Potti Bombshell: He gave lung cancer patients chemotherapy that was not approved

Search words Anil Potti Duke University


There is a bombshell in Sunday's News and Observer: Potti gave a drug cocktail to some lung cancer patients -- a combination of very powerful chemotherapy that was not approved by drug regulators.

✔ The man was out of control.

This is not the bombshell FC has been referring to. On Saturday evening we were offered significant interviews on Monday and Tuesday. So we will delay writing.

On December 3, FC reported on the Potti patient who is the focus of today's N and O article. At that time the patient did not want to be quoted or identified, and we posted significant information from her brother.

For some reason the N and O on-line edition does not give a by-line to its very deserving reporter, Sarah Avery.

Headline: Flawed research appalls cancer patient

RALEIGH A string of what-ifs nags at Joyce Shoffner.

Two years ago, when she was 61, she was diagnosed with breast cancer and agreed to participate in a clinical trial at Duke University that was based on now-discredited science by a now-disgraced scientist.

Although Shoffner received standard chemotherapy, she was steered to the treatment using what was supposed to be a genetic predictor for its success. But the treatment did nothing to stem the tumor's growth, and instead led to blood clots and other problems that complicated her care.

"I can't say how much would be different if I had been on a different chemo," said Shoffner, who lives in Raleigh. "I have nothing but questions and no answers."

Her biggest question is how flawed research advanced to being tested on cancer patients. Many others are asking the same thing in a case that reveals the downside of high-stakes, vastly complicated science.

Shoffner was one of 110 Duke patients enrolled in three clinical trials based on the research of Dr. Anil Potti, who resigned in November as an associate professor at Duke. Potti was found to have embellished his credentials.

(( FC note: We believe 300 people participated in the clinical trials. And many many more were subjected to biopsies and other invasive tests as they were screened for participation. At the time Duke shut them down, there were 109 patients remaining in them. ))

At the same time, his once-heralded scientific work crumbled under scrutiny.

Collaborators - led by Dr. Joseph Nevins, whose lab at Duke has reaped millions of dollars in federal and private research grants - had to retract the work in prestigious peer-reviewed scientific journals, including the Journal of Clinical Oncology and Nature Medicine.

(( FC note: On Saturday, Chancellor Dzau stated more retractions are coming. Duke has had to refund a huge grant to the American Cancer Society. And FC has learned that Duke is giving up a second grant, details and figures not yet known. ))

The clinical trials, which were based on the published science, were shut down.

"These trials should not have been done," said Dr. Michael Cuffe, Duke's vice president for medical affairs.

Cuffe said university officials are reaching out to patients and their families.

(( Whoa. Whoa. FC is going to pin down the timetable on that. Very carefully. We want to make sure this is not cosmetics, because patients told us in early December they got only perfunctory calls from their new doctors and did not hear from officials. ))

In the two breast cancer studies, he said, treatments were standard therapies that doctors would have considered for use anyway. The lung cancer study channeled some patients to a combination therapy that was not yet approved for lung tumors.

In addition, Cuffe said, the university is working to improve its own oversight of increasingly complicated scientific research. It is also cooperating with the Institute of Medicine, a health advisory group affiliated with the National Academy of Sciences, to use the incident as a case study establishing more openness among researchers about their data and methods.

The hope is to safeguard patients such as Shoffner. But critics of Duke contend officials had the wherewithal and yet missed opportunities to act more swiftly and thoroughly when questions about Potti arose more than three years ago.

"I'm aware that at some point in the process, people may have accepted the initial arguments in good faith," said Keith Baggerly, a biostatistician at MD Anderson Cancer Center in Texas who first raised concerns about Potti's research in 2007.

"But once questions have been raised where there are legitimate and demonstrable concerns, I would like to think that drives a very high level of rigor and confirmation."

(FC: we are developing specifics on warnings that Duke got from the MD Anderson scientists, warnings that Duke administrators chose to conceal during Duke's internal investigation last winter. ))

For that reason, Shoffner said, she feels betrayed by Duke, an institution she trusted, and by science, an endeavor she has long championed.

(( FC: Shoffner once worked at Duke, as did her brother. Her father retired from Duke after long service. ))

"I'm devastated by this whole thing," she said. "If you have a very serious cancer and two-and-a-half years later you think you are involved in a study that is cutting edge and [it's discredited], it is devastating."

