These two setbacks for the much-touted "carrotdrug" prompted the editorial board of the Journalof the National Cancer Institute to wonder intheir June issue: "Beta Carotene Didn't PreventCancer: What's Up, Doc?"
Not so quick, say Harvard researchers, whopoint out that the conditions of the two studieswere not general enough to make any conclusionsabout the drug's overall effect.
"In this sort of thing, you'll neverknow...whether it's because the stuff doesn'twork, or whether the stuff wasn't given for longenough and high enough doses," says Mathews-Roth.
Generally, diseases such as those studied inthe Hanover project go though a long promotionstage of 20 years or more, says Mathews-Roth,before the tumor erupts in full force in the body.Any test to measure a drug's effect on skincancer, she contends, whould require monitoringthe patient though both the promotion andinitiation stages, many more years thanGreenberg's study.
"It's wrong to say that beta carotene doesn'twork on skin cancer," says Mathews-Roth, whoseveral years ago found a positive correlationbetween beta carotene and skin cancer in ratswhich have been monitored from birth--indicatingthe opposite of Greenberg's study. "You can justsay it doesn't work under the conditions of thatstudy."
Greenberg's research has prompted Hennekens andhis Harvard colleagues to write a letter whichwill appear next week in The New EnglandJournal of Medicine critiquing the betacarotene study.
"You must have eight to 10 years of treatmentand follow-up," said Hennekens in a phoneinterview. And in the Journal letter, Hennekensand his colleagues take issue with the Greenbergstudy.
"Given the long latency of skin cancer, it isnot surprising that no effect of treatment...wasevident during five years of follow-up," wroteHennekens and his colleagues.
"[W]e believe that the finding of noassociation between beta carotene and skincancer...should not detract from the very stronghypothesis about beta carotene in thechemoprevention of cancer of other epithelialsites."
The Future
Hennekens, Gaziano and their colleagues canonly guess and dream about the outcome of theirstudy. They have decided to conduct the trialunder a "double-blind" standard, which means thatnot only are the subjects unaware of whether theyare taking placebos or the real thing, but so arethe investigators.
The only individuals who have access to all thedata are the members of the Data Monitoring Board,a group which has the power to decide to cut offthe study at any time if data from the studypoints strongly to one hypothesis, as they didwith aspirin two years ago.
Despite their ignorance, though, researchersare already looking past the study, anticipatingpopular reaction to the results. Hennekens fearsthat those who engage in high-riskactivities--smoking in particular--would take betacarotene instead of altering their lifestyle.
Clinical statistics show, says Hennekens, thattobacco raises the risk of cancer by 20 times,while beta carotene would only lower risk by 50percent at most. A smoker taking beta carotene,then, would face a risk 10 times higher than thenormal population.
"If you took beta carotene instead of stoppingsmoking, it would be a real shame," saysHennekens. "It should never be taken as analternative."