Two scientists involved in 25-year, $140 million study say it may underestimate cancer risks for Vietnam vets
By Clark Brooks, Greenville Online
A design flaw in the federal government’s $140 million study of the health effects of Agent Orange on Vietnam veterans has resulted in a quarter-century of inaccurate findings, two scientists involved with the study told The Greenville News.
Begun in 1978 to help settle compensation claims, the Air Force Health Study will end this week as it began, in controversy, with tens of thousands of veterans still seeking answers to chronic illnesses they attribute to herbicides used during the Vietnam War.
Agent Orange and other herbicides sprayed in Vietnam to destroy enemy crops and jungle cover contained cancer-causing dioxin. The U.S. Air Force, however, is closing up shop on the study having found no increased incidence of a serious illness other than diabetes.
The study has compared airmen directly involved with the spraying missions, called Operation Ranch Hand, to Air Force veterans who served in Southeast Asia but had no role in spraying.
However, hundreds in the comparison group spent time in Vietnam and may have been exposed to herbicides, too, said Joel Michalek, who worked on the study from the beginning and was its principal investigator for 14 years until he left in May.
“It spoils everything,” Michalek told The News. “It’s as if you’re running a clinical trial on a new medication, and you found out some of the people who were in your placebo group were actually taking meds. That would spoil your whole study. And that’s what’s going on here in this study.”
Michalek co-authored two articles published in the Journal of Occupational and Environmental Medicine in 2004 and 2005 that found significant rates of cancer in the Ranch Hand and comparison groups.
Air Force spokesman Ed Shannon declined to make officials available for comment. Shannon was asked why Michalek’s analysis published in the Journal showing cancer trends in the comparison group of veterans was not used in the analysis for the final Air Force report published last year.
The Air Force noted in an e-mail reply that a “recently published analysis” showed an increased cancer risk in Ranch Hand and comparison veterans. Shannon said Saturday there would be no further Air Force analysis.
In a follow-up e-mail, the Air Force said the final report included only the veterans who attended the last round of medical tests in 2002 and that all physical examination reports follow the same basic analytical plan.
Michalek’s finding of cancer in the comparison group was not used in the analysis for the Ranch Hand report.
Michalek said he followed up on the cancer articles with an analysis that allowed for the exposed control group and other factors and found a doubling of cancer in the Ranch Hand group.
Further research needs to be done to strengthen these findings and figure out what other diseases the Air Force scientists may have missed because of the exposed comparison group, Michalek said.
The comparison veterans, he said, are similar to average Vietnam veterans, from nurses to truck drivers, who spent most of their time in base camps. The comparisons’ data also should be studied further, he said.
The results could matter greatly to thousands of Vietnam War veterans who’ve never received compensation for debilitating illnesses that earlier Ranch Hand study findings said couldn’t be linked to Agent Orange.
A Department of Veterans Affairs analysis in 1998 found 92,276 Agent Orange claims for compensation had been filed by veterans and their survivors. Of those, 5,908 had been approved.
The analysis was done before diabetes was added to the list of diseases eligible for compensation, which would make both columns much higher today, said Jim Benson, a VA spokesman.
The VA no longer tracks Agent Orange claims because many veterans apply for more than one type of compensation per claim, he said.
The Ranch Hand study has followed about 1,000 Ranch Hand veterans and some 1,300 comparison airmen who served in Southeast Asia.
Although the study will end Saturday for the Air Force, legislation pending in Congress would turn over all the data and specimens to the Institute of Medicine’s Medical Follow-up Agency, which would collaborate on analyses with scientists outside the government.
Michalek left his civilian Air Force job for the University of Texas Health Science Center in San Antonio. He said he will apply on behalf of the school to be a collaborator.
Greer soldier sprayed
The U.S. military sprayed more than 18 million gallons of herbicides over 3.6 million acres of South Vietnam from 1962 to 1971. Nearly two-thirds of it was Agent Orange.
