Pity poor Nick Gonzalez.
Sorry, I couldn’t resist. After having used the same line when discussing the hugely enjoyable humiliation of the Godfather of HIV/AIDS denialism, Peter Duesberg, I couldn’t resist using the same line to introduce my response to Dr. Gonzalez’s woo-ful whine in response to the publication of the disastrous (for him and any patient unfortunate enough to be in the arm receiving his protocol) clinical trial that demonstrated about as unequivocally as it is possible to demonstrate that his “protocol” to treat pancreatic cancer is nothing more than as steaming and stinking a pile of excrement as the, well, “results” of the twice daily coffee enemas that are a part of his treatment, along with all sorts of raw vegetable juices and 150 supplement pills a day.
The first thing I have to wonder upon seeing this, which several of you sent to me and one or two others posted as comments, is: What took Dr. Gonzalez so long? After all, the Journal of Clinical Oncology article reporting the results of the study of the Gonzalez protocol versus the standard of care at the time, gemcitabine chemotherapy, was published nearly a month ago online. Complete and utter silence reigned; that is, until Kimball Atwood, Steve Novella, and yours truly posted deconstructions of this study and pointed out how it should, if there is any science left in academic medicine (or, if you’re a believer, if there’s a just and righteous God in heaven), be the last nail in the coffin of the misbegotten magical, mystical hodge-podge of woo known as the Gonzalez therapy, which turned out to be worse than useless.
It looks as though ol’ Nick is trying to take a crowbar to the coffin and pry open the cover. He begins, appropriately enough, by trying to throw the principal investigator of the study, Dr. John Chabot, under the bus, just as Chabot threw Gonzalez under the bus by publishing the JCO article in the first place:
Recently, to our astonishment we learned that the Journal of Clinical Oncology, considered to be one of the pre-eminent oncology journals in the country, published an article about our NCI-NIH clinical study which claimed that chemotherapy worked better than our treatment with patients diagnosed with inoperable pancreatic cancer.
Though I originally earned the grant from the NCI in 1998, though I was an investigator on this study throughout its existence, and though the clinical trial was set up to compare the efficacy of my treatment with chemotherapy, no one involved with the publication – not the Principal Investigator, Dr. John Chabot of Columbia Presbyterian Medical Center, nor any of his associates, informed me of their intent to publish this article, nor had I seen it. I learned of it serendipitously when the online version appeared on PubMed.
I suppose it’s possible that Gonzalez was so out of the loop that he was unaware that the paper was about to be published, but I tend to doubt it. He’s clearly known that the study was not going to make him look favorable for quite a while, so much so that he has written a book, and had a chapter pre-written and ready to run outlining all the evils he perceives in the study and how it is so badly designed and run, not to mention totally unfair to him. (He also in essence appears to admit elsewhere in his response that he was the one who sent the dogs after Dr. Chabot.) In response, Gonzalez has produced a huge pile of self-serving twaddle, mainly a lot of playing the martyr and complaining that it’s all a conspiracy to make him look bad in order for the principal investigators to save their academic reputations. Actually, Gonzalez may have a point there that I might even agree with. True, it’s just as self-serving of him to bring up this point, but it’s the one comment he makes in the entire torrent of verbiage in his response and all the linked files that isn’t utter twaddle:
In their official determination letter appearing on their website after a two-year investigation, the Office of Human Research Protections, the NIH agency in charge of investigating mismanagement on government funded studies, found that Dr. Chabot, who was in charge of admissions of patients, had improperly approved 42 out of a total of 62 patients, including 40 for whom he had failed to obtain appropriate written informed consent. Furthermore, the determination letter states that the Principal Investigator (Dr. Chabot) admitted he committed the managerial lapses, and in their letter the OHRP requires Columbia set up a program for training in appropriate research methodology – a serious indictment of a major academic medical center.
To my astonishment, the JCO article nowhere mentions the findings of the OHRP, as if this lengthy investigation never existed, leaving the reader with the impression this study was properly managed by Dr. Chabot. In that regard, the article is a gross misrepresentation of what actually transpired during the study’s sad eight year history.
As glad as I was that the results of this idiotic trial had finally been reported, I did retain a bit of ambivalence about it. The reason is, as Dr. Gonzalez discusses, there is no mention of the ethical and regulatory lapses that plagued the trial. Worse, Dr. Chabot, who was clearly an opportunistic fool to have undertaken a trial that was so clearly unethical from the very beginning, bungled the administration of the trial to the point where the Office of Human Subjects Research Protections investigated and issued a determination letter outlining a litany of mismanagement, failure to obtain proper informed consent, and other problems. Finally, after the trial was stopped because it reached a predefined stopping point due to how poorly patients in the Gonzalez arm were doing compared to those in the chemotherapy arm, Dr. Chabot waited nearly four years to publish the results. It rather makes me wonder what happened earlier this year to prod him to submit the results of the trial for publication after all that time. A stench still lies over this whole misbegotten, unethical mess of a trial, and I don’t like it that Dr. Chabot and his coinvestigators get a publication in a high impact clinical oncology journal like JCO to add to their CVs, even though I think the trial had to be published in a respectable journal in order to make sure that oncologists and other physicians take its results seriously.
