Expert Interview

Using (dynamic) ultrasound to make an earlier diagnosis of endometriosis

Author and Disclosure Information


Can you provide some background on endometriosis and the importance of early diagnosis?

Dr. Goldstein: Endometriosis is an inflammatory condition, characterized by endometrial tissue at sites outside the uterus—this definition comes from the World Endometriosis Society.

Endometriosis is said to affect about 10% of women of reproductive age, and if you look at a group, a subset of women with pelvic pain or infertility, the numbers rise to the range of 35% to 50%. It can present in a multitude of locations, mainly in the pelvis, although occasionally even in places like the lung. When it occurs in the uterus, it is known as adenomyosis; when it occurs inside the ovary, it can cause an endometrioma (or what is sometimes referred to as chocolate cyst of the ovary), but you can see endometriotic implants anywhere in the peritoneum—along the urinary tract, rectum, uterosacral ligaments, rectovaginal septum, and even the vaginal wall occasionally.

What I am really interested in is an earlier diagnosis of superficial endometriosis, and it should be apparent to the reader why this is important—the quality of life from pain from endometriosis can be debilitating. It can be a source of infertility, a source of menstrual irregularities, and a source of not only quality of life but also economic consequences. Many women can also undergo as much as a 7-year delay in diagnosis, so the need for a timely diagnosis and initiation of treatment is extremely important.

What is the role of ultrasound in endometriosis diagnostics?

Dr. Goldstein: In an article that I authored 31 years ago, I wrote that there was a difference between an ultrasound examination by referral and examining one’s patients with ultrasound. I coined a phrase: the “ultrasound-enhanced bimanual exam.” I believed that this term should become a routine part of the overall gynecologic exam. I wanted people to think about the bimanual that we had done for at least half a century, which, in my opinion, consists of 2 components:

  1. An objective component: Is this uterus normal? Is it enlarged or irregular in contour, suggesting maybe fibroids? Is an ovary enlarged? If so, does it feel cystic or solid?
  2. A subjective component: Does this patient have tenderness through the pelvis. Is there normal mobility of the pelvic organs?

Part of the thesis was that the objective portion could be replaced by an image that could be produced in seconds, dependent on the operator’s training and availability of equipment. The subjective portion, however, depended on the experience and, often, nuance of the examiner. Lately, I have been seeking to expand that thesis by having the imager use examination as part of their overall imaging—this is the concept of dynamic imaging.

Can you expand on the concept of dynamic ultrasound in this setting?

Dr. Goldstein: Presently, most imagers take a multitude of pictures, what I would call 2-dimensional snapshots, to illustrate anatomy. This is usually done by a sonographer, or a technician, who collects the images for viewing by the physician, who then often does so without holding the transducer. Increasing utilization of remote tools like teleradiology only makes this more likely, and for a minority of people who may use video clips instead of still images, they are still simply representations of anatomy. The guidelines for pelvic ultrasound are the underpinning of the expectation of those who are scanning the female pelvis. With dynamic imaging, the operator uses their other hand on the abdomen as well as some motion with the probe to see if they can elicit pain with the vaginal probe, checking for mobility, asking the patient to bear down. Whether you are a sonographer, a radiologist, or an ObGyn, dynamic imaging can bring the examination process into the imager’s hands.

Can you tell us more about the indications for pelvic sonography for endometriosis and what data can you give to support this?

Dr. Goldstein: There is a document titled “Ultrasound Examination of the Female Pelvis,” that was originally developed by the American Institute of Ultrasound in Medicine (AIUM). In this document, there are about 19 different indications for pelvic sonography (in no defined order), and it is interesting that the first indication listed is evaluation of pelvic pain. Well, I would ask you, how do you evaluate pelvic pain with a series of anatomic images? If you have a classic ovarian endometrioma, or you have a classic hydrosalpinx, you can surmise that these are the source of the pain that the patient is reporting. But how do you properly evaluate pain with just an anatomic image? Thus, the need to use dynamic assessment.

There was a concept first introduced by my colleague, Dr. Ilan Timor, known as the sliding organ sign, that was mainly used to determine if 2 structures were adherent or separate. This involved use of the abdominal hand, liberal use of the probe moving in and out, and under real-time vision, examining the patient with the ultrasound transducer; this is the concept of dynamic ultrasound. This practice can be expanded to verify if there is pelvic tenderness and can be a significant part of the nonlaparoscopic, presumptive diagnosis of endometriosis, even when there is no ovarian endometrioma.

