduce that risk. ... This clearly deserves a prospective study with cancer as the endpoint."
Rothwell and colleagues had sought data from randomized trials with at least four years of treatment with daily aspirin versus placebo and a total study duration of five years or more.
They found eight such trials, with a total of 25,570 participants. Two of the trials assessed aspirin for primary prevention of vascular disease, five recruited patients with vascular risk factors such as diabetes or angina but no history of cardiovascular events, and one was for secondary prevention of events.
Aspirin doses in these studies ranged from 75 to 500 mg daily.
One of the trials showed a significantly reduced rate of cancer deaths by itself -- the UK Transient Ischemic Attack study, which assigned patients to aspirin or placebo in a 2:1 ratio. It found an odds ratio for cancer death of 0.45 (95% CI 0.25 to 0.82) in the aspirin group.
Six of the other trials each showed nonsignificant trends toward lower cancer death rates with aspirin (OR 0.53 to 0.83), and one trial (Early Treatment of Diabetic Retinopathy Study) showed a slight increase in cancer deaths in its aspirin arm.
Pooling the 674 cancer deaths during the study phases of all eight trials yielded an odds ratio of 0.79 (95% CI 0.68 to 0.92) for cancer deaths in the aspirin groups.
When the researchers looked only at data for the first five years of follow-up, the protective effect of aspirin was not statistically significant for all cancer deaths or for deaths from any particular type or class of cancers.
Another notable finding was that the preventive effect was evident mainly in deaths from gastrointestinal cancers, with an odds ratio of 0.46 (95% CI 0.27 to 0.77), compared with 0.76 (95% CI 0.54 to 1.08) for non-GI solid tumor cancers.
The effect was apparent for most types of GI cancers, not just colorectal cancer, for which aspirin had already been shown to have a chemopreventive effect.
In the meta-analysis, lower death rates were seen with aspirin for esophageal and pancreatic tumors as well as for colorectal cancer.
For participants in three of the trials, all of which took place in Great Britain, Rothwell and colleagues were able to count cancer deaths with 20 years of follow-up using public registry data on participants.
In these trials, which had a total of 12,659 participants, there were 1,634 cancer deaths. The hazard ratio for death from solid-tumor cancers associated with assignment to aspirin in the studies was 0.80 (95% CI 0.72 to 0.88) relative to placebo assignment.
There was less difference by cancer type in the strength of the effect in these data. For GI cancer deaths, the 20-year hazard ratio was 0.65 compared with 0.79 for non-GI solid tumors and 1.09 for hematologic cancers.
Among individual non-GI cancer types, only lung cancer deaths were significantly lower in the aspirin groups (HR 0.71 in the 20-year data, 95% CI, 0.58 to 0.89). Deaths from prostate tumors, bladder and kidney cancers, and all other non-GI solid tumors combined were somewhat lower with aspirin but the relationships were not statistically significant.
Rothwell and colleagues also found that deaths from non-adenocarcinoma tumors were not significantly reduced with aspirin (HR 0.87 versus 0.66 for adenocarcinoma deaths).
The researchers didn't address in detail the mechanism by which aspirin may reduce cancer deaths, but they did suggest it probably is not related to platelet aggregation -- other antiplatelet drugs have not reduced deaths from cancer in randomized trials, they wrote.
Instead, they suggested, the mechanism may involve inhibition of cyclooxygenase-2, with "other pro-apoptotic effects early in the development of tumours perhaps also being important."
Posted thanks to kind permission from our friends at MedPageToday.com.