{"paper_id":"be3dca2e-dfc8-40ff-9910-9c213b20818b","body_text":"Research Article\nVolume 9 • Issue 1\n9 \nIs Dienogest More Effective Than Ethinylestradiol/Dienogest or Desogestrel \nin Reducing Ovarian Endometrioma Size? A Sonographic Retrospective \nCohort Study\nDe Cicco Nardone Carlo1, Sangiovanni Maria Cristina1*, De Luca Cristiana1, Plotti Francesco, Montera Roberto1, Luvero \nDaniela1, Martinelli Arianna1, Sangiovanni Gian Mario2, Angioli Roberto1, Terranova Corrado1\nAffiliation:\n1Department of Gynecology and Obstetrics, Campus \nBio-Medico University Hospital, Rome, Italy\n2Department of Statistical Sciences, Sapienza \nUniversity of Rome, Rome, Italy\n*Corresponding author:\nSangiovanni Maria Cristina, Department of\nGynecology and Obstetrics, Campus Bio-Medico \nUniversity Hospital, Rome, Italy\nCitation:  De Cicco Nardone Carlo, Sangiovanni \nMaria Cristina, De Luca Cristiana, Plotti Francesco, \nMontera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, \nTerranova Corrado. Is Dienogest More Effective \nThan Ethinylestradiol/Dienogest or Desogestrel \nin Reducing Ovarian Endometrioma Size? A \nSonographic Retrospective Cohort Study. Obstetrics \nand Gynecology Research. 9 (2026): 09-20.\nReceived: Decmber 16, 2025 \nAccepted: Decmber 24, 2025 \nPublished: January 14, 2026\nAbstract\nObjective: To compare the long-term sonographic effectiveness of \nDienogest (DNG)2mg, Ethinylestradiol/Dienogest (EE/DNG), and \nDesogestrel (DSG) 75 µg in reducing ovarian endometrioma (OMA) \nvolume, and to evaluate their performance relative to untreated patients \nunder active follow-up (A-FU).\nMethods: This retrospective monocentric cohort study included women \naged 16–55 years with transvaginal ultrasound (TV-US)–confirmed \ntypical OMAs. Patients received DNG, EE/DNG, DSG, or no therapy  \n(A-FU). OMAs’ volume was calculated using the prolate ellipsoid formula \nat baseline and at 12 and 24 months; 36-month data were included when \navailable. Longitudinal changes in volume were analysed using a linear \nmixed-effects model (LMM) with random intercepts. A secondary analysis \nemployed log-transformed volumes to account for baseline heterogeneity.\nResults: Sixty-three patients completed at least 12 months of follow-\nup (DNG n=14; EE/DNG n=39; DSG n=28; A-FU n=10). At baseline, \nsignificant differences in age and cyst size were observed, with the DNG \ngroup presenting the largest volumes (87,488 ± 68,211 mm\n3). All hormonal \ntherapies induced progressive volume reduction, while untreated OMAs \ntended to increase.\nThe LMM revealed a significant time × treatment interaction for DNG \nat 24 months (p = 0.020*), indicating a reduction exceeding the natural \ntrajectory despite larger initial cysts. DSG and EE/DNG showed similar \ndownward trends but without statistical significance in absolute-volume \nmodels. In contrast, the log-transformed analysis showed that all three \nhormonal treatments exhibited regression, whereas the A-FU group did \nnot.\nConclusions: Hormonal therapy effectively reduces OMA volume, with \nDNG showing the strongest and statistically significant effect. When \nadjusted for baseline heterogeneity, DSG and EE/DNG demonstrate \ncomparable relative efficacy. Active hormonal therapy should be preferred \nover observation, while treatment choice should remain individualized. \nProspective multicenter studies are needed to confirm these findings.\nKeywords: Dienogest; Ovarian endometrioma.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n10 \nIntroduction\nEndometriosis affects around 10% of reproductive-age \nwomen, as a significant cause of pelvic pain and infertility \n[1]. Endometriosis may be classified into three subtypes: \nsuperficial peritoneal endometriosis (SPE), ovarian \nendometrioma (OMA), and deep infiltrating endometriosis \n(DIE) [2].\nOMA, known as “chocolate cysts” due to the thick \ndark brown appearance of the fluid contained within \nthem [3], affect up to 44% of women with endometriosis \n[4], representing the most frequent manifestation of the \ndisease. With a negative impact on ovarian reserve, the \nmanagement of OMA is particularly relevant in women \ndesiring future fertility [5]. Laparoscopic cystectomy \nwith the stripping technique has traditionally been the \nfirst choice for conservative treatment of OMA [6,7]. \nNevertheless, concerns about ovarian tissue damage, \nparticularly the potential reduction in ovarian reserve and \nits negative impact on assisted reproductive outcomes, have \nhighlighted a paradox [8] in infertile women and led to \ngrowing interest in conservative medical therapy. Medical \ntherapy, which includes, combined oral contraceptives \n(COCs)- and progestin-only pills (POPs), is considered a \nfirst-line treatment to reduce endometriosis-associated pain, \nincluding dysmenorrhea, dyspareunia, and non-menstrual \npelvic pain [9,10].\nAccording to the ESHRE guidelines, COCs are also \nstrongly recommended for long-term prevention of \nendometrioma recurrence after surgery in women not seeking \nimmediate pregnancy [10]. However, in the literature, there \nis a lack of evidence regarding the effect of COC therapy \non OMA diameter [9]. Among POPS, Dienogest (DNG), a \nfourth-generation progestin [11], is also considered effective \nin decreasing the size of endometrioma [12], reducing \nendometriosis-associated pain with a favourable safety and \ntolerability profile [10,13]. Moreover, Desogestrel (DSG) \n[14], a third-generation progestin [15] and a one of the most \nwidely used POPs, is also considered an effective treatment \nfor endometriosis, improving dysmenorrhea in up to 93% \nof cases [16]. Despite the widespread use of COCs, DNG \nand DSG in clinical practice, evidence on their comparative \neffectiveness in reducing OMA size remains inconsistent \nand limited. Furthermore, existing comparative studies \nbetween COCs, DNG and DSG primarily focus only on \nsymptom control, such as pain reduction, without assessing \nchanges in endometrioma volume [17].