Endometriosis: advances and controversies in classification, pathogenesis, diagnosis, and treatment

A. Implantation theory

In 1927, Sampson¹ proposed a retrograde flow of the menstrual mix of blood and full endometrial tissue through the Fallopian tubes into the peritoneal cavity as the first step in the development of the disease. Brosens and Benagiano² suggest that the first retrograde bleeding occurs at birth, when the newborn girl has drastic hormonal deprivation. Tight internal uterine cervix os, thick cervical mucus, or malformations impede the normal external drainage of that mixture, which Brosens and Benagiano consider a source of stem cells. This results in the passage of that content into the abdominal cavity. These first implants will remain dormant because of the lack of estrogens in childhood. They shall grow rapidly after puberty, when the ovaries start to produce sexual hormones.

B. Celomic theory

According to Burney and Giudice³, “celomic metaplasia involves the transformation of normal peritoneal tissue to ectopic endometrial tissue”. Endocrine-disrupting chemicals might play an important role in such transformation. Addressing the theory of Müllerian rests, the authors state that residual cells from the embryonic Müllerian duct migration “maintain the capacity to develop into endometriotic lesions under the influence of estrogens”³.

Endocrine, immune, and stem/progenitor cells and epigenetic modifications “must be considered in the context of genetic background as well as stimulus driven reprogramming of the female reproductive tract”³. Even extrauterine stem/progenitor cells derived from bone marrow are suggested to be possible sources of ectopic endometriotic tissue²⁰.

C. Inflammatory disease

Dmowski (cited by Burney and Giudice³) suggests that there is evidence that endometriosis is, in fact, a pelvic inflammatory condition. A “peritonitis without germs”? The peritoneal fluid has an increased concentration of activated macrophages and an inflammatory profile in the cytokine/chemokine axis. Zimmer, in the review by Burney and Giudice, is reported to link a haptoglobin-like protein (that binds macrophages and reduces their phagocytic capacity) to the genesis of endometriosis. Increased production of interleukin-6 (IL-6), macrophage migration inhibitory factor, tumor necrosis factor-alpha, IL-1b, IL-6, and IL-8 alterations is also described. Gargett et al.²¹ propose that human endometrium regenerates cyclically every month mediated by endometrial stem/progenitor cells such as CD140b⁺, CD146⁺, or SUSD2⁺ endometrial mesenchymal stem cells (eMSCs). N-cadherin⁺ endometrial epithelial progenitor cells and side population cells would also contribute to the pathogenesis of the disease. They are planted retrogradely at the time of birth or at puberty. The authors propose that the eMSCs may have a role in the generation of progesterone-resistant phenotype endometrial stromal fibroblasts. Stem/progenitor cell differences between healthy women and those with endometriosis have been proven.

D. Endometriomas

On the genesis of endometriomas, several theories have been updated by Rizzello and Coccia²²:

E. Aromatases

A rare case of postmenopausal hepatic flexure colon DIE, in a woman who had previously undergone total hysterectomy and bilateral salpingooforectomy, was recently presented by Snyder et al.²³. They propose that autologous aromatase production at the site gave rise to a full-thickness infiltrating nodule from remnants or metaplastic endometrial tissues. They considered this after checking that, without any surgery at all, the nodule vanished after prolonged anti-aromatase treatment.

F. Hormonal receptors

Nisolle et al.²⁴ found “heterogeneity of estrogen receptor α and progesterone receptor distribution in lesions of DIE in untreated women, or during exposure to various hormonal treatments”. This could be because DIE nodules are poor responders to different endocrine treatments.

G. Deep infiltrating endometriosis

Petraglia and Chapron²⁵ consider DIE a different phenotype of the same disease, shared with endometriomas and peritoneal lesions. It includes two locations: anterior compartment disease (bladder) and posterior compartment disease (vagina, uterosacral ligaments, rectum, and ureters).

Some invasive mechanisms characteristic of endometriosis, such as the expression of matrix metalloproteinases and activins, are enhanced in DIE. Also, a very high expression of the different mechanisms of neuroangiogenesis (nerve growth factors, vascular endothelial growth factor, and intercellular adhesion molecule) is present.

For them, other immunological factors (peritoneal macrophages, natural killer cells, and lymphocytes) are critically altered in DIE. The aggressive behavior of DIE may be explained by the highly decreased apoptosis. Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), B-cell lymphoma 2 (Blc-2), anti-Müllerian hormone, and the increased proliferation activity related to oxidative stress (NF-κB, reactive oxygen species, extracellular regulated kinase, and advanced oxidation proteins) also contribute.