Prediction goes awry

Shoffner, now 63, has invasive ductile adenocarcinoma, a form of breast cancer that begins in the milk ducts. When Shoffner's oncologist at Duke mentioned the clinical trial in July 2008, she eagerly volunteered.

"I believe in clinical trials, and if you have the chance to do something to keep someone else from going through this, you ought to do it," Shoffner said.

The chemotherapy was supposed to shrink her tumor before she underwent surgery to remove it. Potti's predictor steered her to a drug that he claimed worked especially well on her type of tumor.

Involvement in the study required surgery. She had to undergo a second biopsy, and nine titanium clips were implanted around the tumor. The clips served as markers, because doctors were confident the chemotherapy would melt the tumor beyond their ability to see it during a subsequent lump removal surgery.

The opposite happened. The cancer grew and spread to lymph nodes, and as a result of the chemotherapy, she developed blood clots. Before she could even have the tumors removed, she had to have surgery in which a filter was implanted in a heart vein to catch blood clots.

Then she had to have another surgery to remove the filter device before she could undergo radiation.

She said subsequent rounds of chemotherapy caused painful nerve damage in her legs, and she was recently diagnosed with diabetes. She continues to get twice-daily injections of an expensive blood thinner; and to pay for it, she moonlights at a hardware store in addition to her full-time job in a chemistry lab.

An avid fox hunter, horseback rider and antiques collector, Shoffner now walks with a cane and can no longer pursue her hobbies.

"I signed my consent to have a legitimate study done," Shoffner said. "I did not sign up for flawed science, which is what I have gotten."

(( FC note: the doctrine of informed consent will play a very important role in any litigation growing out of this mess ))

Shoffner, who said she has nothing but praise for the care she received at the university's cancer clinic, instead blames top Duke officials and Potti's collaborators.

"Who checked the man's credentials?" Shoffner asked. "Who went behind and checked his science? Dr. Nevins is a co-author - why didn't he check? There needs to be some kind of auditor of the data."

(( President Brodhead has stated no changes will be made in the way Duke hires, that it will take credentials at face value. ))

Finding seems promising

The kind of science Potti's team at Duke conducted is mind-boggling in its complexity.

Using sophisticated arrays of genetic information that is run through mathematical algorithms, Potti claimed to have discovered tell-tale genetic characteristics in cancer tumors that correlated to drug susceptibility.

It appeared to be an enviable breakthrough when the first study was reported in October 2006. Using cancer genetics to steer patients to effective treatments has long been the promise of personalized medicine, and research teams around the country have been pursuing the grail.

Among the prizes are lucrative business opportunities to license or market the prediction models. A month after Potti and Nevins published their first big paper in 2006, they were listed as officers in a startup company called Oncogenomics that would capitalize on their science.

The company later became CancerGuide Diagnostics, raising millions of dollars in startup money to market the cancer signature technology.

Duke also had a stake in the company, although Cuffe said the university pulled out in August amid the allegations over Potti's résumé. He said the company "closed shop" late last year.

(( The Duke Ventures office has not answered FC inquiries about Duke's outside partners in this company, how much they paid to get a slice of the action, and how much the value was assigned to the total pie. Shame on them. ))

Beyond sales potential of the Duke team's finding, other cancer doctors were eager to use the predictors for their patients.

A group at MD Anderson in Texas was among the enthusiastic, asking its biostatisticians to check the science and figure out how they could apply it to help their patients.

That's when things began to unravel.

Findings disputed

The key requirement for scientific integrity is replication.

Yet two statisticians at MD Anderson, Baggerly and his colleague Kevin Coombes, immediately ran into problems with the Duke work. They couldn't re-create the findings using Potti's data.

Baggerly asked for more data. Still no success. And subsequent papers from the Duke team about predictors for breast and ovarian cancers yielded equally baffling findings.

Baggerly began to suspect the worst. Some of the data were so fouled up he couldn't help but wonder if the problems went beyond simple error.

"I must acknowledge that during this process, our ability to comfortably ask questions and receive answers from the group at Duke became hindered," Baggerly said. He said he also had trouble getting journal editors who had published the Duke work to acknowledge his concerns.