\n To address this issue, the present retrospective cohort \nstudy evaluates changes in OMA volume when comparing \nDNG, ethinylestradiol/dienogest (EE/DNG) and DSG \ntherapy.\nMaterials and Methods\nStudy Design and Setting\nThis retrospective monocentric cohort study was \nconducted in the Department of Gynaecology and Obstetrics \nat the University \nHospital Campus Bio-Medico of Rome between \nNovember 2018 and October 2025.\nParticipants\nWomen aged 16–55 years with a transvaginal ultrasound \n(TV-US) confirmed typical OMA were eligible for inclusion. \nAdditional criteria were a BMI between 18 and 35, \navailability of baseline and follow-up scans, and a minimum \nfollow-up duration of 12 months. Exclusion criteria included \na history or a suspicion of malignancy in atypical OMA, \novarian surgery performed before or during the study period, \npregnancy, any change in hormonal therapy, or incomplete \nclinical or imaging records.\nTreatment Groups\nPatients were divided into four groups according to the \ntherapy received:\n(A) 2 mg DNG: 1 tablet per day for 28 consecutive days, with \nthe blister pack containing no placebo tablets.\n(B) 2mg DNG/30 μg EE: 1 tablet per day for 28 consecutive \ndays, with the blister pack of 28 tablets containing 7 \nplacebo tablets.\n(C) 75 mcg DSG: 1 tablet per day for 28 consecutive days, \nwith the blister pack containing no placebo tablets.\n(D) Active follow-up (A-FU): control group in which no \ntreatment was administered.\nTreatment duration was at least 12 months.\nClinical Evaluation and Ultrasound Assessment\nAll patients underwent a standard gynaecological \nevaluation, including medical history, physical examination, \nand TV-US. Each patient underwent a baseline scan and a \nfollow-up scan after 12 and 24 months of observation or \ntreatment. When available, data from patients who completed \na 36-month follow-up were also included, acknowledging \nthat not all participants reached this extended timepoint. \nA schematic overview of the study design and follow-up \nschedule is provided in Figure 1.\nFor each patient, the following information was collected: \ntype and start date of hormonal therapy, serial TV-US \nfindings (presence, number of OMAs per patient and their \nthree orthogonal diameters), and changes in treatment during \nfollow-up. All TV-US examinations were performed using \na GE Healthcare Voluson E8TM ultrasound system with \n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n11 \nregular smooth walls with no papillary projections, and \nabsence of vascularized solid components.\nOMAs’ volume can be measured using a prolate ellipsoid \nformula, where the volume is calculated from the three \northogonal diameters expressed in mm (anteroposterior -AP, \ntransverse-TD, and longitudinal- LD) using the formula π/6 \n(AP ×TD×LD), as shown in Figure 2. The volume of the cysts \nis expressed in mm³. This method provides a way to estimate \nthe three- dimensional size of a cyst from bidimensional TV-\nUS measurements.\na RIC 5–9D endovaginal probe (5–9 MHz), by the same \nsenior sonographer with extensive expertise in endometriosis \nimaging.\nOMAs were identified and described according to \nthe International Society of Ultrasound in Obstetrics C \nGynecology (ISUOG)18, to the International Ovarian Tumor \nAnalysis (IOTA) [19]. classification and to the Ovarian-\nAdnexal Reporting Data System Ultrasouns (O-RADS US) \n[20]: Typical OMA is defined as a < 10 cm unilocular or \nmultilocular (less than five locules) cyst with homogeneous \nlow-level “ground-glass’’ echogenicity of the cyst fluid, \n \nFigure 1: Study design and TV-US follow-up timeline. Patients were divided into three groups: (A) continuous dienogest 2 mg (28-day \nregimen without placebo); (B) combined ethinylestradiol 30 μg + dienogest 2 mg (21 active + 7 placebo tablets); (C) continuous Desorgestrel \n75 mcg (28-day regimen without placebo); D active follow-up (without medical treatment). Participants underwent TV-US at baseline, 12 and \n24 months to assess endometrioma size. Data from patients who additionally reached the 3c-month follow-up were also collected, although not \nall participants completed this extended time point.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n12 \nOutcomes\nThe analysis aimed to evaluate the effectiveness of \ndifferent hormonal treatments in reducing endometrioma \nsize by comparing mean volume changes across treatment \ngroups and against untreated controls (A-FU) at 12, and when \navailable, 24 and 36 months. In fact, data from patients who \nalso reached the 24- and 36-month follow-up were collected, \nalthough not all participants completed this extended time \npoint.\nStatistical Analysis\nStatistical analyses were performed using R software \n(version 4.4.3). Normally distributed continuous variables \n(e.g., age) were compared across the four groups using one-\nway Analysis of Variance (ANOVA), while non-normally \ndistributed variables (e.g., baseline cyst volume) were \nanalyzed using the Kruskal–Wallis test. Categorical variables \nwere evaluated using Pearson’s Chi-square test.\nGiven the longitudinal design of the study, in which each \npatient contributed repeated TV-US measurements at baseline \nand at 12, 24, and 36 months, a Linear Mixed-Effects Model \n(LMM) was employed. This approach appropriately accounts \nfor within-subject correlation and accommodates missing \ndata under the Missing at Random (MAR) assumption, \nallowing the inclusion of all patients with at least one follow-\nup assessment.