H. Epigenetic modulators

In a recent “master review”, Gordts, Koninckx, and Brosens elaborate two different pathogenic hypotheses²⁶:

Hypothesis I: pathogenesis of early-onset endometriosis by neonatal uterine bleeding with the cyclic menstruation is the driving mechanism for adenomyotic nodule formation.

Hypothesis II: deep endometriosis is a specific type of abnormal endometrium-like cell benign tumor.

As proof of this, the authors show that deep endometriosis is a specific disease, as reflected by the distribution of deep lesions in all stages of the Revised American Fertility Society classification.

It might share with peritoneal or cystic endometriosis the same cellular origin, but genetic and epigenetic modulators induce distinct presentations of the disease. In some cases, peritoneal endometriosis will prevail. With other epigenetic modulators, DIE will grow.

They have common structures when analyzed by the pathologist. These authors propose that uterine adenomyosis and DIE have common origins, as in both cases glands are seen infiltrating muscle tissue.

A. Anamnesis

Listen to the patient. Carry on a detailed anamnesis in a very slow fashion. This simple action gives us the best approach to the disease. She has so much to tell, to show with her face and expression. In most cases, the disease can be understood just by listening.

The omnipresent symptom is pain: cyclic pelvic pain, dysmenorrhea, periovulatory pain, chronic non-cyclic pelvic pain, dyspareunia (positional or permanent), dyschezia, and dysuria.

There are many other pain presentations that nobody even thinks of until confronted with an endometriosis patient who, incidentally, has exactly “that type of pain”. A young girl we operated last year referred to right shoulder pain at menstruation. At laparoscopy, a large diaphragmatic series of blue and red lesions was excised. She was relieved after surgery.

A similar case was reported recently by Singh et al.²⁷. This publication elucidates the use of MRI for the clinical diagnosis of endometriosis, which will be shown extensively in this review.

Involuntary infertility, even when not the cause for consultation, should also be regarded as one of the frequent symptoms of endometriosis. Less frequently, cyclic nasal bleeding, umbilical bleeding, cyclic hemoptysis, cyclic constipation, and urinary urgency are reported by patients with endometriosis.

B. Pelvic examination

Even today, with the advancement of imaging diagnosis, pelvic examination (in expert hands) continues to be praised as an effective clinical tool for the diagnosis of endometriosis. It should be done with care, slowly, beginning with abdominal palpation. Only after no pain is registered, proceed to pelvic examination. This should be done with extreme delicacy and respect. Bimanual palpation of the uterine/bladder pouch, the Douglas pouch, and adnexa can reveal exquisitely painful sites typical of endometriosis.

Fixed uterine retroversion is frequently due to uterosacral ligament compromise or adhesions at the Douglas pouch. Painful uterine mobilization is another typical sign of endometriosis. Compression of the uterine fundus is frequently painful when adenomyosis is present. Dyspareunia frequently corresponds with extremely painful palpation of the uterine-sacral ligaments.

Always look at your patient’s face during examination. Rictus of pain cannot be avoided. It will tell you exactly where the pain is more intense, helping to clinically determine the extent of the disease. Careful and expert pelvic examination provides a lot of information at a very low cost.

C. Biomarkers

As of today, of the many biomarkers for endometriosis proposed in peripheral blood and endometrium, not one has been validated for endometriosis¹⁸. This could be due to patient selection, sample collection, or analytical procedures. There is a current need to develop a non-invasive test for patients with symptomatic endometriosis.

We still lack a reliable marker for the disease. Ca 125, considered a marker for endometriosis, is helpful only in postoperative follow-up. It usually decreases after surgery and rises when the disease recurs or progresses.

Clinical presentations vary. Signs, symptoms, and markers do not correlate well with the extent of disease, as stated by Taylor et al.¹⁹. In 58 consecutive cases of endometriosis, Hirsch et al.²⁸ found increased values of Ca 125. This group concluded that Ca 125 of at least 30 units per milliliter is “highly predictive of endometriosis” in symptomatic patients¹⁹. The authors propose it as mandatory but consider it “unable to rule out endometriosis”¹⁹.

Many publications describe gene abnormalities in patients with endometriosis. It would take a whole chapter to name them but none has yet been validated for the diagnosis of endometriosis. These alterations have been reported for the last 15 to 20 years. Some are showing ties with the disease. The large number of different approaches shows that the road is still unclear.

In 2016, after a systematic search of the literature, Neil Johnson, Cyndy Farquhar, and the Cochrane Library group found only two biomarkers—PGP 9.5 (neural fiber marker) and CYP19 (hormonal marker)—that showed enough accuracy to replace surgical diagnosis²⁹. Even so, the authors state that “we could not statistically evaluate most of the biomarkers assessed in this review in a meaningful way. In view of the low quality of most of the included studies, the findings of this review should be interpreted with caution. Although PGP 9.5 met the criteria for a replacement test, it demonstrated considerable inter study heterogeneity in diagnostic estimates, the source of which could not be determined”²⁹.

