Effect of Meibomian Gland Expression on Meibomian Gland Regeneration and Function: A Randomized Controlled Trial Authors: Ngoc Van Anh Le, Patraramon Chotikkakamthorn, Ngamjit Kasetsuwan, Huong Luu, Usanee Reinprayoon Published: 2026 ABSTRACT AbstractIntroductionThis study aims to compare the effect of meibomian gland expression (MGX) combined with lid hygiene versus lid hygiene alone on the meibomian gland’s morphology and function for obstructive meibomian gland dysfunction (oMGD).MethodsIn this assessor-masked, parallel-group, randomized controlled trial, 64 oMGD patients with mild and moderate meibomian gland (MG) dropout were randomly allocated 1:1 into the MGX and control groups. All participants were informed to do lid hygiene once daily, while the intervention group received MGX every 2 weeks. The primary outcome was the change in MG dropout. Secondary outcomes included the Ocular Surface Disease Index (OSDI), tear stability, MG functions, changes in tear cytokine levels, and safety of MGX in a 12-week follow-up.ResultsSixty-two eyes of sixty-two patients were enrolled in this study. The dropout area of MG decreased significantly in both groups following treatment at all time points. The recovery was faster in the MGX group at 4 weeks (p<0.001). However, no significant difference was observed between the groups after 8 to 12 weeks. Additionally, the MGX group also showed better improvement in OSDI and meibum quality than the control group, while no significant differences were observed between groups for other measures. No adverse events were reported, and the pain level was mild.ConclusionMGX combined with lid hygiene results in a 13% reduction in MG dropout in the upper and lower eyelids. Moreover, the rate of improvement in MG morphology was faster in patients who underwent additional MGX than in those who performed lid hygiene alone.Keywords: morphology, meibomian gland, dropout, meibomian gland expression, meibomian gland dysfunction, meibography Introduction Meibomian Gland Dysfunction (MGD) is one of the most common conditions encountered in the ophthalmology clinic, with a global prevalence of 35.8%.1 Meibomian gland (MG) atrophy or dropout is one of the pathognomonic characteristics of MGD. The prevailing hypothesis of MG dropout results from obstruction of the gland’s secretory duct, leading to cystic dilation of the ducts and acini, which in turn causes disuse acinar atrophy.2 The obstruction can be caused by hyperkeratinization or changes in lipid composition and quality of meibum.3 In animal studies, meibocytes could be stimulated to proliferate through short-term pharmacological treatment to promote meibocyte renewal, leading to the regeneration of MG.4,5 Appropriate treatment can promote acinar basal cell proliferation, followed by mature meibocyte proliferation. Therefore, treatment promotes the number of acini, increases acinar size, and enhances meibomian gland function in aged mice in histopathological samples.5 Although many clinical trials have demonstrated that physical therapies such as lid hygiene, warm compress, and meibomian gland expression can effectively improve MGD-related symptoms, their impact on human MG morphology remains inadequately elucidated. Meibomian gland expression (MGX) has been described as an efficient and cost-effective therapeutic procedure for obstructive MGD and a complementary approach to lid hygiene.6,7 After removing debris from the lid margin, pressure is manually applied to the eyelids using a rigid object such as a glass rod, cotton swab, forceps, or metal paddle. This technique helps express viscous meibum from the eyelid and restores normal meibum flow. An interventional case series showed significant improvements in patients with mild-to-moderate MG dropout 4 weeks after MGX. In addition to symptoms, tear stability, and meibum expressibility improvements, the meibography score significantly decreased from 3.24 to 2.76 (p<0.01).8 However, the number of clinical studies on the efficacy of MGX combined with lid hygiene is minimal.9 This study aims to assess the impact of combining MGX with lid hygiene on improving meibomian gland morphology in patients with significant MG dropout at baseline by evaluating the reversal of the dropout area to normal gland morphology after 12 weeks of treatment using an imaging system. In addition, we