Back to Index

ABOUT ANECDOTE  

The results of alternative cancer clinics  bear little relationship to the claims that draw patients to them.   We know this from their own published results  (see here),  numerous  formal studies and enquiries, and the very sparseness of the anecdotal material (testimonial,  case reports) they are able to produce.     

Yet the promoters of the methods,   supporters and some clients maintain great faith in them on the basis of the anecdotal material.    They say "we have  seen things with our own eyes" .  Conflicting opinion is attributed to conspiracy, or mindless bias.  Or  perhaps there is something wrong with the way scientists are looking at all the evidence.   

Cancer patients are caught between the opposing viewpoints.  They may be unsure what to think when also badly wanting answers to a potentially mortal illness.   Can we help them better understand it all?    I shall try.  Later we'll look at some better quality anecdotal material, of the type that can  sometimes have sceptics scratching their heads.   Why is that not convincing doctors, government institutions or  health insurers  that alternative methods are regularly curing cancer?   

Firstly,  some  clarification of the role that anecdotal evidence can play in advancing medical knowledge.    I happen to agree with those who think that doctors can sometimes be a little arrogant and heavy handed in their "we know best" dealings with dubious methods (without necessarily being far off in their judgements).

Is "it's only anecdotal evidence" a sufficient reason in itself  to dismiss stories of cancer cures?

The answer,  surely, is definitely not.  A great many important medical advances originated with  a few anecdotal observations.  You only have to think of milkmaids rumoured to be immune to smallpox, or the old woman apparently curing oedema with the foxglove (digitalis). 

Why, then, is anecdotal evidence generally looked down upon?   Well,  frankly, a lot of it is of little value, lying at the lowest possible extreme of evidence worthiness.   Many cancer cures turn out to have no more substance than "I heard somewhere ----".     Most personal testimonials don't hold water even on the information supplied  (See a deceptive breast cancer testimonial  with further  examples in How to Read a Cancer Testimonial. 

Such evidence is deservedly not given much weight,   not because it belongs to an ill-defined category of evidence called "anecdotal",  but because it is obviously not reliable.   We should, however,  be prepared to explain  why that is so.   Dismissing evidence that others find highly significant without offering any reason other than that it is anecdotal,  can raise  hackles, end dialogue, and no one gets  to learn anything.    Inadequate airing of the precise grounds for rejecting anecdotal evidence  may be partly why disputes concerning alternative cancer treatments look much the same now as they did fifty years ago, and, sadly,  often concern the very same treatments.   Allegations of insuperable bias and conspiracy against alternative cancer treatments are surely fuelled by this knowledge gap between the parties.  I hope to show here how apparently overwhelming anecdotal material can paradoxically only serve to raise  doubt.

Another reason for a generally dim view of anecdote is that in very many medical contexts it  IS virtually useless as evidence of a treatment effect.    Some medical conditions, for example anxiety and depression,  can get better by themselves or  fluctuate wildly naturally, and are subject to somewhat arbitrary patient judgements such as "how am I feeling today as compared to three weeks ago?".  It has been proved repeatedly that when given an inert or sham treatment ("placebo") up to fifty per cent of patients with such conditions will report feeling better later, even within the drab, uninspiring environment of the typical clinical trial.       

 Anecdotal treatment successes with such conditions can assume immense personal significance for some patients and practitioners, especially when resonating with other personal beliefs.   But tossing a coin or throwing darts at something will be as reliable a way of directing us towards better treatments,  especially once the influence of  personal biases is allowed for.    Separating true treatment activity from the spurious or from placebo-related phenomena is very difficult with this kind of clinical material..   It requires sophisticated study design, perfect execution, and even then, in the end, it depends upon rule-of-thumb statistical judgements.   

Cancer is completely different.     Most cancerous states are easily measured.  Some are so predictable,  regardless of any placebo usage,  that as many as 99.999% of them will progress to a sad conclusion without effective treatment.    So long as the diagnosis seemed secure  why on earth would  well-documented and verifiable reports of cure or  remission with this kind of cancer ever be dismissed as "just anecdote"?  