Blood, urinary, and endometrial markers—alone or combined with imaging—were analyzed. The authors conclude that none could be evaluated in a meaningful way. For them, there was insufficient or poor-quality evidence. There is a clear final recommendation: “Laparoscopy remains the gold standard for the diagnosis of endometriosis and using any non-invasive tests should only be undertaken in a research setting”²⁹.

D. Genetics

For many years, there has been a search for genetic testing that could identify a population prone to develop endometriosis. A simple literature search identifies more than 3000 publications from 2018 linking genetics to endometriosis. Recently, an Australian group presented a summary of 17,045 cases included in a meta-analysis³⁰. In them, 14 genomic regions were identified, supported by results from multiple studies. The group found that “no independent associations were identified from direct genotyping of common and low-frequency protein-coding variants”³⁰. According to them, the most common genetic factors related to endometriosis risk are located in regulatory DNA sequences. This, they say, alters the regulation of gene transcription. They conclude that the target genes are present in three chromosome regions: “LINC00339 and CDC42 on chromosome 1, CDKN2A-AS1 on chromosome 9, and VEZT on chromosome 12”³⁰.

Using single-nucleotide polymorphism (SNP) array technology, a 2017 publication³¹ describes genomic aberrations linked to the development of endometriosis. These investigators performed SNP array genotyping of pooled DNA samples from 100 patients with endometriosis and 50 controls. The authors detected 49 copy number variation (CNV) loci that were present in patients with endometriosis but that were absent in the control group. Six novel CNV loci in the subtelomeric regions representing gains and losses were identified. An intergenic locus on chromosome 19q12.1 showed a strong association with endometriosis. As with other biomarkers, we still lack a reliable genetic marker for endometriosis, and none of the proposed genes or gene alterations can be used to make a precise diagnosis.

E. Imaging

Ultrasound. In 1979, Walsh et al. presented their findings in 25 patients with surgically confirmed endometriosis or adenomyosis or both³². Sonolucent zones within the uterus representing blood lakes described adenomyosis. Other cases had cystic images, five of which were of mixed characteristics. At that time, “ultrasound alone could not differentiate endometriosis from diseases such as tubo-ovarian abscess, ruptured ectopic pregnancy, other ovarian cysts or tumors”³². The authors stated that the clinical history contributed to the non-surgical diagnosis of endometriosis.

Today, some authors state that TVS “allows a better accurate diagnosis of rectosigmoid endometriosis than MRI”³³. For this group, it is less reliable in the case of uterine, Douglas pouch, and uterosacral ligament disease. Nevertheless, they propose it as a first-line imaging technique because of its low cost and feasibility.

The International Deep Endometriosis Analysis group³⁴, confronting the wide variety of terms and descriptions used to identify endometriosis at TVS, proposes some basic steps that should be followed at the time of examination:

All steps should be performed, though not necessarily in this order, with a small liquid content in the bladder. A dynamic examination assessing the real-time mobility of the pelvic organs is mandatory in these cases.

This article includes a series of drawings and photos that accurately describe the different images related to endometriosis in all of its presentations. For those who practice the steps mentioned above, TVS is the first-line investigative tool in patients with symptoms of endometriosis. The ability demonstrated by them to detect ovarian endometriomas and DIE is well documented. The prediction of the pouch of Douglas obliteration is very accurate. It helps to organize multidisciplinary surgical teams in the most severe cases. They give most importance to the sliding sign since it allows clinicians to predict the severity of the deep pelvic disease. One possible drawback is the issue of experience: only those who have performed more than 2500 scans can achieve real proficiency in the sliding maneuver, after about 40 examinations. Any trained staff can manage this non-invasive diagnostic method for other locations of DIE except for rectovaginal septum DIE. A plight for a standardized nomenclature of US findings is mandatory for this group.

Another group presented clear and sound images of DIE in a prospective study³⁵. They evaluated the wall of the rectum and the lower sigmoid colon with two consecutive TVSs. The first was performed without previous bowel preparation, and the second after a 3-day low-residue diet and two 250-mL enemas (12 and 3 hours before TVS). They demonstrated that TVS after bowel preparation had a higher accuracy, allowing the detection of DIE before surgery.

Transvaginal US is the first option for the imaging diagnosis of ovarian endometriomas. A 2002 meta-analysis performed by Moore and Kennedy et al.³⁶, reviewing seven articles that fulfilled the inclusion criteria, demonstrated that TVS is a useful test in the case of ovarian endometrioma.