are also interested in changes in tear film, MG function, tear cytokine levels, and adverse events during the procedure. Materials and Methods Study Design This assessor-blinded, parallel-group, randomized controlled trial followed the tenets of the Declaration of Helsinki. It was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Thailand (IRB Certificate of Approval 0868/66). The study was registered at ClinicalTrials.gov (NCT06163105) and adhered to the CONSORT statement. The study was conducted between October 2024 and March 2025 at the Chula Refractive Surgery Center of King Chulalongkorn Memorial Hospital (Bangkok, Thailand). Informed consent was obtained in a private and quiet area. Participants Participants were required to be ≥18 years old with a diagnosis of obstructive MGD (oMGD) according to the Japanese MGD working group10 and having visible MG dropout of the upper and lower eyelid on meibography images from grade 1 to grade 2 based on Arita’s meiboscale (Grade 0: no gland loss; Grade 1: area loss < 1/3 of total MG area; Grade 2: area loss between 1/3 and 2/3; Grade 3: area loss > 2/3).11 Stage 3 was excluded because MGX has limited effectiveness in treating MGD with severe dropout.12 If both eyes met the above criteria, only the right eye was chosen for each patient. The exclusion criteria comprised of the following: (1) diagnosis of secondary MGD, (2) any use of eyedrop other than eye lubricant within the last 3 months, (3) use of any eyelid treatment related to MGD within the last 6 months, (4) history of ocular trauma or ocular surface surgery or eyelid surgery or intraocular surgery less than 6 months or ocular diseases other than dry eye; (5) eyelid botulinum toxin injection within the last 6 months (6) contact lens wear; (7) pregnant or breast-feeding; (8) using any hormones therapy, anti-inflammatory drugs, tetracycline, macrolides, fatty acid medication or retinoid acid; (9) refuse to participate in the study. Randomization and Blinding An independent statistician generated a stratified block randomization list (block sizes 2 and 4) with a 1:1 allocation ratio to the MGX or control group. Stratification based on the severity of MG dropout by calculating the combined scores of the upper and lower eyelids based on the Meiboscale: Group 1 (total score 1–2) exhibited mild gland dropout, while Group 2 (total score 3–4) experienced moderate gland loss. Two sets of 64 opaque, sealed envelopes (one per stratum) were prepared in advance. Participants selected an envelope based on their stratum, and the assigned group was revealed and recorded by a research assistant (T.G). Clinical examiner (P.C.) and meibography interpreter (A.L.) were blinded to group allocation. Treatment Protocol The participants in the MGX group were scheduled to do in-office MGX every 2 weeks (6 sessions in total) at Chula Refractive Surgery Center. All MGX procedures were conducted by the same nurse following a standardized protocol: (1) topical anesthesia; (2) cleaning the eyelids; (3) warming the eyelids by heat and moist googles Blephasteam (Spectrum Thea Pharmaceuticals LTD, UK) for 10 minutes; (4) topical anesthesia again; (5) using Mastrota paddle and Q-tip to squeeze the gland; (6) cleaning the lid; (7) instilling Vigamox 0.5% eyedrop (Alcon Lab., Texas, USA). The essential treatment of the two groups included lid hygiene once daily and preservative-free artificial tears (0.18% sodium hyaluronate) 4 times daily throughout 12 weeks. Patients were requested not to use any other eye drops or treatment methods. To standardize lid hygiene, the research assistant demonstrated the procedure and verbally instructed participants on how to perform lid hygiene at home. The patients heated the eye mask in the microwave for 30 seconds. The procedure began with hot compression with a reusable eye mask for 10 minutes. Then, a light massage was applied to both eyelids for 30 seconds. Finally, the participants cleaned their eyelids and lid margins. If they forgot these steps, the participants could watch the tutorial video via a provided QR code. To remind patients to perform eyelid hygiene, they were provided a logbook to record the daily treatment. To assess compliance, patients were divided into two groups: the compliance group, who performed eyelid hygiene at least 5 days per week for over 80% of the treatment