Effects of Edaravone Treatment on Endometriotic Lesion Sizes in a Surgically Induced Rat Model: A Pilot Study

Endometriosis is a chronic, oestrogen-dependent inflammatory gynaecological disorder defined by the presence of endometrial glands and stroma outside the uterine cavity, classified into three main subtypes: superficial peritoneal endometriosis, ovarian endometriomas, and deep-infiltrating endometriosis. 1–4 It affects approximately 5–10% of women of reproductive age globally, rising to 35–50% in those presenting with pelvic pain or infertility. 1 , 2 , 5 Despite its high prevalence, diagnosis is typically delayed by 4–11 years, and effective non-surgical, non-hormonal therapies remain limited. 6 , 7 The pathogenesis of endometriosis is multifactorial, with retrograde menstruation as proposed by Sampson providing the most widely accepted mechanistic framework. 8 Shed endometrial fragments deposited in the peritoneal cavity implant on peritoneal surfaces, establish a blood supply, and evade immune clearance — a process facilitated by aberrant aromatase overexpression driving local oestradiol production in a self-reinforcing feed-forward loop, and by progesterone resistance arising from epigenetic silencing of progesterone target genes. 2 , 8 , 9 A critical and increasingly recognised driver of endometriosis pathogenesis is oxidative stress. The peritoneal fluid of affected women exhibits elevated reactive oxygen species (ROS), increased malondialdehyde (MDA), and diminished antioxidant enzyme activities including superoxide dismutase (SOD), catalase, and glutathione peroxidase. 1 , 10 ROS amplify ectopic implant survival through Raf/MEK/ERK and mTOR pathway activation, promote angiogenesis via VEGF, and sustain the inflammatory microenvironment through macrophage activation. 11 , 12 Current medical therapies — GnRH agonists/antagonists, oral contraceptives, progestins, and aromatase inhibitors — are associated with significant side effects, limited long-term tolerability, and high recurrence rates upon discontinuation, 1 , 6 , 7 underscoring the need for novel therapeutic strategies targeting distinct pathophysiological mechanisms including oxidative stress. 12 , 13 Consistent with this rationale, several antioxidant compounds have been evaluated in experimental endometriosis. For example, gamma-oryzanol, a plant-derived antioxidant, reduced the development of surgically induced endometriotic lesions in a rat model. 14 Such findings support the exploration of clinically approved free radical scavengers, which may offer more direct translational potential, as candidate agents in endometriosis. Edaravone (3-methyl-1-phenyl-2-pyrazolin-5-one, MCI-186) is a lipophilic synthetic free radical scavenger approved for acute cerebral infarction and amyotrophic lateral sclerosis. 15 Its mechanism encompasses direct scavenging of hydroxyl radicals, superoxide anions, and peroxyl radicals; inhibition of lipid peroxidation; suppression of pro-inflammatory cytokines (IL-1β, IL-6, TNF-α); and anti-apoptotic effects. 15 , 16 Of particular relevance, Kara et al demonstrated that edaravone significantly reduced MDA levels, preserved SOD activity, and attenuated histopathological damage in rat ovarian ischaemia-reperfusion injury, 17 while Atallah et al showed improved maternal and foetal outcomes in a placental ischaemia mouse model. 18 To our knowledge, no prior experimental study has evaluated edaravone in an endometriosis model. This small pilot study was therefore designed as an exploratory, hypothesis-generating investigation to obtain preliminary evidence on whether subcutaneous edaravone influences endometriotic lesion size and histological morphology in a surgically induced rat model. Given the limited sample size, it was intended to inform the design of larger, adequately powered studies rather than to establish therapeutic efficacy.

