When the Philadelphia VA Medical Center decided in the late 1990s to start providing a popular prostate cancer therapy, it turned to its longtime, distinguished partner in medicine - the University of Pennsylvania.
That turned out to be a questionable move.
Penn had just published a study in the Journal of the American Medical Association suggesting that the radiation implant treatment, called brachytherapy, wasn't as effective as other therapies. The finding was swiftly attacked by critics who said the real issue was the poor quality of Penn's program.
Penn doctors, they said, lacked the necessary skills and safeguards to perform the radioactive seed implants.
Ten years later, federal investigators cited the VA for the same shortcomings. The VA suspended its brachytherapy program 14 months agoamid revelations that patients were given incorrect and sometimes dangerous radiation doses.
The taint of the VA scandal extends far and wide, touching even federal regulators and professional accrediting groups that overlooked the substandard care. But the origins of the mediocrity, experts and documents suggest, were at Penn.
In response to questions, Penn spokeswoman Susan E. Phillips said in an e-mail, "We have always believed and continue to believe that the quality and effectiveness of Penn Medicine's brachytherapy program is of a very high standard."
Two authors of the 1998 study questioning brachytherapy's effectiveness - Penn radiation oncologist Richard Whittington and Penn urologist S. Bruce Malkowicz - were integral to the VA's program, which Penn contracted to create in 2002.
Whittington mentored Gary Kao, the young radiation oncologist whom Penn selected to lead the new brachytherapy program, according to a former Penn official.
Kao is the only person officials have publicly faulted. He implanted most of the 92 patients - out of a total of 114 - who received insufficient radiation to prostate tumors, excessive amounts to adjacent organs, or both.
But Kao, who has said he feels scapegoated, didn't work in isolation.
Malkowicz assisted on many VA cases. And Whittington did at least three of the VA's subpar procedures, according to the U.S. Nuclear Regulatory Commission, which oversees radiation use.
Whittington also cared for many of Kao's patients before and after their implants, and sat for two years on the VA's radiation safety committee, a monitoring group chaired by a Penn physician.
Neither Whittington nor Malkowicz would comment.
In his textbook on brachytherapy, Kent Wallner, a renowned brachytherapist at the Puget Sound VA in Seattle, offers an explanation for Penn's "remarkably poor" results reported in the 1998 study:
"Isolated poor outcome reports more likely reflect misuse of a technique rather than that the technique is inherently flawed."
Prostate brachytherapy became popular in the late 1980s, thanks largely to improved technology.
With sophisticated scanning tools, internal organs can be imaged before, during, and after seed placement. Hollow needles are used to insert 50 to 100 rice-size pellets into the prostate, where they emit cancer-killing radiation for several months.
Then, computers analyze a final image and calculate the actual radiation doses in and around the prostate. If these vary significantly from the plan, the patient and the NRC are supposed to be informed.
For cancer that hasn't spread beyond the prostate, brachytherapy cure rates are better than 80 percent - comparable to surgery or external radiation, most studies show. External radiation may be used to supplement an implant if the cancer has aggressive characteristics.
The radiation oncologist leads the operation, but his team includes a physicist who plans treatment, and a urologist who inserts an ultrasound probe and places the initial needles.
At the VA, Malkowicz, the urologist, assisted with two patients, the Rev. Ricardo Flippin and Barry Lackro, who have filed claims against the medical center. Both suffered debilitating radiation burns to their rectums - a complication, studies show, in 5 percent of implants.
In Flippin's chart, Malkowicz said he and a resident "assisted in placing 79 seeds . . . throughout the various areas of the prostate."
In all, 35 VA patients received radiation overdoses to tissues outside the prostate, investigators found.
The other 57 flawed treatments underdosed the men's prostates. Eight of these patients had such glaring undertreatment that the Philadelphia VA flew them to Seattle for corrective implants by Wallner.
Wallner declined to be interviewed, but in the textbook he wrote with two other experts, he says gross undertreatment is inexcusable because it is so easily recognized - even by the urologist - on images showing the number and distribution of seeds.
It can be avoided "if even one member of the brachytherapy team has even an ounce of brains," he wrote. "The saving grace for blunders like this is that they are relatively easy to fix" by adding more seeds.
"Probably the worst thing to do," he continued, "is to ignore it - the patient will likely develop cancer recurrence and possibly die from your mistake (and the proof of your guilt is obvious to anyone taking even a cursory look at the films!)."
So far, cancer has come back in six VA patients, and eight others have hints that the treatment failed.
Penn's 25-year-long affiliation with the VA goes way beyond brachytherapy. The VA contracts with Penn's vast health system for maternity, radiology, and other services, while the university's medical students train at the VA.
Indeed, the current issue of Penn's alumni magazine features an article titled, "How the VA Helps Penn Make Better Doctors."
Gillies McKenna, who retired as Penn's head of radiation oncology in 2005, said the VA consulted Penn about offering brachytherapy in the late 1990s.