\nThe primary analysis was conducted on absolute \nendometrioma volumes. Fixed effects included time, \ntreatment group (DSG 75 µg, DNG 2 mg, EE/DNG, and \nA-FU), and their interaction to determine whether volume \ntrajectories differed among groups. The model estimated: \n(i) the baseline mean volume (reference: no therapy), (ii) \nchanges at each follow-up relative to baseline, (iii) differences \nbetween treatment groups at baseline, and (iv) time × group \ninteractions, representing differential longitudinal responses \nto treatment. \nThe statistical significance of fixed effects was \nassessed using Type III ANOVA with Satterthwaite’s \napproximation to derive F-statistics and p-values. Random-\neffects variance components were examined to quantify \nbetween-subject variability. Model assumptions (normality \nand homoscedasticity of residuals) were verified through \ndiagnostic plots.\nAll tests were two-sided, and a p-value < 0.05 was \nconsidered statistically significant. A secondary analysis \nusing log-transformed volumes [log(volume + 1)] was \nperformed to normalize the distribution, stabilize variance, \nand reduce the influence of extreme baseline heterogeneity\nEthical Approval\nThe study was conducted in accordance with the ethical \nstandards of the institutional research committee and in \nadherence to the principles outlined in the Declaration of \nHelsinki. Patient data were anonymized; informed consent \nwas waived due to the retrospective design.\nData Collection and Management\nData were retrieved from the institutional electronic \nmedical record system and entered into a password-protected \nExcel spreadsheet, which was stored on the secure hospital \nserver. All identifiers were removed before analysis, and only \nanonymized data were used for statistical evaluation.\nResults\nFigure 3 illustrates the study flow diagram, showing the \nnumber of patients included in each treatment group and \nthe proportion of who completed the scheduled follow-up \nassessments at 12, 24, and 36 months. The initial cohort \nconsisted of 108 women (DSG 28, DNG n = 15;EE/DNG n = \n45; Control: n = 19).\nAfter exclusions due to non-adherence, treatment \ndiscontinuation, or loss to follow-up, 91 patients remained \neligible for analysis (DSG 28; DNG: n = 14; EE/DNG: \nn = 39; Control: n = 10), all of whom reached at least the \n12-month evaluation. The number of patients available for \nvolumetric assessment declined progressively at later time \npoints, as detailed in Figure 3.\nTable 1 summarises the baseline characteristics of the \nstudy population across the three groups. A total of 14 patients \n(18 cysts) were included in the DNG 2 mg group, 39 patients \n \nFigure 2: Schematic representation of the method used to \nestimate endometrioma volume. The three orthogonal diameters—\nanteroposterior (AP), transverse (TD), and longitudinal (LD)—are \nmeasured on transvaginal ultrasound and used to approximate cyst \nvolume with the ellipsoid formula: π/c × (AP × TD × LD). The \nillustration compares an endometriotic ovarian cyst with a reference \nsphere to visualise the dimensional assessment.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n13 \n(54 cysts) in the EE/DNG group, and 10 patients (11 cysts) \nin the control group. Conversely, a statistically significant \ndifference was observed regarding age (p = 0.026). While \nthe treatment groups (DNG, EE/DNG, DSG) shared a similar \nmean age (range 35– 37 years), the Active Follow-up group \nwas significantly older (mean 44.1 ± 8.9 years). Clinically, \nthis imbalance is expected in observational studies: women \ncloser to menopause (Group 4) often opt for a \"wait and see\" \nmonitoring approach, whereas younger women prioritize \nfertility preservation or long-term symptom control through \nactive therapy.\nTable 2 details the cyst multiplicity per patient. The \nmajority of patients presented with a single endometrioma, \nalthough multiple cysts were observed, particularly in the EE/\nDNG and DSG group.\nTable 3 summarizes the mean endometrioma volume for \neach group at baseline and during follow-up at 12, 24, and 36 \nmonths. At baseline (T0), a statistically significant difference \nin cyst size distribution was observed among the four groups, \nreflecting marked heterogeneity within the study population. \nThe DNG 2 mg group had the largest mean baseline \n \nFigure 3: Study flow diagram illustrating the number of patients included in each treatment group (Dienogest 2 mg, Ethinylestradiol/Dienogest, \nDesogestrel and Control) and the number of women who completed the scheduled follow-up assessments at 12, 24, and 3c months.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n14 \nvolumes, whereas the Active Follow-up (control) group had \nconsiderably smaller lesions. Despite these initial imbalances, \nlongitudinal evaluation revealed clearly divergent volumetric \ntrajectories across groups. All three hormonally treated \ncohorts showed a progressive and substantial reduction in \ncyst volume over time. The DNG group demonstrated the \nmost pronounced decline, decreasing from 87,488 ± 68,211 \nmm³ at baseline to 1,597 ± 1,373 mm³ at 36 months. DSG and \nEE/DNG displayed similar trends of sustained shrinkage. In \ncontrast, the Active Follow-up group exhibited persistence or \nprogression of disease, with mean volumes nearly doubling \nduring the first 12 months and remaining elevated thereafter.