That being said, the rest of Dr. Gonzalez’s little hissy fit boils down to a heapin’ helpin’ of special pleading. But first he threatens Chabot with his patron of woo in Congress, Dan Burton, an antivaccine loon who is also most responsible for prodding the NIH to fund this trial of the Gonzalez protocol:
More recently, Congressman Dan Burton of Indiana and I have requested that the Inspector General of the Department of Health and Human Services begin an investigation to determine if the supervisors of the study committed fraud in the mishandling of the project and its data. We have learned, for example, that according to the published medical literature, Dr. Chabot, who as Principal Investigator should have been a completely neutral manager with no ties to either treatment being evaluated, had worked closely with his Columbia colleague developing the very GTX chemotherapy regimen used against us in the study – an obvious conflict of interest that had never been declared to us. We suspect Dr. Chabot believed it was in his best interest to discredit our alternative therapy and instead prove the value of a treatment he helped develop.
Ah, yes, the “pharma shill” gambit. Personally, I had a hard time finding much evidence to back up this charge. A PubMed search of Dr. Chabot’s publications pulled up mostly articles about surgery, with only a couple of articles about neoadjuvant chemotherapy for pancreatic cancer. However, Gonzalez makes a lot of hay observing that partway through the trial the chemotherapy regimen used in the chemotherapy arm changed. The reason, of course, is that chemotherapy regimens were already evolving early during the course of the study. Gemcitabine alone had only resulted in marginal increases in survival; so it was quite reasonable to want to consider adding additional drugs. At the time the design of the trial was changed from a randomized trial to a nonrandomized design, a chemotherapy protocol known as GTX (Gemzar, Taxotere and Xeloda; Gemzar being the trade name for gemcitabine) had largely supplanted single agent gemcitabine protocols. Consequently, since only three patients had been enrolled, it made sense to change the chemotherapy arm to what was being given at the time at Columbia. Whether this change muddied up the trial (which it probably didn’t; the trial was muddied up enough to begin with) or not, it’s all a smokescreen thrown up by Gonzalez to distract attention from the fact that his therapy did no better than, in essence, untreated pancreatic cancer.
It’s also rather illustrative of Dr. Gonzalez’s “us against them” thinking for him to view someone as having a hopeless conflict of interest if he’s ever studied chemotherapy before. Here’s a clue: It is not a reportable “conflict of interest” to have studied before one modality that you’re studying in a clinical trial now unless there’s a financial interest in that modality. Academic surgeons and physicians study different drugs or treatments all the time. Just because they’ve studied one regimen doesn’t make them hopelessly biased to the point where they are ineligible to head a clinical trial of that regimen against anything else. It would be one thing if Dr. Chabot had expressed unrelenting hostility to the Gonzalez protocol before, but he didn’t. In fact, he put his reputation on the line to head up this study. Also, in marked contrast to Gonzalez’s complaint, Dr. Chabot himself clearly knew CAM-speak pretty well, although he was unhappy over the change of the trial to a nonrandomized design.
Be that as it may, the main strategy for complaining about the clinical trial utilized by Gonzalez is special pleading. Before I get to that, note how vociferously Gonzalez complains about Chabot’s referring patients he considered inappropriate for his trial. For example, Gonzalez points out that the patients who were to undergo the “nutritional” arm of the trial (a.k.a. the woo arm) had to be able to “eat normally.” Well, that’s cherry picking the best patients right there, because few patients with advanced pancreatic cancer can eat normally. I remember well Dr. Gonzalez’s discussion of his initial series of 11 patients who underwent his protocol. He argued again and again that the long survival of these patients compared to historical controls could not be explained by selection bias, but in essence right here he is admitting that he relied on selection bias for his results. He even admits this later when he laments that the chemotherapy protocol, because chemotherapy was given intermittently and could thus be easily given to patients who couldn’t eat while his regimen requires 150 pills a day and that even patients too ill to eat could receive the drugs in the chemotherapy regimen. Does Gonzalez realize that he’s basically saying that chemotherapy can be given to sicker patients and that the only patients who can do his protocol are the patients who are in the best shape and thus most likely to live the longest, regardless of therapy? it’s pure selection bias.
