Volume 18 Number 1 Winter/Spring 2001

back

making the case for trials
by ruthann richter

For the first time, researchers launch a clinical trial to compare pediatric surgeries.

LYLE VACHON'S TINY HEAD RESTS COMFORTABLY AGAINST HIS MOTHER'S BREAST, ONLY THE TUBES IN HIS NOSE AND THE MONITORS ON ONE FOOT BETRAY THE INFANT'S SERIOUS CONDITION. Lyle was born in late August ­ 13 weeks early and startlingly small at 2 pounds, 6 ounces, but his face was pink and he seemed otherwise healthy, says his mother, Paloma Vachon. Within a week, he began to feed on breast milk and to rely less on the respirator for breathing support. He was alert and seemed a happy child, his parents said.

Eight days into his new life, though, Lyle developed a rupture in his intestine, and the organ began to fail. His doctors at Lucile Packard Children's Hospital woke the baby's parents at their Santa Cruz, Calif., home at 5 a.m., asking permission to perform emergency surgery.

"He had looked great. He looked perfect," Vachon says. "Then I woke up to a message that he needed surgery. My husband and I were told by the doctors that there were two surgeries available, and they did not know which was better." Paloma and Bernard Vachon were also told about a third option: to enter their son in a ground-breaking clinical trial comparing the two surgeries. This course would leave the selection of surgical procedure to random chance.

When the Vachons chose to have Lyle participate in the trial, they became pioneers in the field of pediatric surgery. The trial is believed to be the first to compare the value of two operations for the same childhood condition. It was, as Lyle's father would say, by a "throw of the dice" that the surgical team selected which procedure would be used to try to fix his baby's failing intestine, which had been damaged by a rare, life-threatening condition known as necrotizing enterocolitis. Chance determined that Lyle would have major surgery instead of a newer, less-invasive procedure.

In that early September morning phone conversation, Larry Moss, MD, assistant professor of surgery and of pediatrics at Stanford, had told the Vachons there was no way to know from the 20 years of accumulated medical experience which procedure would most likely save their son's life. "There is no precedent for this -- to go to a family with a baby who's dying and say, 'We have no idea which operation is better and that's why we need to do a trial,' " says Moss, who is the principal investigator in the multicenter study, which will compare two treatments for premature newborns afflicted with necrotizing enterocolitis.

"The trial has two missions," he adds. "One is to answer this important question about which procedure is better. The second is to demonstrate to our field [of pediatric surgery] and to the world at large that we're prepared and able to do this kind of clinical trial successfully."

The National Institutes of Health-sponsored study, which began in the summer of 1999, goes against the grain of surgical practice, which historically has relied more on surgeons' individual experience over time than on rigorous scientific studies. Since the 1980s, fewer than 1 percent of published studies in the surgical literature have been randomized clinical trials, even though these trials are considered the gold standard in clinical research generally, Moss and colleagues plan to report to in the June 2001 Journal of Pediatric Surgery.

"When you scratch the surface, it's astounding how much of what we do is based on anecdote," Moss says.

EVEN IN THE ARENA OF ADULT SURGERY, PHYSICIANS HAVE BEEN RELUCTANT TO TAKE PART IN CLINICAL TRIALS FOR A HOST OF REASONS. FOR ONE THING, SURGEONS ARE AN independent sort, prizing their autonomy and their ability to change course mid-stream, if necessary, as a procedure unfolds.

"Surgery historically has been a field of iconoclasts and
individuals," says Thomas Krummel, MD, professor and chair of surgery at Stanford. "For the last 100 years, this was the dominant surgery personality ... . That generates a person who hasn't always been ready to say, 'Maybe my personal experience is not sufficient to answer a question. Maybe there's another idea; let's study it in a systematic way.' "

Another reason for the rarity of clinical trials for surgical procedures is that surgical practice, by its very nature, does not lend itself to rigorous controls. There are inherent difficulties in designing a randomized clinical trial, as each surgeon's hand and each patient's anatomy will differ. Moreover, surgeons can't necessarily perform a "standard" operation in the same way that doctors can give patients a standard dose of a drug, nor can surgeons be "blinded" as to which group they've been assigned.

Though conducting a clinical trial to test a drug makes sense, conducting clinical trials to test a medical device or procedure often doesn't, says Thomas Fogarty, MD, a Stanford professor of cardiovascular surgery who has patents on more than 70 medical devices. "You can't compare a drug to a device or a procedure," says Fogarty. "You can't see what a drug does and you can't see what influences it. With a procedure or a device, a lot depends on the technical expertise of the people doing it. And you can influence it."