\nHowever, cross-sectional comparisons of raw absolute \nvolumes at 12, 24, and 36 months did not reach statistical \nsignificance (p > 0.05). This apparent lack of significance \nis attributable to the very large inter-subject variability \nand wide standard deviations observed within each group, \nwhich dilute the measurable between-group differences in \nsimple unadjusted tests. Moreover, baseline heterogeneity—\nparticularly the substantially larger initial cyst volumes in \nthe DNG cohort—further limits the interpretability of cross-\nsectional statistics.\nTo appropriately address this variability and adjust for \nbaseline imbalances, a Linear Mixed- Effects Model (LMM) \nwas performed. The mixed-effects analysis was performed \non a longitudinal dataset comprising 392 observations \nfrom 91 subjects. Variance component estimates confirmed \nmarked between-subject heterogeneity (random intercept SD \n= 43,684) and substantial residual variability (residual SD \n= 58,194), supporting the use of subject- specific random \nintercepts to model repeated measures appropriately.\nTreatment group N° patients N° cysts \nanalyzed Mean age (years) Mean bmi (kg/m²)\nDienogest 2 mg 14 18 37.6 ± 8.5 26.0 ± 2.1\nEE 30 µg / DNG 2 mg 39 54 35.9 ± 7.5 25.4 ± 2.3\nDesogestrel 75 mcg 28 35 36.7 ± 7.4 25.2 ± 2.3\nActive follow up (no therapy) 10 11 44.1 ± 8.9* 26.5 ± 1.8\np-value 0.026* 0.305\nTable 1: Baseline characteristics of the study population by treatment group, including the number of patients and cysts, mean age, and mean \nBMI.\nTreatment n° patients with\n1\nendometriotic cyst n° patients with 2 \nOMA\nn° patients with \n3 OMA\nn° patients with \n4 OMA\nn° patients with \n5 OMA\nDienogest 2 mg 13 0 0 0 1\nEE 30 µg / DNG 2 mg 27 10 1 1 0\nDesogestrel 75 mcg 21 7 0 0 0\nActive follow up (no therapy) 9 1 0 0 0\nTable 2: Number of patients with one or more endometriotic cysts per treatment group. The table shows the distribution of patients according \nto the number of endometriotic cysts (from one to five) in each treatment arm\nMean (± sd) endometrioma volume (mm³)\nGroup Baseline volume\n(t0)\nVolume at 12 months \n(t12)\nVolume at 24 months \n(t24)\nVolume at 36 months \n(t36)\nDienogest 2 mg 87,488 ± 68,211 21,071 ± 15,019 4,900 ± 3,693 1,597 ± 1,373\nEE 30 µg / DNG 2 mg 14,210 ± 3,538 7,578 ± 2,788 2,108 ± 708 665 ± 261\nDesogestrel 75 mcg 38,462 ± 11,764 10,620 ± 4,917 8,643 ± 3,959 2,910 ± 1,961\nActive follow up (no therapy) 6,008 ± 10,040 10,890 ± 5,286 NS NS\nP-value* 0.011* < 0.001 < 0.001 < 0.01\nTable 3: Mean (± SD) endometrioma volume (mm³) at baseline and at 12, 24, and 3c months in the four treatment groups. P- values refer to \noverall between-group comparisons at each time point. P-value calculated using Kruskal-Wallis test. NS = Not Significant.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n15 \nTable 4 reports the fixed-effect estimates. Type III \nANOVA demonstrated a significant overall effect of Time \n(F(3, ≈304) = 5.77, p < 0.001), indicating meaningful changes \nin endometrioma volume across follow-up for the cohort as \na whole. In contrast, the main effect of Treatment Group was \nnot statistically significant (p = 0.374), likely reflecting the \nhigh inter-subject variability and baseline volume differences \namong groups.\nGiven the marked baseline heterogeneity and high inter-\nsubject variance demonstrated by the mixed-model analysis \nof absolute volumes, a secondary analysis was performed \nusing log- transformed volumes [log(Volume + 1)]. This \ntransformation reduces the influence of extreme values and \nenables a normalized comparison of treatment effects across \ndifferent initial cyst sizes.\nAs illustrated in Figure 4A, the logarithmic trajectories \nshowed clear divergence between treated patients and the \ncontrol cohort. The Active Follow-up group displayed a \nflat or slightly upward trend, consistent with cyst stability \nor progression. In contrast, all hormonally treated groups \n(DSG 75 mcg, DNG 2 mg, and EE/DNG) exhibited similarly \nsteep negative slopes, indicating a consistent rate of volume \nreduction over time, irrespective of baseline cyst size.\n(A) Longitudinal trajectories of log-transformed \nendometrioma volumes [log(Volume + 1)], applied to \nnormalize the marked heterogeneity in baseline cyst size. The \nthree active treatment groups (Dienogest 2 mg, Desogestrel \n75 mcg, and EE/DNG) display nearly parallel downward \nslopes, indicating a comparable rate of cyst regression over \ntime. In contrast, the Active Follow-up group shows a flat \nor slightly upward trend, consistent with cyst stability or \nprogression.\nStatistical testing on the transformed data confirmed \nthat all active treatments significantly outperformed the \ncontrol condition. DSG demonstrated a significant reduction \nas early as 12 months (p = 0.006), with effects maintained \nat 24 months (p < 0.001). Likewise, both EE/DNG and  \nDNG showed highly significant reductions at 24 months  \n(p < 0.001).\nFigure 4B further highlights the relative treatment \nefficacy by illustrating percentage change from baseline at \nthe 24-month endpoint. While untreated patients exhibited \nwide variability with a median tendency toward stability or \ngrowth, all hormonally treated patients clustered within a \npattern of pronounced regression, typically approaching an \n80%–100% reduction in volume. When normalized for cyst \nsize, these findings demonstrate that DSG and EE/DNG \nachieve a therapeutic effect comparable to that of DNG in \nhalting and reversing endometrioma growth.