Adenomyosis is an entity that belongs to the family of estrogen-dependent conditions. The histological definition formulated by Bird in 1972 – “adenomyosis is the benign invasion of the endometrium into the myometrium, producing a diffuse enlarged uterus which microscopically exhibits ectopic non-neoplasic endometrial glands and stroma surrounded by the hypertrophic and hyperplasic myometrium” – remains today the main diagnostic pillar(1). Typically, adenomyosis is associated with a symptomatology dominated by pelvic pain (dysmenorrhea, dyspareunia), abnormal uterine bleeding and infertility. Recent research brought consistent evidence regarding the relationship between this condition and obstetric pathologies: miscarriage, preterm birth, premature rupture of membranes and morbidly adherent placentation(2). The interest of specialists for adenomyosis changed radically during the past 20 years, due to the major advances in the field of imaging techniques and the growing enthusiasm for the study of embryo implantation mechanisms and endometrial malfunction.
Despite the impressive amount of data accumulated during this period, adenomyosis remains a controversial topic. The main difficulties arise from its very polymorphic presentation. The role of surgery is also a matter of debate. Moreover, the characteristics of this tumor (lack of a cleavage plane from healthy tissues) bring about potential surgical complications. Most frequently, professionals raise the issue of hemostasis challenge, which may call for aggressive gestures to control bleeding. This in turn leads to supplementary damage to adjacent structures and, eventually, to the reduction and alteration of endometrial function. For this reason, the indication of surgical interventions, such as adenomyomectomy, must be well argumented, with regard to future clinical benefits.
Complex pathogenesis – polymorphic
Studies assessing the underlying mechanisms of adenomyosis established the presence of abnormalities at the endometrium-myometrium interface, as well as a series of alterations in the local control systems – endometrial cell steroidogenic metabolism, immune system, oxydo-reductive system or interleukin production.
The junction zone is frequently targeted (altered/increased thickness) in patients with adenomyosis. In a study conducted in 2011, Exacustos et al. reported a 84% sensitivity and a 75% specificity for a thickness of the junctional zone of more than 8 mm (documented by 3D ultrasound), and a 88% sensitivity and a 83% specificity, respectively, for structural changes(3). Luciano et al. provide similar data for the correlation between changes in the junctional zone visualized by 3D ultrasound, with increased sensitivity/specificity (92%; 83%) when two types of changes were identified(4).
The local production/action of steroids was modified in patients with adenomyosis – aromatase cytochrome P450, estrone sulfatase, 17-OH steroid dehydrogenase type 2, 17-HSD type 2. Somatic point mutations in genes that encode the estrogen receptor ERa (K231R) had been also described in patients with adenomyosis(5). Kitawaki et al. identified the presence of polymorphisms that target estrogen receptor alpha (ERa) genes in patients with estrogen-dependent diseases (endometriosis, adenomyosis, leiomyoma)(6), and Levi et al. reported the different expression of a series of gene products (estrogen/progesterone receptors) in the same category of patients(7).
The production of interleukines (IL) (6, 8, 10, 18, ITF, TNF-alpha, LIF – leukemia inhibitory factor) was proven to be altered in the endometrium, either in the sense of modified levels, or in terms of loss of their cyclic physiologic variations(8,9). As for IL-22, IL-18, IL-10, TNF-alpha and ITF, increases of different magnitude(10) or decreases in LIF(9) were also reported.
The immune system was also considered to be targeted in patients with this condition. There were more uNK (uterine natural killer) cells in the group of infertile patients with adenomyosis compared to the group of patients without adenomyosis(11). Macrophages were reported to be significantly increased in the stroma of the eutopic endometrium of patients with adenomyosis(11). Regarding lymphocytes, one study attests changes in the balance of TH17/Treg in the endometrium of patients with adenomyosis(12).
Several systems involved in the production of oxide radicals were evaluated in patients with adenomyosis: endothelial nitric oxide synthase, cathalase, superoxide dismutase, xanthine oxidase (XO) and glutathione oxidase. There was an increase in oxide radical levels in the groups of patients with adenomyosis(13). At the same time, the variability in oxide radical production during the menstrual cycle described in the control group was absent in the groups with adenomyosis(14).