Macroscopically visible endometriotic implants were successfully established in all ten animals at the second laparotomy (100% implantation rate), confirming the validity of the surgical induction model ( Figure 1A–D ). All animals remained in good clinical condition throughout the study with no surgical mortality or adverse events requiring early termination. Individual and group-level lesion size data are presented in Table 1 and illustrated in Figures 2 and 3 . At baseline, lesion size was significantly larger in the control group than in the edaravone group (4.47 ± 1.89 mm vs 2.40 ± 0.52 mm; Mann–Whitney U -test, p = 0.019), indicating that the two groups were not comparable at baseline. This between-group baseline comparison is distinct from the within-group progression analysis reported in the Between- and Within-Group Comparisons subsection below. This imbalance is an important limitation and is considered throughout the interpretation of the results (see Discussion). In the edaravone group, mean lesion size decreased from 2.40 ± 0.52 mm at baseline to 1.91 ± 0.54 mm after 28 days of treatment (mean change: −0.50 ± 0.81 mm; −17.0 ± 27.6%). Four of six animals (66.7%) showed a reduction in lesion size, although in one this was only marginal (−0.4%), whereas two animals showed slight increases (+1.8% and +10.3%) ( Figure 2 ). Table 1 Individual Endometriotic Lesion Sizes (mm) and Percentage Changes Before and After the 28-Day Treatment Period Group Animal Baseline (mm) Post-Treatment (mm) Change (mm) Change (%) Edaravone 1 2.28 2.32 +0.04 +1.8 Edaravone 2 1.74 1.92 +0.18 +10.3 Edaravone 3 2.71 2.70 −0.01 −0.4 Edaravone 4 3.09 1.17 −1.92 −62.1 Edaravone 5 2.70 1.70 −1.00 −37.0 Edaravone 6 1.89 1.62 −0.27 −14.3 Mean ± SD (n=6) 2.40 ± 0.52 1.91 ± 0.54 −0.50 ± 0.81 −17.0 ± 27.6 Control 1 3.92 4.20 +0.28 +7.14 Control 2 4.09 4.38 +0.29 +7.09 Control 3 7.15 7.66 +0.51 +7.13 Control 4 2.72 2.91 +0.19 +6.99 Mean ± SD (n=4) 4.47 ± 1.89 4.79 ± 2.02 +0.32 ± 0.14 +7.09 ± 0.07 Notes : Negative values indicate lesion regression; positive values indicate progression. Bold text indicates column headers and the group-level summary rows (mean ± SD); it is used solely to distinguish these from individual-animal values and carries no additional statistical meaning. Abbreviation : SD, standard deviation. Figure 2 Percentage change in endometriotic lesion size per individual animal after 28 days of treatment. A horizontal bar graph showing the percentage change in endometriotic lesion size for each animal. The Y-axis lists animals C-4 to C-1 (control group) in the upper panel and E-6 to E-1 (edaravone group) in the lower panel, separated by a horizontal dashed line. The two panels use independent X-axes so that the small, tightly clustered control values remain legible: the upper axis runs from 5.50 to 7.25 percent and the lower axis from −80 to +20 percent. Control group changes are C-4 (+6.99%), C-3 (+7.13%), C-2 (+7.09%), and C-1 (+7.14%), all showing progression. Edaravone group changes are E-6 (−14.29%), E-5 (−37.04%), E-4 (−62.14%), E-3 (−0.37%), E-2 (+10.34%), and E-1 (+1.75%). A vertical line marks zero on the lower panel, and dashed vertical lines mark the group means: control +7.09% and edaravone −16.95%. The legend distinguishes regression and progression bars for each group, the group means, the panel separator, and the zero reference line. Statistical results shown on the figure are Mann–Whitney U, p = 0.114; rank-biserial r = 1.00; and Wilcoxon (within-edaravone), p = 0.313. Horizontal bar graph showing the percentage change in lesion size for each individual control and edaravone animal. Figure 3 Paired endometriotic lesion sizes before and after 28-day treatment in edaravone and control groups. Two paired dot and line plots compare groups at two time points. Left plot: edaravone group. Y-axis: Lesion Size (mm), 0.0 to 3.5 in 0.5 steps. X-axis: Second Laparotomy Day 28 and Third Laparotomy Day 56. Six paired trajectories: Day 28 values ~3.1, 2.7, 2.3, 1.9, 1.9, 1.6 mm; Day 56 values ~2.7, 2.3, 1.9, 1.7, 1.6, 1.2 mm. Mean: 2.40 ± 0.52 at Day 28, 1.91 ± 0.54 at Day 56. Right plot: control group. Y-axis: Lesion Size (mm), 0 to 8 in 1 steps. X-axis: Second Laparotomy Day 28 and Third Laparotomy Day 56. Four paired trajectories: Day 28 values ~7.3, 4.1, 4.0, 2.6 mm; Day 56 values ~7.6, 4.3, 4.2, 2.7 mm. Mean: 4.47 ± 1.89 at Day 28, 4.79 ± 2.02 at Day 56. Statistics: Wilcoxon signed rank test within groups and Mann Whitney U test between groups, p = 0.0095. A paired dot and line plot of endometriotic lesion size across two time points for two