In a retrospective observational study, an Italian group³⁷ used TVS to evaluate 250 women of reproductive age (20 to 40 years) presenting endometriomas larger than 20 mm in diameter. The mean endometrioma diameter was 40 mm. Bilateral disease was found in 25.5% of the patients, posterior rectal DIE in 21.5%, and the thickening of at least one uterosacral ligament in 35.4%. Seventy-three percent of the patients showed adhesion signs, and 53% had concurrent uterine adenomyosis.

Only 15% of the studied population presented a single isolated endometrioma with a mobile ovary and no other signs of further disease. This publication highlights the utility of TVS for the diagnosis of endometriomas, adjacent coexisting disease in other locations, and adherences. In 50 patients operated by laparoscopy, TVS diagnosis was confirmed.

In 85% of the cases, endometriomas were associated with other locations of endometriosis. Left-side cysts were more frequently associated with same-side uterosacral ligament infiltration and DIE. Bilateral endometriomas usually obliterate the pouch of Douglas.

Computerized axial tomography. “Computed tomography has no role in the routine evaluation of endometriosis except in very few particular scenarios”³⁸. An inguinal endometriotic nodule and a case of round ligament endometriosis that looked like a hernia were the only references found after a quick search of different databases, including Medline, linking endometriosis and computerized axial tomography (CAT) scans. Contrast studies might be of use for the diagnosis of ureteral stops, stenosis, or deviations in the case of lateral pelvic side-wall DIE.

CAT virtual colonoscopy can also be of help. A recent study describes its use before surgery for DIE³⁹. Associated with MRI, the preoperative diagnosis was confirmed in 71 patients who presented a total of 105 endometriotic bowel lesions. This group found 71.2% rectal nodules and 60% sigmoid nodules that infiltrated the muscularis propria in extensions varying from 25% to 50% of the circumference. Stenosis was present in 73% to 96% of the cases.

“The concordance between intraoperative and preoperative findings concerning the presence of rectal nodules was high, at 0.88 when associating CTC [computed tomography of chemiluminescence] with MRI, whereas each imaging technique taken individually provided lower concordance coefficients”³⁹. In this study, 80.3% of patients underwent the procedure that had been preoperatively planned. These authors propose that the association of both techniques improves the accuracy of preoperative assessment of colorectal DIE.

Magnetic resonance imaging. In 1999, a pioneer article described the use of MRI for the preoperative diagnosis of endometriosis⁴⁰. The authors described, in 20 patients, MRI findings of DIE at the uterosacral ligaments, the pouch of Douglas, the rectum, and the bladder that were histologically proven at surgery. Diagnosis was accurate except when contrast was not used (two of three patients with rectal endometriosis).

A decade before, Arrivé et al.⁴¹ published the first report of MRI use for the clinical diagnosis of endometriosis. Using only 0.35 Tesla, they prospectively studied 30 consecutive women with symptomatic disease. In 25 cases, endometriosis was confirmed by surgery. A sensitivity of 64% and a specificity of 60%, with an accuracy of 63%, were shown. Most endometriomas were correctly identified. Only 14 of 29 cases of adhesions and 6 of 45 cases of peritoneal implants were diagnosed by MRI. “MRI findings did not correlate with the surgically determined severity of the disease”⁴¹. In 1989, the authors concluded that MRI could not be used as the first study to detect endometriosis. For them, laparoscopy was the procedure of choice.

Last year, at the XIII World Congress on Endometriosis, our group presented three posters^42–44 that established a perfect correlation of high-resolution contrast MRIs with laparoscopic findings, using 2- and 3-Tesla devices. MRIs and intraoperative images from one of the posters are displayed in Figure 1.

Figure 1. High-resolution magnetic resonance imaging and laparoscopic findings.

Using a high-resolution technique (1-mm slices), intravenous contrast (for bladder visualization), and vaginal and rectal gel contrast (for better visualization of the rectovaginal septum), we were able to stage the disease before laparoscopy. We demonstrated the special ability of this method to visualize superficial implants, adhesions, uterosacral ligament infiltration, rectovaginal septum infiltration (including the depth of rectal invasion), bladder wall infiltration, and ovarian disease. Images of ureteral compromise were also obtained (unpublished). Whenever possible, MRI would be mandatory before laparoscopy.

A recent publication⁴⁵ shows an interesting algorithm that allows clinicians to predict the probability of bowel resection at the time of laparoscopy for DIE using MRI. In 52 patients studied preoperatively, a positive predictive value of 87% and a negative predictive value of 83% were demonstrated. This group calculated the impact angle and lesion size by using a mathematical algorithm.

In 2009, the PRISMA (Preferred Reporting Items of Systematic reviews and Meta-Analyses) group proposed an evidence-based minimum set of items required for reporting in systematic reviews (http://www.prisma-statement.org). A recent publication⁴⁶ evaluated the use of TVS and MRI for the diagnosis of adenomyosis, reviewing evidence in accordance with PRISMA requirements.