They shouldn't be and wouldn't be.  Nevertheless, even with anecdotal data of adequate quality,  occurring in an appropriate  clinical setting,  there is still  the need for a minimum quantity.   A single apparent cure of a bad cancer might  spark some interest, but if attributed to a  treatment that otherwise has no track record  of ever curing cancer,   doctors would probably wait to see if a second case can be produced before letting themselves get too excited    The isolated case might yet prove to be a so-called spontaneous remission, or a rare misdiagnosis, or as yet unknown factors,  or even an entirely different treatment that the patient forgot to mention when multiple measures are being applied, as is usual in alternative medical practice..  

In alternative cancer practice most patients use multiple treatments, sometimes dozens of them.   This is a further problem for those saying that it was their specific treatment  that cured a cancer.  It is not an insuperable problem, so long as they are able to do it again, and preferably yet again.     I elsewhere challenge  any alternative agency  to produce even two or three cases with the desired qualities using any combination of alternative methods they wish.     They should be able to produce them by the dozen, if some prevalent claims were true.

 We have arrived at a rough description of what good anecdotal evidence for a cancer cure might look like    It would include patients with definite, measurable and normally very predictable cancer states.   The cancer would be shown to regress or disappear with the treatment, as demonstrated by photography, x-ray, scans, or other reliable tests.    This happens to be precisely what doctors look for when testing a potential new cancer treatment.   All known active treatments against cancer can do this.  Complicated clinical trials only come later, when treatments of known effectiveness have to be compared,  or when measures directed purely at symptom relief,  or at producing less obvious effects such as prolonging survival are being assessed.      

I have also suggested that rejecting anecdotal evidence without further explanation  can be an unsatisfactory response to dubious  claims.   Let us now look at an example of some better quality anecdotal material.

 Some quality anecdotal evidence

Dr Seymour Brenner MD presented the following five cases to the American Office of Technology Assessment   when that body was looking for alternative cancer treatments worthy of further scientific assessment [1].    Brenner was a very experienced radiation oncologist (radiotherapist) and was very  impressed by some of the results of the Revici treatment.    

This is fairly typical of the best that believers are seeing " with their very own eyes".   There is the added bonus of endorsement by an experienced oncologist.  Read them carefully. 

These patients all supposedly had serious, even "incurable" cancer,  and Revici made it go away.  At first sight that is what the cases seem to show.  But they don't -- not quite. 

Every case depends to some degree on assumptions  as to the very presence of cancer, or its state when the alternative treatment was commenced, or as to the patient's prognosis, or as to what was truly responsible for the favourable outcomes.  They are not  unreasonable assumptions, mind you, which is why the cases are impressive.    If everything was as the stories are  implying,  such patients would nearly always fare badly without effective treatment.  A lot of such cases (and Brenner said he had five more) might carry considerable weight -- if it were not for one other thing about them.

And that is the kind of case that is missing!   Missing are the kind of cases we described above.   We are being asked to infer that Revici made advanced cancer go away without precise documentation of the "before, during and after" states in any case.  This is a critical matter, because  most patients with serious cancer have easily measurable cancer at some stage during their illness.   During the period of Brenner's association with Revici  hundreds of such patients must have been treated.     Brenner would surely have produced cases documenting the actual shrinkage or disappearance of cancer, as it happened,  if any were available to him.  He would be as aware of the importance of such cases as anyone.  Or,  are we to believe that for some peculiar reason the method didn't work with the cancers  that might make for the most convincing test of the method?    

It is this peculiarity that weakens the evidence.  Even taking the cases at near face value, we are left unsure whether such "best cases" hand-picked out of a presumed  very large number of failures (See footnote) are merely reflecting cancer's sometimes erratic behaviour at the fringes of normal oncological experience!   For oncologists regularly get unusual outcomes too.  Untreated cancer is predictable,  but only over a range of from about 70%  to about 99.999% for its various types and stages.    Cancer can mostly be accurately diagnosed, categorised and staged,  but we know mistakes are made in those quite regularly [4].   

