Device-based treatment of meibomian gland dysfunction: where the evidence stands

Meibomian gland dysfunction (MGD) is the leading driver of evaporative dry eye, and four in-clinic devices now target the glands themselves: intense pulsed light (IPL), vectored thermal pulsation (LipiFlow), thermomechanical treatment (Tixel) and low-power high-frequency electrotherapy, marketed as quantum molecular resonance (Rexon-Eye). They differ in mechanism and in the maturity of their evidence. This review sets them side by side, with the numbers.

Vectored thermal pulsation (LipiFlow)

Hu and colleagues pooled ten randomised trials (761 patients) comparing LipiFlow with lid hygiene (Graefes Arch Clin Exp Ophthalmol 2022). Pooled mean differences favoured LipiFlow on symptoms — OSDI −7.4 (95% CI −11.06 to −3.74), SPEED −2.7 (−3.95 to −1.45) — and signs: corneal staining −0.42 (−0.75 to −0.10), glands yielding liquid secretion +1.3 (0.78 to 1.82), gland secretion score +4.09 (1.18 to 6.99). Adverse events were comparable. The authors attributed much of the heterogeneity to unit-of-analysis errors (eyes counted as independent patients), which widens the true uncertainty; a 7-point OSDI advantage also sits at the lower end of what patients notice.

Durability rests on Blackie and colleagues’ nine-site US trial of 200 adults (Clin Ophthalmol 2016): one 12-minute treatment against three months of twice-daily warm compresses and lid hygiene, then crossover. At three months, gland secretion score (0–45) and OSDI improved more with LipiFlow (P < 0.0001 and P = 0.0068, intention-to-treat); 93% completed 12 months. The 12-month phase was observational: among completers with one treatment and no other prescribed therapy (86% of that arm), gland score rose from 6.4 ± 3.7 to 17.3 ± 9.1 and OSDI fell from 44.1 ± 20.4 to 21.6 ± 21.3. Transient, non-serious device-related events occurred in 5.1% of treated subjects versus 7.1% with control therapy. The trial was open-label, meibography was not used, no confidence intervals were reported, and two of three authors were manufacturer-affiliated. Milder baseline gland scores and shorter time since diagnosis were associated with larger gains — an association, not proof that earlier treatment causes better outcomes. Patient-facing coverage is here.

Intense pulsed light (IPL)

Toyos and colleagues randomised 88 adults with moderate-to-severe MGD (TBUT ≤ 7 s, OSDI ≥ 23) at three sites to four sessions of IPL or sham light, each followed by gland expression; patients and examiners were masked and follow-up was four weeks after the last session (PLoS One 2022). In 82 completers, TBUT rose 2.0 s (95% CI 1.4 to 2.6) with IPL against 0.7 s (0.3 to 1.2) with sham, a difference of about 1.2 s (P = 0.0076 baseline-adjusted). Gland secretion score improved 18.5 versus 5.2 points (difference 13.3, P < 0.0001). OSDI, however, fell 25.9 points in both arms with no between-arm difference (P = 0.998): expression alone relieved symptoms substantially. Adverse events occurred in 8.9% with IPL and 20.9% with sham, none serious; all six dropouts were in the IPL arm, one for procedure pain. The control arm had worse baseline gland function and meibography, and the sponsor performed the post-hoc analyses.

D’Souza and colleagues’ single-centre, double-masked trial randomised 100 patients with chronic MGD to three sessions of IPL plus low-level light therapy or sham, 15 days apart, without gland expression (Indian J Ophthalmol 2023). From a baseline OSDI of about 37 and TBUT of about 5 s, the treated group improved on both at one and three months (P < 0.0001 and P < 0.005), gaining further between visits after treatment had finished; the sham group improved non-significantly. Gland expressibility matched sham and Schirmer was unchanged; no burns or pigmentation occurred. Post-treatment values appear only in graphs, analysis was within-group, and the manufacturer funded the study, so no effect size can be quoted. Patient-facing coverage is here.