Identifying patient phenotypes which might benefit from surgery
The complexity of its pathogenesis confirms the theory according to which, under the umbrella of adenomyosis lie entities with different onset and evolutive course. Histologic data follow the same pattern: adenomyosis includes a large number of features – ectopic endometrium (glandular/stromal structures) with behavior either similar or very different from that of eutopic endometrium, fibrotic, respectively mitotic reaction. Thus, adenomyosis has a focal and a diffuse form, solid or cystic, affecting the junctional zone, but also the peritoneal serosa.
In recent years, the importance of establishing a common terminology and of developing a classification that would be helpful for practitioners has become evident. In this respect, we note the classifications proposed by Gords and Brosens(15,16) which, unfortunately, have not been tested in clinical studies. Kishy(17) was the first author to test a clinical classification based on three different mechanisms of development: subtype I – intrinsic, defined by the development of adenomyosis in direct connection with the junctional zone; subtype II – extrinsic, defined by the development of adenomyosis in contact with the serosa layer, without impairment/alteration of the junctional zone; subtype III – intramural, defined by the distinct development of adenomyosis, without affecting the junctional zone or the serosa; and subtype IV – including forms that do not fit into any of the previous subtypes, which is virtually equivalent to the extension of adenomyosis to all the layers of the uterus. Despite the similarities in terms of symptomatology, the three subtypes showed a very different behavior, starting from the context of the onset, the area affected, the adjacent structures interested or the type of surgical intervention used, thus justifying their separate evaluation.
The Kishy classification, however, ignores a series of variants which, in turn, my have individuality – cystic and respectively polypoid structures. Brosens suggests a distinct classification for these forms: A1 type – submucous or intramural cyst adenomyomas; type A2 – polypoid lesions; type B1 – subserous cyst adenomyomas; type B2 – subserous exophytic cysts; type C – uterine type masses within the uterus. Finally, in 2018, Bazot put together all the preexisting definitions into a common formulation, which has not yet received clinical validation(18).
Evidence regarding the impact
of conservative surgery on fertility
At present, in the absence of objective data, surgery is considered a valid option for certain forms of adenomyomas, respectively for adenomyomatous polyps. Regarding the opportunity of hysteroscopic resection of polyps, there is a unanimous agreement (without solid evidence), based on the experience with usual polyps. On the other hand, regarding adenomyomas, the existing data may be resumed as follows.
Based on the objective of this research, we looked for studies that evaluated the clinical effect of conservative surgery on fertility, and studies that clarified the potential pathogenetic links through which tumor resection could improve various local dysfunctions (the production of aromatase by endometrial cells, peristalsis of the junctional zone, expression of various cytokines involved in implantation).
We identified four studies that compared medical and surgical treatments, six case series studies that assessed the consequences of surgery on patient’s complaints and fertility outcome and a meta-analysis of studies conducted in literature up to 2013. Unfortunately, we did not find any studies that analyzed the impact of surgery on local dysfunctions.
Studies comparing medical and surgical treatment used the natural conception rate as a main outcome. Fedele (1993) was the first to document the superiority of the surgical treatment in subfertile patients with adenomyomas compared with the medical treatment(19). In a subsequent prospective study, Wang (2009) compared the isolated medical treatment (GnRH Ag) with the surgical treatment (plus medical treatment) in patients with symptomatic adenomyomas. The results of the study established the superiority of the surgical treatment with respect to natural conception rate(20). In a retrospective study, the same research team evaluated the effectiveness of the medical versus surgical approach in patients with diffuse adenomyosis, obtaining a rate of pregnancies/births which was clearly superior for the surgical treatment: 46.4% versus 10.8% (p=0.002), and 32.1% versus 8.1% (p=0.02) respectively at 3 years of follow-up(21). Al Yahami (2014) also confirmed the superiority of cytoreductive surgery associated with GnRH Ag compared with isolated medical treatment, obtaining a natural pregnancy rate of 44.4% versus 13.6% (p=0.06) after 36 months of follow-up(22).
The case series studies that address the role of surgery in the treatment of patients with adenomyosis evaluated the effectiveness of the treatments in terms of fertility (natural conception/IVF), and of disappearance/reduction of clinical complaints. The results on fertility vary greatly among authors, but there is unanimous agreement on the significant improvement in clinical complaints.
