I am grateful to Arie Brand for this, admittedly, rather long blog posting. He now lives in Australia, and shares my passion in medical research – and how it is often, far too often, twisted for non-scientific reasons. He was stimulated to write this article in response to my blog on ‘losing faith’ with medical research.
For many years Arie has looked at the work of Linus Pauling – double Nobel prize winner, and genius. In his later life Pauling turned his attention to vitamin C, and the benefits that it could have in heart disease and cancer.
I have written on the subject of Pauling’s many times, and feel that he, and the infamous Matthias Rath, were well on the way to making some important discoveries. However, for various reasons, the mainstream research world turned on Pauling, and his work in this area.
Anyway, here is a tale of how Linus Pauling’s work was undermined, discredited and , eventually, dumped. I hope you like it. It is long, and it is packed with facts, but it represents a forensic attempt to get at the truth. Something which I completely support.
‘A lie gets halfway around the world before the truth has a chance to get its pants on.’ Winston Churchill
1. How to discredit unwelcome research
As far as losing faith in medical research is concerned I have got to tell a tale as well.
A few years ago a nephew of mine who is involved in cancer research in the Netherlands sent me a newspaper column by an eminent Dutch oncologist, an emeritus Professor Piet Borst, in which this worthy spoke in a rather scathing tone about Linus Pauling’s preoccupation with vitamin C, and particularly Vitamin C and cancer.
The general thrust of his little article was, yes, of course, this double Nobel laureate was a brilliant man but as far as vitamin C was concerned he had a screw loose. In the case of cancer, the professor thought that this had been proven through work done by ’ bona fide’ oncologists which appeared to contradict the results of Dr. Cameron. Dr Cameron had done some very positive studies on the use of vitamin C in cancer in the Scottish Vale of Leven hospital – a treatment in which Linus Pauling had been Dr.Cameron’s intellectual ally.
Now Pauling (according to the journal New Scientist one of the ten greatest scientists who ever lived) has been rather a hero of mine for many years. Starting when I managed to cope with the rather rough climate at Aberdeen many years ago, thanks to his advice on vitamin C (or so I thought). Coincidentally, I had been appointed at Dr.Kendrick’s alma mater, Aberdeen University. Because of my personal interest in Pauling’s work, I went a bit deeper into the matter.
The essence of research in the natural sciences, including medicine, is to come up with results that can be replicated. But what if this replication is a sham designed to discredit the original inquiry? I have the feeling that this is what happened with Dr.Cameron’s (and Pauling’s) results.
There is a short summary of Dr.Cameron’s research here:
After achieving his rather remarkable results naturally he, and even more so Pauling, looked for a bona fide replication of the Vale of Leven tests by a seemingly independent research institute. It took Pauling a lot of effort to persuade the American National Cancer Institute to arrange for this (he even managed to involve President Carter). The arrangement was finally made, through the Mayo clinic. However, it is clear there never was genuine replication.
The article I have linked to quotes part of Pauling’s and Cameron’s own opinion on why this was the case. But this does not provide the whole story. In particular, it does not address the question whether one can justifiably suspect that the Mayo Institute, never intended to come up with a genuine replication in the first place.
2 The first Mayo test
Before the Mayo test started its chief investigator, the late Dr.Charles Moertel, had an exchange of letters with Pauling, in which the former claimed that he would try to replicate Dr.Cameron’s tests as exactly as possible. Pauling wrote him in reply on the 9th August 1978: “In my last letter to you I pointed out to you that the patients studied by Dr.Cameron had not received chemotherapy. The cytotoxic drugs damaged the body’s protective mechanisms, and vitamin C probably functions largely by potentiating these mechanisms. Accordingly, if you hope, as you stated in your letter, to repeat the work of Cameron as closely as possible, you should be careful to use only patients who have not received chemotherapy… On page 2 (of your letter) there is no mention of earlier chemotherapeutic treatment as a contraindication for patient eligibility.’
I think that this question might be important however, and I recommend that you be sure that there are a sufficient number of patients enrolled who have not received chemotherapy. Otherwise the trial cannot be described as repeating the work of Cameron.”
