Re: How scientific are orthodox cancer treatments? by Walter Last

From: Orac (orac_at_wabcmail.com)
Date: 08/14/04 

Date: Sat, 14 Aug 2004 17:19:57 GMT

In article <hKidnc0ZH-eqp4PcRVn-rA@nildram.net>, "Anth" <anon@anon.com>
wrote:

> http://educate-yourself.org/cancer/kellypancreaticcancerpatients.shtml

Even for a case series, this is pretty poorly designed. Also, the doctor
who wrote this article makes a couple of whopper idiotic statements that
I find it hard to believe she ever completed medical school. Some
examples of egregiously incorrect statements:

1. Cancer is the easiest of all the major degenerative disease
conditions to properly treat and cure. COMMENT: Really? Then why have we
not cured it?

2. All persons who have Cancer die of starvation, unless they are first
killed - usually by their Physician. COMMENT: Bull***. Many patients
die from the local effects of their disease. Some die of liver failure
from liver metastases or lung failure due to lung metastases. Many die
from infections due to immunosuppression (particularly true of
hematologic malignancies, which even when untreated--can't blame it on
chemotherapy here--will suppress the immune system.) Some die of
bleeding (again, particularly true of hematologic malignancies, which
can suppress the bone marrow and decrease the platelet count to
dangerous levels, and eventually platelet transfusions do little to
help, as the patient develops extensive antibodies to donor platelets).
The cancer cachexia syndrome (producing the characteristic emaciation)
only occurs for some, not all, cancers.

3. Legally the definition for a Cancer Cure is, a patient must be free
of Cancer five years after initial diagnosis. COMMENT: "Legally"? No,
it's a medical convention for most cancers. It is not a legal definition.

But, back to the article:

First off, it is quite wrong to claim a 100% success rate if only 5 out
of 22 patients could complete the regimen. He could claim 23% (or
5/12=42% if we exclude the patients who never started the regimen).
Also, the requirement that the patient must have at least 6 months of
time to get well into the Kelley Program automatically skims the best
50% of patients in terms of prognosis right off the top (the median
survival for unresectable, but not metastatic, pancreatic cancer is on
the order of 6-12 months and on the order of 3-5 months for metastatic
pancreatic cancer).

Second, if only 23% of patients can complete a regimen, then that
certainly implies a selection bias for patients who are in
physiologically better shape and thus able to complete such a rigorous
program. These patients would be expected to have a better prognosis.
(At the very least it implies that, even if it were every bit as
effective as claimed, the Kelley protocol would only potentially benefit
around 1 in 4 pancreatic cancer patients.) Add to that the fact that we
are not told HOW these 22 patients were selected from the bigger set of
33,000 patients. It is not reported how many of the 33,000 patients had
pancreatic cancer, but it's not unreasonable to expect that it was way
more than 22 patients, given that pancreatic cancer is not an uncommon
cancer. (If even only 1% of the patients in the database were pancreatic
cancer patients, that would mean there were 330 pancreatic cancer
patients. If it were 10%, that would mean 3,300 patients. What happened
to all the other pancreatic cancer patients in the series of 33,000?)

Third, no mention is made of histology. Were these biopsy-proven
pancreatic cancers? How were they diagnosed? What was the histology?
There are different kinds of pancreatic cancer. The most common kind,
adenocarcinoma, is the most deadly one, the one that is usually meant
when one says "pancreatic cancer." However, there are other, albeit much
less common, varieties that are far less deadly than the usual
adenocarcinoma of the pancreas and eminently curable with conventional
therapy. For instance, cystadenocarcinoma can be "cured" around 65% of
the time with surgery, and endocrine tumors can be "cured" over 90% of
the time by simple enucleation. Admittedly, these are pretty uncommon
tumors, but, given how few patients there were reported relative to the
complete set of 33,000 patients, one has to wonder if they didn't pick
up at least a couple of these less common tumors. Impossible to say from
what's reported.

Bottom line: This may represent a real effect or it may represent
selection bias and/or sloppy design. It is not adequate evidence to say
one way or the other. 

-- 
Orac        |"A statement of fact cannot be insolent."
            |
            |"If you cannot listen to the answers, why do you
            | inconvenience me with questions?"