Thermomechanical treatment (Tixel)

Safir and colleagues treated 40 patients (80 eyes) at one Israeli centre with three Tixel sessions two weeks apart in an uncontrolled pilot (Cont Lens Anterior Eye 2022). Participants (over 45, TBUT under 5 s, corneal staining) were recruited from cosmetic periorbital-rejuvenation seekers; settings were 6–8 ms at 400 µm protrusion. Thirty-two (80%) reached the final visit (mean 2.1 months): SPEED II 16.5 to 11.8, staining 2.0 to 0.5, MGD score 2.7 to 1.2, drop use 3.4 to 1.9 a day, TBUT 2.7 to 6.5 s (all P < 0.001); two minor events occurred. Examiners were unmasked, so regression to the mean and expectation cannot be separated from device effect, as the authors acknowledge; it was unfunded.

Sadri and colleagues extend to six months a five-site US pivotal trial that randomised three Tixel sessions against one LipiFlow treatment, assessor-masked, and claimed non-inferiority on TBUT change (J Ocul Pharmacol Ther 2025). It took 21 Tixel-treated subjects (42 eyes) who had gained at least 2.5 s of TBUT at one or three months; the Tixel arm’s original size is not stated, no control was carried forward, and the assessor was unmasked. Six-month changes were: TBUT +5.2 ± 3.8 s (95% CI 4.0 to 6.3), gland score +18.2 ± 10.9 (14.8 to 21.6), OSDI −24.3 ± 26.5 (−36.3 to −12.2), not significantly different from three months, though conjunctival staining had returned to baseline; no ocular adverse events occurred. Novoxel funded the study and writing; two authors are its employees. A responder-only sample of unknown denominator cannot, by construction, say how often treatment works; one LipiFlow session against three Tixel sessions is a lenient comparator. Patient-facing coverage is here.

Quantum molecular resonance (Rexon-Eye)

Shemer and colleagues’ single-centre, double-masked trial randomised 40 adults with dry eye to four weekly Rexon-Eye sessions or identical zero-power sessions — imperceptible to patients, with examiners also masked — on top of usual drops, an adjunct design (Cornea 2024). The primary outcome, OSDI change, favoured treatment: −8.6 ± 7.8 versus +0.5 ± 9.4 (P = 0.002); MGD score and staining also differed (P = 0.007 and 0.044); TBUT (P = 0.097) and Schirmer did not; no adverse events occurred. Outcomes were measured only immediately after the last session; no confidence intervals are reported; the treated arm began with worse symptoms; and 20 per arm was powered only for a 5-point OSDI shift. A symptom gain without a tear-film change leaves open whether the glands did anything. Kavroulaki and colleagues’ 51-patient single-arm series in mixed-aetiology dry eye (Cureus 2023) warrants scepticism: only 32 patients (63%) returned at two months; NIBUT moved from 7.50 to 7.59; tear meniscus height rose then fell below baseline; the scale labelled OSDI appears to be a 0–28 SPEED-type score; and there was no comparator. It establishes tolerability (no adverse events in 102 eyes), not efficacy. Patient-facing coverage is here.

Interpretation: the governing principle is measurement

First, the evidence gradient is real: LipiFlow rests on pooled randomised data, IPL on short sham-controlled trials, Tixel on an uncontrolled pilot and a responder-only extension of a pivotal trial, and Rexon-Eye on one 40-patient trial plus an uncontrolled series. Second, symptoms and signs diverge: against active comparators, IPL and LipiFlow improved gland function more convincingly than symptoms; the Rexon-Eye trial showed the reverse. Each device treats a mechanism; the trials succeeded in measured populations, and without measurement any may underperform.

Device choice should therefore follow a measured phenotype. Thermal and light-based approaches have their evidence in evaporative disease with recoverable glands; electrotherapy’s one controlled trial pooled subtypes, so its use outside MGD rests on unsubdivided or uncontrolled data. This measure-first logic governs the dry eye clinic at Rose Optometry and the network practices of the Dry Eye Specialist Group.

