From the Web Editor’s Desk
Dr. Fred Grannis, Associate Professor and Staff Surgeon Thoracic Surgery, City of Hope National Medical Center in Duarte, California weighs in on “Politics, Profits and Public Health: The C Scanning Controversy”. We are posting Dr. Grannis’ letter in full and hope that it encourages open discourse on this burning issue.

Kamlesh Asotra, Ph.D.

More Fuel for a Burning Issue
September 21, 2009

Although the article on lung cancer screening by M.F. Bowen in the August issue of the TRDRP Burning Issues newsletter is far more nuanced and balanced than most of what has been published on the subject in the media to date, I believe that your readers should be provided with further information that will provide additional balance to this controversy.

First, there are circumstances when a randomized controlled trial is inappropriate. We cannot, for example, prove that asbestos causes mesothelioma with prospective randomized trial data, because it would be unethical to give a toxic substance, like asbestos, to research subjects. Yet, without RCT evidence, all of medicine understands and agrees the asbestos causes mesothelioma. The same is true for causation of lung cancer by tobacco products. We have an obligation to make clinical guideline decisions based on the best-available evidence even when RCT data is unavailable.

Although the NLST is a prospective randomized study, there is serious question whether it is an appropriate or even an ethically sound RCT. I believe that it is not. Spokespersons from the National Cancer Society have repeatedly, over two decades, stated that chest roentgenographic lung cancer screening is ineffective and possibly dangerous. The real research and public health question is not whether computerized tomograms are better than chest x-rays, but whether CT screening is better than no screening. The NCI’s National Lung Screen Trial (NLST) will not answer this question. The decision to include a chest-x-ray arm by NCI epidemiologists continues the work they began with the Prostate, Lung, Colon, Ovary screening study (PLCO). The fact that PLCO, 16 years after its inception, still has not told us whether lung cancer mortality is reduced by a now obsolete test raises serious questions about the competence of the NCI investigators. NCI’s media campaign against screening appears to this observer to be cynical turf protection.

Third, given the enormous 5-6 fold difference between the actuarial 10-year survival in the IELCAP study (>80%) and that in the unscreened U.S. population (15%) investigators in the IELCAP study (myself included) did not believe that it was possible to achieve ethical equipoise in a RCT of computerized tomogram vs. observation alone. I was a principal investigator in the IELCAP study but was also (because of past smoking) a high-risk study participant. If I would not agree to participate in a control arm (I would not), I certainly could not ask research participants to do so ethically. In addition, such a study would be difficult to accrue sufficient numbers of control group study participants and would predictably result in many participants obtaining chest x-rays and computerized tomograms outside of the study (contamination). The NLST was crafted despite a clear understanding of these research compromises.

Fourth, a guideline recommendation for population CT LCS or a larger demonstration project at this time, based upon best available evidence, would have no effect whatever on the NLST study. It is already fully accrued. If and when the study is ever published, it will be evaluated on its own merits.

Fifth, survival is a perfectly acceptable endpoint. 99% of cancer research uses survival as an endpoint. RCT chemotherapy studies that report improvements in lung cancer survival measured in weeks or months are considered clinically significant and drive guideline recommendations that result in enormous public expense with minimal improvement in survival and little or no reduction in mortality. Mortality is nothing but the reciprocal of survival, in the absence of overdiagnosis bias. Although NCI spokespersons persist in postulating high percentages of slow-growing, non-lethal lung cancers, comparable to the situation with prostate cancer screening, there is no credible evidence to support the possibility that overdiagnosis bias accounts for the survival difference between IELCAP 80% and SEER 15%. This hypothesis is so unlikely that it is misleading and irresponsible to persistently expound upon it in public media statements.

“Epidemiology; medicine without the tears.” Arthur Holleb

With reference to breast cancer, the same epidemiologists who are responsible for the faulty design of the NLST also appear to be doctrinaire opponents of mammographic screening. A review of the bibliography of Anthony Miller in PubMed identifies an individual who has been a doctrinaire critic of mammographic screening for three decades. It is, accordingly, worrisome that Miller, who is not a clinician, is chair of the validation evaluation committee which determines cause of death in NLST. Although cause of death decisions are supposed (by protocol) to be determined by committee, NLST minutes show that in as high as 80% of cases, determination of cause of death is made by single individuals. The minutes do not inform how often Miller is that sole individual. Individuals with a strong bias against screening who do not have clinical expertise in the care of lung cancer patients, could make erroneous or biased cause of death decisions that could skew the end results of NLST markedly.

The same NCI epidemiologists now use the exact same arguments against lung cancer screening that they have used in the past in the case of breast cancer screening. Their opposition to breast cancer screening probably resulted in the unnecessary deaths of many women. They are wrong again in the case of lung cancer screening and many men and women are dying and will die in coming years for this perverse stubbornness.

The information on the Vector story is more accurate than prior published accounts and also correctly points out that the New England Journal of Medicine was provided by the Cornell U. authors with patent information but chose not to publish it, and then criticized the authors in the press for supposedly not providing it.

The media statement that this journal has never published anything by authors funded by tobacco is also not true. The New England Journal published a LCS review article by three Duke authors Edward Patz, Phillip Goodman and Gerold Bepler in 2006. Patz and Bepler had each applied for Council for Tobacco Research and been denied. Bepler, on behalf of the cancer center director at Duke, gave a paid seminar to R.J. Reynolds. Goodman has twice testified for the tobacco industry against lung cancer screening and was paid approximately $350,000 for his testimony. He has also performed lucrative medicolegal work from Womble, Carlyle a prominent Durham NC tobacco industry law firm. (Legacy Library) I notified the NEJM of these facts in a letter to the editor which they declined to publish. I copied this letter to Marcia Angel a reforming NEJM editor and again received no reply.

I joined IELCAP in 2001 and did not know about the Vector grant until 2007, but David Burns of the University of California San Diego, who is a respected tobacco control expert, was responsible for setting up the grant. He did so on the basis that the grant address three important problems in tobacco funded research. First, the Vector grant was public information; announced in press releases in two newspapers. Second, the tobacco company had no control over the type of research or its publication. Finally, the tobacco company was not allowed to make any public relations comment on its financial contribution. All of the other tobacco companies refused these terms. It is important also to remember that Liggett is the only tobacco company specifically exempted from Judge Kessler’s RICO guilty verdict against the other tobacco companies.

With respect to the NCI’s virtually sole emphasis on molecular screening for lung cancer, such molecular screening may be useful some day, but it is important to remember than any person who has a positive molecular screen must next have radiographic screening. Meaningful molecular research cannot be conducted outside of an imaging research program.

Controversy between IELCAP and NLST investigators antedates 2006 at a minimum going back to 2001 when NCI investigators were asked to combine studies with IELCAP. They refused to attend an open public debate on the design of lung cancer screening research.

For the sake of balance, I believe readers should be informed that the University of California has no policy against acceptance of tobacco industry grant funding and that a number of branches and multiple investigators including Aberle of UCLA have accepted tobacco money!

Regards,

Fred Grannis, M.D.
Associate Professor and Staff Surgeon
Thoracic Surgery
City of Hope National Medical Center
1500 East Duarte Road
Duarte, CA 91010
FGrannis@coh.org