period, and the non-compliance group, who did not meet these criteria. Outcome Measurement Examinations were scheduled at baseline, week 4, week 8, and week 12 after treatment. Demographic data, including age and sex, were obtained for all patients before the baseline examination. Each examination conducted by an ophthalmologist was composed of 5 parts: (1) symptom evaluation: Ocular Surface Disease Index (OSDI); (2) tear function assessment: lipid layer thickness (LLT) and fluorescein tear breakup time (FTBUT); (3) lid margin assessment; (4) meibomian gland function assessment: meibum quality (MQ), meibum expressibility (ME); and (5) Meibography. Primary Outcome The primary outcome was changing the MG dropout area using non-contact infrared meibography (Keratograph 5M; Oculus OptikgeräteGmbH, Wetzlar, Germany) with consecutively upper and lower eyelids everted. To calculate the dropout area percentage of each eyelid, the investigator analyzed the images by utilizing the ImageJ software (version 1.53k). These areas were defined according to Pult’s protocol13 (Figure 1). The dropout area of each eyelid is calculated by (Area of MG dropout) / (Area of tarsal plate)*100%. Interpretation of MG dropout by Image J. The dropout area was defined by (1) the actual endpoints of the glands (A), (2) the width of the region, which extended at least from the tear punctum (C) to the temporal boundary marked by the most clearly visible tarsal conjunctiva of the everted eyelid (B), and (3) the maximum depth, estimated based on the expected termination of glands in normal MG (D). In measuring the total area, reference lines “B,” “C,” and “D” remained unchanged, with only the border “A” adjusted by mouse clicks to align with the lid margin. Secondary Outcomes To evaluate the symptoms of MGD, participants were interviewed with the OSDI questionnaire (Allergan Inc., Irvine, CA, USA). Lipid layer thickness was measured by Lipiview interferometer (TearScience Inc., Morrisville, NC, USA). Patients were instructed not to use artificial tears for at least 4 hours and ointments for at least 24 hours before the LipiView assessment. FTBUT was determined by the mean value of 3 consecutive measurements with fluorescein dye and cobalt blue illumination. LMS in the upper and lower eyelids were summed according to the criteria of Arita et al.14 The ME and MQ scores were conducted by the Evaluator (TearScience Inc., North Carolina, USA) by applying a persistent pressure of 1.2 g/mm2 just below the lash line, specifically at the nasal, central, and temporal parts of the eyelid margin. MQ and ME were assessed according to the International Workshop on Meibomian Gland Dysfunction15 (Figure 2). Meibum quality and meibum expressibility assessment. Yellow circle: indicates the eight glands located in the central third of the lower eyelid used for MQ assessment. Blue box: indicates the five central glands of the lower eyelid used for ME assessment. Green line: indicates the division of the eyelid into three equal parts, with the central third used for gland assessment. MQ was obtained by evaluating each of eight glands (yellow circle) of the central third of the lower eyelid on a scale of 0 to 3 for each gland (0: clear; 1: cloudy; 2: cloudy with debris (granular); and 3: thick, like toothpaste). ME was evaluated on a scale ranging from 0 to 3 for the lower eyelid’s five central glands (blue box). The assessment is based on the number of glands that can be expressed (0, all glands; 1, three to four glands; 2, one to two glands; and 3, no gland). Safety The safety of MGX was assessed based on any occurrence of adverse events and pain levels during the study. Unfavorable symptoms or signs will be evaluated at each examination or self-reported by patients. Pain level was evaluated through an interview within 15 minutes after the procedure using a numeric rating scale-11 (NRS-11). Pain levels range from 0 to 10: 0, no pain; 10, worst possible pain; 1 to 3, mild pain; 4 to 6, moderate pain; and 7 to 10, severe pain. Tear Collection The Schirmer test I was conducted following the guidelines of TFOS DEW II16 and then transferred to a 2 mL centrifuge tube. The value of the soaked length in millimeters was also recorded. After that, this tube was immediately stored on ice for 20 minutes to 1 hour before being transferred to be stored in the −80°C