Apr 16 2008
This is a slight departure from the usual fare of pseudoscience, but a matter that should concern us because of the vulnerability this matter confers on medicine – the borderline practices of major medical centers. The article can be viewed here.
Several days ago the San Francisco Chronicle printed a second article about the plight of a 37 year old woman (EP) with an inflammatory breast cancer who was denied insurance coverage for an expensive treatment, high-dose chemotherapy with autologous bone marrow (or stem cell) transplant or infusion (HDCT/BMT or SDI.) The institution is the MD Anderson Cancer Center in Houston. The problem is that although the treatment is effective, it is no moreso than moderate dose HDCT without the marrow or stem cell infusion, and also is more expensive and has significant morbidity.
Inflammatory breast cancer is a highly aggressive form that is usually regarded as “advanced” when diagnosed, that is, spread beyond the breast and regional lymph nodes. One cannot tell from the article whether EP’s cancer spread is documented or implied. But because of the poor prognosis and presumed incurability in either case, options are limited. In the 1980s -90s, HDCT/BMT was thought to be a promising method on the basis of studies that showed a prolonged disease-free and overall survival compared to results of prior studies using more conservative treatment. The problem then was that the studies were uncontrolled.
Such claims are often found to be wrong because the cases making up the statistics had been selected and more favorable than average. So when randomized controlled trials of the procedure were finally carried out and reported, out of a group of 5 studies, 4 of the 5 showed no difference between the treated and the control groups. The fifth study showed prolonged disease-free survival, but the data were found to have been manipulated, and the investigator left his country and went into hiding. It was a scandal at the time.
Until then, previously advanced and incurable cases had been treated with HDCT/BMT on a case by case basis and the cost, at several hundred thousand dollars each, was placing a severe burden on the insurance system (most cases were women in pre-Medicare age.)
For the past ten years, HDCT/BMT use for breast cancer has been progressively minimized. Since moderately HDCT without the marrow or stem cells gave equivalent results, variants of HDCT have been treatment of choice for inflammatory breast Cancer.
The MD Anderson Center formed an independent study on HDCT/SCI for selected cases in order to find out if there is a sub-group of patients who still might benefit from the method, but which had escaped detection in prior studies. Most researchers would find this to be improbable based on what is known. The positive predictive value (PPV) would be still too low to justify the outlay in funds and in personnel and equipment effort. Although mortality rate has been reduced in experienced centers, it can run between five and fifteen percent. The study is not supported by a funding agency, so individual patients must pay.
EP, is covered by Kaiser/ Permanente (KP) in northen California, which pays for standard and some experimental treatments but not for an unfunded private experimental one. After the original 2007 SF Chron article, a public collection paid enough for EP’s initial HDCT/SCI procedure. But the study policy requires frequent bone and other scans that Anderson requires be done only at its facilities. K/P refuses to pay for for scans, Xrays, lab tests, etc., outside of its own. So patient EP is apparently accepting donations again.
And, K/P is receiving another figurative black eye for lacking humanity and so forth.
So one an see the problem here. Consider MD Anderson; is it proper to offering a controversial, low-yield treatment as an uncontrolled study? How much can be learned from a such a trial? What are the motivations? Patient attractition? Hope of finding positivity contrary to current opinion? Commercial support? Add to that the fact that inflammatory breast cancer can respond well to standard moderately high doses of chemotherapy anyway.
Then, consider patient EP. She is only 37. Can she be blamed for searching for better answers? Should she and her family be prepared tp pay the price if she is justified? Consider K/P. Is it being too rigid? Does it not have a duty to its subscribers to keep costs minimal and its premiums reasonable?
Then consider the Press. Does it have a responsibility in presenting such cases for popular interest when the major effect is to increase sympathy for unfortunate cancer patients, while casting an implied blame on insurance companies and HMOs? MD Anderson comes off well in the article, which does not mention the problematic ethics.
How closely does this situation compare to ones that involve proposed pseudo-methods that have even less plausibility? At least this one works. But where does one draw the line? Legally? Ethically?
Disclosure: In 1996 or so, HDCT/BMT was becoming increasingly common, and Blue Cross/Blue Sheld had already funded its study of the method to satisfy its subscribers. I had previously reviewed cases applying for HDCT/BMT for several years, and approved none, because the mortality rate approached the prolonged survival rate. I was soon excluded from review panels. By 1996, the Milan group submitted its study of a modified HDCT regimen for advanced breast Ca, the results of which were the equal of any BMT study, with less mortality. Asked to review the paper for JAMA, I recommended accelerated publication. So you have my reasoned bias on the matter.
But I still see a problem…it persists and is not easily dismissed. But I don’t find any answer really satisfying. But the best one to my mind is that one cannot repeatedly look for and test sub-groups of completed negative trials on the possibility one might have missed something. This is a lesson the NCCAM also needs to learn.
13 Responses to “Borderlines in research”