Contact Lens and Anterior Eye xxx (xxxx) xxx

Contents lists available at ScienceDirect

Contact Lens and Anterior Eye

journal homepage: www.elsevier.com/locate/clae

Interventions for Demodex blepharitis and their effectiveness: A systematic

review and meta-analysis
Dayron F. Martínez-Pulgarín a, b, *, Marcel Y. Ávila a, b, Alfonso J. Rodríguez-Morales c
a
Department of Ophthalmology, School of Medicine, Universidad Nacional de Colombia, Bogotá, Colombia
b
Basic and Clinic Ophthalmology Research Group, School of Medicine, Universidad Nacional de Colombia, Bogotá, Colombia
c
Grupo de Investigación Biomedicina, Faculty of Medicine, Fundacion Universitaria Autonoma de las Americas, Pereira, Risaralda, Colombia

A R T I C L E I N F O A B S T R A C T

Keywords: Purpose: A systematic review and meta-analysis was performed to evaluate the effectiveness of interventions in
Blepharitis the treatment ofDemodex blepharitis in adult patients.
Demodex Methods: A systematic review and meta-analysis of studies reporting the efficacy of treatments forDemodex
Treatment
blepharitis in the main databases (PubMed / Scopus / Cochrane / EMBASE / Science Direct / WOS / Scielo /
Inflammation
Google Scholar / metaRegister of Controlled Trials / ClinicalTrials.gov/ WHO ICTRP) until November 24, 2020
was performed according to the PRISMA statement for meta-analysis.
Results: Overall, 18 studies were included for 29 different interventions in 1195 participants with 1574 eyes that
were positive for Demodex Spp. Demodex counts, total eradication, clinical improvement, Ocular Surface Disease
Index, Tear Break-Up Time, cylindrical dandruff, Schirmer test, osmolarity and adverse reactions were analysed,
and stratified sub-analyses conducted. The overall effects for Demodex count (mean difference), total eradication
(risk ratio) and adverse reactions (risk difference) were -2.07 (95 % CI -3.99 to -0.15) p = 0.03, 1.84 (95 % CI
1.27–2.66) p = 0.001 and 0.24 (95 % CI 0.08 to 0.41) p = 0.005, respectively. The most frequent interventions
evaluated in the included studies were tea tree oil (TTO) and its derivatives, such as terpinen 4-ol.
Conclusion: Multiple therapeutic choices were evaluated in this meta-analysis. Pharmacological interventions
were superior to non-pharmacological (mechanical, thermal and pulsed light) interventions. It was not possible
to establish significant differences between TTO and non-TTO-derived treatments. Adverse reactions were more
frequent in TTO-derived treatments, however all were mild. It is necessary to execute studies with longer follow-
up times to determine whether re-infestation occurs after the administration of different treatments.

1. Introduction age, rosacea, alcohol intake, sun exposure, smoking, stress, local or
systemic immunosuppression, and poor hygiene [6]. Pathogenesis of
Blepharitis is a common entity in which a multifactorial inflamma­ Demodex blepharitis can be described through three mechanisms: direct
tory process affects the eyelid, where meibomian gland dysfunction damage (micro-abrasions caused by the mite’s claws, meibomian gland
(MGD) is almost always present, leading to symptoms including burning orifice block and consumption of epithelial cells), a vector for bacteria
and foreign body sensation, tearing, eyelid and conjunctival redness, (Demodex can carry bacteria like Streptococci and Staphylococci on its
itching, as well as many other inflammatory symptoms in the ocular surface and Bacillus Oleronius inside their abdomen, producing super­
surface [1]. Blepharitis and MGD can be present in 27.5 %–68.3 % of dry antigens and inducing immune responses) and hypersensitivity reaction
eye disease cases [2]. One of the etiologies of blepharitis is the infesta­ (the debris or waste of Demodex can induce inflammatory responses in
tion by Demodex mites, either the Folliculorum or Brevis species [3], the host via a delayed hypersensitivity or an innate immune response)
reaching a prevalence of 84 % in the population older than 60 years and [7].
even 100 % in people older than 70 years [4]. In patients with ble­ The annual treatment costs for patients with dry eye disease are USD
pharitis, 41.6–81.25 % are positive for Demodex infestation [5]. 3.84 billion in the US [8] and USD 0.27–1.1 million/1000 patients in
The main risk factors associated with Demodex infestation include Europe [9]. In Asia, estimated annual work productivity losses of USD

* Corresponding author at: Department of Ophthalmology, Universidad Nacional de Colombia, Bogota, 11001, Colombia.
E-mail address: dfmartinezpu@unal.edu.co (D.F. Martínez-Pulgarín).

https://doi.org/10.1016/j.clae.2021.101453
Received 26 September 2020; Received in revised form 23 April 2021; Accepted 28 April 2021
1367-0484/© 2021 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.