freezers until cytokine profiling. The Bio-Plex Pro Human Cytokine 8-plex (Bio-Rad Laboratories Inc., Hercules, CA) used in this study was composed of the following interleukins (IL): IL-2, IL-4, IL-6, IL-8, IL-10, tumor necrosis factor-alpha (TNF-α), granulocyte-macrophage-colony stimulating factor (GM-CSF), and interferon-gamma (IFN-γ). The cytokine extraction complied with VanDerMeid’s protocol.17 However, we did not dilute the sample after centrifugation because only 50 µL of the sample was required. Samples with cytokine concentrations below the limit of detection (LOD) were assigned a value of “LOD/2” in the statistical analysis.18,19 Sample Size The standard deviation of the meibomian gland dropout area in obstruction MGD patients is 16.7%.20 For a power of 80% to detect a difference of 10% at a significance level of 0.05, 28 participants were needed. This was inflated for a 10% dropout, giving a total sample size of 32 per group (64 in total). Statistical Analysis Stata/SE 18 software (StataCorp, College Station, TX) was utilized. All statistical analyses were undertaken by a researcher blinded to group allocation. Demographic data was interpreted by descriptive analysis. The primary outcome was the change in MG dropout percentages of each eyelid from baseline over all follow-up visits. The change was calculated using a linear mixed model (LLM) with random intercept and random slope for patient, adjusting for baseline MG percentage, and modelling time as a discrete variable. For secondary outcomes, continuous variables (OSDI, FTBUT, LLT, LMS, MQ, ME, and MG morphology) were analyzed by LMM. The Wilcoxon signed-rank test and Mann–Whitney-U test were performed for cytokine levels. A significance level of 5% was used in all statistical tests. Results A total of 64 patients met the eligibility criteria and were randomized into the study (Figure 3); 2 patients discontinued before the 4-week visit. Finally, 62 participants completed the study with MGX (n = 31) or control (n = 31). Most of the participants were female (85.5%) and had a mean age (SD) of 47.3 (14.1) years. The rate of eyelid hygiene compliance at home is 79%, 79%, and 66% at 4, 8, and 12 weeks, respectively. Detailed baseline outcomes were described in Table 1. CONSORT diagram of the enrollment, allocation, follow-up, and analysis of the study participants. Table 1Baseline Characteristics of the Two GroupsMGXControlDemographic dataAge (years)Mean (SD)46.7 (14.8)48.0 (13.6)Gendern (%)Male4 (12.9)5 (16.2)Female27 (87.1)26 (83.8)Meiboscalen (%)Grade 113 (41.9)12 (38.7)Grade 218 (58.1)19 (61.3)Good compliancen (%)24 (77.42)25 (80.65)Clinical dataMG dropout percentageMean (SD)51.41 (15.80)53.32 (14.1)Upper eyelid dropoutMean (SD)27.25 (8.26)28.28 (8.31)Lower eyelid dropoutMean (SD)24.16 (9.88)25.04 (9.00)OSDIMedian (IQR)40 (20, 57)21 (14, 35)LLTMedian (IQR)52 (32, 64)50 (40, 61)TBUTMedian (IQR)5 (4, 6)5 (4, 7)LMS (0–10)Median (IQR)Upper lid2 (1, 3)3 (1, 3)Lower lid3 (2, 3)3 (2, 4)MQ (0–24)Median (IQR)21 (19, 24)20 (16, 22)MEMean (SD)1.81 (0.13)1.83 (0.12) Baseline Characteristics of the Two Groups Primary Outcome At week 4, both the MGX and control groups showed significant reductions in MG dropout area, with the MGX group exhibiting a greater mean change of −9.45% (95% CI: −11.94, −6.96) compared to the control group’s −4.06% (95% CI: −6.56, −1.57). The between-group comparison revealed a significant difference of −5.47 (95%: CI −8.13 to −2.82) in favor of the MGX group. By week 8, both groups continued to show significant reductions in MG dropout, with the MGX group improving by −11.81% (95% CI: −14.75 to −8.86) and the control group by −10.29%. At week 12, the reductions were −13.23% (95% CI: −16.80, −9.65) for the MGX group and −12.35% (95% CI: −15.93, −8.78) for the control group (both p-values <0.001). However, the between-group difference was not significant at week 8 and week 12 (p > 0.05). Overall, while the MGX group demonstrated a faster improvement early on, by the end of the 12-week period, both treatments showed similar reductions in MG dropout, with no significant difference between the groups (Figure 4). Mean difference in MG dropout percentage of both eyelids from baseline at each visit. Values were presented as mean ± 95% CI. (†p < 0.001, within-MGX group differences. *p < 0.001, within-control group differences. §p = 0.002, between-group difference). Meibomian Gland Functions OSDI A better improvement in OSDI was found in the MGX group than in the control group at all time points, with p-values < 0.05. The between-group mean differences were −6.82 (95% CI: −11.67, −1.97), −9.75 (95% CI: −17.21, −2.30), and −12.65 (95% CI: −23.05, −2.26) at weeks 4, 8, and 12, respectively. There are no significant differences between pre- and post-treatment in the control group at all time points (Table 2). Table 2Comparison of Mean Changes in Study’s Outcomes Between MGX and Control Groups Over TimeVariablesAt WeekMGX GroupControl GroupBetween GroupMean Change(95% CI)p-valueMean Change(95% CI)p-valueMean Difference(95% CI)p-valueMG dropout area (%)4−9.45 (−11.94, −6.96)<0.001−4.06 (−6.56, −1.57)0.001−5.47 (−8.13, −2.82)<0.0018−11.81 (−14.75, −8.86)<0.001−10.29 (−13.23, −7.35)<0.001−1.31 (−3.65, 1.03)0.3712−13.23 (−16.80, −9.65)<0.001−12.35 (−15.93, −8.78)<0.001−0.96 (−5.45, 3.54)0.68MG dropout of upper eyelid (%)4−5.90 (−7.25, −4.55)0.001−2.97 (−4.32, −1.62)<0.001−2.94 (−4.84, −1.03)0.0038−7.94 (−9.67, −6.21)<0.001−6.62 (−8.34, −4.88)<0.001−1.32 (−3.77,1.12)0.2912−7.87 (−10.01, −5.68)<0.001−7.23 (−9.41, −5.04)<0.001−0.65 (−3.74, 2.45)0.68MG dropout of lower eyelid (%)4−3.55 (−5.0, −2.09)<0.001−1.10 (−2.55, 0.36)0.24−2.45 (−4.51, −0.39)0.028−3.87 (−5.78, −1.96)<0.001−3.68 (−5.59, −1.77)0.001−0.19 (−2.89, 2.50)0.8912−5.35 (−7.81, −2.90)<0.001−5.13 (−7.81, −2.90)<0.001−0.23 (−3.69, 3.24)0.89OSDI (score)4−9.74 (−13.74, −5.75)0.001−1.16 (−5.16, 2.83)0.57−6.81 (−11.67, −1.97)0.018−14.00 (−19.59, −8.41)0.001−2.48 (−8.07, 3.10)0.38−9.75(−17.21, −2.30)0.01 12−19.35 (−26.88, −11.83)0.001−4.94 (−12.46, 2.59)0.2−12.65 (−23.05, −2.26)0.02TBUT (second)40.81 (−0.09, 1.71)0.080.45 (−0.45, 1.35)0.330.35 (−0.61, 1.24)0.980.94 (−0.05, 1.92)0.060.71 (−0.27, 1.69)0.160.18 (−0.87, 1.24)0.48121.13 (0.02, 2.24)0.0460.39 (−0.72, 1.49)0.490.73 (−0.52, 1.98)0.08LLT (nanometer)47.10 (1.17, 13.02)0.026.10 (0.18, 12.02)0.040.64 (−5.98, 7.26)0.8585.13 (−1.19, 11.45)0.11−0.77 (−7.09, 5.54)0.815.54 (−2.12, 13.20)0.16126.29 (−0.763, 13.21)0.082.10 (−4.83, 9.02)0.553.83 (−5.11, 12.78)0.4LMS of upper eyelid (0–10)40 (−0.42, −0.42)1−0.42 (−0.84, −0.001)0.0490.31 (−0.14, 0.76)0.188−0.65 (−1.12, −0.17)0.01−0.65 (−1.12, −0.17)0.01−0.11 (−0.66, 0.45)0.712−0.84 (−1.39, −0.29)0.003−0.68 (−1.22, −0.13)0.02−0.27 (−0.95, 0.42)0.44LMS of lower eyelid (0–10)4−0.81 (−1.21, −0.40)<0.001−1.10 (−1.50, −0.69)<0.0010.18 (−0.28, 0.64)0.278−1.23 (−1.70, −0.75)<0.001−1.42 (−1.90, −0.94)<0.0010.08 (−0.52, 0.68)0.5712−1.52 (−2.09, −0.94)<0.001−1.23 (−2.03, −0.88)<0.001−0.18 (−0.94, 0.59)0.88Meibum quality (0–24)4−3.29 (−5.42, −1.16)0.002−3.00 (−5.13, −0.87)0.006−0.23 (−1.97, 2.43)0.858−5.77 (−8.10, −3.44)<0.001−3.77 (−6.10, −1.44)0.001−1.48 (−4.05, 1.09)0.1712−6.77 (−9.41, −4.14)<0.001−3.13 (−5.76, −0.49)0.02−3.13 (−6.23, −0.03)0.048Meibum expressibility (0–3)4−0.48 (−0.80, −0.17)0.003−0.58 (−0.89, −0.27)<0.0010.10 (−0.21, 0.41)0.538−0.94 (−1.29, −0.59)<0.001−0.94 (−1.29, −0.59)<0.0010.003 (−0.35, 0.35)0.9912−1.10 (−1.49, −0.71)<0.001−0.71 (−1.10, −0.32)<0.001−0.38 (−0.81, 0.04)0.08 Comparison of Mean Changes in Study’s Outcomes Between MGX and Control Groups Over Time LLT LLT only significantly increased at week 4 in both the MGX and control groups, with values of 7.10 (95% CI: 1.18, 13.02) and 6.10 (95% CI: 1.18, 12.02), respectively. There was no significant difference between the two groups at any time point (Table 2). FTBUT There was no significant difference in FTBUT in the control group (p>0.05), while FTBUT significantly increased by 1.13 seconds in the MGX group at 12 weeks (p = 0.046). However, the between-group differences were not significant (Table 2). Lid Margin Score In the MGX group, the LMS of the upper eyelid decreased by 0.65 (95% CI: −1.12, −0.17) at week 8 and 0.84 (95% CI: −1.39, −0.29) at week 12 compared to baseline. The control group also showed significant reductions in upper eyelid LMS at weeks 4, 8, and 12. However, no statistically significant between-group differences were found at any time point (all p > 0.05) (Table 2). For the lower eyelid, both groups showed a significant decrease in LMS from baseline at all follow-up visits (p<0.001), but there was no significant difference detected between the two groups at any follow-up time point (p > 0.05 