Moertel answered to this: “Certainly in any presentation of this (sic) data I can assure you we will call attention to the fact that the majority of our patients had had prior chemotherapy, whereas in the study conducted by you and Dr.Cameron it was clearly stated that none of the patients had had prior chemotherapy.” 1
In the first report on their research Moertel and his associates do indeed clearly indicate this (New England Journal of Medicine 1979 Vol 301 No 13). They wrote: “It should be noted, however, that only nine out of our 123 patients had not previously received chemotherapy or radiation therapy.
It is therefore impossible to draw any conclusions about the possible effectiveness of vitamin C in previously untreated patients. In Cameron and Campbell’s report of a 10 per cent regression rate in 50 patients with widely disseminated cancer, none had received definitive prior treatment and presumably were more immunocompetent than our patients.”
At this point, one can justifiably ask the question why Moertel and his associates started on this bit of research in the first place because the supposed intention was that they would replicate Cameron’s tests. So, understandably, Pauling protested against the title of their report in the New England Journal, which was “Failure of high-dose vitamin C (ascorbic acid) therapy to benefit patients with advanced cancer”. They should have indicated more clearly, he thought, that the researchers had been dealing here mainly with patients who had already undergone chemotherapy.
There was, however, another marked difference with the original tests by Cameron. In the Vale of Leven hospital the patients involved in his test received their vitamin C dose intravenously during the first ten days. This did not happen in the Mayo-test.
So it was clear, as Moertel himself acknowledged, that the Vale of Leven tests had not been replicated. Pauling kept urging therefore for a second test that was finally done, years later. But this only made matters worse. Not only was Cameron’s test again not replicated but also there were now grounds, given Moertel’s research design and his various statements and actions, to entertain doubt about his good faith.
3. The second Mayo test – Again no replication
At this point I will attempt to answer three questions:
- Were the second Mayo tests an exact replication of the Vale of Leven tests?
- Was the research strategy followed by Mayo in this second test “methodologically sound”- as sound as claimed in a guest editorial in the relevant issue of the NEJM by a certain Dr.Robert Wittes, then Associate Director of the Cancer Therapy Evaluation Program of the NCI
- Are there reasons for doubting the good faith of Dr.Moertel and his associates?
Replication of the Vale of Leven tests?
Pauling and Cameron had never claimed that vitamin C could cure cancer. Their claim was far more modest. They claimed, and thought they had proved, that it could lengthen the survival time of patients deemed incurable, and enhance whatever remained of their quality of life.
Neither did Pauling and Cameron claim that vitamin C could stop the growth of the tumor, only that it could possibly delay it. For these reasons Cameron gave vitamin C to his patients until the very end (or in some cases cure).
Though they had made this very clear in the articles they published about it in the Proceedings of the National Academy of Sciences, the Mayo team followed a different strategy. It gave patients vitamin C until these were no longer able to take this orally, or until it became clear that the cancerous tumor kept growing. So they dealt with vitamin C as if it were a cytotoxic drug that should no longer be administered after it became clear that it was not preventing tumour growth, rather than an ordinary nutritient that has no dangerous side effects even at high doses. Pauling objected to this.
The Mayo team’s haste in stopping with vitamin C was strange for another reason. They wrote in their 1985 article in the NEJM:
“For this trial we elected to study patients with advanced cancer of the large bowel because this was the most frequent tumor type in the study of Cameron and Pauling and one for which they claimed a striking improvement in survival with vitamin C therapy. We felt ethically justified in studying this group of patients without first offering cytotoxic drugs because in our opinion there is no known form of chemotherapy for colorectal cancer that has been demonstrated to produce substantive palliative benefit or extension of survival.”
So why then this haste to stop administering a relatively innocuous nutritional element. They had, on their own testimony, no other effective medication to use. There was, moreover, a strange contrast with their first test. In their 1979 NEJM article about this test they wrote:
“Treatment was continued until death or until the patient was no longer able to take medications by mouth.”
It is true that they were dealing here with patients who had already undergone chemotherapy but in reality this didn’t make any difference – for their new test group chemotherapy had been deemed to be useless anyway. So either way there was no alternative course of action for which vitamin C had to be stopped.
Once again, Cameron’s test was not replicated.