For TVS, they found a high heterogeneity between studies. The pooled positive likelihood ratios for adenomyosis were 0.72–0.82, 0.85–0.81, and 4.67–3.7, a clearly diffuse variety of information. In contrast, the pooled sensitivity for MRI was 0.77, the specificity 0.89, the positive likelihood ratio 6.5, and the negative likelihood ratio 0.2 for all subtypes of adenomyosis and publications. This suggests that MRI is more useful than TVS for the diagnosis of adenomyosis.

Revised American Society for Reproductive Medicine classification

In Figure 2, the rASRM classification is deployed. It was originally thought of by Acosta et al.⁴⁸ in 1973 while working with V. C. Buttram and after analyzing the results of 107 infertile patients operated for endometriosis. This group designed a simple scheme that evolved into the first ASRM classification. This classification is intended mainly to be used for those endometriosis patients consulting for infertility. DIE is not considered in this scheme.

Figure 2. American Society for Reproductive Medicine revised classification.

Enzian

The Enzian proposal, which includes DIE, is shown in Figure 3. Clear drawings help the surgeon to better stage the disease. This classification addresses the issues of the posterior compartment DIE and bladder (anterior) and ureteral (lateral) compartments.

Figure 3. The Enzian classification for deep infiltrating endometriosis^51–53.

Fertility index

Adamson considered the poor fertility prognosis derived from the exclusive use of the ASRM modified classification. In 2010, he introduced the EFI, shown in Figure 4⁴⁹, as a complement that allows a better diagnosis of the endometriosis-associated fertility status. Validated by several other authors, such as Hobo et al.⁵⁰, this index includes not only the laparoscopic findings (least functional score at the end of surgery) but also other issues that affect fertility, such as the length of infertility, patient’s age, and previous pregnancies.

Figure 4. Endometriosis fertility index.

Drawbacks

A major drawback of all existing classifications is that no one of them links the severity of the pain with the findings. Some patients who would be classified as “severe” by ASRM revised charts experience little pain but do not get pregnant. Others, with only superficial red and blue lesions and minor adhesions, experience severe pain and a poor QOL. They probably would be entered as “mild” or even “minimal” using this proposal. The same occurs with DIE and pain. The question is: Should we include pain as a new item when establishing a definitive classification? Adamson, in a recent update on classifications⁴⁹, reports that the American Association of Gynecological Laparoscopists is developing a categorization system that will be more focused on pain.

Koninckx and Wattiez et al.

In 2011, Koninckx and Wattiez et al.⁵⁴ presented a proposal that included adenomyosis, peritoneal pocket lesions, and subtle endometriosis plus the three traditionally recorded lesions (peritoneal, cystic, and deep). It considers the size of each lesion and includes pain as an issue. They state that the significance of the subtler lesions is not clear since they have not “been demonstrated to be associated with pain or infertility”⁵⁴.

For them, subtle lesions and DIE (any lesion deeper than 5 mm under the peritoneal surface) should be classified apart. In regard to pain, they cite authors who link pelvic pain with lymph node involvement in the case of DIE. That is, lymph compromise is a marker of more intense pain. In Figure 5, we depict their classification proposal, which is not yet validated.

Figure 5. A new proposal by Koninckx and Wattiez et al.⁵⁴.

Other proposals

In 2011, Charles Miller and Mauricio Abrao presented an article entitled “Six Good Reasons for a NEW Endometriosis Classification”⁵⁵. They propose that a classification should do the following:

These concepts should be included if (ever) a single, global classification is finally created.

Overview

Vercellini and Somigliana et al. summarize current available medical treatments for symptomatic endometriosis in a recent publication⁵⁶. According to them, they all act by inhibiting ovulation, reducing serum estradiol levels, and diminishing uterine blood flow. They state that several drugs can be used “with a similar magnitude of effect, in terms of pain relief”⁵⁶. They can be categorized by price, as low-cost drugs (OCs and most progestogens) or high-cost drugs (dienogest, GnRH agonists, and the newly introduced elagolix). They recommend starting treatment with low-cost drugs and step up to high-cost ones only in case of “inefficacy or intolerance”⁵⁶. For them, OCs are useful for superficial peritoneal disease or endometriomas smaller than 5 cm in diameter, and progestogens have a better effect in severe dyspareunia associated with deeper infiltrating lesions.

Back in the late 1970s, danazol NR was “the” drug. It relieved pain after diagnostic laparoscopy or conventional ovarian cystectomy—not much more was accomplished in the operating room (OR) at that time. Patients gained weight, grew a beard, and had elevated hepatic enzymes, but pain was gone, and progression probably arrested.