In a prospective study with follow-up period of two years, Osada (2011) reported obtaining 12 pregnancies through IVF and 4 naturally, out of 26 patients with focal/diffuse adenomyosis who underwent conservative surgical treatment (uterine wall reconstruction using the triple flap method). However, the number of IVF procedures performed in order to determine the actual pregnancy rate was not specified(23).
In another prospective study, Dai (2012) evaluated the effects of the local excision of adenomyosis in 86 patients. He obtained a natural conception rate of 2.3%, given the patients’ desire to conceive had not been evaluated(24).
Saremi (2014) prospectively assessed the impact of adenomyomectomy in 103 patients documented with adenomyosis, of which 79.45% with infertility over a follow-up period of two years. The authors reported obtaining 21 clinical pregnancies, 7 by natural conception (which corresponds to 33.3%) and 14 after assisted conception. The number of IVF procedures performed was not specified in this study either(25).
Kishy (2014) retrospectively evaluated 102 patients with adenomyosis in an attempt to identify the type of patient who would benefit the most, in terms of fertility (IVF), from the surgical treatment (adenomyomectomy). The data obtained in this study showed a clear benefit for patients younger than 39 years old – the pregnancy rate by natural conception was 41.3%, compared to those above 40 years old (3.7%)(17).
Sun (2011) assessed the long-term effects of conservative interventions (modified wedge resection – adenomyomectomy) in terms of quality of life and reproductive potential. The figures reported are very good: 45.5% pregnancies through IVF (the number of procedures was not specified), and a 22% rate of pregnancies obtained by natural conception at the 3-year follow-up(26).
By contrast, Nishida reported poor results after conservative surgery in diffuse adenomyosis (one pregnancy by IVF/one interstitial pregnancy out of 44 patients)(27).
Grimbizis (2014) performed a systematic review of data from literature regarding conservative surgery for adenomyosis following a very careful selection based on the quality of the existing studies. The results support a clearly favorable impact on the natural conception rate (48.89% for complete adenomyomectomy, and 46.88% for partial adenomyomectomy, respectively), and a very modest rate for ART procedures (9.34% of all pregnancies). The existing data have not objectified notable differences between the various techniques – even in the case of the two broad categories (complete versus partial excision) there have not been recorded statistically significant differences. This piece of information is extremely important from the perspective of our review. While the complexity and safety of the intervention are essential elements for the practitioner, with regard to the pathogenetic dilemmas of this disease, it is very important to analyze comparatively the cytoreductive interventions (partial excision) and those which aim at complete excision(28).
Even though the evolution of symptoms following conservative surgery was not a direct objective of the present study, we considered it useful to address this issue as well. Literature up-to-date allows us to speculate that the improvement of adenomyosis-related symptoms is an indirect sign for the correction of a series of dysfunctions which also underlie infertility problems, such as endometrial aromatase production, functionality of the junctional zone, and immune system cell cytokins production. The existing data in literature were unequivocal. All authors, without exception, reported a clear improvement both regarding pain complaints and the degree of menorrhagia, for cytoreduction or complete excision surgery. Grimbizis also confirmed the control of menorrhagia – decreased blood volume by 50% after partial resection and by 68% following total resection, and decreased dysmenorrhea by over 80%, regardless of the type of surgery performed(28).
The data obtained allowed us to make some observations. Clearly, there are certain groups of patients who seem to particularly benefit from surgery. Unfortunately, there are numerous missing elements in the phenotype of this type of patients. We identified a group of young patients with symptomatic adenomyoma, but with no other pathologies, who showed a very good rate of natural conception after surgery. There is also a group of patients with adenomyomis and multiple failed IVF procedures, who showed noticeably improved postoperative pregnancy rates by IVF. However, we cannot ignore the significant intra- and postoperative risks of these interventions due to the “invasive” nature of the adenomyotic tissue. Even though the figures reported in literature are very encouraging (the rate of uterine rupture <1%, and no synechiae/major postoperative alterations of the endometrium)(28), we need to point out that these data belong to centers of excellence. Therefore, we can anticipate that, used on a large scale, such interventions may cause significant morbidity. We believe that the choice of this therapeutic approach should be reserved to very experienced surgeons in carefully selected clinical circumstances.
Conflict of interests: The authors declare no conflict of interests.