The National Institutes of Health is in the midst of a $200 million trial to determine if routine CT scans of smokers' and former smokers' chests can identify lung cancer in its earliest stages and save lives through early intervention. The Wall Street Journal reports this morning (subscription required) that two of the researchers involved in the study recently served as paid experts for tobacco companies in civil suits brought by patients who want the tobacco industry to pay for the screenings.
The issue was raised by the Washington-based Lung Cancer Alliance, which supports screening. This patient advocacy group, the story notes, has received $100,000 from General Electric, which would profit from a massive expansion in cancer CT scans.
The House Energy and Commerce Committee has launched an investigation into the conflicts of interest, and whether the researchers who designed the NIH study tilted its design in ways that would discourage mass screening. There are about 90 million smokers and ex-smokers in the U.S.
Like all cancer screening tests, lung cancer screening would inevitably turn up thousands -- perhaps millions -- of suspicious-looking abnormalities. Biopsies would get done. Surgeons would operate. Drugs would get taken. Costs would skyrocket. The cancer rate, obviously, would rise, since more cancers would be found earlier. And the survival rate -- defined as persons living at least five years with the disease -- would probably improve.
But would it have done anything for overall lung cancer mortality? Perhaps not. There is no evidence that earlier intervention for this disease saves lives, and living longer with the knowledge that you're going to die from lung cancer -- even if it goes over the arbitrary five-year time frame -- isn't much of a benefit.
The two researchers with ties to Big Tobacco are University of California Los Angeles radiologist Denise Aberle, who testified in a 2003 case in Louisiana on behalf of American Tobacco Co., and Dartmouth College radiologist William Black, who testified in a New York case for Philip Morris. Aberle, one of the national leaders of the NIH trial, said UCLA got the money from her testimony and Black subsequently ended his relationship with tobacco companies.
NIH defended using scientists with previous ties to the tobacco industry:
In a written response to the Lung Cancer Alliance, the NCI said the expert witness work was appropriate. "Service as an expert witness, presenting independent analyses based on published medical literature, is a commonly accepted activity for physicians, researchers, and other experts and in the instance of the specific circumstances described did not violate the required disclosure guidelines of the organizations involved," NCI director John Niederhuber wrote.
Moreover,
In an interview, an NCI spokesman said the institute has no way of knowing whether any other investigators in the 30-center study have financial ties to tobacco or scanning companies because it does not examine such potential conflicts.
So there you have it. Rival industrial interests are funding researchers and patient advocacy groups who square off over the utility of a wildly expensive technology of questionable benefit, and NIH turns a blind eye to how conflicts of interest might undermine its much-needed objective study of the issue. This is a classic example of everything that ails the health care technology assessment field, made more graphic by the presence of the tobacco industry, everybody's favorite whipping boy and for good reason.
The steps forward are clear. Congress should create a new institute to test new technologies and compare them to existing technologies, a move that has been endorsed by leaders from both political parties. The new agency must scrupulously avoid all conflicts of interest, with industrial interests kept as far as possible away from the decisions of what to study, how to design the trials, and how to interpret the results. And, the researchers who conduct the studies must follow a simple rule: You can't have had any financial conflict of interest within the past five years to participate in such a study.
Meanwhile, NIH must end its laissez-faire attitude toward monitoring the conflicts of interest of the extramural researchers based at the nation's universities, who absorb 80 percent of its annual $30 billion budget. Universities have proven that they're aren't up to the task of monitoring their professors. Indeed, the commercialization imperative that now dominates thinking at most major universities (they earn revenue from licensing the patents that come from NIH-funded research) has created a structural conflict of interest that has blinded them to improprieties like what just took place at UCLA in the CT study.
NIH catalogues every extramural research grant in a database that is open and available to the public (it's called the CRISP database for those who want to google it). One of the fields in that database should include a complete disclosure by the grantee of all industry affiliations and financial conflicts of interests over the past five years. That will make it much harder to hide the kind of conflicts of interest that now cloud this $200 million cancer screening trial.
The results of the trial are due in 2009. Should it show that screening is not worth the candle, the trial will undoubtedly be seen as fatally flawed by patient advocates, who have their own conflicts of interest.
NIH insists the trial is well-designed. But when it comes to conflicts of interest, appearances matter. The expenditure of $200 million in taxpayer money should have resulted in a definitive answer. Now all we'll get is endless bickering over the meaning of the results by self-interested parties whose first line of attack will be the financial ties of the researchers or advocates on the other side.
NIH has once again been caught with its pants down on the conflict-of-interest issue. It should move quickly to change its rules. Besides disclosure, it should not allow researchers to participate in the major technology assessment trials that the agency funds if those researchers have relevant conflicts of interest.
And the agency should set up an independent committee to review whether all investigator-driven trials funded by the federal government should have similar restrictions. The argument against this radical step is that it will divide the world of research into two camps: those who take money from industry and those who don't. Maybe that isn't such a bad idea.