"I remember wheeling down this long, narrow corridor into my office, just a confused, scared mess, but I had to confront him." Phil Vardy was a young medical researcher in Sydney that day in 1982. The man he had to confront was his boss, one of Australia's most prominent doctors, William McBride.
He had just seen an article McBride published suggesting an ingredient in the drug Bendectin caused birth defects - with fabricated defects and experimental pregnant rabbits that didn't exist either.
Vardy's confrontation with McBride did not go well, and "small battles led to bigger ones which led to bigger ones." It would cost Vardy his job and his marriage. After a decade of media storms and public enquiries establishing fraud, McBride lost his license to practice medicine.
I was a maternity consumer advocate in Australia then, and watching what happened to Bendectin and the women who needed it was a long slow revelation. This month, the FDA added a hopeful chapter to the saga. But that's jumping too far ahead. Let's go back first.
McBride's part began in 1961 when he sent a letter to The Lancet reporting limb malformations associated with the drug Thalidomide. The pendulum swung overnight from under-awareness and inadequate protection to hyper-vigilance, emotion and hype about drug safety in pregnancy.
In the wake of Thalidomide, Bendectin inevitably came in the firing line. In some countries up to a third of all pregnant women had taken the drug - over 30 million by the early 1980s.
That number of women would be expected to give birth to over a million babies with some kind of birth defect - by coincidence alone. So the potential for people to point the finger of blame at this drug was sky-high.
The National Enquirer and other media spread the word. The Public Citizen's Health Research Group (unsuccessfully) petitioned the FDA in 1980 and 1982 to remove its marketing approval. Bendectin, they argued, didn't work and was "unsafe for human use."
Bendectin did work and was never proven to cause birth defects, but it caused a lot of litigation. McBride - his fraud not yet revealed - was a prominent medical expert for many plaintiffs.
The case against the drug was made up of flimsy science and selective choices of studies. Even without junk and fraud, it's risky to look at some studies in isolation, rather than systematic reviews of all strong evidence. Litigation over Bendectin led the Supreme Court to set a new standard for scientific evidence and testimony.
But the drug was off the market by 1983 purely for commercial reasons. Defending it was just too expensive. This quote in the NY Times shows what Bendectin was up against: "Hundreds of thousands of pregnant women and their unborn children will be spared the risk of exposure to this questionably effective and unsafe drug."
With Bendectin gone, the rate of birth defects didn't fall. The ratio of hospitalization of pregnant women for nausea and vomiting almost doubled in the US (from about 7 per thousand to 13). There aren't very effective non-drug alternatives, so many women used drugs not approved safe for pregnancy.
This month, 30 years after the end of Bendectin, Diclegis was approved by the FDA for the treatment of nausea and vomiting in pregnancy. It's a new formulation of Bendectin's key ingredients. But we still have a far larger problem.
We don't have adequate evidence of the effects of drugs women need to use when they're pregnant. We seem to be more concerned about protecting people from the risks of research than the risks of not doing research.
Animal studies can't resolve the problem. Some drugs cause birth defects only in humans, or only in some animals and not in humans. Surveillance only gets us so far. Such as finding out too late that a particular type of hypertensive drug in pregnancy was sometimes fatal for the baby.
The problem leaked out to partially affect all women. Women of childbearing age or even all women have often been excluded from drug research - just in case. Yet differences in body weight, metabolism and more between men and women can affect drug response and ideal dosing in critical ways.
Children are another group with extensive additional safeguards for research. By the late 1990s, children were described as "therapeutic orphans", because there were so few safety and efficacy studies in children. And legislative efforts to try to improve the situation began.
It's the Catch-22 of clinical trials. For too long we have been trying to protect women and children from the risks of untested drugs...by not testing their drugs adequately. Here's hoping the arrival of Diclegis is a sign of wider progress.