Sunday, 6 January 2013

Lessons from history #5: Bone marrow transplant for breast cancer


In the 1990s, giving patients with breast cancer massive doses of chemotherapy or radiotherapy followed by a bone marrow “rescue” was thought to improve survival and even cure the disease. It was complex, expensive and risky, which only raised the perceived effectiveness. The treatment spread, and insurance companies had to fall in line to cover the treatment due to legal, government and public pressure. The randomised trials that were eventually done showed that it was no more effective than standard treatment. Here is the story and the lessons that can be learned.

Background
  • Chemotherapy can provide marginal survival benefits in patients with solid tumours, such as breast cancer. The same goes for radiotherapy.
  • The dosage of chemotherapy or radiotherapy is limited by its toxicity, particularly to the bone marrow.
  • There is a commonly held belief in medicine that if some treatment is good, more treatment is better.
  • It was thought that better survival would result from high doses of chemotherapy/radiotherapy, knocking out the bone marrow, and then injecting previously harvested bone marrow cells in order to regenerate the bone marrow (a bone marrow ‘rescue’ or autologous bone marrow transplantation).


What happened
  • Some guys tried this and said that patients did well. Really well.
  • The practice spread, without scientific experiments (randomised trials).


What happened next
  • Pressure from doctors, patients and other interest groups grew, demanding funding for this complex and expensive procedure.
  • In one case, a health insurance company settled for $US89 million for not providing the treatment, even though the patient died soon after receiving the treatment (from money raised by donations).
  • Health insurers then fell into line, and covered the procedure.
  • The practice spread further, with further reports of good results (“8 times better than conventional treatment").
  • One randomised study was reported in 1995 that showed excellent results (although a later inspection could find no trace of whether the control patients in this study ever existed).


What finally happened
  • After the practice was established, some doctors decided to perform a randomised trial to scientifically test the treatment.
  • They had difficulty recruiting for the studies because people assumed it worked.
  • The studies were also criticised as being unethical.
  • Four randomised studies were finally done and presented in 1999.
  • The studies showed that the treatment didn’t work.
  • This treatment is not done any more.


Lessons
  1. People (doctors and patients) assume that new, bigger, more extreme, desperate and expensive treatments work better.
  2. People are quick to criticise the scientific evaluation of a treatment as unethical or unnecessary, and are slow to criticise the acceptance and introduction of unproven and possibly harmful treatments.
  3. Governments, insurers and other health agencies respond to public pressure and are often confused by the science.
  4. In medicine, the 'customer' (public pressure) is not always right.


Reference
This story is told in greater detail in a BMJ article by H Gilbert Welch, which is where I got the information above.

4 comments:

  1. This reminds me of the FDA's approval of brain stents as a "humanitarian exception" to their usual review process (i.e., without an RTC). Turned out they CAUSED strokes and FDA had to withdraw its approval.

    Is it still a "lesson" if we refuse to learn from it?

    ReplyDelete
    Replies
    1. Thanks John, good example.

      It is still a lesson, but the message seems to go against our natural way of thinking, so it takes a very open, skeptical mind to learn it.

      Delete
  2. Dr. Otis Brawley (How We Do Harm) discusses this (among many other non-evidence based treatments)at length during his presentations ... many are available online.

    ReplyDelete
    Replies
    1. Thanks Ken, I will look up his presentations.

      Delete