FOGARTY SAYS THERE ARE SITUATIONS IN WHICH CLINICAL TRIALS ARE APPROPRIATE, AS IN THE CASE OF A DEVASTATING DISEASE LIKE NECROTIZING ENTEROCOLITIS, BUT that randomized trials are not always called for, especially when the clinical value of a given procedure or device seems intuitively apparent. Randomization is only one format of a clinical trial and is most appropriate for drugs, but it is often inappropriate for surgical procedures and devices, he believes. He also notes that randomization can be time-consuming and in the case of a new device or technology, the technology may well change by the time the study results are available two or three years later.

"I think people use randomized clinical trials as a surrogate to good clinical judgment," Fogarty says. "People are using randomized trials as a last recourse in an attempt to resolve a lot of issues. Randomization doesn't always do it. It doesn't do it in terms of loss of lives and needless complications -- because the longer the study takes to perform, the more people will have suffered in the process. So the quicker we get the answer, the better off our patients would be."

Moss acknowledges that clinical trials can be challenging and cumbersome to perform, but he believes many surgical approaches to treatment still cry out for scientific validation.

"People argue that by the time a trial is done, the technology will be more refined and won't mean anything anymore," Moss says. "But I believe that technology does not necessarily equal true innovation. We can't focus on improving a technology before we even know if it works."

He notes that over the years many procedures, such as coronary artery bypass, have come into universal acceptance without being tested in a rigorous fashion. And in some cases, he adds, widely accepted procedures have been shown to be ineffective when studied in a randomized trial. For instance, for more than a decade neurosurgeons performed a procedure known as extracranial-intracranial bypass on patients with blockages in the carotid artery, which carries blood to the brain, in the belief that it would reduce the patients' risk of stroke. But results of a multicenter randomized trial published in November 1985 in the New England Journal of Medicine found that patients who had the operation fared no better than those who did not.

As another example, many operations nowadays are being done laparoscopically so that surgeons don't have to make a large incision but instead work through a tiny opening, significantly reducing patient recuperation times. Surgeons now routinely use a laparoscopic approach to remove diseased gall bladders even though
injuries to the bile duct, which can be devastating, occur twice as often with this less-invasive approach, Moss notes.

"I'm not saying laparoscopic cholecystectomy [gall bladder removal] is wrong; rather I am arguing that we still don't really know its true benefits and risks because we have not subjected it to scientific study," Moss says. "Here's a complete change in practice involving the most common operation in general surgery without a shred of evidence that it's better."

Krummel notes that surgical treatment for some conditions, such as gunshot wounds or transplantation for kidney failure, are intuitively obvious; with other conditions, however, the solution may be less apparent.

"Many new surgical procedures have been thought to be so obvious that we've never gone to the trouble of doing a trial. But what seems intuitive is not always correct. Do I think it's reasonable to do a trial before it's broadly practiced? Yes," Krummel says.

IN THE CASE OF NECROTIZING ENTEROCOLITIS, THE LITERATURE HAS NEVER BEEN CLEAR ON WHICH SURGICAL INTERVENTION IS MOST LIKELY TO SAVE LIVES, MOSS SAYS. The disease, responsible for some 5,000 to 9,000 newborn deaths a year, occurs as a result of reduced blood flow and bacterial overgrowth in the fragile intestine of the premature infant. The organ can then tear and begin to fail. The symptoms are readily apparent, as the infant will show obvious signs of infection -- vomiting, bloody stools and a distended belly.

For years, doctors treated the disease as they would in adults, performing major surgery on these tiny babies. The surgery involves opening the abdomen, removing the dead portion of the intestine and then creating a stoma, or bag, by bringing out the ends of the intestine through the abdominal wall. The ends then are reattached and the hole closed several months later, once the organ has healed. But some surgeons were reluctant to perform such major surgery on the smallest of preemies -- i.e. those weighing less than 1,000 grams. In the mid-1970s, doctors at the University of Toronto's Hospital for Sick Children tried an alternative procedure as a last resort to save these infants. This alternative involved the simple insertion of a drain through a 1/4-inch incision on the baby's lower right side, the most common site of infection. The drain allows the body to rid itself of the offending bacteria and relies on the body's ability to break down and reabsorb the damaged tissue in the intestine and effectively heal itself. Much to the physicians' astonishment, many of the babies treated in this minimally invasive fashion survived.