In 1988, Taymor et al.⁵⁷ questioned its clinical efficacy in infertile women. This prospective randomized study showed danazol to be ineffective in improving pregnancy rates over doing nothing at all in patients with minimal endometriosis. Years later, the World Consensus for the Current Management of Endometriosis¹¹ recommended it only before IVF in severe endometriosis.

Danazol is a drug of the past. OCs, progestogens, GnRH agonists, and lately GnRH antagonists (elagolix) and (to some extent) hormonal receptor modulators (such as ulipristal) are current specific medications. The role of antiestrogens is not clear, nor is that of natural origin substances (such as resveratrol), anti-aromatases, anti-angiogenic molecules, and immunomodulators.

Pain

Pain is the most common reason for consultation; moreover, infertile patients often have chronic pelvic pain as well. Arrest of pain is always a priority.

We presented a poster at the XI World Congress on Endometriosis (Montpellier, France, 2011) and posters at the XII World Congress on Endometriosis (São Paulo, Brazil, 2014) showing the immediate pain relief and positive changes in over 100 studied patients. Their QOL improved immediately after a correct laparoscopic surgery in superficial endometriosis of different severities.

Two surgeons—Jose M. Curto and myself—participated. The quality of surgery was standard and reproducible. Surprisingly, 100% of the patients indicated an immediately better QOL after surgery. In all cases, postoperative treatment was allocated to control pain or infertility or both.

The traditional management of pain includes different approaches and different pathways. The group headed by Catherine Allaire states that, other than non-steroidal anti-inflammatory drugs (NSAIDs), “further work is required for nonhormonal therapies such as antiangiogenic and immune-modulating drugs”⁵⁸. They list as “traditional hormonal therapies” estrogen-progestin contraceptives, progestins, and GnRH agonists. Other hormonal treatment options are androgens and aromatase inhibitors. “Research also suggests a possible role for GnRH antagonists and selective progesterone receptor modulators, always discussing with each patient side effects and/or desire for pregnancy to ensure personalized treatment and optimal outcomes”⁵⁸.

Curiosities still occur in this field, such as the method used by George et al.⁵⁹ to treat persistent endometriosis chronic pain in a 32-year-old patient. That group administered 10 daily, 20-min sessions of 2-mA anodal transcranial direct current stimulation over the left primary motor cortex. In their experiment, visual analog scale pain symptoms were reduced by 60%, and at the 4-month follow-up, the patient still expressed an overall decrease in pain symptoms of 30%.

The European Society for Human Reproduction and Embryology (ESHRE)⁶⁰ proposes a clear and simple guideline. It recommends, as a Good Practice Point, to “counsel women with symptoms presumed to be due to endometriosis thoroughly, and to empirically treat them with adequate analgesia, combined hormonal contraceptives or Progestagens”⁶⁰. It nevertheless states that “it is clearly a paradox that by recommending empirical treatment in symptomatic (young) women” we might induce a delay in diagnosing the disease⁶⁰.

Hormones. The guideline recommends prescribing hormonal treatment—combined OCs, progestagens, anti-progestogens, or GnRH agonists—“as one of the options, as it reduces endometriosis-associated pain”⁶⁰. It also “recommends that clinicians take patient preferences, side effects, efficacy, costs and availability into consideration when choosing hormonal treatment for endometriosis-associated pain”⁶⁰.

Combined estrogen and progestin OCs are recommended, as they reduce endometriosis-associated dyspareunia, dysmenorrhea, and non-menstrual pain, in a continuous protocol. Also, vaginal contraceptive rings or transdermal estrogen/progestin patches are suggested.

Recommended progestogens are medroxyprogesterone acetate (oral or depot), dienogest, cyproterone acetate, and noretisterone acetate. Anti-progestogens such as gestrinone are considered. The different side effect profiles of each one of those drugs, especially thrombosis and androgenism, should be regarded. The use of a levonorgestrel-releasing intrauterine device is also an option. Aside from difficult cycle control in some users, dienogest in doses of 2 mg per day results in very effective pain reduction and control of the disease progression. It can be used over prolonged periods of time provided that no recurrent bleeding occurs. This is the main cause of discontinuation.

Anti-hormones. According to the guideline, evidence regarding dosage or duration of GnRH agonists is limited. “Clinicians are recommended to prescribe hormonal add-back therapy to coincide with the start of GnRH agonist therapy, to prevent bone loss and hypoestrogenic symptoms during treatment” since this will not reduce the effect of the pain treatment⁶⁰. The guidelines recommend giving “careful consideration to the use of GnRH agonists in young women and adolescents, since these women may not have reached maximum bone density”⁶⁰.