Gonzalez also complains ad nauseam that patients in the nutritional arm were not adequately screened for ability and motivation to follow the protocol. However, if one wants to avoid bias creeping into a trial, all patients would have to be screened using exactly the same criteria, regardless of which group they entered. That’s really hard to do with a trial in which patients can choose which arm of the study they want to be in. The patients choosing the chemotherapy arm would quite reasonably ask why they should be screened for the Gonzalez protocol, and screening too closely those choosing the Gonzalez protocol would allow the very cherry picking of the least debilitated patients that must be avoided. Of course, part of me wonders whether investigators intentionally cut Gonzalez out of the patient qualification and selection process for this trial because they knew he’d try to cherry pick the best patients. I also note that self-selection would similarly tend to funnel the least debilitated and most motivated patients to the nutritional arm, which would in turn tend to mean that the patients most likely to survive the longest would be most likely to end up in that arm. In essence, you’d expect that there would be an apparent survival advantage in the nutritional for that reason alone, but the results of the study were exactly the opposite–resoundingly so. Whatever shortcomings there were in the design and administration of the trial, they were not enough to explain why the patients in the nutritional arm had a median survival of only 4.3 months, in essence the expected survival of patients with untreated advanced pancreatic cancer. Robbed of his ability to pick the best patients, Gonzalez’s results were no better than no treatment at all, and certainly not the equal of chemotherapy.
Gonzalez reached his zenith of disingenuousness here:
Clinical trials lacking a lead-in period often – though not always – adopt an “intent-to-treat” format. With such a provision, researchers agree that all patients qualified and entered into the study for any of the treatments under scrutiny will be considered as having been treated, regardless if they actually proceed with the prescribed therapy or not. Though such an approach on first glance might not make much sense, researchers justify such an “intent-to-treat” rule as necessary to evaluate fully a new drug. For example, if in a study 100 patients receive some new medication but 50 drop out after a week because of serious side effects, certainly it would seem prudent to include these patients as treatment failures rather than discount them, since they quit because of some negative reaction to the drug. On the other hand, such a design can be disastrous for a lifestyle intervention trial such as ours, since patients who might initially be enthusiastic but who can’t or choose not to proceed with the self-administered dietary/nutritional regimen will be counted as having been fully treated.
In essence, Gonzalez is engaging in special pleading here. He is saying that the normal guidelines for what constitutes good clinical research shouldn’t apply to his protocol. Intent-to-treat analyses are very important because if a patient stops a treatment it can be because of disease progression or because the treatment is toxic or difficult. Either way, it’s important to know. Gonzalez seems to think that “lifestyle interventions” should be exempt from such an analysis for…no reason whatsoever. Excluding patients who couldn’t make it through Gonzalez’s protocol, which is, as has been pointed out before, quite onerous, would introduce bias in that the more debilitated patients, who couldn’t swallow 150 pills a day, along with the raw juices and various other dietary woo, and undergo coffee enemas twice a day, would be excluded, leaving patients in better shape for analysis.
I was also heartened, believe it or not to read that not a single oncologist referred a patient to the trial. This is truly good news because it tells me that there are actually still oncologists left in New York who haven’t bought into CAM:
Ultimately, only the oncology team at Columbia cooperated in any way only after much prodding by Dr. Antman and Chabot, and only for the admission of chemotherapy subjects to form the comparison “control” arm as we shall see. Even for this group their referrals proved not helpful.
Oncologists not only refused to refer patients to the trial, but at times actively discouraged their patients who might express an interest from seeking entry. A number of candidates suitable for the study who had learned about our treatment on their own informed our office that their oncologist had strongly argued against considering the project. One well-known Memorial oncologist warned a candidate interested in joining the study that I was a “quack” and the study a “fraud.”
Oncologists also frequently discouraged patients who actually entered the nutritional arm of the study from continuing with the prescribed regimen.
I’d love to know who that well-known Memorial Sloan Kettering Cancer Center oncologist was who called Gonzalez a quack and his trial a fraud, as I’d love to take him out to dinner and shake his hand. He called it exactly right, in my opinion. I’m also heartened that not even the oncology team at Columbia wanted anything to do with referring patients to this trial. It shows that there is at least some sanity at that institution. But it wasn’t just that oncologist at MSKCC. There are a lot of oncologists like him:
Unfortunately, a protocol provision against which we argued and that ultimately caused enormous damage required that each patient assigned to the nutrition arm consult with a physician monthly for an examination and blood work. On the surface, such visits would hardly seem to be the source of potential catastrophe, since, one might think, how can a visit with a doctor be a problem? And trials involving chemotherapy drugs often require frequent physician assessments to monitor closely the toxic side effects of the medications being tested, such as severe anemia or immune suppression.