To support this theory, I would point you toward a classic article by E Okaro and colleagues in the British Journal of OB-GYN. This study took 120 consecutive women with chronic pelvic pain who were scheduled for laparoscopy, but performed a transvaginal ultrasound prior, and they looked for anatomic abnormalities and divided this into hard markers and soft markers. Hard markers were obvious endometriomas and hydrosalpinges, while soft markers included things like reduced ovarian mobility, site-specific pelvic tenderness, and presence of loculated peritoneal fluid in the pelvis. These were typical of chronic pelvic pain patients that ranged from late teens to almost menopausal, as the average age was about 30 years old.

Patients had experienced pain for anywhere from 6 months to 12 years, but the average was about 4 years. At laparoscopy, 58% of these patients had pelvic pathology, and 42% had a normal pelvis. Of the 58% with pathology, the overwhelming majority—about 51 of 70 women—had endometriosis alone, and another 7 had endometriosis with adhesions. A normal ultrasound, based on the absence of hard markers, was found in 96 of 120 women. Thus, 24 of the 120 women had an abnormal scan based on the presence of these hard markers. At laparoscopy, all 24 women had abnormal laparoscopies. Of those 96 women who would have had a normal ultrasound, based on the anatomic absence of some pathology, 53% had an abnormal scan based on the presence of these soft markers while the remaining women had no soft- or hard-markers suggesting any pelvic pathology. At laparoscopy, 73% of the patients with soft markers had pelvic pathology and 27% had a normal laparoscopy. Of 45 patients who had a normal, transvaginal ultrasound, 9 were found to have small evidence of endometriosis without discrete endometriomas at laparoscopy.

To summarize the study data, 100% of patients with hard markers and chronic pelvic pain had abnormal anatomy at laparoscopy, but 73% of patients who had soft markers but otherwise would have been interpreted as normal anatomic findings had evidence of pelvic pathology. Such an approach, if used, could lead to a reduction in the number of unnecessary laparoscopies.

What it really boils down to is, if you have 100 women with chronic pelvic pain, are you willing to treat 100 patients without laparoscopy, knowing that 73 are going to have a positive laparoscopy and will require treatment anyway? You would treat 27% with a pharmaceutical agent that may provide relief of their pain, or may not, depending on what the true etiology was. I would be willing to do so, as a positive predictive value of 73% makes doing that worthwhile, and I believe a majority of clinicians would agree.

Do you have any other tips or ways to improve the reader’s understanding of transvaginal ultrasound?

Dr. Goldstein: Pelvic organs have mobility. If a premenopausal woman is examined in lithotomy position, if the ovaries are freely mobile, by gravity, they are going to go lateral to the uterus and are seen immediately adjacent to the iliac vessels. But remember, iliac vessels are retroperitoneal as they are outside the peritoneal cavity. If you were to turn that patient onto all fours, so that the ovaries are freely mobile, they are going to move somewhat toward the anterior abdominal wall. When an ovary is seen in a nonanatomic position, it could be normal or it could be held up by a loop of bowel, but it may indicate adhesions. This is where this sliding organ sign and liberal use of the other hand on the lower abdomen can be extremely important. The reader should also understand that our ability to localize ovaries on ultrasound depends on the amount of folliculogenesis. Follicles are black circles that are sonolucent, because they contain fluid, so they make it easy to localize ovaries, but also their anatomic position relative to the iliac vessels. However, there is a caveat—which is, sometimes an ovary might look like it is behind the uterus and not in its normal anatomic location. When dynamic imaging is used, you are able to cajole that ovary to move lateral and sit on top of the iliac vessels, which can enable you make the proper diagnosis.

Recommended Reading

How effective is elagolix treatment in women with fibroids and HMB?
MDedge ObGyn
Appendix may be common site of endometriosis
MDedge ObGyn
New hormonal medical treatment is an important advance for AUB caused by uterine fibroids
MDedge ObGyn
Endometriosis-associated ovarian cancer
MDedge ObGyn
Endometriosis: Disease burden and the problem of missed or delayed diagnosis
MDedge ObGyn