Please cite this article as: Dayron F. Martínez-Pulgarín, Contact Lens and Anterior Eye, https://doi.org/10.1016/j.clae.2021.101453
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

6160 per employee have been reported in Japan [10] and in Singapore, metaRegister of Controlled Trials (www.controlled-trials.com), Clin­
annual expenditure on dry eye treatment was USD 1,520,798 in 2009 [2, icalTrials.gov (www.clinicaltrials.gov) and WHO International Clinical
11]. Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en) using
Extrapolating the prevalence of blepharitis and Demodex in these these search strategies: (Blepharit* OR inflamm*) AND (demodex OR
cases, it is possible to deduce treatment costs of USD 0.43–2.13 billion/ demod* OR "cylindrical dandruff" OR folliculorum OR brevis OR mite*)
year in the United States and Europe. For patients with blepharitis, AND (treat* OR therap* OR drug* OR efficacy OR effect* OR manage­
blepharoconjunctivitis and chronic conjunctivitis, the 5 year costs are ment) AND (eye* OR ocular OR meibom*) for PubMed, Cochrane Li­
estimated to be around USD 1428 ± 1752/patient [12]. In Latin America brary, EMBASE, WOS, Scielo and Google Scholar; (Blepharitis) AND
and Colombia, there are no studies that determine these variables, and (demodex OR folliculorum OR brevis) AND (treatment OR therapy OR
there are no data about specific costs of Demodex blepharitis. efficacy) AND (eye OR ocular) for Science Direct, Scopus, ClinicalTrials.
The diagnosis of Demodex blepharitis depends on a good clinical gov and metaRegister of Controlled Trials; “Demodex blepharitis” for
evaluation. High magnification and direct observation with the slit lamp ICTRP. The assessment of adverse reactions was performed by reviewing
can be helpful in identifying signs suggestive of infestation by Demodex, the articles for efficacy analysis and reporting these data. The search was
such as cylindrical dandruff at the base of the lashes that represents a limited to articles written in English, Spanish and Portuguese. Preferred
pathognomonic sign [6,13–15]. Direct microscopic visualization of the Reporting Items for Systematic reviews and Meta-Analyses (PRISMA)
mite from an eyelash sample obtained by lash epilation can provide a was followed, except for the registration as a review protocol in an in­
definitive diagnosis. in vivo confocal microscopy allows rapid visuali­ ternational register. The study protocol was approved in 2018 by the
zation of the eye surface with almost histological resolution, allowing Institutional Review Board of the Universidad Nacional de Colombia
examination of the follicle and the detection of mites [13,16]. (Facultad de Medicina), Colombia. Microsoft Excel 2013® was used to
Multiple therapeutic choices have emerged in an attempt to eradi­ manage the identified records and eligibility status.
cate Demodex mites, with descriptions of resistance to options like
alcohol 75 %, iodopovidone 10 % and baby shampoo [17]. Other 2.3. Study selection
choices described in literature such as lid hygiene, pilocarpine gel, some
antibiotics, mercury oxide or sulphuric ointment vary in their reported One author executed all literature searches (DM) and collated the
efficacy [18,19]. Recently, other promising alternatives have been abstracts. Two authors (DM and MA) separately reviewed the abstracts
described, including tea tree oil (TTO) and its derivatives, such as ter­ and decided the suitability of the articles for inclusion, based on the
pinen 4-ol, which exhibit broad spectrum antimicrobial activity against selection criteria. If there was any doubt as to the eligibility of a study, it
bacteria and fungi through disruption of the cytoplasmic membrane, was obtained and read in full by a third author (AR). Finally, all authors
although the exact mechanism against Demodex viability has not been reviewed the eligible articles.
found [20].
Non-TTO-derived treatments include permethrin, which shows a 2.4. Risk of bias assessment
disrupting effect on nerve cell membrane polarization [21], ivermectin
(with or without metronidazole), which increases the permeability of Risk of bias was assessed with the Cochrane Risk of Bias tool. Se­
the cell membrane to chloride ions, inducing paralysis and death of the lection bias, performance bias, detection bias, attrition bias, reporting
mite [22–24], microblepharoexfoliation, whose mechanical action re­ bias and other biases were evaluated for each included study.
duces Demodex counts [25] and intense pulsed light (IPL), which de­
creases inflammatory mediators and reactive oxygen species [26]. 2.5. Data collection, synthesis and analysis
Given the high prevalence of the disease, the repercussions for
quality of life [27], as well as the lack of integrative studies analysing Review Manager Software 5.3® was used for the statistical analyses.
therapeutic choices, a systematic review and meta-analysis was con­ Dichotomous outcome data were analysed by calculating risk ratios
ducted to evaluate the effectiveness of treatments for Demodex (RRs) or difference of risks (for adverse reactions only), and continuous
blepharitis. outcome data were analysed by using mean differences (these data have
a negative value when they show a decrease in the evaluated variable).
2. Material and methods Pooled estimates of effects were calculated by using random effects
models. Subgroup analyses were performed in some outcomes to
2.1. Eligibility criteria compare TTO-derived treatments versus others, and pharmacological
interventions versus non-pharmacological interventions (non-pharma­
Randomized and non-randomized controlled studies comparing the cological measures refer to those with a mechanical, thermal or lumi­
efficacy and/or adverse reactions of medication X versus medication Y, nous effect). 95 % confidence intervals (CIs) were presented on forest
or medication X versus placebo/absence of treatment, in eyes of patients plots, and the significance was set at p < 0.05. I-squared (I2) >60 %
older than 16 years, with a diagnosis of Demodex blepharitis were defined heterogeneity between studies.
included. The primary outcome was the mites per eyelash count at the
end of treatment. Secondary outcomes included total eradication rates 3. Results
(total absence of mites at the end of treatment), Tear Break-Up Time
(TBUT), Ocular Surface Disease Index (OSDI), Schirmer test, osmolarity, A total of 3426 references were found after a systematic literature
cylindrical dandruff score, adverse reactions rate and stratified sub- search, and 18 randomized clinical trials, that fitted the inclusion
analyses comparing either TTO-derived treatments versus others or criteria, were included. A flow diagram of the search strategy and se­
pharmacological interventions versus non-pharmacological in­ lection is given in Fig. 1.
terventions, in terms of mite count and total eradication of mites at the The included studies evaluated 1195 participants and 1574 eyes that
end of treatment. were positive for Demodex infestation and differed from each other with
respect to the length of follow-up (from 28 to 90 days) and the type of
2.2. Search methods evaluated interventions. In total, 4 studies compared a therapeutic
intervention versus placebo [28–31], 12 studies assessed an intervention
A systematic literature search was conducted, with a final search on versus other interventions [17,23,25,26,32–39], and 2 studies compared
November 24, 2020, in PubMed, Scopus, Cochrane Library, EMBASE, an intervention versus the absence of treatment [40,41]. Only one study
Science Direct, Web of Science (WOS), Scielo, Google Scholar, was specifically designed to evaluate adverse reactions [38].