groups. Individual Endometriotic Lesion Sizes (mm) and Percentage Changes Before and After the 28-Day Treatment Period Notes : Negative values indicate lesion regression; positive values indicate progression. Bold text indicates column headers and the group-level summary rows (mean ± SD); it is used solely to distinguish these from individual-animal values and carries no additional statistical meaning. Abbreviation : SD, standard deviation. Percentage change in endometriotic lesion size per individual animal after 28 days of treatment. Paired endometriotic lesion sizes before and after 28-day treatment in edaravone and control groups. In the control group, mean lesion size increased from 4.47 ± 1.89 mm to 4.79 ± 2.02 mm (mean change: +0.32 ± 0.14 mm; +7.09 ± 0.07%). All four control animals showed progressive and notably consistent lesion growth ( Figure 3 ). The post-treatment macroscopic appearance of lesions in both groups is illustrated in Figure 1E and F . The between-group difference in absolute lesion size change (mm) was statistically significant (Mann–Whitney U -test: U = 0, p = 0.0095), with a large effect size (rank-biserial correlation r = 1.00), reflecting complete separation of the two groups on this measure ( Figure 4 ). When expressed as percentage change from baseline, the between-group difference did not reach statistical significance (Mann–Whitney U -test, p = 0.114). Final post-treatment lesion sizes also differed between groups; however, because baseline lesion size was substantially larger in the control group (see Lesion Size Changes and Discussion), this comparison is confounded and is not interpreted as a treatment effect. Figure 4 Change in endometriotic lesion size (mm) after 28 days of treatment. A box-and-whisker plot comparing the change in endometriotic lesion size between the edaravone group (n=6) and the control group (n=4), shown in two panels separated by a vertical dashed line. The two panels use different Y-axis scales so that the smaller control values remain legible: the left axis runs from −2.0 to +0.5 mm and the right axis from 0.0 to +0.5 mm. A vertical arrow between the panels indicates that downward values represent regression and upward values progression. In the left panel, the individual changes for the edaravone animals are −1.92, −1.00, −0.27, −0.01, +0.04, and +0.18 mm, with a mean ± standard deviation of −0.50 ± 0.81 mm. The box spans the interquartile range with a horizontal line at the median, and the whiskers extend to the mean ± standard deviation limits. In the right panel, the individual changes for the control animals are +0.19, +0.28, +0.29, and +0.51 mm, with a mean ± standard deviation of +0.32 ± 0.14 mm. All control values lie above zero, while most edaravone values lie below it, so the two groups do not overlap. Dashed horizontal lines mark the zero reference level in each panel. The figure legend reports the treatment protocol (edaravone 3 mg/kg/day subcutaneously; control 0.9% saline 2 mL/day, both for 28 days) and the statistics: Mann–Whitney U test, p = 0.0095, rank-biserial r = 1.00. Box and whisker plot showing the change in lesion size after 28 days in the edaravone and control groups. Change in endometriotic lesion size (mm) after 28 days of treatment. Within-group analysis showed that all four control animals underwent lesion progression over the treatment period, although with n = 4 this within-group change did not reach statistical significance on the exact Wilcoxon signed-rank test (p = 0.125). Within the edaravone group, the pre- to post-treatment change was likewise not statistically significant (p = 0.3125), consistent with the small sample size and the heterogeneity of individual lesion responses. Histological examination of H&E-stained sections revealed marked differences between the two groups ( Figure 5 ). In control animals, endometriotic implants demonstrated well-preserved endometrial glands with cuboidal to columnar glandular epithelium lining clearly identifiable lumen structures, embedded within vascularised fibrous stroma — consistent with active, viable endometriosis ( Figure 5A and B ). Multiple glandular units per section were observed, with intact cellular morphology and no signs of degenerative change. Figure 5 Representative histological sections of endometriotic lesions from the control and edaravone-treated groups (haematoxylin and eosin staining). ( A ) Control group, overview (scale bar = 200 µm): multiple active endometrial glands with clear lumen structures embedded within fibrous stroma, consistent with established and viable endometriosis. ( B ) Control group, detail (scale bar = 100 µm): well-preserved glandular epithelium lining endometrioid glands surrounded by endometrial stroma, confirming active implant viability. ( C ) Edaravone group, overview (scale bar = 100 µm): marked reduction in glandular structures with predominant fibrotic tissue replacing the endometriotic implant. ( D ) Edaravone group, detail (scale bar = 100 µm): dense collagenous fibrous stroma with sparse stromal cells and absence of identifiable glandular epithelium, indicative of lesion regression following edaravone treatment. Four histopathology images, pink/purple stain, control vs edaravone, overview/detail, with scale bars. Representative histological sections of endometriotic lesions from the control and edaravone-treated groups (haematoxylin and eosin staining). ( A ) Control group, overview (scale bar = 200 µm): multiple active endometrial glands with clear lumen structures embedded within fibrous stroma, consistent with established and viable endometriosis. ( B ) Control group, detail (scale bar = 100 µm): well-preserved glandular epithelium lining endometrioid glands surrounded by endometrial stroma, confirming active implant viability. ( C ) Edaravone group, overview (scale bar = 100 µm): marked reduction in glandular structures with predominant fibrotic tissue replacing the endometriotic implant. ( D ) Edaravone group, detail (scale bar = 100 µm): dense collagenous fibrous stroma with sparse stromal cells and absence of identifiable glandular epithelium, indicative of lesion regression following edaravone treatment. In edaravone-treated animals, implants showed a substantially different histological profile. Glandular structures were markedly reduced or entirely absent, replaced predominantly by dense collagenous fibrous tissue with sparse stromal cell populations ( Figure 5C and D ). This fibrotic pattern, accompanied by the absence of identifiable glandular epithelium, is consistent with reduced lesion activity and parallels the macroscopic size data. These histological observations are qualitative and descriptive: no formal histological scoring (gland counts, epithelial integrity, fibrosis grading, or inflammatory-cell assessment) was performed, and the sections shown are representative rather than quantified. They should therefore be interpreted as supportive descriptive findings rather than as definitive evidence of regression.

Ten adult female Wistar Albino rats (200–250 g, 6–8 weeks) were obtained from the Experimental Research Application and Research Centre, Erciyes University, Kayseri, Türkiye. Animals were housed under controlled conditions (20–22°C, 12-hour light/dark cycle) with ad libitum access to standard chow and water. All experimental procedures were approved by the Local Ethics Committee for Animal Experiments of Erciyes University, and were conducted in accordance with the 3Rs principles. Endometriosis was surgically induced using the established method of Vernon and Wilson, 19 validated for experimental endometriosis research 20 ( Figure 1A and B ). Animals were anaesthetised with ketamine hydrochloride (60 mg/kg i.p.) and xylazine hydrochloride (10 mg/kg i.p.). Approximately 0.5 cm of the right uterine horn was excised, opened longitudinally, and sutured onto the vascularised left lateral pelvic peritoneum. All animals received cefazolin (50 mg/kg/day i.m.) for five postoperative days. Figure 1 Intraoperative macroscopic photographs documenting the three-stage surgical protocol, showing representative images from different animals. ( A ) First operation — endometriosis induction: general view of the operative field following midline laparotomy, with the excised uterine horn segment being prepared for peritoneal implantation using the Vernon and Wilson technique. ( B ) First operation — suturing: the opened uterine horn segment being sutured onto the vascularised pelvic peritoneum, establishing the experimental endometriosis model. ( C ) Second operation — baseline measurement (edaravone group): digital calliper measurement of an established endometriotic implant (2.28 mm) four weeks after induction, immediately prior to treatment allocation. ( D ) Second operation — baseline measurement (control group): digital calliper measurement of an established endometriotic implant (4.09 mm), demonstrating successful model establishment and inter-animal variability in baseline lesion size. ( E ) Third operation — post-treatment appearance (edaravone group): macroscopic view of an endometriotic implant after 28 days of edaravone treatment, demonstrating a small, pale, fibrotic