Richard Leoffels of Greer saw the planes spraying overhead when he was an Army infantryman with the 1st Cavalry Division in 1968-69. Sometimes the wind blew it onto him and his buddies as they set up for ambushes, he said.
He didn’t give it much thought, he said, even as he occasionally crawled through areas saturated with herbicides. He was more concerned about the enemy.
“I didn’t know anything about Agent Orange until I came back, did some reading and saw a couple specials on TV,” he said.
Red blotches appeared on his legs in 1969, just a minor annoyance, he said. Later, he would suffer a litany of more serious conditions.
The Air Force has announced in periodic updates since 1984 that the Ranch Hand veterans are about as healthy as the comparisons and have no significant increase in cancer or heart disease or any other serious illness except diabetes.
Ranch Hand and comparison veterans were thoroughly examined every three to five years, beginning in 1982. The results were recorded in thick Air Force reports.
The final one of those, published last year, presented the results from the sixth and last round of testing, conducted in 2002. It concluded the cancer analysis “did not suggest an adverse relation between cancer and herbicide exposure.”
Ron Trewyn, a biochemist and member of the Ranch Hand study advisory committee, reviewed that report’s cancer chapter.
He argued strongly during advisory committee meetings that the cancer chapter should include all the cancer data used to write the 2004 and 2005 articles in the Journal of Occupational and Environmental Medicine. It didn’t happen, he said.
“They referenced those papers, but they left all the data out from those cancer papers that were done that showed the cancer effects,” he said. “It’s huge, because then the conclusion is there’s no cancer effect, when as part of the study, the same investigators, just analyzing the data in a different way, found that when they did that, lo and behold, then there were significant cancer effects.
“And so for the final report to say there’s no cancer effect when the investigators themselves published papers saying there is a cancer effect, that’s just flat scientifically wrong.”
Without factoring in the new information about the comparison veterans, Trewyn said, the Air Force got the same, predictable results.
“When they use an exposed control group and they say the two groups have roughly the same amount of cancer and so forth, what is that finding good for? Nothing,” said Trewyn, vice provost for research and dean of the graduate school at Kansas State University.
And it doesn’t take a scientist to figure that out, he said.
“This is common sense now, a lot of it,” he said. “It’s like now wait a minute. This just does not pass the smell test or the common sense test.”
Trewyn, who said he began wondering about exposures in the comparison group in 1999, did cancer research for 20 years.
Because many comparisons were exposed to the same environmental conditions as the Ranch Hand veterans, all major health outcomes need to be re-examined, he said.
“There have been industrial studies related to dioxin where as they looked back at it they thought they had a few exposed in the control group and so the statistics went to hell,” he said.
In the Ranch Hand study, it’s more than a few. At least 600 members of the comparison group spent time in Vietnam, Michalek said.
New rates found
Michalek said the breakthrough that led to the new data analysis came when he started to look not just at the numbers but at the men behind them. Where in Southeast Asia did the Ranch Hand and comparison veterans serve? For how long?
He learned some Ranch Hand veterans didn’t take part in spraying because none was done while they were there, and those who served earlier in the war had higher levels of dioxin.
When he factored in that information along with the exposed comparison group, Michalek said he found a doubling of cancer among Ranch Hand veterans with the highest dioxin exposures. He also found cancer increasing with dioxin exposure, the first time such a trend has been seen in the Ranch Hand study, he said.
Michalek said he also found a stronger showing than previously for diabetes.
Advisory committee members wanted him to get the new cancer and diabetes findings published in a scientific journal, and he told them he intended to, according to minutes from the June 2005 committee meeting.
However, Col. Karen Fox said during the committee’s final meeting this month in Rockville, Md., that the Air Force has no plans to publish the new findings in any Air Force report or scientific journal, The News reported earlier this month.
Fox, responding to extensive questioning from advisory committee members, said the Air Force told Michalek to destroy the data.