\nDiscussions\nOMAs occur in 17–44% patients with endometriosis \n[21]. Ovaries are the most common sites of endometriosis \n[22], with an interesting left lateral predisposition [23,24,25]. \nVercellini et al. [26] and Sznurkowski et al. [27] attribute this \nleft-sided predominance of OMAS to the mechanical effect of \nthe sigmoid colon, which reduces peritoneal fluid clearance \non the left side. Chapron et al. [28]. additionally suggest \nhormonal and microenvironmental differences between \nhemipelves. On the other side, Matalliotakis et al. [29] \nParameter\nEffect \nestimate \n(mm³)\n95% confidence \ninterval P-value\nTime effects  \n(ref: baseline)\nTime: 12 months −20,411 −66,048 to +28,226 0.41\nTime: 24 months −11,470 −76,117 to +53,178 0.73\nTime: 36 months −10,810 −136,263 to +117,674 0.87\nTreatment effects \n(relative to no \ncontrol group at \nt0)\nDSG 75 mcg +6,765 −41,677 to +61,508 0.71\nDNG 2 mg +73,337 +16,563 to +130,081 0.01*\nEE/ DNG −10,864 −56,586 to +37,801 0.66\nSignificant \ninteractions\nDNG at 24 months −63,670 −173,002 to −14,638 0.02*\nTable 4: Fixed effects estimate from the Linear Mixed-Effects \nModel regarding endometrioma volume (mm³). Values are presented \nas effect estimates with corresponding S5% Confidence Intervals \n(CI) and p-values. The model reveals that the DNG 2 mg group had \na significantly higher baseline volume compared with the control \ngroup (p = 0.01) and exhibited a significant negative interaction at \n24 months (p = 0.02), indicating a greater rate of volume reduction \nover time relative to no therapy.\nExamination of specific contrasts revealed clinically and \nstatistically relevant differences in treatment trajectories. The \nDNG 2 mg group had a significantly higher estimated mean \nbaseline volume than the Active Follow-up group (Estimate \n≈ +73,337 mm³; p = 0.012). Importantly, this group exhibited \na significant negative interaction at 24 months (Estimate ≈ \n−93,970 mm³; p = 0.020), indicating a greater reduction in \ncyst volume over time relative to the natural progression \nobserved in untreated patients. Thus, despite larger initial \ncyst volumes, Dienogest-treated patients demonstrated a \nmore pronounced volumetric decline.\nFor the DSG and EE/DNG groups, the observed \nreductions did not reach statistical significance compared \nwith the control group (p > 0.05), a finding consistent with \nboth the smaller baseline cyst dimensions in these groups and \nthe substantial residual variance captured by the model.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n16 \npropose the “female varicocele theory,” linking left-sided \nvenous stasis to increased susceptibility to endometrioma \nformation.\nTraditionally, laparoscopy was considered the gold \nstandard for the examination of endometriotic lesions [30]. \nNowadays, guidelines [10] have shifted their focus to non-\ninvasive imaging-based diagnosis of deep endometriosis in \npreference to surgery, with a great attention to TV-US, which \nshould be performed in a standardised manner, as described \nby the International Deep Endometriosis Analysis (IDEA) \ngroup [31].\nA typical endometrioma on TV-US is described as \na unilocular or multilocular (less than five locules) cyst, \nwith homogeneous low-level echogenicity (ground glass \nechogenicity) of the cyst fluid. Because endometriomas \nare usually poorly vascularised32. Resonance Magnetic \nImage (RMI) for OMAs is requested only in selected \ncases if ultrasound outcome is inconclusive, if malignant \ntransformation is suspected, or both [32].\nThe therapeutic approach to endometriosis should be \nindividualised, considering the clinical presentation (whether \npain or infertility), the patient’s age, ovarian reserve, \nreproductive desire, and the specific disease phenotype, \nincluding the presence of endometriomas, adenomyosis, and \nthe overall extent of the lesions [33].\nVolumetric reduction achieved through medical therapy \nmay decrease the need for surgical intervention, thereby \nsupporting ovarian function preservation and improving \nsymptom control34. POPs and COCs are both well-tolerated \nmedical treatments and represent safe, long-term, first-\nline therapeutic options33, 35. COCs, administered orally, \ntransdermal or via vaginal ring, and POPs are strongly \nrecommended for reducing endometriosis-associated \ndyspareunia, dysmenorrhea and non-menstrual pelvic pain \n[10].\nAmong POPs, DNG, 19-nortestosterone derivative, \nfourth-generation progestin, used at dosage, 2 mg per day, \nprovides a significant local effect on endometriotic lesions and \nit is as first-line medical treatment for pain in endometriosis \n[35].\nMoreover, previous studies demonstrated that DNG 2 mg/\nday also leads to a significant reduction in both the diameter \nand volume of OMAs36. After six months of therapy, mean \ncyst volume may decrease by approximately 66%, with \nreductions exceeding 76.19% after twelve months36. Recent \nevidence indicates that, in cysts measuring ≥4 cm, the mean \ndiameter can decrease from around 50.5 mm to 34 mm within \none year, accompanied by a marked volumetric reduction \n(from 37.8 mL to 11.8 mL in 12 months) [37]. DNG is also \neffective in reducing the size of recurrent OMAs38, while \nmaintaining a favourable safety profile, including minimal \nimpact on bone health39 even with treatment durations of up \nto one year. This is a notable advantage compared with other \nmedical therapeutic options (such as Gonadotropin- Releasing \nHormone Antagonists -GnRH- agonist and antagonist) that \nare associated with substantial bone loss [39].\nAlso COCs, and in particular 2mg DNG/30 μg EE have \nalso been associated with a significant reduction in OMAs \nsize, particularly when used continuously and over more \nextended treatment periods exceeding 6–12 months40.\nFrom this perspective, the present study aimed to evaluate \nand compare the efficacy of continuous DNG, DSG and \na COC containing 2 mg DNG/30 µg EE in reducing the \nsonographic size of OMAs.