"It does seem to be the case that there is a very big difference in the way the body of a neonate responds to a serious abdominal infection compared with an adult. If you or I had this infection, we would have to be opened up," says Lawrence Mathers, MD, PhD, Stanford associate professor of pediatrics and of surgery and former chair of the ethics committee at Packard. "The mystifying and fascinating part of all this is that it has been shown that, with newborns, there's comparably good outcome to just putting in a soft rubber drain -- similar to a finger from a rubber glove -- to drain the pus out of the abdomen."

OVER THE YEARS, INDIVIDUAL SURGEONS HAVE COME TO ADVOCATE ONE APPROACH OR ANOTHER, BUT THE DATA HAVE NEVER PROVED CONCLUSIVELY WHICH IS BETTER. Moss and his colleagues recently reviewed all medical literature on the subject since 1978 and found an overall survival rate of 67 percent for infants who had the major operation, compared with 55 percent for those who had the drainage procedure. However, the results were skewed by the fact that the larger, and presumably more sturdy, babies were more likely to be treated with the major procedure. Moss and his colleagues, who presented the results of their review at an October 2000 meeting of the American Academy of Pediatrics, concluded that the data failed to show which procedure was superior.

Moss decided the issue was ripe for a clinical trial. He began on a small scale in the summer of 1999, using what limited resources he had. But clinical trials are expensive to run and funding for surgery trials is particularly scarce, as private industry has little self-interest in supporting these endeavors, Moss notes. So he applied for a grant from the National Institutes of Health and was delighted when, in June 2000, the NIH informed him that it would provide more than $1 million for a large-scale trial at multiple medical centers around the country. In addition to Stanford, the centers include University of California-San Francisco, Baylor College of Medicine, Columbia University, the University of Pittsburgh, the University of Texas, the Medical College of Wisconsin, the Children's National Medical Center in Washington, D.C., and others.

The three-year trial, which already has enrolled 11 patients, will include 130 children. Half will undergo the major operation, while the other half will have the drainage procedure. Participating physicians will literally pull an envelope at random in the neonatal intensive care unit to determine which child gets which procedure, Moss says.

Enrolling patients is always a challenge in clinical trials, but it is even more so in a trial of this sort, which involves surgery on critically ill babies. Parents essentially are being asked to entrust their child's fate to an unknown.

"People want a course they can be confident in. 'I don't know' is not an answer they want to hear when they walk into an office," Krummel notes. That is one of the many reasons, he says, that it has been so difficult for clinical trials in surgery to gain acceptance.

Lyle's father, who was asked to make an urgent decision about his son's treatment in the wee hours of the morning, said he saw no alternative but to sign up for the trial. Here, after all, was a physician telling him that he did not know which course of action would be better.

"In retrospect, we think Lyle got the right surgery, and he seems to be doing better," Bernard Vachon said five weeks after the procedure. "Maybe the less-invasive surgery would have been better. Who knows? It would be nice to have a definitive answer. ... At the time, I had some thought that it would be a noble thing to do for science. Now my feeling is more along the lines of 'I don't want my son to be a guinea pig.' "

Indeed, Lyle has done well since the surgery in which doctors removed some seven inches of his small intestine. He has had some ups and downs that have been emotionally tortuous for his parents, but they were thrilled to be able to take him home on New Year's Day 2001.

To help parents in the decision-making process, Moss and his colleagues are making unprecedented use of the Web to disseminate materials about the two surgical procedures and the trial itself. Shawn Rangel, MD, a general surgery resident, has been working with Moss to develop a Web site that includes a 20-minute video showing both operations, parental consent information and answers to often-asked questions. It also includes a video-based protocol for participating physicians and instructions for patient care after the initial surgery.

"With surgery, there are so many steps and variations. This is an effort to standardize the process as much as possible," says Rangel, who is developing the site as part of a postdoctoral research fellowship in pediatric surgery.

The hope is that the Web site, the first of its kind for a clinical trial, will become a model for other surgical studies, he says.

Moss' ultimate goal is to conduct a trial with methods that are unassailable and to show the surgery world that clinical trials for surgeries can be done -- and done well. He hopes to make clinical trials the norm in pediatric surgery, just as they have become the norm for children with cancer.

"There are a huge number of constraints against this kind of thing that are slowly breaking down," Moss says. "I think of it as a global, glacial, cultural change that needs to happen. We will be successful." SM

by ruthann richter