Fox, who succeeded Michalek as principal investigator of the study, declined to be interviewed by The News during breaks in the meeting.
She said during the meeting the Air Force “tried to enter into a relationship” with Michalek to write the cancer and diabetes papers, but “he elected not to do that.”
Michalek said the Air Force told him he would have to contract with Science Applications International Corp., which does data analysis for Ranch Hand study reports. He said he negotiated with SAIC but wasn’t hired.
Maurice Owens, a project manager for SAIC, told The News the company decided it would be a conflict of interest to work with Michalek because he had been a scientist for the Air Force.
There is precedent for such a hire, however. Col. George D. Lathrop, who helped design the Ranch Hand study, moved to SAIC during the 1980s after he retired from the Air Force.
Owens said he couldn’t comment on that.
Michalek said he began writing the cancer paper without pay. He said he finally gave up when he got a letter from the Air Force dated July 6, 2006, ordering him to delete the data.
Rick Weidman, who has monitored the Ranch Hand advisory committee meetings for Vietnam Veterans of America, said he believes the Air Force had no intention of letting Michalek write the cancer paper on his own.
“They didn’t want him to publish because they wanted to be able to censor it,” Weidman said. “That’s just plain as day to us.”
Because Ranch Hand study reports had said the health of the Ranch Hand and comparison veterans was about the same, some members of Congress sought other ways to settle compensation claims. The Agent Orange Act of 1991 established a compensation list.
The first entries were non-Hodgkin’s lymphoma, soft-tissue sarcoma and chloracne, a skin condition. The act also authorized the National Academy of Sciences to evaluate dioxin research from a host of studies, mostly of civilians.
Using the results of that research, the Department of Veterans Affairs has added nine diseases, mostly cancers.
Leoffels suffered his first of three strokes in 1998. They were minor as strokes go, but for a time, he couldn’t control his left leg.
He was working as a letter carrier for the post office, a good job, he said, but not one a person can stagger through.
“People were calling the post office and saying, ‘Hey, the mailman is walking around drunk,'” he said.
Circulatory disorders are on the long list of diseases and conditions for which the NAS has not found enough evidence of a dioxin association to be included for compensation.
Leoffels, 58, does receive compensation for type 2 diabetes, he said, $112 a month. It’s the one illness on the list that might owe its spot to the Ranch Hand study, said David Tollerud, an epidemiologist who headed the NAS research during the 1990s.
Spina bifida, a birth defect, is the only other condition on the list that received an assist from the Ranch Hand study, he said.
Tollerud, a professor of public health at the University of Louisville, chaired the IOM panel that recently recommended the Ranch Hand data and specimens be saved for study outside the Air Force.
He briefed the Ranch Hand advisory committee during a meeting in February. He called the biological specimens accumulated over 25 years “a trove of valuable research material,” according to the minutes from that meeting.
Tollerud also pointed out some study limitations, including the study’s “flawed design and execution” and “potential herbicide exposures in the comparison populations,” the minutes show.
In an interview with The News, Tollerud said his comments were not meant to be condemning but to recognize limitations that future researchers need to take into account.
As for the exposed comparison group, he said, “The general result of that kind of a complication in a study design would be to do what we call bias it toward the null, meaning that it might make it less likely that you would observe findings that were really there.”
Leoffels said he is in favor of continuing the Ranch Hand study as long as it is done outside the Air Force.
“Why throw away $140 million?” he said.
Leoffels said he lost his job as a letter carrier to post-traumatic stress disorder. The VA compensates him for it, offsetting what he believes he should be getting for Agent Orange damage, but isn’t.
He helps other vets navigate the VA, though many get discouraged the first time they are turned down and never go back, he said.
Leoffels said it shouldn’t be so difficult for veterans to get the help they need.
“I think what the government wants is for us to die off so they don’t have to pay us anything,” he said.