Combining Ultrasound and Biomarkers to Diagnose Superficial Endometriosis

Introduction Endometriosis is a common benign disease characterized by the growth of uterine cells outside of the uterus, causing various forms of pelvic pain and/or difficulty with conceiving a pregnancy, amongst other symptoms. Endometriosis most commonly occurs in the forms of Superficial Endometriosis (SE), Deep Endometriosis and Ovarian Endometriosis (Endometrioma) (OE). Ultrasound diagnosis of DE and OE has become more reliable with advances in ultrasound technology, technique and expertise leading to decreased diagnosis time for patients and allowing for better optimization of surgeries if required. SE, however, lacks any reliable non-invasive diagnosis methods.

SE is the most common form of endometriosis and is defined as a disease that lines the peritoneum and is small and superficial in nature, leading to chronic inflammation, infertility, and pain. SE is difficult to visualize on ultrasound due to its size and alignment with tissue, requiring fluid in order expand the pelvis and partially suspend these small lesions, allowing them to be diagnosed through ultrasound. Leonardi et al. observed that in some patients, a physiologic change occurs whereby fluid fills the pouch of Douglas (POD) allowing increased visualizing during ultrasound. This led to the development of Saline-infused sonoPODgraphy (SPG), a novel method utilizing modified commonly used ultrasound techniques called saline-infusion sonohysterography (SIS) and hysterosalpingo-contrast-sonography (HyCoSy), to reliably visualize the POD. The ability to visualize the POD is important as SE is often deposited in the POD.

Biomarkers are also an active area of research in endometriosis diagnosis. There are possible biomarkers that have been identified in the peritoneal fluid of people with endometriosis, including Tumour Necrosis Factor (TNF) and interleukin-6 (IL-6). Indeed, a landmark biomarker study in 2002 demonstrated that TNF in peritoneal fluid had a 100% sensitivity for identifying women with endometriosis. However, this finding has never been further evaluated or validated, even though there is evidence of TNF involvement in endometriosis development. Recently, insulin-like growth factor 1 (IGF-1) has been noted to be elevated in the peritoneal fluid of people with endometriosis compared to those without. Further, IGF-1 has also been found to contribute to the development of abnormal nerves, which may explain the link between endometriosis and pain.

The previous work with POD imaging combined with the knowledge of endometriosis biomarkers leads to the question: Can these non-invasive diagnostic methods be combined to diagnose superficial endometriosis accurately and consistently. We hypothesize that using a combined approach of SPG and biomarker analysis to diagnose superficial endometriosis will provide sufficient evidence to support a large multi-centre international diagnostic accuracy trial.

Technical Study Summary

Study Design

This diagnostic accuracy feasibility study is a Health Canada-sponsored study whereby fluid will be injected through the cervix among recruited and consented participants. During routine clinical visits at the McMaster Women's Health Clinic, participants will be recruited on a rolling basis. The study will start in October 2022 and will continue until the intended sample size is reached and/or January 2024

Sample size

In this feasibility trial, we aim to recruit all patients who meet the eligibility criteria over 14 months from the McMaster University Medical Center. Given the previous publication on SPG and SE by Leonardi et al (2020), based on an estimated prevalence of SE of 60 % and sensitivity of 80 %, and a half-width of 95 % CI of 0.15 (i.e. the sensitivity cannot be lower than 0.65), a total of 50 participants will be required.

Variables

In addition to data regarding index and reference standard tests, demographic and clinical information that is routinely collected will be obtained, including age, BMI, gravidity, parity, history of infertility (as defined by 1 year of trying to conceive without success), past medical history, past surgical history, use of hormonal agents for contraception or treatment of gynecological concern, menstrual characteristics (last menstrual period (at time of TVS-C and surgery), cycle frequency, duration of menses, abnormal uterine bleeding). Previous sexually transmitted and blood-borne infections (STBBI) results will be assessed as potential differential diagnoses when adhesions are present in the absence of endometriosis upon surgery.

During TVS, the disease will be characterized for each patient using the Leonardi et al (2020) modified descriptors for SE under TVS, including size, location, hyperechoic projections, hypoechoic areas, cystic areas, filmy adhesions, peritoneal pockets, and jiggling of the lesion. These characteristics will then be compared to surgical findings using the validated and internationally recognized WERF Surgical questionnaire as categorical data. Due to previous findings in the Leonardi et al (2020) study suggests that SPG may be unsuccessful (SPG fluid filling the POD) in a small portion of the population (5.6%; excluding those with POD obliteration), and the rate of success will be recorded and reported as part of the diagnostic accuracy and indicative of the efficacy of the technique. However, in the event of unsuccessful SPG, tubal patency assessment will be performed at laparoscopy with dye chromotubation to assess whether the bilateral tubal blockage was the reason for the failure of SPG. As there is a slight possibility of having physiological fluid accumulated in the POD before SPG, an estimate of volume will be made under TVS for those with fluid present (calculated using a length x width x height approach, which is the generally accepted method to calculate the volume of pelvic fluid). The volume of SPG fluid injected will be recorded as a continuous variable for each participant. Upon the completion of the study, the volumes will be categorized into groups, allowing for subgroup analysis.

Following the TVS-C, participants will be asked to complete a short questionnaire on their tolerability and acceptance of the procedure. Similarly, following laparoscopy, patients will be asked to answer the same series of questions regarding the laparoscopy and to re-evaluate the TVS-C, having had both procedures to compare.

Statistical analysis

Sensitivity, specificity, positive, and negative predictive values will be calculated for SPG and identified biomarkers in the aspirated fluid. The demographic and clinical variables, peritoneal fluid features and TNF levels, TVS and surgical findings will be described descriptively using standardized terminology for all participants per routine clinical guidelines and assessed as categorical data upon data cleaning. The success of the acquisition of peritoneal fluid will be described. Subgroup analysis will be done with those who had serosanguinous peritoneal fluid.