lesion consistent with lesion regression. ( F ) Third operation — post-treatment appearance (control group): macroscopic view of an endometriotic implant after 28 days of saline treatment, demonstrating a large, hyperaemic, well-vascularised lesion consistent with progressive active endometriosis. The macroscopic contrast between panels E and F corroborates the quantitative size data and histological findings. Composite figure of six labelled intraoperative photographs (A through F) arranged in three rows. The left column shows the edaravone group and the right column the control group; row labels along the left indicate the 1st Operation, 2nd Operation, and 3rd Operation. Panel A: first operation — following a midline laparotomy, the abdominal cavity is opened and a segment of the excised uterine horn is held with forceps, prepared for implantation onto the pelvic peritoneum. Panel B: first operation — the opened uterine horn segment is being sutured onto the vascularised pelvic peritoneum to establish the experimental endometriosis model. Panel C: second operation, edaravone group — a digital calliper measures an established endometriotic implant, showing a lesion size of 2.28 mm. Panel D: second operation, control group — a digital calliper measures an established endometriotic implant, showing a lesion size of 4.09 mm. Panel E: third operation, edaravone group, after 28 days of treatment — the endometriotic implant is small, pale, and fibrotic, consistent with lesion regression. Panel F: third operation, control group, after 28 days of saline treatment — the endometriotic implant is larger, dark red, and well vascularised, consistent with persistent active endometriosis, appearing larger and more vascularised than in panel E. Six-panel intraoperative photographs showing endometriosis induction, lesion measurement, and post-treatment appearance in rats. Intraoperative macroscopic photographs documenting the three-stage surgical protocol, showing representative images from different animals. ( A ) First operation — endometriosis induction: general view of the operative field following midline laparotomy, with the excised uterine horn segment being prepared for peritoneal implantation using the Vernon and Wilson technique. ( B ) First operation — suturing: the opened uterine horn segment being sutured onto the vascularised pelvic peritoneum, establishing the experimental endometriosis model. ( C ) Second operation — baseline measurement (edaravone group): digital calliper measurement of an established endometriotic implant (2.28 mm) four weeks after induction, immediately prior to treatment allocation. ( D ) Second operation — baseline measurement (control group): digital calliper measurement of an established endometriotic implant (4.09 mm), demonstrating successful model establishment and inter-animal variability in baseline lesion size. ( E ) Third operation — post-treatment appearance (edaravone group): macroscopic view of an endometriotic implant after 28 days of edaravone treatment, demonstrating a small, pale, fibrotic lesion consistent with lesion regression. ( F ) Third operation — post-treatment appearance (control group): macroscopic view of an endometriotic implant after 28 days of saline treatment, demonstrating a large, hyperaemic, well-vascularised lesion consistent with progressive active endometriosis. The macroscopic contrast between panels E and F corroborates the quantitative size data and histological findings. Four weeks post-induction, all animals underwent a second laparotomy. Endometriotic lesions were identified and measured using a digital calliper ( Figure 1C and D ). Animals were then randomly allocated to: Edaravone group (E+EDA; n=6): Subcutaneous edaravone 3 mg/kg/day for 28 days. This dose was selected on the basis of previous experimental studies in which edaravone at doses of approximately 1–6 mg/kg exerted antioxidant and tissue-protective effects in rodent models of reproductive-organ oxidative injury, 17 , 18 and lies well below the no-observed-adverse-effect level (NOAEL) of 10 mg/kg/day reported for repeated-dose administration in rats in the FDA pharmacology review of edaravone, 21 supporting its tolerability over the 28-day treatment period. Control group (E+SF; n=4): Subcutaneous 0.9% saline 2 mL/day for 28 days. After 28 days, all animals underwent a third laparotomy. Lesion sizes were re-measured using the same digital calliper ( Figure 1E and F ). Endometriotic implants were excised and fixed in 10% neutral-buffered formalin for histological processing. Animals were subsequently euthanised