4. Sub B. Mayo’s methodology – Double Blind?
Moertel et al’s article 1985 NEJM article about the second test is only five pages long. However, we are assured exactly nine times by the use of one or all of the following adjectives that we are dealing here with a “prospective randomized double blind” study – a fact designed to impress us with the great methodological superiority of their research strategy above that of Cameron and Pauling (who took as their control group cases that matched from the files).
But I have my doubts about that “double blind” character, and also about the way the control group was controlled – or rather not controlled. Here was a group of patients that, in all likelihood, was told that chemotherapy offered no hope in their situation. They were also told explicitly that there would be a test with vitamin C. Moertel et al. wrote:
“The nature and purpose of the study was fully discussed with each patient, with specific emphasis on the fact that they would receive either vitamin C or an inert placebo on the basis of random assignment.”
So, patients whose hope for any other treatment had disappeared were told here that there would be a test with a potential medicine that can be easily obtained by anyone, and that could be easily self-administered because in the test that would only be done orally anyway.
They also knew that they would either be in the test- or in the placebo group (which stops the test from being double blind). Wouldn’t at least some of those people think, to insure myself against the chance of merely being in the placebo group I will get hold of that stuff myself.
That at least some of them would think so would seem fairly obvious. Which meant that strict monitoring of the control/placebo group have been a prime requirement. There were only about fifty people in this group, so control should not have been a major task. But the test was deficient on this point as well.
Moertel et al. wrote:
“To further ensure compliance, 11 consecutive patients were selected during the course of our study for urinary assays of ascorbate … Patient selection was made at an arbitrarily chosen time in our study and without knowledge of the drug assignment of individual patients. The laboratory was also blinded as to drug assignment. Patients were not told the purpose of the urine collection; they were simply asked to submit a 24-hour specimen. Five patients who were later determined to have been assigned vitamin C all had high urine levels of ascorbate (more than 2 g per 24 hours.). Of the six assigned to placebo, five had negligible levels that were within the range of normal controls for our assay method (less than 0.55 g per 24 hours). A single patient had an intermediate value between these two ranges.”
I conclude from this that this test was only done once, and that it was done on urine (whereas Dr.Cameron directly measured the serum level of vitamin C). There was also a patient in the control group who was above the (arbitrarily chosen) upper limit of more than 0.55 g.vitamin C per hour whereas another patient in that small test group of 11 people was at that upper limit. No further information has been provided about measurements among the others.
5. Sub B. (contin.) No effective control of the control group
The figures clearly demonstrate that at least part of the control group were self-administrating vitamin C, as any one with common sense could have expected. Unless they were taking extra vitamin C, cancer patients should not have more than 0-10 mg in their urine (Richards). But here the upper limit for non-administration was supposed to be 550 mg.
Wikipedia provides (or provided) the following substantive criticism:
“ … the dose given was oral, not intravenous (uptake of vitamin C from oral dosage is very low), of short duration (averaging 72 days rather than lifelong, roughly 2 years more) stopped the vitamin C abruptly (generally bad) after which death rates rose, used a different oral form (dry AA caps vs neutralized AA-DHA-sorbitol solution) and … the Mayo patients died after stoppage of vitamin C usually on or after chemotherapy.
The ascorbate monitoring was poor and compliance controls of outpatient treatment were flawed, despite Ewan Cameron’s specific request to Moertel for better monitoring….] Moertel (1985) tested only 6 of 49 placebo patients for a very high threshold of urinary residuals of ascorbate spillage, over 550 mg/day. This urinary threshold might roughly correspond to 2g-4g/day in healthy patients or about 8+g/day in less severe cancer cases vs ~0.06g/day intake (RDA then) assumed with a placebo. Still one of only six placebo “controls” measured higher than this threshold, belatedly dismissed as only a potential measurement interference. In one of Klenner’s rare comments on cancer, intravenous administration of 17g/day ascorbate for 92 days yielded no measurable urinary residual in a severe cancer case. The Moertel (1985) tests were poorly analyzed in terms of radically changing test conditions (initial delays, brief vitamin C; abrupt stop & repeated chemo) and quality of life improvements …”
So this whole exercise was problematic: “a proportion of the controls was clearly medicating with the substance under evaluation” and this whereas Dr.Cameron had urged effective monitoring of this control group.
