The New York Early Lung Cancer Action Program (NY-ELCAP) has come under fire from researchers who claim the study has fundamental design flaws, could harm participants by overdiagnosis, and is a poor use of public funds.
Estimated to cost $10 million over two years, NY-ELCAP was funded primarily by tobacco settlement money. It is already halfway to its target of providing spiral CT screening for 10,000 people who are at risk for lung cancer.
The critique, led by Dr. Steven Woloshin, a researcher at Dartmouth University's VA Medical Center, charged that without a control group the study cannot establish whether screening with CT saves lives. It appeared in the June 15 Lancet.
"The flaw is not in ELCAP itself, but in drawing inferences about the benefit of screening from it, namely, that it reduces one's chance of lung cancer death. NY-ELCAP is advertised this way -- the original Lancet paper didn't make this inference," Woloshin said.
In this type of study, lead, length, and overdiagnosis biases favor the screened group, he said. A trial, on the other hand, would overcome these problems by assessing mortality from the time of randomization -- and not from time of diagnosis.
"The ELCAP study -- and as we note in our paper, NY-ELCAP -- can teach us a lot: how often spiral CT scans demonstrate abnormalities suspicious for cancer, how often patients with these abnormalities undergo various follow-up tests, and how often lung cancer is detected. It just can't tell us if screening saves lives," Woloshin said.
The Dartmouth researchers contend that had the study undergone scientific review like many projects of this scope, peer reviewers would have insisted it be randomized.
NY-ELCAP lead investigator Dr. Claudia I. Henschke, chief of chest imaging at New York Weill Cornell Medical Center, dismissed the critique by saying it misses the point of the study.
"The project was specifically funded because it was an alternative to the randomized trial. We have introduced a new paradigm on evaluating screening," Henschke said.
NY-ELCAP attempt to validate the preliminary results of the original ELCAP study in a larger segment of the population. It will also examine the efficacy of annual follow-up screening, provide data on the curability of different size lung cancers, and help determine how frequently screening should be performed based on the subject's risk factors.
Participants may also benefit from new treatments that become available. A randomized trial can't do that, she said.
Results of the study will provide the necessary database to define standards of care for the work-up of CT-detected abnormalities, including computer-aided diagnostic techniques. These data will be relevant to the formulation of public policy on CT screening for lung cancer, Henschke said.
"The stance taken by Woloshin as to how to conduct screening studies is no longer the mainstream view, as both the National Cancer Institute and the American Cancer Society have sponsored several meetings and endorsed the need for alternative studies to be done," she said.
The primary force behind funding for NY-ELCAP was the Academic Medicine Development Company (AMDeC), a nonprofit corporation created in 1997 by a statewide consortium of medical institutions to help obtain federal funding for medical research in New York.
NY-ELCAP was thoroughly reviewed by the AMDeC board, which includes the deans and presidents of many of New York's medical schools, academic health centers, and research institutions, Henschke said.
"To attack an ongoing study that has been thoroughly vetted by many experts is quite surprising," she said.
Woloshin concluded that an effective screening program would be beneficial, but to move forward with spiral CT screening for lung cancer before a randomized trial has confirmed its safety is premature and possibly dangerous. He mentioned two randomized controlled trials under active consideration, one by the National Cancer Institute, the other in the U.K.
For her part, Henschke could never participate in a randomized lung cancer screening trial.
"Ethically, I couldn't say I don't know that a chest x-ray is any better or any worse than a CT," she said. "I know definitively from our study that CT picks up six times more small stage I lung cancers, those that are curable, than chest x-ray."
For more information from the Diagnostic Imaging archives:
