Abstract
Objective: This study aimed to evaluate the effects of nano-curcumin on markers of endometrial receptivity in a rat model of adenomyosis, focusing on lesion characteristics and gene expression related to embryo implantation.
Materials and methods
An experimental study was conducted using 30 female Wistar rats (Rattus norvegicus) at the Biomedical Laboratory, Faculty of Medicine, Universitas Riau, from March to October 2024. The rats were randomly assigned into five groups: control, adenomyosis model, and adenomyosis treated with nano-curcumin at doses of 25, 50, and 75 mg/kg body weight. Adenomyosis was induced via oral administration of tamoxifen (2.7 µmol/kg) from postnatal day 2 to 5. Nano-curcumin was administered orally for 15 consecutive days starting 75 days after induction. Adhesion scores, lesion number, and lesion volume were evaluated, while expression levels of HOXA10, Leukemia Inhibitory Factor (LIF), and integrin β3 were quantified using polymerase chain reaction assays during days 4–6 of pregnancy.
Results
Nano-curcumin reduced lesion count, lesion volume, and adhesion scores in a dose-dependent manner. However, these changes were not statistically significant (all p > 0.05). In contrast, nano-curcumin demonstrated a clear dose-dependent enhancement of endometrial receptivity. Expression of HOXA10, LIF, and integrin β3, markers suppressed in the adenomyosis group, was progressively restored toward control levels with increasing doses, indicating improved endometrial readiness for implantation despite minimal changes in lesion morphology.
Conclusion
Nano-curcumin enhances the expression of key endometrial receptivity markers without significantly affecting lesion characteristics in rats with adenomyosis, suggesting its potential as a therapeutic agent to improve fertility outcomes.
Introduction
Adenomyosis is a benign gynecological condition characterized by the infiltration of endometrial glands and stroma into the myometrium, typically accompanied by surrounding myometrial hyperplasia.1 The molecular mechanisms involved in its pathogenesis include abnormal tissue remodeling, chronic inflammation, and dysregulated hormonal signaling.2,3,4,5 The exact incidence of adenomyosis remains uncertain. In Italy, it was reported to affect approximately 1.22 million women, with the highest prevalence observed among those aged 15 to 59 years between 2011 and 2013.6,7 In Asia, the estimated prevalence is around 5%, which is notably lower than that in Australia, where it reaches up to 19%.8 In Indonesia, the precise incidence is unknown; however, estimates range between 2.39% and 11.7%.9 Clinically, adenomyosis manifests with a wide spectrum of symptoms, including abnormal uterine bleeding (AUB), pelvic pain, and dysuria. In many cases, dysmenorrhea is also a frequent complaint.10 In addition, infertility represents another important clinical manifestation of adenomyosis. Several mechanisms have been proposed to explain adenomyosis-associated infertility, including impaired endometrial receptivity, abnormal uterine contractility, and chronic inflammatory changes that alter the local hormonal and molecular microenvironment.8,9 The invasion of endometrial glands into the myometrium and subsequent disruption of the junctional zone lead to defective embryo implantation and reduced expression of key receptivity markers such as HOXA10, LIF, and integrin ß3.11,12,13 Consequently, women with adenomyosis often experience lower implantation rates and poor pregnancy outcomes, particularly in assisted reproductive settings.8,9 Management strategies depend on the severity of symptoms and the patient's proximity to menopause. Treatment options include hormonal contraceptives, pharmacologic therapies, and hysterectomy. Women with mild symptoms often choose conservative management until they reach menopause, whereas those with more severe symptoms may require hysterectomy.14 Considering the morbidity associated with definitive hysterectomy, there is a growing need for alternative therapeutic approaches. These therapies should ideally exert anti-inflammatory, antineoplastic, and antioxidant effects, aiming to downregulate the NF-B signaling pathway and suppress key inflammatory mediators such as integrins, IL-6, IL-8, TGF, and TNF-.15,16 Endometrial receptivity refers to the capacity of the endometrium to support embryo implantation and development. The implantation window typically occurs within 3 to 6 days during the secretory phase in healthy women. Curcumin has been investigated for its potential to modulate estrogen responsiveness.17
Materials and methods
The study protocol was reviewed and approved by the Health Research Ethics Committee of the Faculty of Medicine, Universitas Riau, Indonesia (Approval No. B/064/UN19.5.1.1.8/UEPKK/2024). This experimental study was conducted on thirty healthy female white rats (Rattus norvegicus, Wistar strain), aged 8-10 weeks and weighing 150-200 g. The rats were randomly allocated into five groups (n=6/group): healthy control, adenomyosis, and three nano--curcumin treatment groups (25, 50, and 75 mg/kg). Sample size was determined using the Federer formula (t-1)(n-1)=15, where t represents the number of treatment groups and n the number of animals per group, ensuring adequate statistical power. Randomization was performed using a simple randomization technique. A computer-generated random number table was used to assign the animals into five groups to reduce allocation bias.
The study was carried out at the Laboratory of Biomedical Sciences, Anatomical Pathology and Histology, Faculty of Medicine, Universitas Riau, between March and October 2024. Adenomyosis was induced during the prenatal period by administering oral tamoxifen at a dose of 2.7 µmol/kg body weight per day to the lactating dams from prenatal day 1 to 5. This exposure allowed tamoxifen to reach the neonatal female pups through maternal milk, which has been demonstrated to induce adenomyotic lesions by disrupting uterine development. Following induction, all pups were weaned and maintained under standard laboratory conditions until they reached maturity. The rats were observed for 75 days post-induction to allow the establishment of adenomyosis lesions.
On day 75, two randomly selected rats from the adenomyosis group were euthanized to confirm the success of adenomyosis induction through histopathological examination of the uterus. After confirmation, the remaining rats were used for the treatment phase. From day 76 to day 90, the treatment groups received daily oral nano-curcumin at doses of 25, 50, or 75 mg/kg body weight for 15 consecutive days, while the adenomyosis group received the vehicle only. The control group did not undergo any induction or treatment.
After completion of the treatment period, all female rats (except those in the control group) were paired with proven fertile males at a ratio of 2:1 to induce pregnancy. The presence of a vaginal plug or spermatozoa in vaginal smears was recorded as gestation day 0. This mating step was essential because the main objective of the study was to evaluate endometrial receptivity markers, which are maximally expressed during the implantation window in early pregnancy. On gestation days 4 to 6, corresponding to the peri-implantation period, pregnant rats were euthanized by cervical dislocation. The uteri were carefully dissected and examined for adenomyotic lesion characteristics, including lesion count, lesion volume, and the degree of pelvic adhesions. The severity of adhesions was assessed using the Blauer adhesion scoring system, which ranges from 0 (no adhesions) to 4 (dense adhesions causing complete uterine fixation).
Results
AND DISCUSSION
Lesion count, lesion volume, and adhesion score
In this study, the administration of nano-curcumin in rats with adenomyosis did not result in statistically significant changes in lesion count, adhesion score, or lesion volume compared to the untreated adenomyosis group. The mean lesion count in the adenomyosis group was 8.7 ± 1.1, while the values in the 25 mg/kg, 50 mg/kg, and 75 mg/kg nano-curcumin treatment groups were 8.4 ± 1.2, 8.6 ± 1.0, and 8.2 ± 1.3, respectively. Statistical analysis showed no significant differences in lesion count between these treatment groups and the adenomyosis group (p=0.278, 0.395, and 0.151, respectively). Similarly, adhesion scores did not differ significantly among the groups. The adenomyosis group had a mean adhesion score of 3.8 ± 0.6, while the 25 mg/kg, 50 mg/kg, and 75 mg/kg groups had scores of 3.5 ± 0.7, 3.6 ± 0.8, and 3.3 ± 0.6, respectively (p=0.193, 0.337, and 0.128) (Table 1).
For molecular analysis, tissue samples weighing approximately 60 mg were fragmented into pieces not exceeding 1 mm³. These fragments were homogenized in 1 mL of TRIzol reagent and incubated for 10 minutes at room temperature to allow for cell lysis. The samples were then centrifuged at 12,000 rpm for 10 minutes, after which 200 µL of chloroform was added. The mixture was vortexed for 15 seconds, incubated at room temperature for an additional 10 minutes, and centrifuged again at 12,000 rpm for 15 minutes.
Lesion volume also showed no statistically significant differences. The adenomyosis group exhibited a mean lesion volume of 11.2 ± 1.0 mm³, compared to 10.9 ± 1.1 mm³, 11.0 ± 0.9 mm³, and 10.7 ± 1.2 mm³ in the 25 mg/kg, 50 mg/kg, and 75 mg/kg groups, respectively (p=0.261, 0.442, and 0.116). These findings suggest that nano-curcumin, at the administered doses, did not significantly alter the macroscopic characteristics of adenomyotic lesions in this experimental model (Table 1).
The aqueous phase containing RNA was then collected for further analysis. The extracted RNA was reversetranscribed into complementary DNA (cDNA) using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, USA) according to the manufacturer's instructions. Quantitative PCR was performed using SYBR Green Master Mix (Bio-Rad, USA) on a CFX96 Real-Time PCR Detection System (Bio-Rad). Specific primers targeting Hoxa10, LIF, and integrin ß3 were designed based on published sequences (primer sequences are available upon request). Gene expression levels were normalized against the housekeeping gene GAPDH. The relative expression levels were calculated using the 2^-ΔΔCT method.
HOXA-10 gene expression
A significant reduction in HOXA-10 gene expression was observed in the adenomyosis group compared to the control group (0.50 vs. 1.00, respectively). Treatment with nano-curcumin led to a dose-dependent increase in HOXA-10 expression, reaching 0.75, 0.85, and 0.95 in the groups receiving 25 mg/kg, 50 mg/kg, and 75 mg/kg, respectively (Table 2).
Data distribution normality was assessed using the Shapiro-Wilk test. For normally distributed data with homogeneous variances, comparisons between groups were performed using one-way ANOVA followed by Tukey's post hoc test. For non-normally distributed data or unequal variances, the Kruskal-Wallis test with Dunn's post hoc test was applied. Statistical analyses were conducted using SPSS version 27, with p-values < 0.05 considered statistically significant.
LIF gene expression
Leukemia inhibitory factor (LIF) expression was significantly lower in the adenomyosis group (3.1 ± 0.8) compared to the control group (5.2 ± 1.0). Administration of nano-curcumin increased LIF expression to 4.0 ± 0.9 at 25 mg/kg, 4.5 ± 0.7 at 50 mg/kg, and 4.8 ± 0.6 at 75 mg/kg (Table 2).
| Groups | Lesion Count (mean ± SD) | Adhesion Score (mean ± SD) | Lesion Volume (mm³ ± SD) | p-value (Lesion Count) | p-value (Adhesion Score) | p-value (Lesion Volume) |
|---|---|---|---|---|---|---|
| Control | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | - | - | - |
| Adenomyosis | 8.7 ± 1.1 | 3.8 ± 0.6 | 11.2 ± 1.0 | - | - | - |
| Curcumin 25 mg/kg | 8.4 ± 1.2 | 3.5 ± 0.7 | 10.9 ± 1.1 | 0.278 | 0.193 | 0.261 |
| Curcumin 50 mg/kg | 8.6 ± 1.0 | 3.6 ± 0.8 | 11.0 ± 0.9 | 0.395 | 0.337 | 0.442 |
| Curcumin 75 mg/kg | 8.2 ± 1.3 | 3.3 ± 0.6 | 10.7 ± 1.2 | 0.151 | 0.128 | 0.116 |
| Note: |
| Groups | n | HOXA-10 Expression (Relative Units) | LIF Expression (Mean ± SD) | Integrin β3 Expression (Mean ± SD) |
|---|---|---|---|---|
| Control | 6 | 1.00 | 5.2 ± 1.0 | 3.5 ± 0.5 |
| Adenomyosis | 6 | 0.50 | 3.1 ± 0.8 | 2.0 ± 0.4 |
| Adenomyosis + Curcumin 25 mg/kg | 6 | 0.75 | 4.0 ± 0.9 | 2.5 ± 0.6 |
| Adenomyosis + Curcumin 50 mg/kg | 6 | 0.85 | 4.5 ± 0.7 | 3.0 ± 0.5 |
| Adenomyosis + Curcumin 75 mg/kg | 6 | 0.95 | 4.8 ± 0.6 | 3.2 ± 0.5 |
Curcumin, a bioactive constituent extracted from the medicinal plant Curcuma longa (turmeric), is widely recognized for its broad spectrum of pharmacological effects. Multiple studies have examined its potential therapeutic applications, particularly in regulating gene transcription, and its anti-inflammatory, antineoplastic, and anti-aging properties. In the current study, the impact of curcumin was assessed in a rat model of adenomyosis. The results indicated no statistically significant differences in lesion number, lesion volume, or adhesion scores among the control group, the untreated adenomyosis group, and the nano-curcumintreated adenomyosis groups. These findings are in agreement with the study by Jie Ding et al., who also observed no significant changes in adhesion scores, lesion counts, or lesion volumes between control, untreated, and curcumin-treated adenomyosis rats.18 In support of these findings, Chowdhury et al. reported that curcumin inhibits the NF-B signaling pathway, thereby reducing proangiogenic and proinflammatory factors in human eutopic endometrial stromal cells. This suggests an anti-inflammatory mechanism of action, though not necessarily resulting in direct reductions in lesion size or severity within adenomyosis models.16 Additionally, a review by Arablou and Kolahdouz-Mohammadi outlined the molecular mechanisms by which curcumin influences inflammation and angiogenesis in endometriosis. Despite limited evidence regarding its direct effects on lesion morphology, the authors highlighted its promising therapeutic potential, thus warranting further investigation in related gynecological disorders such as adenomyosis.20 The Homeobox A10 (HOXA10) gene has been closely linked to endometriosis, although its precise role in disease pathogenesis remains inadequately defined due to limited empirical data. Prior research has demonstrated decreased expression of HOXA10 in women with endometriosis compared to healthy controls. Among the ten most recent studies, half reported an association between reduced HOXA10 expression and infertility. In this study, significantly lower HOXA10 expression was found in the adenomyosis group without nano-curcumin treatment, compared to both the nano-curcumin-treated group and the control group.
HOXA10 is known to be a critical modulator of endometrial receptivity. It promotes endometrial implantation capability by enhancing Ecadherin promoter binding and directly upregulating Ecadherin expression. Furthermore, HOXA10 influences endometrial responsiveness to estrogen and progesterone, contributing to a receptive environment for embryo implantation. The gene also regulates several key biomarkers, including vascular endothelial growth factor (VEGF), osteopontin (OPN), and cyclooxygenase-2 (COX-2). Diminished HOXA10 expression has been associated with reduced endometrial receptivity, lower implantation rates, and elevated infertility risk.11,12 This study also observed elevated expression levels of leukemia inhibitory factor (LIF) and integrin ß3 in the adenomyosis group treated with nano-curcumin. These results are consistent with findings from Guan et al., who demonstrated that melatonin improved pregnancy outcomes in a murine adenomyosis model by upregulating endometrial receptivity markers, including LIF and integrin ß3.13 Similarly, the current findings correspond with the work of Zutautas et al., who reported decreased LIF expression in women with endometriosis, particularly during the mid-secretory phase of the menstrual cycle. This decline was accompanied by elevated levels of IL-6 and IL-1, proinflammatory cytokines known to adversely affect female fertility and closely associated with endometriosis. LIF has been identified as an important mediator of anti-inflammatory effects and a regulator of three principal signaling pathways: Janus kinase/signal transducer and activator of transcription (JAK/STAT3), mitogen-activated protein kinase (MAPK), and phosphatidylinositol-3 kinase (PI3K), all of which are involved in cellular proliferation and tissue regeneration.21 Accordingly, nano-curcumin administration may offer a promising therapeutic strategy for the prevention and management of adenomyosis.
Integrin ß3 gene expression
Integrin ß3 expression also decreased in the adenomyosis group (2.0 ± 0.4) compared to the control group (3.5 ± 0.5). Treatment with nano-curcumin improved integrin ß3 expression to 2.5 ± 0.6, 3.0 ± 0.5, and 3.2 ± 0.5 for the 25 mg/kg, 50 mg/kg, and 75 mg/kg groups, respectively (Table 2).
To our knowledge, this is among the first studies to examine the effects of nano-curcumin on endometrial receptivity markers within an adenomyosis model. Nevertheless, several limitations should be acknowledged. The findings were obtained from an animal model and may not fully reflect human pathophysiology. In addition, the sample size was relatively small, and the long-term effects of nanocurcumin were not investigated. Future research should include clinical trials and more extensive molecular analyses to confirm these preliminary observations and further evaluate the therapeutic potential of nanocurcumin in reproductive health.
Conclusion
This study found no significant differences in lesion morphology among all groups. However, nanocurcumin treatment in the adenomyosis model led to increased expression of HOXA10, LIF, and integrin ß3, similar to levels observed in the control group. These findings indicate that nano-curcumin may enhance endometrial receptivity by modulating molecular markers critical for implantation. Further studies are recommended to confirm these effects in clinical models and to explore the therapeutic potential of nanocurcumin in reproductive health.
DISCLOSURES
The authors sincerely thank the Faculty of Medicine, Universitas Riau, for providing the necessary facilities and laboratory support throughout this study. We also acknowledge the valuable contributions of the research team and laboratory staff during data collection and analysis. Furthermore, we appreciate the constructive feedback and guidance from the doctors in the Department of Obstetrics and Gynecology at the Faculty of Medicine, Universitas Riau, which significantly enhanced the quality of this research.
Additional Section
Majalah Obstetri & Ginekologi 2026;34(1):9-15 Sihotang et al. : Nano-curcumin and endometrial receptivity in adenomyosis Journal of Obstetrics & Gynecology Science p-ISSN: 0854-0381; e-ISSN: 2598-1013.
Funding Information
This study was financially supported by the DIPA LPPM Universitas Riau for the fiscal year 2024, under contract number 9499/UN19.5.1.3/AL.04/2024. The funding institution had no role in the study design, data collection, analysis, or interpretation of the results.
Data Availability
The data supporting the findings of this study are not publicly available due to confidentiality and ethical restrictions.
Conflict of interest
All authors declare that they have no conflict of interest related to this study.
Author contribution
All authors significantly contributed to the research process, including study conception and design, data collection and analysis, manuscript drafting, and final approval of the version to be published.
Copyright & License
© 2026 Jojor Sihotang, Imelda EB Hutagaol, Adriansyah, Muhammad Yusuf.
This work is licensed under a Creative Commons Attribution 4.0 International License.1. Copyright of the article is transferred to the journal, by the knowledge of the author, whilst the moral right of the publication belongs to the author.2. The legal formal aspect of journal publication accessibility refers to Creative Commons Attribution-Non Commercial-Share alike (CC BY-NC-SA), (https://creativecommons.org/licenses/by-nc-sa/4.0/)3. The articles published in the journal are open access and can be used for non-commercial purposes. Other than the aims mentioned above, the editorial board is not responsible for copyright violationThe manuscript authentic and copyright statement submission can be downloaded ON THIS FORM.
Dason E, Yong P, Smith C, et al. Guideline No. 437: Diagnosis and management of adenomyosis. J Obstet Gynaecol Can. 2023;45(6):417–29.
Vannuccini S, Tosti C, Carmona F, Huang SJ, Chapron C, Guo SW, Petraglia F. Pathogenesis of adenomyosis: an update on molecular mechanisms. Reprod Biomed Online. 2017 Nov;35(5):592-601. doi: 10.1016/j.rbmo.2017.06.016.
Antero MF, Ayhan A, Segars J, Shih IM. Pathology and Pathogenesis of Adenomyosis. Semin Reprod Med. 2020 May;38(2-03):108-118. doi: 10.1055/s-0040-1718922. Epub 2020 Oct 20. PMID: 33080632; PMCID: PMC7987203.
Ahn SH, Monsanto SP, Miller C, Singh SS, Thomas R, Tayade C. Pathophysiology and Immune Dysfunction in Endometriosis. Biomed Res Int. 2015;2015:795976. doi: 10.1155/2015/795976. Epub 2015 Jul 12. PMID: 26247027; PMCID: PMC4515278.
Upson K, Missmer SA. Epidemiology of adenomyosis. Semin Reprod Med. 2020;38(2–3):89–107.
Morassutto C, Monasta L, Ricci G, Barbone F, Ronfani L. Incidence and estimated prevalence of endometriosis and adenomyosis in Northeast Italy: a data linkage study. PLoS One. 2016;11(4):e0154227.
Mishra I, Melo P, Easter C, Sephton V, Dhillon-Smith R, Coomarasamy A. Prevalence of adenomyosis in women with subfertility: systematic review and meta-analysis. Ultrasound Obstet Gynecol. 2023;62(1):23–41.
Dian T, Adhi P, Darojat ARS, Shofiatun NA, Tono D. Comparison of oocyte count, fertilization, and pregnancy rates in adenomyosis patients undergoing in vitro fertilization with short and long protocol controlled ovarian stimulation: a retrospective study. Ann Med Surg (Lond). 2022;82:104620.
Chen Q, Li YW, Wang S, Fan QB, Shi HH, Leng JH, et al. Clinical manifestations of adenomyosis patients with or without pain symptoms. J Pain Res. 2019;12:3127–33.
Gunther R, Walker C. Adenomyosis. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2023.
Benagiano G, Habiba M, Brosens I. The pathophysiology of uterine adenomyosis: an update. Fertil Steril. 2012 Sep;98(3):572-9. doi: 10.1016/j.fertnstert.2012.06.044. Epub 2012 Jul 21. PMID: 22819188.
Peng Y, Ao M, Dong B, et al. Anti-Inflammatory Effects of Curcumin in the Inflammatory Diseases: Status, Limitations and Countermeasures. Drug Design, Development and Therapy. 2021 ;15:4503-4525. DOI: 10.2147/dddt.s327378. PMID: 34754179; PMCID: PMC8572027.
Lessey BA, Young SL. What exactly is endometrial receptivity? Fertil Steril. 2019;111(4):611–7.
Ding J, Mei S, Cheng W, Ni Z, Yu C. Curcumin treats endometriosis in mice by the HIF signaling pathway. Am J Transl Res. 2022;14(4):2184–98.
Chowdhury I, Banerjee S, Driss A, Xu W, Mehrabi S, Nezhat C, Sidell N, Taylor RN, Thompson WE. Curcumin attenuates proangiogenic and proinflammatory factors in human eutopic endometrial stromal cells through the NF-κB signaling pathway. J Cell Physiol. 2019 May;234(5):6298-6312. doi: 10.1002/jcp.27360. Epub 2018 Sep 27. PMID: 30259980; PMCID: PMC6344303.
Arablou T, Kolahdouz-Mohammadi R. Curcumin and endometriosis: Review on potential roles and molecular mechanisms. Biomedicine & Pharmacotherapy = Biomedecine & Pharmacotherapie. 2018 Jan;97:91-97. DOI: 10.1016/j.biopha.2017.10.119. PMID: 29080464.
Bi Y, Huang W, Yuan L, Chen S, Liao S, Fu X, et al. HOXA10 improves endometrial receptivity by upregulating E-cadherin. Biol Reprod. 2022;106(5):992–9.
Lazim N, Elias MH, Sutaji Z, Abdul Karim AK, Abu MA, Ugusman A, et al. Expression of HOXA10 gene in women with endometriosis: a systematic review. Int J Mol Sci. 2023;24(11):12869.
Guan X, Liu D, Zhou H, Dai C, Wang T, Fang Y, et al. Melatonin improves pregnancy outcomes in adenomyosis mice by restoring endometrial receptivity via NF-κB/apoptosis signaling. Ann Transl Med. 2022;10(24):1317.
Zutautas KB, Sisnett DJ, Miller JE, Lingegowda H, Childs T, Bougie O, et al. The dysregulation of leukemia inhibitory factor and its implications for endometriosis pathophysiology. Front Immunol. 2023;14:1089098.
Copyright (c) 2026 Jojor Sihotang, Imelda EB Hutagaol, Adriansyah, Muhammad Yusuf
This work is licensed under a Creative Commons Attribution 4.0 International License.
1. Copyright of the article is transferred to the journal, by the knowledge of the author, whilst the moral right of the publication belongs to the author.
2. The legal formal aspect of journal publication accessibility refers to Creative Commons Attribution-Non Commercial-Share alike (CC BY-NC-SA), (https://creativecommons.org/licenses/by-nc-sa/4.0/)
3. The articles published in the journal are open access and can be used for non-commercial purposes. Other than the aims mentioned above, the editorial board is not responsible for copyright violation
The manuscript authentic and copyright statement submission can be downloaded ON THIS FORM.
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.