Also, every alternative method seems to be able to accumulate a few such cases over time, no matter how foolish, implausible or fraudulent they may appear.    This reinforces suspicion that very occasional impressive cases are unrelated to the treatment.

The "missing data" is an almost universal weakness of alternative cancer anecdotal material.    We are constantly required to give the benefit of the doubt in small ways and wonder why.    If a method can make so many presumed advanced cancers go away,  why can't it ever  be shown making a few actual advanced cancers go away (more on this in http://www.users.on.net/~pmoran/cancer/how_to_read_a_testimonial.htm )?      The failure to document accurately what is going on within the patients denies us certainty as to the exact state or even the presence of active cancer when the alternative treatment was commenced,  and also, importantly, whether the timing of any cancer remission is such as to allow it to be reasonably attributed to the method claimed.    Indeed,  "best cases" such as these from such a widely applied method strongly suggest that the method  cannot cure patients with well-established cancer -- useful information for many..   

 Perhaps this is why Brenner himself goes on to express some uncertainty as to whether the Revici treatment can cure established cancer.  He is mainly suggesting that the method should be investigated further..    We, on our part, are unable to say dogmatically that  Revici doesn't ever  work, merely that the data doesn't establish that it does.    With alternative methods  such as Revici's decades then typically go by without better material ever emerging, reinforcing negative opinion.   Since even rumours of cancer cures attract intense public and media attention,  it is most unlikely that methods that have been in widespread usage for a decade or more are ever now going to show that they can cure established cancer.

That is how it looks to science.   Frightened cancer patients may understandably look at things quite differently, and want to give dubious treatments a trial.  But it is unquestionably in the best interests of all cancer patients that they should be demanding much higher standards of evidence from  those offering them alternative cancer treatments..

The Assumptions

We have so far been accepting hat everything is more or less as the stories imply, but I should give some examples of the assumptions I have suggested are being made in these cases. 

 As an example of how cancer is less predictable than many assume,   we can now largely discount the fourth case, the patient with inoperable lung cancer that lived on for nine years after palliative radiotherapy.    We now know that palliative radiotherapy (i.e. not expected to cure) of lung cancer unexpectedly  does lead to some ten-year survivors [2,3].   .   The first of these studies  was not published until 2000.  Brenner, even as a radiation oncologist,  would not know this  when he stated that this could not happen, in 1990,. 

That case illustrates something else.  We may not always have at the time all the information needed to assess the true significance of  occasional remarkable cases.  They can have a strong hold on our minds,  but they in truth can never stand alone as  automatic proof of  the claimed treatment effect,  even when of superb quality.   Thus,  I don't know quite know why the ovarian cancer patient did as well as she did and I would like more details about that case and to know whether any comparable cases exist.   Two or three cases the same would certainly arouse interest and undoubtedly prompt further research,  although there might still be that nagging doubt as to why equally advanced but more easily measured and followed up cancers than ovarian cannot be shown responding. 

 Brenner did not describe the actual bladder pathology in the first case, as we would expect a doctor talking to other doctors to do automatically if he had any personal involvement with the case.   He merely offers "what the doctors are supposed to have said".   I have pointed out elsewhere that this is a common source of weakness in testimonial.  It is very often a grossly distorted version of the facts, designed to make the testimonial more impressive.  In this case,  the surgery may have been advised to a patient with pre-invasive  bladder papillomata  recurring despite regular episodes of endoscopic destruction.  These have a strong potential to develop into invasive cancer, but they can also be erratic in their behaviour.  A single such case cannot establish that it was the Revici treatment that stopped them forming over the period described    Also, for all we know also, the patient may have eventually required the recommended surgery or even died of bladder cancer.     

The second case is probably the weakest for me,  since we have no direct knowledge as to the state of the cancer when Revici treated her.  