Limitations of this evidence

• Follow-up is short. Four weeks to three months in every controlled study; Blackie’s 12-month and Sadri’s six-month data are uncontrolled.
• Masking and comparators. Blackie was open-label; the Tixel pilot, Sadri’s extension and Kavroulaki had no comparator; only the IPL trials and Shemer were sham-controlled.
• Selected samples. Sadri enrolled responders only; Safir recruited cosmetic-clinic patients, losing 20%; Blackie’s 12-month figures are completers; Kavroulaki lost 37%.
• Sponsor involvement. Manufacturer employees authored Blackie and Sadri (Novoxel-funded); the Toyos sponsor ran post-hoc analyses; D’Souza was manufacturer-funded. Hu, Safir, Shemer and Kavroulaki appear independent.
• Proxies. Gland scores and TBUT are surrogates; symptom gains without sign changes may not mean durable benefit.

Jagrut Lallu — BOptom (Hons), MSc Specialty Lenses (Hons), FIAOMC. Therapeutic Optometrist & Contact Lens Specialist. Founder, New Zealand Eye Research Centre; Partner, Rose Optometry; WCO Asia-Pacific Myopia Ambassador; International Myopia Institute NZ Ambassador; Clinical Senior Lecturer, Deakin School of Optometry; Honorary Teaching Fellow, University of Auckland.

References

Sourced and verified via PubMed.

• Hu J, Zhu S, Liu X. Efficacy and safety of a vectored thermal pulsation system (LipiFlow®) in the treatment of meibomian gland dysfunction: a systematic review and meta-analysis. Graefes Arch Clin Exp Ophthalmol. 2022;260(1):25-39. PMID 34374808. https://doi.org/10.1007/s00417-021-05363-1
• Blackie CA, Coleman CA, Holland EJ. The sustained effect (12 months) of a single-dose vectored thermal pulsation procedure for meibomian gland dysfunction and evaporative dry eye. Clin Ophthalmol. 2016;10:1385-96. PMID 27555745. https://doi.org/10.2147/OPTH.S109663
• Toyos R, Desai NR, Toyos M, Dell SJ. Intense pulsed light improves signs and symptoms of dry eye disease due to meibomian gland dysfunction: a randomized controlled study. PLoS One. 2022;17(6):e0270268. PMID 35737696. https://doi.org/10.1371/journal.pone.0270268
• D’Souza S, James E, Koul A, et al. A randomized controlled study evaluating outcomes of intense pulsed light and low-level light therapy for treating meibomian gland dysfunction and evaporative dry eye. Indian J Ophthalmol. 2023;71(4):1608-1612. PMID 37026310. https://doi.org/10.4103/IJO.IJO_2834_22
• Safir M, Hecht I, Ahimor A, et al. The effect of thermo-mechanical device (Tixel) treatment on evaporative dry eye disease: a pilot prospective clinical trial. Cont Lens Anterior Eye. 2022;45(6):101741. PMID 35864019. https://doi.org/10.1016/j.clae.2022.101741
• Sadri E, Verachtert A, Parkhurst GD, et al. Six-month extension of a randomised, masked pivotal study of thermomechanical (Tixel) treatment for dry eye disease. J Ocul Pharmacol Ther. 2025;41(5):237-243. PMID 40238710. https://doi.org/10.1089/jop.2025.0033
• Shemer A, Altarescu A, Nusbaum L, et al. Quantum molecular resonance electrotherapy (Rexon-Eye) for dry eye disease: a double-blind randomised placebo-controlled trial. Cornea. 2024;43(9):1144-1149 (online December 2023). PMID 38166187. https://doi.org/10.1097/ICO.0000000000003443
• Kavroulaki D, Konstantinidou E, Tsiogka A, et al. Quantum molecular resonance electrical stimulation as a beneficial and safe treatment for multifactorial dry eye disease. Cureus. 2023;15(5):e39695. PMID 37398748. https://doi.org/10.7759/cureus.39695