for all) (Table 2). Meibum Quality and Meibum Expressibility After 12-week treatment, ME decreased by −1.10 (95% CI −1.50, −0.69) and −0.71 (95% CI −1.11, −0.31). However, no statistically significant between-group differences were observed in ME at any time point (p>0.05). Similarly, MGX and control significantly improved MQ at all time points (p<0.01). MGX group had greater reductions in MQ were only observed in the intervention group at week 12 with a between-group mean difference of –3.13 (95% CI: –6.23 to –0.03) (Table 2). Tear Cytokines IL-2, IL-4, IL-10, GM-CSF, TNF-α, and IFN-γ were undetectable in more than 50% of samples and were excluded from further analysis. A statistically significant reduction was found in IL-6 in the MGX and control group over 12 weeks by 87% and 82%, respectively (p<0.001). In addition, IL-8 level was reduced by 47% after MGX (p=0.02), whereas the control group had no significant change in IL-8 level after treatment (Table 3). Intergroup comparison at each time point did not reveal statistically significant differences in either IL-6 or IL-8 (p=0.28 and p=0.56, respectively). Table 3The Cytokine Levels Found in the MGX Group Compared to the Control GroupMGXControlBetween GroupGM (95% CI)GM (95% CI)GMR (95% CI)p-valueIL6Baseline13.73(5.29, 35.63)26.30(9.50, 72.81)12 weeks1.77(0.61, 5.16)4.81(1.70, 13.61)0.59 (0.22, 1.55)0.28GMR0.13(0.05, 0.34)0.18(0.07, 0.48)p-value<0.001<0.001IL8Baseline757(363.13, 1578.10)831.17(345.56, 1999.19)12 weeks375.22(162.17, 868.15)447.19(190.10, 1051.96)0.85 (0.49, 1.46)0.56GMR0.53(0.31, 0.91)0.63(0.36, 1.07)p-value0.020.09Abbreviations: GM, geometric mean; GMR, geometric mean ratio; CI, confident interval; IL6 interlukin 6; IL8, interlukin 8. The Cytokine Levels Found in the MGX Group Compared to the Control Group Abbreviations: GM, geometric mean; GMR, geometric mean ratio; CI, confident interval; IL6 interlukin 6; IL8, interlukin 8. Safety Over the 12-week visit, no adverse events were reported in all participants. Although the patients underwent the MGX procedures with topical anaesthesia, they generally experienced no pain or mild discomfort, with the median (IQR) of the total pain score being 2.00 (0, 3). In the first procedure, 77.4% of patients experienced no or mild pain. In subsequent procedures (rounds 2 to 6), the percentage of patients with no or mild pain increased, ranging from 83.9% to 93.5% (Figure 5). Graph of pain level after each MGX treatment. Discussion This randomised controlled trial demonstrated that both office-based MGX and lid hygiene effectively promoted improvement in MG morphology by reducing the MG dropout area. Especially, the findings indicated that MGX offers faster therapeutic benefits than lid hygiene alone during the early treatment period, particularly in the first 4 weeks, while lid hygiene alone caught up by 12 weeks. We defined clinical significance as a 10% improvement in gland dropout area based on expert opinion, since there is no consensus on how much gland opening is considered acceptable Although our primary outcome has not yet been reached, this serves as proof of concept that the MG can regenerate. Although clinicians commonly interpret gland dropout regions as permanent MG atrophy, these areas have the potential to recover, signifying a restoration of normal physiological function. This reversal suggests that the meibocytes may not be entirely nonviable, or that the condition has not yet progressed to an irreversible stage. Furthermore, other clinical signs and symptoms support this finding, showing positive improvement (Figure 6). The meibography images of the upper and lower eyelid before and after treatment. Symptomatically, patients in the MGX group experienced a faster improvement in OSDI scores, implying superior relief of dry eye symptoms. This early symptom relief is likely attributable to the quicker restoration of gland function with MGX. In spite of stratified randomization, two groups had a notable baseline difference in OSDI. This imbalance may reflect, at least in part, the lack of participant masking and the potential influence of expectation or perceived therapeutic benefit associated with the more intensive, procedure-based MGX intervention. In addition, the higher baseline symptom burden in the MGX group raises the possibility of regression to the mean contributing