The FDA held hearings on Dec 8 and 9 on two birth control methods: oral contraceptives containing a progestin called drospirenone (DRSP) a spironolactone analogue) and a contraceptive patch marketed in the USA as Ortho EVRA. Several post market studies suggest that pills containing drospirenone impose a higher risk of venous thromboembolic events (VTE) than earlier versions of the pill. The patch has also been associated with a higher risk.
On the first day, the FDA affirmed, voting 15 to 11, that the benefits of the pills with drospirenone (DRSP) in preventing unwanted pregnancies outweighed the risks. The panel then voted 21 to 5 to upgrade the warnings in the label to more clearly convey that VTE with DRSP pills may be as much as 74% higher than with some older versions of the pill.
The following day, the FDA reviewed the findings on the patch. The panel voted 19-5 that the benefits of the patch outweigh the risks, despite a 60-100% increase in the incidence of VTE. One consideration was that EVRA is unique. It is essentially the only drug in class, with a transdermal delivery system that medicates over a 7 day time span. Health care providers testified that the patch is particularly useful for women who have trouble remembering to take a pill every day.
Both hearings allocated time for public comments. Sadly, parents of several young women, whose deaths were attributed to hormonal contraceptives, gave heartfelt and heartrending testimony, describing the achievements, joys, and great love they shared with their daughters. And then they tearfully related the pain of losing a child.
When I participate in pharmaco-vigilance, looking at cold, hard statistics, I do not see that “pain.” I am handed a huge printout and, at the end, there is a line summing all the adverse events. And I am soothed and reassured that very bad things happen very rarely. A doubling of a rare event is still a rare event. The background rate of VTE in women age 25-35 is 3/10,000, and with hormonal contraception, the rate is between 6-10/10,000. Attributable mortality is extremely low. In a study from Sweden, the VTE mortality rate in current birth control pill users was 7.5/ million user-years. But a death is a death. Though the numbers are vanishingly small, I am insulated from the fact that each death attaches to a real person, and that the vines of grief entwine around an entire family and perhaps an entire community.
I want to emphasize as I did in an earlier blog, that in the life sciences, the sands under our feet are constantly shifting. “Truth” is the last best data set. In the US, we do not have the best data set for our drugs. Our ability to track adverse events is hampered by a lack of good surveillance.
In Europe, contraceptive (and other) drug safety information accrues from medical and hospital records. Since there is universal health care coverage, the data pool includes everyone and events can be attached to a real person, and to their entire medical history. In the US, with the fragmentation of health care, and with the large segment of the population lacking health care coverage, we cannot do that. We have some large population based studies like Boston Collaborative Drug Study, one of the best sources of drug safety information. But often, and as evidenced at the recent FDA hearings, drug safety data in the US comes from drug sales information generated by commercial services like the IMS, a provider of informational services for the healthcare industry. Much of the data presented came from drug sales records, not medical records. In using this kind of commercial data, diagnoses cannot be verified, patient demographics cannot be confirmed, concurrent medications and confounding illnesses cannot be assessed. Marketing data is a poor substitute for clinical data.
Our post market surveillance system is voluntary and haphazard. As a result, it takes longer in the US to identify a “signal,” a pattern of adverse events associated with a drug, than it does in the EU. The poor quality of post market tracking after drug approval impacts the initial drug approval process. More drugs are approved in EU each year, and drugs are brought to market faster, because adverse events can be identified and compiled more reliability and more rapidly than in the US.
Fortunately, the people working in drug safety in industry are a dedicated and diligent group. They do the best they can with the information they have. In concert with FDA mandated post market safety studies, industry does a relative good job protecting the public, while providing valuable treatment options.
The biggest challenge is deciding how much risk is acceptable. Drug development nowadays is focusing on a group of conditions referred to as “lifestyle” problems: annoying, inconvenient, but not life threatening conditions. No one dies of a leaky bladder. Profuse sweating is not lethal. Constipation may be approached with diet.
When developing drugs for “lifestyle” disorders, you are not allowed to kill patient. And you are not supposed to maim them either. Tolerance for adverse events is very low, making drug development a really delicate balancing act.
While attending these hearings, a constant question was what is an acceptable level of risk? And what do you use as a baseline or a comparator? Do you use the background age specific rate of VTE, the rate for other hormonal methods, the rate for pregnancy, the rate for barrier contraceptives plus the morbidity from method failures and unplanned pregnancies? The FDA committee enlisted a wide range of experts — epidemiologists, statisticians, clinicians, public advocates, and other health professionals to define an acceptable level of risk. Very challenging….
If you feel up to the challenge, consider joining us!