© The Author 2007. Published by Oxford University Press.
CORRESPONDENCE |
Re: Extended Lung Cancer Incidence Follow-up in the Mayo Lung Project and Overdiagnosis
Affiliation of authors: Center for the Biology of Natural Systems, School of Earth and Environmental Sciences, Queens College, City University of New York, Flushing, NY
Correspondence to: Alfredo Morabia, MD, PhD, Center for the Biology of Natural Systems, School of Earth and Environmental Sciences, Queens College, City University of New York, 163-03 Horace Harding Expressway, Flushing, NY 11365 (e-mail: alfredo.morabia{at}qc.cuny.edu).
Marcus et al. (1) are to be commended for successfully carrying out the follow-up of the 1983 survivors of the Mayo Lung Project. We question, however, whether overdiagnosis of screening-detected lung cancer is a necessary or sufficient explanation of the observed results.
If we assume that successful randomization resulted in two study arms with a similar risk of lung cancer throughout follow-up, the existence of screening-related accelerated lung cancer diagnoses in the intervention group before 1983 should have led to a deficit of lung cancer incidence in the same arm after 1983. This redistribution is explained by the lead and length time advantages conferred by screening. However, not only was no lung cancer deficit observed in the screened group after 1983 but also there was an absolute excess of lung cancers in this screened group compared with the control arm. Thus, the true excess of lung cancers that was observed in the screened group after 1983 was larger than it appeared because it included both the lung cancer cases that erased the expected deficit and the observed cases in excess of the control arm. Marcus et al. note the latter component of the excess, but not the former, and describe this result as "perplexing." When the true magnitude of this excess is considered, these results are more than just perplexing and require reconsideration of the overdiagnosis explanation.
We propose that oversurveillance of the intervention group provides a simpler explanation that reconciles all observations. The intervention group apparently had closer medical surveillance than the usual-care group, both before and after 1983. Evidence thereof can be found in Table 3 of Marcus et al. (1): more lung cancer cases were identified, 30 in the Mayo Clinic records and 18 in death certificates, for the intervention arm than for the usual-care arm. It is reasonable to assume that this medical oversurveillance already existed before 1983. Swensen et al. (2) have shown that lung cancer deaths may have been misattributed to other causes in the Mayo Lung Project. In the presence of oversurveillance, this misclassification is likely to have been higher in the usual-care than in the intervention group. Thus, oversurveillance from 1971 to 1996 could be responsible for a constant excess in lung cancers identified in the intervention group that would explain a fraction of the lung cancer case difference before 1983, do away with the deficit expected after 1983 attributed to lead and length time, and explain the observed excess in lung cancer cases in the intervention group after 1983.
Table 1 and Fig. 1 simulate a situation of oversurveillance. If we subtract 12.5% oversurveillance as indicated by Table 3 in Marcus et al. (48/379 = 0.127) for the 27 years of follow-up, we would expect an excess of 27 lung cancer cases in the intervention group before 1983 from accelerated diagnoses, followed by a post-1983 deficit of 21 cases, which nearly compensates the pre-1983 excess. Thus, oversurveillance provides a simple explanation to the observations. Under this scenario, positing an overdiagnosis in lung cancer screening becomes superfluous.
|
|
NOTES
Editor's note: Drs A. Morabia and S. Markowitz are expert witnesses for the plaintiff in a lawsuit (Caronia v. Philip Morris) that involves the use of helical computerized tomography for lung cancer screening. Both declare that this correspondence was not supported or instigated by their involvement in this lawsuit.
REFERENCES
(1) Marcus PM, Bergstralh EJ, Zweig MH, Harris A, Offord KP, Fontana RS. Extended lung cancer incidence follow-up in the Mayo Lung Project and overdiagnosis. J Natl Cancer Inst (2006) 98:748–56.
(2) Swensen SJ, Jett JR, Hartman TE, Midthun DE, Mandrekar SJ, Hillman SL, et al. CT screening for lung cancer: five-year prospective experience. Radiology (2005) 235:259–65.
Response to this Correspondence
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2007 99: 899.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||