to the observed improvement. To overcome this limitation, we analyzed the change from baseline and subsequently compared the magnitude of change between the groups. However, we acknowledge that residual bias related to treatment intensity, patient expectation, and attention effects may still have influenced subjective outcomes such as OSDI, and therefore the symptomatic benefits of MGX should be interpreted with caution. Moreover, MQ improvement in the MGX group was significantly higher than in the control group only by week 12, suggesting that MGX directly targets the underlying mechanism of MGD, which is obstruction of the MG. This intervention may improve the quality of meibum. While both groups demonstrated significant reductions in ME and LMS in both upper and lower eyelids, no significant differences were observed between groups, indicating that lid hygiene and MGX offer comparable improvements in expressibility of MG and lid margin health. A meta-analysis published at the end of 2024 on the efficacy of MGX combined with home-based therapy in managing dry eye related to MGD concluded that there was no difference in FTBUT after treatment,9 which is consistent with the findings of our study. The study by Aketa, in which MGX was performed weekly, reported a significant improvement in FTBUT of approximately 2.43 seconds after 4 weeks of treatment.7 In contrast, Kaiserman et al, applying MGX once per month, found no statistically significant difference In our study, MGX was administered biweekly, resulting in a mean FTBUT improvement of 1.13 seconds; however, this change did not reach statistical significance. These findings suggest that the frequency of MGX treatment may play an important role in therapeutic effectiveness, with more frequent sessions potentially contributing to greater improvements in tear film stability. Our therapeutic intervention shares a similar underlying mechanism with thermal pulsation therapy, as both approaches target the meibomian glands through the application of heat and gentle pressure to the eyelids, thereby facilitating glandular clearance and restoring physiological function. LipiFlow has been shown to improve meibomian gland function, as measured by the Meibomian Gland Yielding Secretion Score (MGYSS), and to reduce ocular surface staining; however, its effects on patient-reported symptoms (OSDI) and tear film stability (FTBUT) appear to be limited.21 Consistent with these findings, our results also showed improvements in symptoms, MQ, ME, and LMS; however, the effects on FTBUT and LLT were limited. A study by Hye Jee Kim showed that both LipiFlow alone and MGX after LipiFlow are effective for obstructive MGD, but adjunctive MGX provides greater and more sustained benefits.22 One possible explanation is that the maximum pressure delivered by LipiFlow (6 psi) may be insufficient, given that approximately 25 psi is required to evacuate all expressible gland contents.23 MGX may therefore serve as a useful adjunct to treatments such as LipiFlow. MGX enhanced the improvement of FTBUT driven by intense pulsed light (IPL) therapy.24 Additionally, since higher temperatures may be required to fully liquefy gland contents in severe MGD, we applied warm compresses prior to MGX to enhance treatment efficacy. IL-8 is a chemokine that plays a crucial role in attracting T cells and neutrophils to the site of inflammation, which can lead to further damage to the meibomian glands.25 On the other hand, IL-6 is produced by various inflammatory cells, including macrophages, neutrophils, T lymphocytes, and epithelial cells of the ocular surface. Higher IL-6 levels in tear fluid were found in the MGD group with dry eye compared to dry eye patients without MGD, indicating that measuring IL-6 could serve as a means to assess treatment effectiveness in MGD associated with dry eye.26 Among the remaining analytes, both groups demonstrated substantial reductions in IL-6 levels over the 12-week period, with no statistically significant difference observed between groups. This finding suggests that the observed decrease in IL-6 may reflect a general treatment or time effect rather than a specific effect of MGX. For IL-8, a statistically significant reduction was observed within the MGX group, whereas the control group showed a non-significant downward trend; however, the between-group