This was even obvious to the guest editorialist Dr. Robert Wittes, who was, as we saw, was keen to give this study his imprimatur as a “prospective randomized double blind study”. He wrote that, if even more patients had been controlled “surreptitious ingestion of this easily obtainable vitamin among the control patients might have been excluded with somewhat more confidence”.
All the above returns us to the question as to whether one was dealing here with fools rather than knaves. Honest errors, as Dr. Cameron was inclined to think, or rather with a deliberate effort to make this test fail, as Linus Pauling was inclined to think. An opinion Dr.Cameron eventually shared. Was Pauling’s suspicion about the lack of good faith here justified? That is the question I will deal with next
6. Good faith?
Moertel et al wrote a propos of their second test:
“We had previously attempted to validate the results of Cameron and Pauling in a prospective randomized trial that was double blinded to prevent any inadvertent bias. We selected patients for study according to the published criteria of Cameron and Pauling: i.e.all the patients had proven terminal cancer and “all were treated initially in a perfectly conventional manner by operation, use of radiation therapy, and administration of hormones and cytotoxic substances” “.
The bit between quotation marks is supposed to come from Pauling’s and Cameron’s first article (1976) in the Proceedings of the National Academy of Sciences. But let us look again at what Moertel wrote to Pauling a propos of his and Cameron’s 1976 article:
“Certainly in any presentation of this (sic) data I can assure you we will call attention to the fact that the majority of our patients had had prior chemotherapy, whereas in the study conducted by you and Dr.Cameron it was clearly stated that none of the patients had had prior chemotherapy.”
And in their 1979 – NEJM artikel Moertel et al. wrote as I quoted earlier:
“In Cameron and Campbell’s report of a 10 per cent regression rate in 50 patients with widely disseminated cancer, none had received prior treatment and presumably were more immunocompetent than our patients”
It is therefore rather strange, to put it mildly, that six years later they asserted that in that earlier test they had included people who had already received chemotherapy in the group to be tested. And that this had been guided by the “published criteria of Cameron and Pauling”.
Had they forgotten their earlier statements, or did they not believe that anybody would go back to the earlier issue of the NEJM to check up (no articles online then).
Moreover there is a snake hidden in the grass of that alleged quote from Cameron and Pauling. We saw that Moertel quoted them as follows: ”all were treated initially in a perfectly conventional manner by operation, use of radiation therapy and administration of hormones AND cytotoxic substances.” (emphasis added AB)
The suggestion is here that that “conventional manner” of Cameron always included the use of cytotoxic substances (= chemotherapy). However, Cameron and Pauling had written “OR cytotoxic substances”. Is this replacement of “OR’ by “AND” due to a simple oversight? Perhaps.
What Moertel et al. were, however, also silent about is that, immediately following that passage about that conventional treatment, Cameron and Pauling gave a few examples that make clear that virtually nobody in their test group had undergone chemotherapy. They wrote:
“For example, all of 11 breast-cancer patients in the ascorbate-treated group, with the exception of one who first presented in a grossly advanced state, had already had mastectomy and radiotherapy and all, including the exception, had been given hormones, sometimes with considerable benefit; but all had relapsed by the time ascorbate supplementation was commenced, and it seemed clear that their tumors were escaping from hormonal control. Similarly, all of the seven bladder-cancer patients in the ascorbate-treated group, with one exception because of her frailty, had received megavoltage irradiation and several had had a partial cystectomy (one total) before ascorbate treatment was commenced when it seemed that these standard procedures had failed.”
Thus none of the patients referred to here had undergone chemotherapy. Chemotherapy in the treatment of cancer patients was apparently then comparatively rare in Scotland – compare Richard’s remark about the vitamin C strategy:
“ Its therapeutic promise, as Cameron and Pauling initially represented it, was that it would not replace or threaten conventional treatments but actually improve their efficacy by strengthening the intrinsic defence mechanisms of the patient and containing the spread of the tumour … This strategy, however, while appropriate to the context of conventional Scottish methods of cancer treatment where surgery and radiotherapy were the established therapies, did not, as we shall see, translate readily to the American medical context, where cytotoxic chemotherapy had become the norm in conventional cancer treatment”
Moertel et al. do mention that Pauling had mentioned in the retrospective discussion that only four persons in their test group had undergone chemotherapy, but they only did so after they had suggested misleadingly that they, in their choice of persons for the test group, had been led by the “published criteria” of Pauling and Cameron.