Cuffe said Duke officials initially considered the challenges by Baggerly and Coombes to be a scientific spat - a common difference of opinion where one side hews to a certain viewpoint and the other clings to the opposite.

But Baggerly and Coombes were relentless in their criticism. Eventually they published a paper and raised their fears in a newsletter called Cancer Letter.

Baggerly said they heightened their calls for action when they learned patients were being enrolled in studies based on Potti's science.

With patients' safety publicly questioned, Duke halted the trials in November 2009 and enlisted outside reviewers to examine Potti's data.

By then, Shoffner had already been through the round of chemotherapy she blames for her declining health. She said she was told nothing of the investigation and had no idea there were concerns about the trials.

Trials resume

Although Duke's review team acknowledged some errors in the data, which Potti and Nevins said they had corrected, the team found no serious problems with how the experiments were conducted.

As a result, the clinical trials resumed last January.

Cuffe said the reviewers were not asked to comb through all of the underlying data, even though Baggerly and Coombes had recently raised their most serious allegations about its integrity, or "providence."

"The information we had at the time was not about research misconduct," Cuffe said. "It was patient safety, and someone needs to look at the methodology."

He said the reviewers tested the methodology using "the premise that the core data was correct. They didn't go back and check the providence of the data."

(( FC here: Let's look at the premise that the core data was correct. Compare please: we have a building that is toppling over and we investigated each floor, but we did not look at the foundation. ))

But everything changed in July. Allegations surfaced that Potti puffed his credentials - he falsely claimed to have won a Rhodes scholarship amid other academic honors - and Duke officials grew alarmed.

"He appears to have been dishonest in a portion of his professional life," Cuffe said. "That's a pretty rare event. It just gives me chills thinking about it. The ability to misrepresent so grossly in one part of life gives me concerns about misrepresenting in other areas.

"Now I have additional worries. ... I'm no longer confident that the outside review was deep enough."

Misconduct charges

A charge of scientific misconduct is a serious matter governed by federal rules and regulations through the Office of Research Integrity. If fraud is suspected - either through plagiarism, manipulation of data or outright lies - institutions are required to mount a thorough investigation.

The United States is one of only a few countries that has such rules governing scientific integrity. And the process of investigation can take years.

Meanwhile, Potti remained on Duke's payroll until his resignation in November, despite an August finding by university officials that he had inflated his accomplishments.

Cuffe said the misrepresentations appeared on Potti's academic résumé, not on a clinical application he submitted when he applied to Duke as a medical resident. The clinical application was thoroughly vetted, Cuffe said, requiring copies of his diplomas, medical licenses and a criminal background check.

In his resignation in November, Potti took responsibility for "a series of anomalies" in the data. But the scandal still reverberates through scientific circles. Three papers have been retracted or put under review, and more may fall.

Additionally, the misconduct investigation against Potti continues. He could face punishment by the N.C. Medical Board, which licenses and disciplines doctors.

(( FC: What has Duke reported to the Medical Board? The state website indicates nothing to this date. There is also the issue of prosecution: the US Attorney, the District Attorney and the Inspectors General of federal agencies that funded this fake research should be following up ))

Duke's handling of the case also remains under scrutiny, both internally and by the Institute of Medicine.

(( FC There is no guarantee that either investigation how underway will touch the admnistrative blunders in not putting a stop to Potti earlier. See FC next week ))

Cuffe said he hopes some good comes of the ordeal, even if the pace of science is slowed to accommodate the additional examination of the underlying assumptions.

"It's clear additional steps need to be taken," Cuffe said. "We owe it to patients ... as these things move forward, to go back and test the providence of the data."

For Shoffner, that approach is heartening, if late.

"I have been told there was no harm done to us," she said. "But no good was done to us, either. I got all the side effects from chemo and none of the benefit. Duke needs to stand up to this thing and say 'We failed. We didn't check [Potti's] credentials properly; we did not check his papers.'

"They need to step up to the plate and admit that, and come up with a new way of doing things so nothing like this happens again."

(( FC: To participate in Potti's trials, patients agreed to forgo other forms of treatment that might have helped them. As of last Friday, Duke was contending no patients were harmed; stay tuned. ))