comparison did not reach statistical significance. Taken together, these findings suggest a possible signal of inflammatory modulation associated with MGX, but provide insufficient evidence to conclude a superior anti-inflammatory effect compared with the control treatment. MGX has a long-standing history as a treatment for oMGD dating back to 1888.27 It has been proven to be effective and safe.6–8,28 Most patients experienced no pain or only mild discomfort, consistent with the findings of a previous study by Wang et al, where the mean pain level reported was 24 ± 1.0.29 From a clinical perspective, the more rapid improvement in MG morphology and symptoms observed with MGX suggests it may be an appropriate early intervention for patients with oMGD, particularly those presenting with significant symptoms or ocular inflammation. Additionally, the comparable outcomes at 12 weeks suggested that lid hygiene alone could be an effective long-term management, especially for patients who are unable to access in-office MGX treatments. Lid hygiene training is essential for all patients with MGD. A structured lid hygiene training program should be implemented in all ophthalmic centers. To the best of our knowledge, this is the first RCT to evaluate the changes in MG morphology after MGX combined with lid hygiene, providing valuable insights into the structural recovery of MG in response to treatment. In addition to the limitations discussed above, several other limitations should be acknowledged. First, this was a single-center study with a predominantly female sample, limiting generalizability. Second, the study focused only on patients with mild and moderate MG dropout, underscoring the need for future studies of severe MGD cases with a longer follow-up period. Multiple secondary outcomes were repeatedly assessed over time without formal multiplicity adjustment. Although linear mixed-effects models were applied, the risk of type I error cannot be excluded. Therefore, secondary outcome findings should be considered exploratory and interpreted cautiously. Conclusion Adding office-based MGX to lid hygiene may accelerate early improvement in meibomian gland morphology, dry eye symptoms and meibum quality in patients with oMGD. However, the lack of sustained between-group differences in the structural outcomes suggests that the benefit of MGX may mainly reflect earlier morphological improvement rather than long-term superiority over lid hygiene alone. Acknowledgment The authors would like to acknowledge Prof. Stephen John Kerr, Ph.D., and Yuda Chongpison, Ph.D., M.S., MBA, for consulting on statistical analysis. We thank Chula Refractive Surgery Center, Department of Ophthalmology, King Chulalongkorn Memorial Hospital, Thailand, for providing a study site and personnel. Funding Statement This work was financially supported by the Quality Improvement Fund, King Chulalongkorn Memorial Hospital, and The Thai Red Cross Society (Grant number: HA-67-3300-C1-052) and the 2024 Global Health Research Promotion Scheme by the School of Global Health, Faculty of Medicine, Chulalongkorn University. The funder has no role in study design, data collection, analysis, manuscript writing, or publication. Abbreviations MGX, meibomian gland expression; oMGD, obstructive meibomian gland dysfunction; MG, meibomian gland; MGD, meibomian gland dysfunction; OSDI, Ocular Surface Disease Index; LLT, lipid layer thickness; FTBUT, fluorescein tear breakup time; LMS, lid margin score; MQ, meibum quality; ME, meibum expressibility; NRS-11, numeric rating scale-11; IL, interleukins; TNF-α, tumor necrosis factor-alpha; GM-CSF, granulocyte-macrophage-colony stimulating factor; IFN-γ, interferon-gamma; LOD, limit of detection; LMM, linear mixed model; IPL, intense pulsed light. Data Sharing Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Ethics Approval and Informed Consent The study protocol followed the ethical guidelines regarding human subjects given by the Declaration of Helsinki and was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Thailand (IRB Certificate of Approval 0868/66). All participants provided written informed consent prior to enrollment in the study. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors declare that there is no conflict of interest in this work. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.