\nThe patient flow analysis (Figure 3) clearly outlines the \nstudy's retention profile over time. An essential limitation \nof the longer follow-up period must be acknowledged. The \nlongitudinal evaluation is affected by the fact that not all \npatients reached the 24- and 36-month assessments. This \nprogressive reduction in sample size (attrition rate) was \nparticularly pronounced in the Active Follow-up group, \nwhich, already smaller at baseline compared with the \nactive treatment arms, was substantially eroded over time, \nresulting in only one evaluable patient at the final follow-up. \nConsequently, while short-term comparisons (12 months) are \nreliable, interpretation of long-term trajectories, especially \nfor the untreated group, requires caution, given the limited \ndata available.\nA significant demographic finding emerged from the \nbaseline comparison: patient age differed among groups (p = \n0.026). Women receiving hormonal therapy had comparable \nmean ages, whereas the Active F-U cohort was significantly \nolder (44.1 ± 8.9 years). This discrepancy likely reflects a \nwell-recognized observational pattern in clinical practice: \nperimenopausal women frequently opt for a conservative \n“wait-and-see” approach, avoiding hormonal treatments \nin anticipation of the natural decline in ovarian activity. \nConversely, younger patients typically require active \nmedical management to preserve fertility and maintain long-\nterm symptom control. Notably, there were no significant \ndifferences in BMI across groups, supporting anthropometric \nhomogeneity of the sample.\nAnother key aspect concerns the baseline distribution \nof cyst volume. Patients treated with DNG 2 mg presented \nwith substantially larger cysts at enrolment compared with all \nother groups, particularly the control group. This imbalance \nmay suggest the presence of “channelling bias” [41]: DNG \nmay be preferentially prescribed to women with more \nsevere disease, higher symptom burden, or larger OMAs. \nThis underscores the importance of employing statistical \nmethods that appropriately account for significant baseline \nheterogeneity.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n17 \nThe longitudinal raw-volume trends demonstrated \ndivergent behaviours across groups. All hormonal treatments \nshowed progressive reductions in endometrioma volume, \nwhereas the Active F-U group displayed stable or slightly \nincreasing volumes over time. However, high intersubject \nvariability and imbalanced baseline volumes limited the \nstatistical significance of cross-sectional comparisons at \neach time point, despite a clear clinical signal: untreated \nendometriomas tend not to regress spontaneously.\nThe linear mixed-effects model with random intercepts \nrepresented the appropriate analytical approach to address \nboth repeated measurements and baseline heterogeneity. The \nmodel demonstrated a significant time × treatment interaction \nfor the DNG group at 24 months. The model showed that the \nvolumetric decline in the DNG group diverged significantly \nfrom the natural course observed in the untreated cohort. \nNotably, the interaction at 24 months indicates that, despite \ninitial disadvantage, patients receiving DNG experienced a \nsubstantially greater reduction in OMAs’ volume over time, \nexceeding what would be expected based on baseline size \nalone.\nThis result underscores that DNG not only reduces cyst \nvolume but also actively modifies the trajectory of disease \nprogression, a conclusion that could not be reliably inferred \nfrom analysis of raw absolute volumes alone.\nOther treatments did not reach statistical significance \nin the absolute-volume model, likely because these groups \nstarted with substantially smaller cysts, which diluted the \ndetectable statistical signal.\nOverall, the log-transformed analysis effectively \nnormalized baseline size heterogeneity and confirmed a \nclear therapeutic advantage of all active treatments over \nobservation alone. Compared with the Active F-U group, \nwhich showed a flat or slightly increasing trajectory, DNG, \nDSG, and EE/DNG consistently exhibited a downward trend, \nindicating progressive and sustained reduction in OMAs’ \nvolume. These findings reinforce the positive impact of \nhormonal therapy in modifying the natural history of OMAs.\nThe findings of this study highlight several clinically \nmeaningful aspects of endometrioma management. In the \nabsence of treatment, OMAs generally persist or enlarge \nover time, reinforcing the chronic, progressive nature of the \ndisease. By contrast, all hormonal therapies evaluated induced \nsubstantial regression, both in absolute terms and in relative, \nsize-adjusted analyses. DNG emerged as the treatment \nthat produced the most pronounced effect in the adjusted \nlongitudinal model, significantly altering the natural evolution \nof the cysts, despite being prescribed to patients with the largest \nbaseline volumes. Yet, when baseline heterogeneity was \nnormalized through log-transformed and percentage-change \nanalyses, DSG and EE/DNG demonstrated a therapeutic \nimpact comparable to that of DNG, suggesting that their \nefficacy should not be underestimated. Similar to our study, \nthe survey conducted by Angioni et al. [34] demonstrated a \ngreater reduction in OMAs size in patients treated with DNG \nalone, with mean cyst diameter decreasing from 54± 22 mm \nto 32 ± 12 mm after six months, a reduction corresponding \nto approximately 75% of cyst volume. In contrast, combined \nEE/DNG therapy did not result in a significant dimensional \nchange despite symptomatic improvement, indicating that \nonly DNG monotherapy exerted a meaningful effect on \nendometrioma shrinkage [34].