"We first proposed to the VA that cases could be performed at [Penn's hospital], but they preferred to establish a program based at the VA."
Kao, who had been working at the VA, volunteered to lead the program, McKenna said. Kao's lack of brachytherapy experience was not an obstacle.
"Dr. Kao was felt by everyone to be an outstanding clinician," McKenna recalled, "and we made arrangements for him to receive additional training at a program in Seattle. His initial implants at the VA also were mentored" by Whittington.
That program, at Northwest Hospital in Seattle, offered a two-day course.
This spring, investigators looking into the "root causes" of the VA problems said the primary radiation oncologist had "limited experience." Both radiation oncologists suffered from "inadequate preparation" and supervision of their implants. Names were omitted, but Kao and Whittington were the only radiation oncologists in the program.
Alvaro A. Martinez, an eminent brachytherapist at William Beaumont Hospital in Royal Oak, Mich., questioned the prudence of the VA's partnership with Penn.
"My interpretation is that they [VA] did not do their homework," he said in a recent interview. "They figured, 'Penn is a well-known medical center. They're good.' But if you do your homework, you say, 'Time out. This is probably not the group you want to be subcontracting to.' "
Experience is key to a good implant, experts agree. And Penn had relatively little in 1998, when its results appeared in JAMA.
From 1989 through 1997, Penn had done 218 implants - fewer than 30 a year.
"If you don't do 100 a year, you shouldn't do the procedure," said Gregory S. Merrick, director of the Schiffler Cancer Center at Wheeling Hospital in Wheeling, W.Va.
The JAMA study, written with Harvard researchers, found that brachytherapy was not as effective as surgery or external radiation except for nonaggressive tumors - the kind studies say may not need treatment at all.
After an average of three years, Penn's implants failed to control cancer in half the patients with moderately aggressive tumors. Eighty percent with aggressive tumors had relapses. These failure rates were far worse than with the other two treatments.
Penn spokeswoman Phillips called the paper "a seminal publication" and said the findings continue to stand.
Leading brachytherapists across the country disagree.
After the paper came out, seven of them wrote JAMA to rebut it.
Radiation oncologist Louis Potters, then at Memorial Sloan-Kettering Cancer Center in New York, said the cancer-free, three-year survival rate for comparably risky implant patients there was 78 percent. He warned against giving credence to conclusions drawn by researchers "who perform too few procedures."
San Diego radiation oncologist Donald B. Fuller wrote, "Tricks, traps and subtleties need to be mastered before a brachytherapist is skilled enough to perform the procedure. . . . The 'learning curve' may have contributed to the higher failure rate."
The critics also faulted Penn for a quality assurance problem that, years later, would compromise treatment at the VA: not using CT scans to calculate the actual radiation dosage and compare it to the planned dosage.
Penn used outmoded X-ray technology - Martinez called it "primitive" - on two-thirds of the patients.
At the VA, CT scans were taken, but no one analyzed the images of 23 patients implanted over 14 months. Why? Basically, the scans couldn't be transmitted to the planning computer because it was unplugged from the hospital network, investigators found.
Penn's smallish brachytherapy program got even smaller after the 1998 JAMA paper came out, hospital records show.
Yet James Boyle, 77, of Colmar, Montgomery County, recalls feeling reassured because his doctor - Whittington - said he was very experienced.
Boyle said he didn't realize that the quality of Penn's brachytherapy program had been questioned.
"If we had known about this study, we definitely would have gone somewhere else," Boyle said.
Whittington implanted Boyle at Penn on May 4, 2001 - before the VA program began. Malkowicz assisted on the case.
Boyle's outcomes have not been good. He developed serious rectal burns that, despite surgical treatment, cause occasional bleeding and chronic diarrhea.
Worse, his cancer has come back.
Boyle assumed he was simply unlucky until he read stories in The Inquirer about the VA. Now, he's convinced his implant was bad, although his medical records don't include his actual doses.
Despite Whittington's experience, he did not always report significant dosage errors, according to records in the federal investigation of the VA's brachytherapy program.
In 2005, for example, Whittington and a resident accidently put 13 percent of the prescribed seeds into the bladder, leaving the patient's prostate with only 30 percent of the intended radiation dose, records show.
Whittington performed a corrective implant, but even then, the prostate dose was 20 percent less than optimal - just within the NRC's acceptable range. The VA's Radiation Safety Committee discussed the case, but did not report it, as required, records show.
This implant, like almost all of the 92 deficient ones, was identified only after all VA patient records were reviewed beginning last May.
Last month, Penn's current chair of radiation oncology, Stephen M. Hahn, said that as part of an overhaul of its brachytherapy program, Penn has switched to a technique that allows individual seed placement and simultaneous dosage calculations. He also said Penn would "of course assist in modernizing" the VA's technique if it resumes brachytherapy.
An interactive graphic of how brachytherapy is done and a video of an actual procedure are at http://go.philly.com/