For those subjects who lived in the New York area, Dr. Isaacs and I could satisfy this rule by meeting with the patient ourselves monthly. We had no problem with such an arrangement, of course. But as it turned out, only three of the patients ultimately entered into the nutrition arm lived in the New York area, with the great majority residing at great distance. Consequently, nearly all subjects assigned to us for treatment were followed by a local doctor, most frequently an oncologist completely unfamiliar with our treatment approach and usually hostile toward it, with only a few exceptions.
Repeatedly, we heard from our patients that during the required monthly meetings, the local physicians aggressively discouraged them from continuing their treatment with us, instead urging them to proceed with some standard approach – despite the fact that the conventional therapies for inoperable pancreatic cancer have proven largely worthless.
Help, help, I’m being repressed!
Put yourself in the position of one of those oncologists. What would you do? You took an oath to do your best for patients. Your training correctly tells you that this ridiculous regimen advocated by Gonzalez is based on no science and indeed so incredibly unlikely to do any good that medical ethics demands that you try very hard to persuade your patients not to engage in a course of action that your professional knowledge and understanding of science tell you to be harmful. Make no mistake, even if the Gonzalez protocol did not hasten the deterioration of the patient, it put the patient through hell for no benefit. Dr. Gonzalez scoffs at doctors who pointed out to patients who chose the Gonzalez protocol that they were choosing to spend the last months of their lives following a restricted diet, swallowing 130-170 pills a day, and subjecting themselves to coffee enemas, a protocol that couldn’t possibly help their disease, instead of enjoying themselves as much as they could with pizza and ice cream. I find nothing to criticize these doctors for; they were absolutely correct.
Finally, the most disturbing part of Dr. Gonzalez’s defense of his protocol is his admission to something I alluded to as a possibility in my previous post, namely that perhaps there were differences in palliative care between the groups. One of the most common causes of death from pancreatic cancer is biliary sepsis, namely infection of the backed up bile that accumulates behind the bile duct obstruction caused by the cancer. That’s why biliary obstruction is treated aggressively by drainage with stents, which can be placed endoscopically via the stomach and duodenum or through the skin directly into the main bile ducts in the liver. Infections need to be treated aggressively. Failure to do so can result in a patient dying even sooner than he had to.
Guess what? Gonzalez in essence admits that there were huge differences in supportive care between the two arms of the trial:
Of all the nutrition patients, only one – who ultimately survived 3.5 years – received anywhere near the level of intensive supportive care and encouragement given the chemotherapy patients. In this unique situation, the local doctors coordinated their treatment with me, realizing full well that he was sustaining a most unusual response. In no other case did the local doctors encourage aggressive intervention to keep the patients alive and also on the nutritional therapy.
Dr. Gonzalez just admitted a horrific lapse in clinical trial ethics. This lapse is not just his fault but the fault of each and every investigator in the trial. That both groups did not have access to the same level of palliative care is criminal–yes, criminal. The patients in the Gonzalez arm were condemned to suffer symptoms of progressive pancreatic cancer that were not treated with the latest and most aggressive palliative care: stents, antibiotics, laparoscopic gastrojejunostomy to bypas gastric outlet obstruction. If the investigators were unwilling or unable to make sure that patients in both arms had equal access to palliative care, then the trial should have been shut down until this glaring problem could be fixed. If the investigators couldn’t find a way to fix this disparity between groups, then the trial should have been scrapped. The reason? Simple. Medical ethics demands it. For example, the Helsinki Declaration, the international agreement governing human subjects research, which states, “In medical research involving human subjects, the well-being of the individual research subject must take precedence over all other interests.” The Belmont Report, the guiding document for medical research in the U.S. states: “In this document, beneficence is understood in a stronger sense, as an obligation. Two general rules have been formulated as complementary expressions of beneficent actions in this sense: (1) do not harm and (2) maximize possible benefits and minimize possible harms.”
Gonzalez, while trying to cover his tail, just admitted that this trial failed to maximize possible benefits and minimize possible harms–and failed miserably.
In a way, it’s fun to watch the flurry of charges and countercharges flying fast and furious back and forth between Dr. Chabot and Columbia University on the one side and Dr. Gonzalez on the other. However, we should never forget one thing, namely who suffered because of this trial. In the name of testing a ridiculously implausible “alternative medicine” therapy and an open-mindedness so wide that the investigators’ brains fell completely out, patients with a terminal illness were denied therapy that would have palliated their suffering. As much schadenfreude as I feel for Gonzalez’s discomfiture and frustration that Dr. Chabot managed to notch another publication in a high impact journal with apparently no harm to his career from his career, remember that it’s not about Gonzalez or Chabot or any other investigator. It’s about the patients with pancreatic cancer who were harmed in this study, which I view as the most unethical study done since the Tuskegee syphilis study. Never forget that as you’re buried in self-serving twaddle.