\nThe present study has several strengths that contribute to \nits clinical relevance. It provides one of the few longitudinal \nevaluations comparing different hormonal regimens for \novarian endometriomas, using repeated sonographic \nmeasurements over a follow-up of up to 36 months, thereby \ncapturing both short- and long-term therapeutic trajectories. \nThe integration of absolute, log-transformed, and percentage-\nchange analyses offers a nuanced interpretation of treatment \nefficacy, overcoming the inherent heterogeneity in baseline \ncyst volumes and providing a more robust representation of \nactual clinical effectiveness.\nNevertheless, several limitations must be acknowledged. \nFirst, the retrospective and monocentric design introduces \ninherent risks of selection bias and limits the ability to account \nfor unmeasured confounders. Second, the overall sample \nsize remains relatively small, particularly at longer follow-\nup intervals due to attrition, which reduces statistical power \nand may affect the stability of subgroup comparisons. Third, \nunequal baseline characteristics between groups, especially \ndifferences in age and initial cyst volume, reflect real-world \nprescribing patterns and complicate direct comparisons \ndespite statistical adjustment. Finally, the generalizability \nof these findings remains limited, and larger prospective \nmulticenter studies are warranted to validate these results, \nrefine comparative estimates of treatment effectiveness, and \nbetter guide individualized therapeutic decision-making.\nConclusions\nThis study provides new evidence supporting the role of \nhormonal therapy in the long-term conservative management \nof OMAs. Among the therapies evaluated, DNG 2 mg \ndemonstrated the most pronounced therapeutic effect, \nachieving a statistically significant time × treatment interaction \nat 24 months, indicating a volumetric reduction that clearly \ndiverged from the natural course of untreated cysts. Notably, \nthis effect emerged despite the DNG group presenting with \nsubstantially larger cysts at baseline, suggesting that DNG is \nremarkably effective even in the context of advanced disease \nburden. These findings reinforce the drug’s\nability not merely to reduce cyst size, but to modify the \ntrajectory of OMAs progression, a result that would not have \n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n18 \nbeen detectable through simple analysis of raw volumes \nalone.\nThe secondary log-transformed analysis further \ndemonstrated that all three hormonal regimen, including DSG \nand EE/DNG, were superior to observation alone, showing a \nconsistent downward trajectory of OMA’s volume regression \nover time. When baseline heterogeneity was normalized, \nDSG and EE/DNG achieved a degree of relative regression \ncomparable to DNG, underscoring that multiple medical \noptions can provide meaningful disease control.\nTaken together, these findings confirm that active \nhormonal therapy should be preferred over observation \nin reproductive-age women with OMAs. Therapy choice, \nhowever, should remain individualized, considering patient \nage, reproductive goals, symptomatology, baseline cyst \ncharacteristics, and treatment tolerability.\n Given the limitations inherent to the retrospective, \nmonocentric design and the attrition observed at longer follow-\nup intervals, larger prospective multicenter studies are needed \nto validate  these results, refine comparative effectiveness \nestimates, and guide evidence-based personalization of \nmedical therapy for ovarian endometriomas.\nReferences\n1. Taylor, H. S., Kotlyar, A. M. C Flores, V. A. Endometriosis \nis a chronic systemic disease: clinical challenges and \nnovel innovations. Lancet 3G7 (2021): 839-852.\n2. Imperiale, L., Nisolle, M., Noël, J.-C. C Fastrez, M. Three \nTypes of Endometriosis: Pathogenesis, Diagnosis and \nTreatment. State of the Art. J Clin Med 12 (2023): 994.\n3. Hoyle, A. T. C Puckett, Y. Endometrioma. in StatPearls \n(StatPearls Publishing, Treasure Island (FL), 2025).\n4. Billow, M., Sridhar, S. C Mintz, G. Endometrioma: \nContemporary Approach to Diagnosis and \nManagement. Semin Reprod Med https://doi.\norg/10.1055/s-0045-1810430 (2025).\n5. Perrone, U. et al. Endometrioma surgery: Hit with your \nbest shot (But know when to stop). Best Practice & \nResearch Clinical Obstetrics & Gynaecology G6 (2024): \n102528.\n6. Bafort, C., Beebeejaun, Y., Tomassetti, C., et al. \nLaparoscopic surgery for endometriosis. Cochrane \nDatabase Syst Rev (2020): CD011031.\n7. Pados, G., Daniilidis, A., Keckstein, J., et al. A European \nsurvey on the conservative surgical management of \nendometriotic cysts on behalf of the European Society \nfor Gynaecological Endoscopy (ESGE) Special Interest \nGroup (SIG) on Endometriosis. Facts Views Vis Obgyn \n12, 105-108.\n8. Dias Jr, J. A. et al. The endometrioma paradox. JBRA \nAssist Reprod 2G (2025): 145-149.\n9. Muzii, L. et al. Expectant, Medical, and Surgical \nManagement of Ovarian Endometriomas. Journal of \nClinical Medicine 12 (2023): 1858.\n10. Becker, C. M. et al. ESHRE guideline: endometriosis†. \nHum Reprod Open 2022, hoac009.\n11. Schindler, A. E. Dienogest in long-term treatment of \nendometriosis. Int J Womens Health 3 (2011): 175-184.\n12. Huang, Y., Zhang, D., Zhang, L., et al. Clinical efficacy \nof dienogest against endometriomas with a maximum \ndiameter of ≥4 cm. Ann Med 56 (2024): 2402942.\n13. Uludag, S. Z., Demirtas, E., Sahin, Y. C et al. Dienogest \nreduces endometrioma volume and endometriosis-related \npain symptoms. J Obstet Gynaecol 41 (2021): 1246-1251.