LUNA Explains: Imagery In Endometriosis

You lie on your back, sometimes in a gynaecological chair.

The examination starts with an external examination, through the skin of the abdomen, with concomitant analysis of the kidneys and bladder.

Then the examination continues with the endovaginal (=internal) route, with an adapted probe, disinfected between each patient according to codified procedures, and protected by a single-use probe guard, by default made of latex (an allergy to latex must be reported before the examination).

This examination should not be painful, it is important to inform the doctor of the presence of pain.

Images are acquired of the various organs, in 2D and sometimes in 3D.

There is no official recommendation for specific preparation for ultrasound. Some teams recommend a rectal enema beforehand, others a filling of the bladder.

You also lie on your back for this examination. The stomach is in the centre of the ring, allowing the patient’s head to remain outside the machine in most cases.

The examination is noisy, due to the technology used (magnet rotating in the machine to create the images with electromagnetic waves). There are no X-rays used (which is different from X-ray or CT scan).

The examination takes about 15 to 20 minutes, and you may be given an injection of products to improve the quality of the image (glucagen, gadolinium).

MRI is a precise examination, like a magnifying glass that allows you to explore a specific area precisely, but it does not allow a global analysis of the whole body.

This examination should ideally be preceded by a rectal preparation consisting of an anal enema, in order to eliminate all the material that tends to cause artefacts and strongly degrade the images.
Gel opacification of the vagina in non-virgin patients and/or rectal opacification may be performed, but are not mandatory.

Endometriosis can occur in a number of different locations and imaging tests can detect some of these.

Superficial peritoneal involvement is very difficult to diagnose on ultrasound, but also on MRI.

The appearance is that of hyper-echogenic spots on ultrasound, and fine punctiform images of haemorrhagic signal on MRI. The preferential location is pelvic (behind the uterus, peri-ovarian, vesico-uterine recess) but it can be found throughout the peritoneal cavity, and particularly in the peri-hepatic area, giving rise to cyclical right subcostal or peri-gastric pain.

Ovarian endometriotic cysts, or endometriomas, are haemorrhagic cysts with “chocolate” contents located in the ovaries. They can be bilateral, i.e. on both sides of each ovary, in about a third of cases. Ultrasound can reliably diagnose endometriomas when they are larger than 10 mm. MRI is a powerful examination for diagnosing endometriomas (sensitivity* 95%, specificity** 91%). Endometriomas can sometimes present clots or pseudo-fibrin walls. Their appearance is typical on MRI. On ultrasound, the appearance is suggestive (light grey in colour, rather homogeneous) but there are other cysts which may have the same appearance (mucinous or dermoid cysts for example).

The ovaries may be adherent, and displaced in contact with the uterus, usually behind it, giving a “kissing ovaries” appearance

* Sensitivity: probability of obtaining a positive test on an individual with the disease
** Specificity: probability of obtaining a negative test on an individual without the disease

Endometriosis of the fallopian tubes may be associated with infertility and may result in haemorrhagic dilatation (haematosalpinx) or fluid dilatation (hydrosalpinx).

The appearance is that of a dilated tube, resulting in a tubular image, para-ovarian in location. It is sometimes difficult to differentiate the tube from the ovary when the adhesions are significant and create a tubo-ovarian complex.

This is the presence of endometriosis below the peritoneal surface, which may be posterior (torus, uterosacral ligaments, rectum, vagina), lateral (sacro-genitopubic blades, sciatic notch) or anterior (round ligaments, bladder).

For deep involvement, pelvic MRI is more sensitive (Sensitivity = 95%) than pelvic ultrasound.

Thus, a negative pelvic MRI can exclude deep pelvic endometriosis lesions with a performance close to that of surgery (sensitivity greater than 90%).
Deep endometriosis is shown on ultrasound and MRI as an infiltrating nodule with irregular, fibrous contours, with the possibility of haemorrhage on contact.

Ultrasound has a sensitivity of 80% and a specificity of 85%, MRI has a sensitivity of 86% and a specificity of 84% for the diagnosis of endometriosis of the torus and uterosacral ligaments. Invasion of an uterosacral ligament is diagnosed in the presence of a thickened ligament of more than 3 mm, with a nodule of regular contours. It is not very specific when this aspect is isolated. It is important to correlate an isolated thickening of the torus or uterosacral ligaments with the examination, in order to avoid false positives (symptoms and data from the clinical examination, particularly the vaginal touch).

The diagnosis is made in the presence of a nodule or thickening of the vaginal wall or cul-de-sac, sometimes haemorrhagic.

It is defined by the involvement of the muscularis propria of the digestive, colonic or small intestines. The location in the rectum or colon is by far the most frequent, representing 90% of digestive tract infections. It is multifocal in 40% of cases, i.e. it may be located in several parts of the digestive tract.

Ultrasound is effective in seeing digestive lesions located close to the probe, i.e. in the region of the rectum and the recto-sigmoid hinge, with a high sensitivity and specificity (sensitivity 97%, specificity 60%). The most accurate examination for the assessment of low rectal or sigmoidal digestive lesions is rectal echo-endoscopy, performed after local anaesthesia. This examination is limited by the accessibility of the lesion in depth.

MRI is also a good examination for the assessment of digestive endometriosis, especially in the recto-sigmoid region (sensitivity 92% specificity 96%). Certain regions are more difficult to analyse (caecum, transverse colon, small bowel), and are better assessed by entero-MRI, requiring digestive preparation and specific acquisition.

Ultrasound and MRI findings can be seen as a nodule or parietal thickening.

Anterior involvement is rarer than posterior involvement (6% of patients).

The vesico-uterine recess may be affected superficially and in isolation. In this case, only MRI will reveal lesions, visible as simple spots without nodules.

An attack on the bladder wall, visible on ultrasound and above all on MRI, results in a thickening or a nodule of the bladder wall, either at the level of the dome or laterally. Lateral involvement may obliterate the ureteral meatus and be responsible for ureteropylocal dilatation which may impact on renal function.

Only thoracic or diaphragmatic MRI can explore this location, which is suggestive on clinical examination. MRI performed during the menstrual period increases diagnostic performance. Involvement is discrete, with spots most often located on the right diaphragmatic dome.

It is usually located in a surgical scar, laparoscopy or caesarean section, or in the inguinal recesses or umbilicus. It appears as a fibrous nodule with irregular contours. The appearance is non-specific, and the main differential diagnosis is desmoid tumour. The examination is very suggestive, with a menstrual recrudescence of pain.

Only MRI allows exploration of this location, and results in a spiculated nodule, with spicules that may come into contact with lumbosacral nerve endings.

Other rare locations are possible (muscles, brain, bronchi, pericardium, etc.). The diagnosis is essentially clinical, and will be evoked in the presence of various symptoms of menstrual recrudescence.

There are no official recommendations for systematic follow-up imaging. Some teams follow up every one or two years, others request imaging when there is a new symptom.

MRI and ultrasound are complementary, and may be prescribed alone or in synergy, depending on the time, the history of the disease, the locations, or the treatments.

Written and scientifically validated by Dr Corinne BORDONNE

Radiologist specialist, Imagerie Femme Enfant Centre, IMPC Bachaumont and Hotel Dieu Hospital in Paris