under general anaesthesia. Excised endometriotic implants were processed using standard paraffin-embedding techniques, sectioned at 4–5 µm, and stained with haematoxylin and eosin (H&E). Sections were examined under light microscopy (Olympus BX51, Olympus Corp., Tokyo, Japan) and photographed at ×100 and ×200 magnification. Histological assessment focused on the presence and morphology of endometrial glands, surrounding stromal composition, and degree of fibrosis. Data are expressed as mean ± standard deviation (SD). Non-parametric methods were employed throughout given the small sample sizes. Between-group comparisons of lesion size changes were performed using the Mann–Whitney U -test. Within-group pre- versus post-treatment comparisons used the exact Wilcoxon signed-rank test. Effect size for the between-group comparison was estimated using the rank-biserial correlation (r), a non-parametric measure appropriate for these sample sizes. A p-value of <0.05 was considered statistically significant. Analyses were performed using IBM SPSS Statistics v22.0 (IBM Corp., Armonk, NY, USA).

This pilot study provides preliminary experimental evidence that edaravone may influence endometriotic lesion behaviour in a surgically induced rat model. Over the 28-day treatment period, edaravone-treated animals showed a mean lesion size reduction of 17.0%, with regression in four of six animals, whereas all control animals showed lesion progression; the between-group difference in absolute lesion size change was statistically significant (p = 0.0095, rank-biserial r = 1.00). This macroscopic signal was accompanied by a consistent histological difference: control implants retained active glandular epithelium within vascularised stroma, whereas edaravone-treated implants were predominantly fibrotic with markedly reduced or absent glandular structures ( Figure 5 ). Two important caveats temper this interpretation. First, baseline lesion size was significantly larger in the control group (p = 0.019), so the two groups were not comparable at baseline; the between-group difference in absolute change may therefore be influenced by this imbalance and by regression-to-the-mean effects. Second, the within-edaravone reduction was not statistically significant (p = 0.3125), and the between-group difference in percentage change did not reach significance (p = 0.114). These observations should therefore be regarded as hypothesis-generating rather than as evidence of a definitive treatment effect. The pathophysiological basis for this observation is grounded in the central role of oxidative stress in endometriosis. Ectopic endometrial tissue is characterised by elevated ROS production, increased lipid peroxidation, and reduced antioxidant enzyme capacity. 1 , 10 , 11 These oxidative perturbations activate key molecular pathways promoting implant survival and growth, and sustain the pro-inflammatory peritoneal microenvironment. 12 Edaravone’s dual capacity to directly scavenge ROS and to suppress upstream cytokine-mediated inflammation positions it mechanistically to disrupt multiple nodes of this pathological network. 15 , 16 The reproductive organ protective effects of edaravone have been directly demonstrated by Kara et al, who showed that intraperitoneal edaravone significantly reduced MDA levels and preserved SOD activity in rat ovarian ischaemia-reperfusion injury, with histopathological protection of follicular architecture 17 — establishing that edaravone reaches gynaecological tissue in effective concentrations. Atallah et al further confirmed its anti-apoptotic and antioxidant efficacy in a reproductive context through the RUPP placental ischaemia model. 18 The anti-inflammatory dimension of edaravone’s action is also pertinent. The peritoneal microenvironment of endometriosis is characterised by elevated IL-1β, IL-6, IL-8, TNF-α, RANTES, and VEGF. 1 , 7 Edaravone has been shown to reduce concentrations of IL-2, IL-6, IL-1β, and TNF-α across multiple experimental systems, 16 suggesting it may attenuate the cytokine-mediated mechanisms that sustain implant growth. The fibrotic histological pattern observed in treated animals ( Figure 5C and D ) is compatible with, but does not establish, reduced pro-inflammatory cytokine signalling and diminished angiogenic support for the lesion. It must be emphasised that no oxidative stress markers, cytokines, angiogenic factors, or apoptotic indices were measured in the present study. The specific cytokines and molecular pathways discussed above are therefore proposed as plausible mechanisms drawn from other experimental systems, and remain speculative in this model; they are presented to frame future mechanistic work rather than to explain the present findings. The heterogeneity of size responses within the edaravone group — ranging from −62.1% to +10.3% ( Figure 2 ) — reflects the well-recognised biological variability of endometriotic lesions. 