Aromatase inhibitors. These options are considered for those who have pain from rectovaginal endometriosis, refractory to other medical or surgical treatment. They should be prescribed in combination with OC pills, progestagens, or GnRH analogs.

Analgesics. The guideline asks a simple question⁶⁰: “Are analgesics effective for symptomatic relief of pain associated with endometriosis?” Then it tells us that evidence on the use of NSAIDs for endometriosis is scarce. The authors of the guidelines present a referenced publication from 1985 and one study on the cyclooxygenase-2 (COX-2) inhibitor rofecoxib (withdrawn from the market in many countries because of severe side effects). For the guideline, NSAIDs have a “favorable effect on primary dysmenorrhea and are widely used as a first-line treatment of endometriosis-associated pain”⁶⁰. However, it recommends the use of NSAIDs or other analgesics to reduce endometriosis-associated pain. Clinicians should discuss the associated side effects with the patient.

The off-label use of a COX-2 inhibitor, etoricoxib, available in many countries could replace rofecoxib. As was demonstrated long ago⁶¹, its tolerance and fewer upper gastrointestinal clinical events versus the traditional NSAID diclofenac are notable, although this difference was not reported in relation to the rare and very serious adverse side effects that both drugs can produce. Petraglia et al.⁶² studied rofecoxib back in 2004, before it was banned, in 16 patients versus 12 in a placebo group in a dose of 25 mg per day for 6 months. In this group of young patients, no significant side effects occurred. Pain relief was significant and persistent, making this a “safe and low-cost therapy for endometriosis associated pain”⁶². A future preclinical randomized controlled trial (RCT) is needed to replace COX-2 inhibitors in the public awareness as an option, using alternative drugs such as the previously suggested etoricoxib. The Practice Committee of the ASRM states that “first-line medical treatment for pain due to endometriosis is often a nonsteroidal anti-inflammatory drug”⁶³. This everyday practice, difficult to change, affects women without a certified diagnosis of endometriosis.

The Cochrane Library⁶⁴ recently reviewed the use of NSAIDs for endometriosis-associated pain. The Cochrane Gynecology and Fertility Group Specialized Register of Controlled Trials (October 2016), MEDLINE (January 2008 to October 2016), and the Embase project (January 2016 to October 19, 2016) were analyzed. All RCTs describing the use of NSAIDs for the management of endometriosis-associated pain in women of all ages were included. The available evidence was very low, measured by the GRADE system (Grades of Recommendation Assessment, Development and Evaluation, a public domain system that describes evidence as “very low” when the true effect is probably very different from the estimated effect, “low” when the true effect might be very different, “moderate” if the true effect is probably close to the estimated effect, and “high” when the true effect is similar to the estimated effect). Comparison of NSAIDs (naproxen) versus placebo revealed no evidence of a positive effect on pain relief, although women taking NSAIDs (naproxen) were less likely to require additional analgesia. Those studies provided no data on QOL, effects on daily activities, absence from work or school, need for more invasive treatment, or participant satisfaction with treatment. No judgment as to whether NSAIDs (naproxen) are effective in managing pain caused by endometriosis can be made. There is no evidence that one NSAID is more effective than another. As shown in other Cochrane reviews, women taking NSAIDs must be aware that these drugs may cause unintended effects.

Opioid derivatives such as codeine and tramadol have rarely been studied for the treatment of endometriosis-associated pain. Many doctors prescribe them to these women⁶⁵: “Obstetrician–gynecologists reported prescribing a median of 26 opioid pills across all indications combined”. “Opioid-related deaths recently exceeded motor vehicle accidents as the leading cause of injury-related death in the United States”. “In 2015, 2 million Americans had a prescription opioid use disorder, and more than half of those who reported prescription opioid misuse obtained the drugs through diversion of prescribed medications”. The use of this type of analgesics should be limited to very exceptional cases in which other drugs have failed and during the least span of time.

Elagolix

The elagolix phase III clinical trial⁶⁶ introduced a new and promising treatment for endometriosis. Oral anti-gonadotrophic agents have a sound future. They arrest the progression of the disease and dramatically reduce pain. Relief is fast and significant. New clinical trials for similar drugs are ongoing. In the publication of reference, elagolix compared with placebo showed a significant decrease from baseline in the mean pain score. This significant effect was seen at months 3 and 6 of treatment.

One adverse side effect, bone density damage⁶⁷, was dose-dependent. Decrease in lumbar spine bone mineral density (BMD) following 6 months of treatment with elagolix compared with placebo was significantly different for each elagolix dose. “The proportion of participants with lumbar spine BMD decrease from baseline greater than 3% was also dose-dependent”, as was the case of decreases in BMD of the total hip and femoral neck⁶⁸.