The fifth case is to do with a very rare tumour and made obscure by language that skirts around the question as to just what pathology was being dealt with later i.e. "nodules", "renal mass".   It is being implied that this was "metastatic cancer" but why not just say so?  We are never told what the "renal mass" was.  If this was actually a  primary renal cancer and the lung nodules were secondary to that rather than the giant cell tumour of bone then this could well be an example of the well-known spontaneous remissions that occur with renal cancer, even when it is metastatic.

Brenner may well have had answers to some of these questions.  The critical point I wished to get across was that even without such questions about individual cases there remains a  significant gap between the very best  "alternative" anecdotal material and the criteria that doctors use themselves  when testing out a cancer treatment.  

But what about -------?

I can't think of an example right now, but I think I have encountered other alternative case reports that do show clear cut cancer remissions of serious cancers at the same time as an alternative treatment is applied, and with no other obvious reason.   These are exceedingly rare.  They  also don't favour any particular one of the dozens of alternative treatments now in use.   Until an individual alternative method, or for that matter an alternative practitioner using any combination of methods they wish,  is able to produce a small series of even two or three such cases out of their recent experience it remains difficult to believe that any of them are regularly curing established cancer.     They should be able to do this quite easily if the usual claims were true. 

Blazing Away at the Barn Wall

The weaknesses in alternative anecdotal material arises from the lack of method and forethought in the collection of evidence.   This is, of course,  one defining feature of anecdotal material   Patients are treated in a haphazard fashion in the hope that cases will emerge that  look good or can be made to look good on  loose,  ad hoc  criteria.    When there are no predefined criteria for determining treatment success  there will be less effort at defining what exactly is being treated in the first place,  or at keeping track of what is going on within the patient, even though both of these should be easily achieved with present technology.    Most alternative cancer clinics also have very poor standards of record-keeping and follow-up [5]..  

As a result of the general lack of rigour,   the involved parties are able to imprint their own biases, hopes  and expectations on the data.     In other scientific contexts such unplanned, ad hoc approaches to research have been  likened to blazing away at the side of a barn with a shotgun and then drawing targets around the best-looking  clusters.   That is precisely what is happening here.   It is what sustains all quackery, when it is not frankly fraudulent.   If you have ever wondered how dozens of cancer cures could possibly be false even though supported by honest belief, this is why.

Another potential consequence of the lack of method is that even if a treatment sometimes had an impact on cancer there may well be too many deficiencies in the data for it to be convincing.   Quality clinical research is not primarily about all that baggage of control groups, placebos, double-blinding etc.  It is about first deciding on the precise question you want answered,  and then planning how to most reliably test it out.  Cancer treatments are much easier to investigate than many other treatments, because we are dealing with a mostly predictable condition whose progress can be easily measured in most patients    There is no real excuse for not being able to produce the top quality semi-anecdotal data we would expect from any effective cancer treatment.    I offer some advice on how to do this elsewhere.

Footnote

 A group of doctors were appointed to assess the Revici method, and published results of a small prospective (planned) study in 1965 [6]   Of  33 cases with established cancer 22 died.    8 left the study because they were unimproved by the treatment.  At the close of the study 15 months later only 3 patients were still under Revici's care but all had signs of tumour progression.   Back
 

References
 

1.  Brenner's speech is recorded in full at http://gerson-research.org/docs/OTA-1990-2/index.html

2.  Michael P. Mac Manus, Jane P. Matthews, Morikatsu Wada, Andrew Wirth,
Valentina Worotniuk, David L. Ball.  Unexpected long-term survival after
low-dose palliative radiotherapy for non small cell lung cancer


3  Quddus AM, Kerr GR, Price A, Gregor A. Long-term survival in patients
with non-small cell lung cancer treated with palliative radiotherapy. Clin
Oncol. 2001; 13: 95-98.
 

4. These matters are dealt with in more detail in sections l and ll of What's so Hard about Showing a Cancer Cure Works? 

5.  Some data on this with references in What do Alternative Methods Really Do (the Inside Story)

6. David Lyall and Others: "Treatment of Cancer by the Method of Revici" . JAMA Oct 18, 1965;194:p165-166

Back to Index