6 , 7 Some animals with only modest macroscopic size reduction nonetheless showed a fibrotic histological pattern. We note this observation cautiously; because histology was assessed qualitatively on representative sections and individual-animal histological data were not quantified, any suggestion that histological change precedes macroscopic reduction remains speculative and would require formal, blinded histological scoring to test. Several limitations must be acknowledged. The sample size is small, limiting statistical power and generalisability. Formal histological scoring of gland number, epithelial integrity, and inflammatory cell infiltrate was not performed. Oxidative stress biomarkers (MDA, SOD, CAT, GPx) and inflammatory cytokine levels were not measured. The study did not include a naïve (unoperated) control group and did not assess serum oestrogen levels. The Vernon and Wilson rat model, while well-validated, 19 , 20 does not fully recapitulate human disease since rats do not menstruate. Several additional methodological limitations should be noted. Although animals were randomly allocated to groups, the specific randomisation method was not stratified by baseline lesion size, and the resulting groups differed significantly at baseline; future studies should use stratified or baseline-adjusted designs. The oestrous cycle stage of the animals at the time of surgery and treatment was not determined; because endometriosis is oestrogen-dependent and ectopic implant growth is influenced by hormonal fluctuations, uncontrolled cycle stage may have contributed to the observed variability and should be monitored in future work. Outcome assessment was not blinded: the investigator performing digital calliper measurements was aware of group allocation, which may have introduced measurement bias. Histological evaluation was based on a single representative section per animal, without a predefined section-selection protocol. Finally, edaravone was administered subcutaneously, whereas the clinically approved formulations are given intravenously or orally; the subcutaneous route was chosen for practicality of repeated dosing in rats, but pharmacokinetic differences between routes mean that these findings may not directly translate to approved formulations. In addition, the study design required a second laparotomy to obtain baseline lesion measurements, which introduces a local surgical and potentially pro-oxidative stimulus; although this was applied identically to both groups and is therefore unlikely to bias the between-group comparison, minimally invasive or imaging-based measurement approaches may reduce this confounder in future studies. These limitations notwithstanding, the results add to the growing body of evidence supporting antioxidant-based approaches as adjunctive strategies in endometriosis, a disease in which high post-surgical recurrence rates 22 and limited long-term tolerability of hormonal therapies underscore the unmet need for novel mechanistic interventions. Future studies should employ power analysis-guided sample sizes, implement formal histological scoring, measure tissue-level oxidative stress markers, and investigate edaravone’s molecular effects on aromatase activity, NF-κB signalling, and apoptotic indices in ectopic tissue.

In this pilot study, edaravone-treated animals showed a mean lesion size reduction of 17.0%, with regression in four of six animals, whereas all control animals showed progression; the between-group difference in absolute lesion size change was statistically significant (p = 0.0095, rank-biserial r = 1.00). It should be emphasised, however, that the within-edaravone reduction did not reach statistical significance (p = 0.3125), that baseline lesion size differed significantly between groups (p = 0.019), and that no mechanistic markers were measured. Accordingly, these findings should be interpreted with caution and regarded as hypothesis-generating rather than as evidence of a definitive therapeutic effect. Within these limitations, the histological and macroscopic observations are compatible with a possible effect of edaravone on endometriotic tissue and support the conduct of larger, adequately powered, and mechanistically comprehensive studies to test this hypothesis.