That’s the headline the media should have used when publicizing the results of a recent study comparing Depo-Provera use among women ages 20-44 with or without breast cancer. Because that’s exactly what the study’s authors found.
There was no difference in exposure to Depo-Provera among women with breast cancer compared to controls – about 11% of women in each group had ever used Depo-Provera.
Small Sub-group analysis raised a question
However, the researchers also did what is called sub-group analysis – looking at smaller and smaller groups within the breast cancer study population to see if they could find any women who might have a higher risk from Depo-Provera use, even if the overall group did not.
Before I get to those results, I need to point out that cases and controls, and Depo-Provera users and non-users had significant differences in body mass index, ethnicity, family history, age and pregnancy history – all factors related to breast cancer risk. In addition, controls had had less mammograms than their counterparts who had breast cancer. There is also a large issue of recall bias at play in studies such as this, since we know that subjects with the disease or condition being studied tend to remember drug exposures better than controls do.
Despite these marked differences between the groups, no statistically significant difference was found between Depo-Provera use in breast cancer patients compared to controls.
BUT, when subgroup analysis was done, there was one group that showed a difference – women who had used Depo-Provera for a year or more within the past 5 years. Here’s the finding that led to the headlines –
However, recent users of DMPA (Depo-Provera) for 12 months or longer had a 2.2-fold increased risk of breast cancer (95% CI: 1.2-4.2).
That was all the media needed to hear, and they were off and running, with headlines such as “Depo-Provera shots tied to breast cancer ” and “Depo-Provera shots double breast cancer risk.” Such headlines fail to convey the overall results of the study in a way that is meaningful to anyone but the lawyers.
The Sub-Group analysis that raised concerns was conducted in less than 100 women
What the media did not tell you was that in this “large study”, the actual number of women in the sub-group that showed this statistically significant and concerning result was extremely small – only 34 controls and 36 cases had used Depo-Provera in the past 5 years. Among these, only 15 and 32 respectively had used Depo-Provera for a year or more in that time!
So there you have it – headlines based on a subgroup analysis of less than 100 women who were poorly matched to start with.
These kinds of studies raise far more questions than they answer
Subgroup analysis is a research tool to look for potential topics for further investigation or to inform questionable findings in the overall analysis. For researchers, this type of analysis can be valuable in defining areas for future larger study. But publicizing the results of a small subgroup analysis as if it were the findings of the larger group mis-represents the strength and reliability of the findings. And in this case, that publicity does nothing more that to frighten women away from using hormonal birth control.
I for one am not changing my prescribing practices around Depo-Provera based on this study. Depo-Provera remains an important contraceptive option for many women, although potential side effects (irregular bleeding, weight gain) limit its acceptability as a first line method for the majority of them. Overall, about 3% of women who use contraception in the United States currently use Depo-Provera.
The discussion around breast cancer and current hormonal contraception is not new
The magnitude of the risk cited, and it’s relationship to current use of Depo-Provera in younger women, is similar to that which has been reported in the past in studies of oral contraceptives and breast cancer risk. Specifically with Depo-Provera, the risk that appears in current users in some studies actually disappears with longer term use, suggesting that the mechanism may be acceleration of growth of pre-existing tumors, rather than induction of new cancers over time.
Overall, neither oral contraceptives nor Depo-Provera appear to increase lifetime risk of breast cancer. Both reduce the risk of endometrial cancers, and the pill reduces ovarian cancer risk.
Future research may elucidate better what the effect of medroxyprogesterone acetate, the drug in Depo-Provera and the one studied as hormone replacement in the Women’s Health Initiative, may be on breast cancer risk. For now, that risk has best been defined in post-menopausal women, and not in users of hormonal contraception.
_________________________________________________
Further reading on Breast Cancer Risks and Depo-Provera
- Depot medroxyprogesterone acetate and breast cancer. A review of current knowledge. Drug Safety Review, 1996
- WHO Study showed similar results in subgroup analysis for current use in women under age 35.
- WHO Memorandum on Depo-Provera and Cancer risks.
- New Zealand study showing risk with longer term use in younger women, with same limitation of small numbers
- Pooled Analysis of WHO and New Zealand Studies . JAMA 1995
Conflict statement – In the late 90’s, I published two studies of Depo-Provera acceptance and continuation, one of which was partially funded by small investigator-initiated grant from Upjohn that paid part of my research assistant’s salary, along with an NIH training grant. I have no current conflicts to report.
Seems like breast cancer was a concern the first time Depo was on the market and was pulled off (early 1080’s I think). Except it turned out it was breast cancers in beagle dogs.
The data on breast cancer in users of hormonal contraception have been studied a fair amount – a few, but not all studies have suggested the possibility of small increases in women under age 35 on the combination pill as well as Depo Provera. One large New Zealand study of depo users found increases in current users, but that risk disappeared with longer term use. The numbers have never been large enough to do more than raise the question without answering it. The explanation usually given for possible increases in younger women who are current users is acceleration of the growth of pre-existing tumors, but to my knowledge it is just a hypothesis.
Interestingly high dose depo provera has been studied as a possible treatment for advanced breast cancer. http://www.asco.org/ascov2/Meetings/Abstracts?&vmview=abst_detail_view&confID=102&abstractID=81017
Studies have also shown that Depo Provera protects against endometrial cancer.
So the answers are not so simple…
The data are certainly worth a discussing, but the discussion needs to be framed in a meaningful way for the lay reader. The media unfortunately does not do that as well as they could.
Thanks for your comments.
Peggy
Thanks for that post. There are far too many missconceptions about hormonal birth control methods out there. The types of headlines you mention may add fuel to the arguements of those who are opposed to universal coverage of contraception for women.
Thanks, that’s helpful information.
Re the change in cancer risk. I seem to remember reading at some time that using BCPs for some amount of time also reduced the risk of ovarian cancer for women? Does Depo also change that risk?
(Or am I totally remembering wrong about BCPs?)
bardiac- You are correct. The pill does prevent ovarian as well as endometrial cancer. The data on depo had not been there yet, but a quick search on Pub Med today finds a new article in the British med journal this may that shows a similar reduction with use of Depo provera (83% reduction with > 3 yrs use). Think I’ll post it as a separate blog post today. Thanks for asking!
http://www.ncbi.nlm.nih.gov/pubmed/22489761
[…] is all important information for women using Depo Provera, particularly in light of recent media attention to the potential for breast cancer in users under age […]
I’m glad I found this article. Women may take Depo-Provera for more reasons than birth control and birth control is the only thing that is ever discussed. As a mother of a 16 year old girl, I need real info for my daughter. She isn’t on the shot for birth control. Her doctors have put her on it, because the depo shot has helped decrease her epileptic activity. She had infant seizures (grand mal/tonic clonic), but they went away and returned when she hit puberty. After that, her grand mal seizures occurred on the third day of her periods. Even before the seizures occurred she had so many memory issues, disorientation, migraines, moodiness that essentially shut her down 2 weeks each month during PMS week and her cycle week. The seizure meds stopped the seizures, but not these other side effects. The Depo shot has made these other issues disappear. Even though my daughter is only a teenager, her doctors anticipate having her on Depo for 5 years or more. They are hoping her hormone levels will calm down and balance more naturally when she enters her 20s. I can’t seemed to find any studies of women on the Depo for 5 or more years. My concern is that breast cancer runs in my family. My mother survived a very aggressive breast cancer 2 years ago. It was caught early and she had surgery and radiation. She did not need chemo. That puts my daughter and I at higher risk. When my daughter complains of breast pain and tenderness, it makes me a little apprehensive and I don’t know if I should worry about it or not. I am being told it is a side effect of Depo.