\n14. Desogestrel. in Drugs and Lactation Database \n(LactMed®) (National Institute of Child Health and \nHuman Development, Bethesda (MD), 2006).\n15. Scala, C., Leone Roberti Maggiore, U., Remorgida, V., \net al. Drug safety evaluation of desogestrel. Expert Opin \nDrug Saf 12 (2013): 433-444.\n16. Weisberg, E. C Fraser, I. S. Contraception and \nendometriosis: challenges, efficacy, and therapeutic \nimportance. Open Access J Contracept 6 (2015): 105-115.\n17. Yurtkal, A. C Oncul, M. Comparison of dienogest or \ncombinations with ethinylestradiol/estradiol valerate \non the pain score of women with endometriosis: A \nprospective cohort study. Medicine (Baltimore) 103 \n(2024): e38585.\n18. Van Holsbeke, C. et al. Endometriomas: their ultrasound \ncharacteristics. Ultrasound in Obstetrics & Gynecology \n35 (2010), 730-740.\n19. Cohen Ben-Meir, L., Mashiach, R. C Eisenberg, V. H. \nExternal Validation of the IOTA Classification in Women \nwith Ovarian Masses Suspected to Be Endometrioma. \nJournal of Clinical Medicine 10 (2021): 2971.\n20. (PDF) O-RADS US Risk Stratification and \nManagement System: A Consensus Guideline from \nthe ACR Ovarian-Adnexal Reporting and Data System \nCommittee. ResearchGate https://www.researchgate.\nnet/publication/337045294_O-  RADS_US_Risk_\nStratification_and_Management_System_A_Consensus_\nGuideline_from_the_ACR_Ovarian-Adnexal_\nReporting_and_Data_System_Committee (2018).\n21. Gałczyński, K., Jóźwik, M., Lewkowicz, D., et al. Ovarian \nendometrioma – a possible finding in adolescent girls \nand young women: a mini-review. Journal of Ovarian \nResearch 12 (2019): 104.\n\nDe Cicco Nardone Carlo, et al., Obstet Gynecol Res 2026\nDOI:10.26502/ogr0200\nCitation: De Cicco Nardone Carlo, Sangiovanni Maria Cristina, De Luca Cristiana, Plotti Francesco, Montera Roberto, Luvero Daniela, Martinelli \nArianna, Sangiovanni Gian Mario, Angioli Roberto, Terranova Corrado. Is Dienogest More Effective Than Ethinylestradiol/Dienogest \nor Desogestrel in Reducing Ovarian Endometrioma Size? A Sonographic Retrospective Cohort Study. Obstetrics and Gynecology \nResearch. 9 (2026): 09-20.\nVolume 9 • Issue 1\n19 \n22. Lee, H. J., Park, Y. M., Jee, B. C., et al. Various anatomic \nlocations of surgically proven endometriosis: A single-\ncenter experience. Obstet Gynecol Sci 58 (2011), 53-58.\n23. Sznurkowski, J. J. C Emerich, J. Endometriomas are more \nfrequent on the left side. Acta Obstet Gynecol Scand 87 \n(2008): 104-106.\n24. Matalliotakis, I. M. et al. Arguments for a left lateral \npredisposition of endometrioma. Fertil Steril G1 (2009): \n975-978.\n25. Eshkoli, T., Weintraub, A. Y., Erenberg, M., et al. The \nlaterality of endometriosis. Journal of Endometriosis and \nUterine Disorders 3 (2023): 100039.\n26. Vercellini, P. et al. Is cystic ovarian endometriosis an \nasymmetric disease? Br J Obstet Gynaecol 105 (1998): \n1018-1021.\n27. Sznurkowski, J. J. C Emerich, J. Endometriomas are more \nfrequent on the left side. Acta Obstetricia et Gynecologica \nScandinavica 87 (2008): 104-106.\n28. Chapron, C. et al. Does deep endometriosis infiltrating \nthe uterosacral ligaments present an asymmetric lateral \ndistribution? BJOG: An International Journal of Obstetrics \n& Gynaecology 108 (2001), 1021-1024.\n29. Matalliotakis, I. M. et al. Arguments for a left lateral \npredisposition of endometrioma. Fertil Steril G1 (2009): \n975-978.\n30. Guerriero, S. et al. Addendum to consensus opinion \nfrom the International Deep Endometriosis Analysis \n(IDEA) group: sonographic evaluation of superficial \nendometriosis. Ultrasound in Obstetrics & Gynecology \n66 (2025): 541-547.\n31. Guerriero, S. et al. Systematic approach to sonographic \nevaluation of the pelvis in women with suspected \nendometriosis, including terms, definitions and \nmeasurements: a consensus opinion from the International \nDeep Endometriosis Analysis (IDEA) group. Ultrasound \nObstet Gynecol 48 (2016): 318-332.\n32. Exacoustos, C., Manganaro, L. C Zupi, E. Imaging for \nthe evaluation of endometriosis and adenomyosis. Best \nPractice & Research Clinical Obstetrics & Gynaecology \n28 (2014): 655-681.\n33. Petraglia, F., Vannuccini, S., Santulli, P., et al. An update \nfor endometriosis management: a position statement. \nJournal of Endometriosis and Uterine Disorders 6 (2024): \n100062.\n34. Angioni, S. et al. Is dienogest the best medical treatment \nfor ovarian endometriomas? Results of a multicentric case \ncontrol study. Gynecological Endocrinology 36 (2020): \n84-86.\n35. Treatment of endometriosis: a review with comparison of \n8 guidelines | BMC Women’s Health | Full Text. https://\nbmcwomenshealth.biomedcentral.com/articles/10.1186/\ns12905- 021-01545-5.\n36. Vignali, M. et al. Effect of Dienogest therapy on \nthe size of the endometrioma. https://air.unimi.it/\nhandle/2434/726511 (2020).\n37. Huang, Y., Zhang, D., Zhang, L., et al. Clinical efficacy \nof dienogest against endometriomas with a maximum \ndiameter of ≥4 cm. Ann Med 56 (2024): 2402942.\n38. Aizzi, F. J. Recurrent Endometrioma; Outcome of \nMedical Management with Dienogest. European Journal \nof Experimental Biology 7, 0–0.\n39. Momoeda, M. et al. Long‐term use of dienogest for the \ntreatment of endometriosis. J of Obstet and Gynaecol 35 \n(2009), 1069-1076.\n40. Carrillo Torres, P. et al. Clinical and sonographic impact \nof oral contraception in patients with deep endometriosis \nand adenomyosis at 2 years of follow-up. Sci Rep 13 \n(2023): 2066.\n41. Petri, H. C Urquhart, J. Channeling bias in the interpretation \nof drug effects. Stat Med 10 (1991): 577-581.\nThis article is an open access article distributed under the terms and conditions of the  \nCreative Commons Attribution (CC-BY) license 4.0","source_license":"CC0","license_restricted":false}