7. More dishonesty
We are dealing with another example of dishonesty when Moertel et al: emphasize the virtues of their “prospective randomized double blinded” test. They wrote in their 1985- NEJM article:
“Randomization and double blinding served to protect against any possible conscious or unconscious bias on the part of the investigators as patients were selected for treatment assignment and as their results were evaluated. There was no such protection against bias for Cameron and Pauling as they selected and then reselected the patients they decided to evaluate for their first and second reports” (emphasis added by author).
It is however quite clear from Cameron and Pauling’s 1976-PNAS article that their criterion for inclusion in the experimental group was “terminal” cancer – that is cancer incurable with the conventional therapies then found in Scotland. Whether patients had indeed reached this terminal stage was decided by ”at least two independent clinicians”. Cameron and Pauling wrote:
“Fifty of the treated subjects are those described in ref.4 and the other 50 were obtained by random selection from the alphabetical index of ascorbate treated patients in Vale of Leven District General Hospital …”
Unfortunately ref.4 refers to an earlier article of these authors that I haven’t been able to track down.
The control group of one thousand patients was obtained by “a random search of the case record index” in the hospital of similar patients – similar in terms of sex, of approximate age, cancer in the same organ and histological type of tumor. The services of a medico who had no further involvement in the research was obtained to establish survival time since the patients had been declared to be “terminal”.
Thus when Moertel et al. wrote “they” in the quote I gave above they falsely suggested that only Cameron and Pauling were involved in this selection. And what is also conspicuous here is that Moertel et al.go on, from their high horse, about their own excellent “randomized sample” but in their report on how this randomization had taken place they do not get any further than saying:
“All the patients selected for study had histological proof of advanced adenocarcinoma established as having originated in the colon or rectum. All were beyond any reasonable hope of potentially curative surgery or radiation therapy. All were ambulatory and capable of taking oral medication. None had received any chemotherapy.”
But who actually established that these patients were “beyond any reasonable hope”? And if the total number of such patients was greater than the ultimate vitamin C and placebo group (it is not very likely that their number was exactly one hundred) how then took the selection from this larger number place? Moertel et al. keep up a discrete silence about such questions.
Let us see whether there are more reasons to doubt the good faith of these researchers. Pauling asked Moertel in writing to be sent the manuscript with the report on the second Mayo test a few weeks before publication so that he could prepare himself for questions by the media. What actually happened was according to the academic historian of science Richards the following. She writes in her book:
“ On 30 November Moertel confirmed in a letter to Pauling that the paper had indeed been submitted to the New England Journal of Medicine. He undertook to make certain that a true copy of the manuscript would be in Pauling’s hands a short time before publication date. But it was the press who got the pre-publication copies first, three days before Pauling received his on the actual date of publication – 17 january 1985.”
Feelings ran high at the Institute, with Pauling and Cameron besieged by reporters. Pauling was outraged by what he construed (and denounced in the media) as Moertel’s unprofessional conduct in deliberately withholding the paper from him and making it impossible for him to comment on a study he had not seen. By the time Cameron managed to obtain a copy from a United Press International reporter (on 16 January), the news was already going stale. Moertel had appeared on all major TV networks to denounce vitamin C as “worthless” in cancer treatment and the Pauling-Cameron studies as biased.” Pauling wrote letters with questions to all of the six authors – nobody answered him. Subsequently he wrote to the editor of the NEJM asking information about the way this article had been accepted by the journal. That letter also remained unanswered 2.
Moertel can no longer be called to account. He died in 1994 – in the same year as Pauling and just like Pauling from cancer. But Pauling was 93 when he died and Moertel 66. Pauling’s cancer was, moreover, in the prostate, an organ in which almost any man can get cancer if only he gets old enough.
1: see Pauling’s letter in the New England Journal of Medicine 1980 Vol.302 No.12.
2: Richards, Vitamin C and cancer – medicine or politics?, 1991:142