Topics of Interest

How Common is Uterine Leiomyosarcoma? (LMS)?

There is a lot of confusing information circulating about morcellation and uterine leiomyosarcoma in the media since the FDA hearing on this issue a few weeks ago.  I plan to post my analysis of the information, in segments, over the next week or two.  This first post provides commentary about the numbers used in research articles about the risk of having leiomyosarcoma and the numbers cited in the FDA hearings.

Some gynecologists and major medical centers had been quoting the risk of a woman having surgery for what are thought to be “fibroids” and being found to, in fact, have a leiomyosarcoma (LMS) as 1 woman out of 10,000.  However, this number is incorrect.  The rate of 1 in 10,000 is the incidence of LMS in the entire population of women, meaning that for every 10,000 women in the US (or elsewhere) 1 will have LMS.   This is a very different number than will concern women with fibroids.  Although many women will have some fibroids in their lifetime (77% of women by age 50), only a small minority (20%) will ever need treatment.  For these women who have symptoms bothersome enough to need treatment for fibroids (medication, surgery, UAE, HIFU), the risk of having LMS had been previously been stated as 1 in 1,332, based on a study we published in 1994.  So, that’s the first difference: 1 in 10,000 women in the entire population versus 1 in 1332 women who actually have fibroids and need surgery because of fibroids.

However, the FDA recently calculated the risk of LMS to be 1 in 482 women having surgery for fibroids and presented this data to the FDA hearing a few weeks ago. (I will discuss other uterine cancers in a future post)  At the same FDA hearing, Dr. Elizabeth Pritts, a reproductive endocrinologist from Wisconsin, presented another analysis showing that the incidence of LMS in women having surgery for fibroids was 1 in 7,450. (disclosure: I am one of the co-authors on this paper that Dr. Pritts will soon be submitting for publication)  Why are these numbers so different, and which number is accurate?   Below is a comparison of how the two numbers were determined.

In order to calculate 1 in 482, the FDA researchers drastically limited their search of the medical literature.  The FDA required that any studies showing up in their computer search have the words “uterine cancer” included in the terms used by journals to categorize articles in Pub Med or other medical search engines.  Therefore, all studies of women who had surgery for fibroids, but where no one was found to have cancer were excluded from the FDA analysis. Let me say this clearly: any study of surgery for fibroids where no LMS was found was excluded from the FDA calculation! This is a very large number of studies since LMS is a very rare disease and is infrequently found.

Dr. Pritts looked at ALL articles where surgery was performed for fibroids and the pathology of all the removed tissue was both analyzed and reported in the paper.  So, studies where the incidence of LMS was “0” were included, as they should be.

The FDA limited their search to articles only written in English.  This may have been acceptable 20 years ago, but now the rest of the world actively publishes quite good medical research, and free computer translators can generate an understandable version of an article quickly.  Many articles pertinent to this issue but published in languages other than English were readily available, but not included in the FDA analysis.

Dr. Pritts reviewed hundreds of articles in other languages and translated them to see if they met the criteria for analysis.

There are two major types of research studies for the data we need in order to figure out the risk of LMS: retrospective and prospective.  A retrospective study often begins with this scenario: someone operates on a woman thought to have a fibroid and the pathology later shows the tissue to be a LMS. The doctor then decides to go back through their hospital’s records to see if other similar cases have occurred.   But, on day one the incidence of LMS is 100%.  Depending on how many years the authors go back in the records, the 100% incidence will be diluted by the many cases without LMS.  But, the incidence will never be 0% because they started counting when they had the first case.

A prospective study follows this scenario:  the authors decide to study women who are going to have surgery for fibroids in the future, and among other things they will study whether any women had a LMS.  So, on day 1 the incidence of LMS is 0% and the research data starts with a clean slate. This is the gold standard of this type of research.

The FDA literature search yielded 18 studies of which 9 were used for their analysis.  One was a prospective study, and 7 were retrospective studies. The last “study” was a letter to the editor, which is not subject to peer review of data by outside experts. The 9 studies included 9,160 women and 19 were found to have a LMS (1 in 482 or 0.2%).

Dr. Pritts’ much more complete search yielded 1,020 papers, of which 365 were peer reviewed and felt to be of good enough quality to be included in the data analysis.  In the 67 prospective studies (best quality),  5,951 women had surgery for fibroids and 2 women were found to have leiomyosarcomas (1 in 2,975 or 0.03%).  Dr. Pritts also found  66 retrospective studies, where the first case may have prompted the study, and of the 23,926 women having surgery for fibroids, 22 were found to have LMS (1 in 1087 or 0.09%).

I am shocked, embarrassed and disgusted by the FDA’s lazy and irresponsible analysis of the “literature”.  I don’t know how long it took the FDA to do their analysis, but Dr. Pritts spent months collecting and then analyzing the papers.  For me, there is no question as to which analysis is more complete and representative of how rare this disease is.  It is a terrible, aggressive cancer, but it is rare.

Some women may feel that any risk of LMS is more than they want to take. Other women may find Dr. Pritts’ analysis of the available data reassuring, since the risk of LMS is so low.  But I absolutely believe that it is your right to decide.  All medical treatments are associated with some risk.  It is the doctor’s job to inform you of the risks (informed consent), but it is your job to decide what risks you are willing or not willing to take.

The next post will deal with how often morcellation involves cancers other than LMS.


Disclaimer: The ideas, procedures and suggestions contained on this web site are not intended as a substitute for consulting with your physician. All matters regarding your health require medical supervision.

Fibroid Doctor William H. Parker

Dr. William H. Parker is a board-certified Fellow in the American College of Obstetricians and Gynecologists. Dr. Parker is an internationally recognized expert in fibroid surgery and research. Based in San Diego, California, he is considered one of the best fibroid surgeons for abdominal and laparoscopic myomectomy in the United States and abroad. He has been chosen for Best Doctors in America and Top Doctors every year beginning in the late 90's.

Free Email Updates

Powered by Google FeedBurner : Receive free updates right to your inbox every time new content is added to this website.

Schedule an Appointment Today!
Read Dr. Parker's Blog

Blog Categories