Current protocols for similar drugs include addback hormonal therapy to minimize the effect on bone density. How long can the patient be treated with GnRH antagonists that produce medically induced menopause-like status? This is a question for the near future since trials provide only up to 12 months of treatments. With addback therapy, these drugs could probably be used for prolonged periods of time.

On July 24, 2018, the FDA approved elagolix for the management of moderate to severe pain associated with endometriosis.

Recent studies^69,70 address other advantages: dyspareunia and pain reduction across a range of baseline characteristics have been shown to be consistent.

During the first 12 weeks of treatment with either leuprolide acetate (LA) or elagolix, subjects had lower estradiol (E2) levels compared with women who received placebo. Estradiol levels were lowest among LA-treated women. The results were placebo = 0 (0%) reduction, elagolix 150 mg = 0 (0%), elagolix 250 mg = 6 (14%), and LA = 29 (67%). This indicates a need for lower hormonal addback doses, or none, depending on the dose administered.

Proellex

Selective progesterone receptor modulators are a class of drugs with progesterone antagonist activity that may confer therapeutic benefit for reproductive disorders in premenopausal women. Endometrial structure, which is dynamically controlled by circulating sex hormones, is likely to be perturbed by progesterone receptor modulators through their progesterone antagonist properties.

A selective progesterone receptor modulator, CDB-4124, telapristone acetate (Proellex), was clinically studied recently⁷¹. Its performance as a treatment for endometriosis was proven. It is a drug with progesterone antagonist activity. A major concern was the alteration of the eutopic endometrium. The structure of this tissue, dynamically controlled by circulating sex hormones, is altered by progesterone receptor modulators through their progesterone antagonist properties. A group of pathologists examined endometrial histology in 58 premenopausal women whose endometriosis was treated with the progesterone receptor modulator CDB-4124 in daily oral doses of 12.5, 25, or 50 at 3 and 6 months’ follow-up⁷². Most of the endometrial biopsies (103 of 174 biopsies) contained histologic changes that are not seen during normal menstrual cycles. The endometrium was generally inactive or atrophic and, less frequently, proliferative or secretory. Some presented anomalies, including formation of cystically dilated glands, and secretory changes coexisting with mitoses and apoptotic bodies. Increasing daily dose and duration made the cysts predominant and their lining inactive or atrophic. None of the patients who received CDB-4124 developed endometrial carcinoma or hyperplasia while on therapy.

Proellex has been questioned by the FDA because of liver enzyme increases shown by patients who received treatment in the endometriosis clinical trial. Clinical trials were halted and the company that developed the drug is now redirecting future studies to vaginal administration. A partial halt of the studies has not yet been reverted by the drug authority.

Ulipristal

In his doctoral thesis, Simpson⁷², after a 3-month treatment course with ulipristal acetate (Esmya) for endometriosis-associated pain, found a good clinical response in 56% of the studied cohort. Post-treatment histological and immunohistochemistry changes were correlated to changes in the macroscopic appearance of the disease and changes in symptom severity. In this descriptive observational cohort study, ulipristal acetate appears to offer an effective treatment for endometriosis with histological changes in the eutopic endometrium that should be carefully observed. “The safety of this compound remains to be elucidated but the results from this pilot study are encouraging and should prompt further exploration”, wrote Simpson⁷².

A possible relationship with endometrial malignancies and severe liver damage reported excludes it from currently available endometriosis drugs. The FDA recently (August 2018) refused, once again, to authorize Esmya (brand name for ulipristal acetate) for human use. The drug had been preliminarily approved by the EMA in 2015. In May 2018, the EMA issued a warning about the rare occurrence of liver complications. In June 2018, the agency approved its use in the preoperative treatment of fibroids (that is, for a short period of time). There could be a possible role for ulipristal acetate in the management of endometriosis: a surgical pretreatment. In my opinion, long-time use is not guaranteed to be safe at the present time. We must keep in mind that endometriosis is a benign condition and care should be taken when prescribing medications with dangerous side effects, even if rare and infrequent.

In regard to hyperplasia, some authors explain it as a singular endometrial alteration induced by progesterone receptor modulators and not a real hyperplasia. This modification of the endometrial structure would regress after drug cessation.

Resveratrol

Resveratrol, a natural drug derived from grape wine, induces apoptosis in endometrial stromal cells via the suppression of survivin expression. Ines Baranao and her group at IBYME–CONICET (Buenos Aires, Argentina) are working in animal experimentation with endometriosis surgical implants in rats, where they demonstrate the suppressive effect of resveratrol on the progression of the disease. Makabe et al.⁷³ describe how it enhances apoptosis in endometriotic stromal cells. There is a long journey yet to be accomplished but these very preliminary results are promising⁷⁴.