2
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

Fig. 1. Flow diagram of search strategy and selection.

TTO-derived treatments were evaluated in 11 studies [17,25,28,29,31,

33,35–38,40], 9 studies evaluated non-TTO-derived treatments [23,25,
26,30,32,34,38,39,41], only 1 study evaluated systemic medications
[23], and 3 studies evaluated non-pharmacological interventions [25,
26,39].

3.1. Quality of studies

Six studies (33.3 %) included in this review were industry funded.

Four (22.2 %) of the included studies included 30 or fewer participants
with Demodex blepharitis. All articles were written in English, except for
one that was in Spanish [32]. In terms of the risk of bias: for selection
bias, a low risk was identified in 30.6 % of articles and an unclear risk in
69.4 % ; for performance bias, a low risk was established in 27.8 % of
articles and a high risk in 55.6 % ; for detection bias, a low risk was Fig. 2. Risk of bias assessment of included studies.
identified in 55.6 % of articles and a high risk in 16.7 % ; for attrition
bias, a low risk was identified in 88.9 % of articles and a high risk in 5.6 interventions with a significant effect on Demodex/eyelash reduction
% ; for reporting bias, an unclear risk was identified in 94.4 % of articles. included artificial tears + topical steroid drops + eye shampoos with 5%
The risk of bias assessment of the included studies is provided in Fig. 2. TTO [37] and eyelid scrubs with 50 % TTO weekly + 10 % TTO daily
[28] (− 1.71 and − 1.00 Demodex/eyelash, respectively) (Fig. 3). Strat­
3.2. Demodex count ified meta-analysis did not show significant subgroup differences be­
tween TTO-derived and non-TTO-derived treatments (p = 0.34),
Eight studies were included for this analysis [23,25,26,28,30,31,35, although the evaluation of this variable exclusively in TTO-derived
37]. The overall change in total Demodex/eyelash count was -2.07 (95 % treatments showed a statistically significant effect of − 0.90 Demo­
CI -3.99 to -0.15) p = 0.03, with the highest effects being observed for dex/eyelash (95 % CI − 1.39 to − 0.40) p = 0.0004 versus a
systemic and topical ivermectin-metronidazole (− 5.10 Demodex/­ non-significant statistical effect in non-TTO-derived treatments of − 2.76
eyelash and -7.50 Demodex/eyelash, respectively) [23,30]. Other Demodex/eyelash (95 % CI − 6.56 to 1.04) p = 0.15. Comparing

3
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

Fig. 3. Forest plot comparing the mites’ count in eyes randomized to intervention or comparative groups. CI¼ Confidence Interval; IV¼ Inverse Variance;
SD¼Standard deviation; IVM¼ Ivermectin; MTZ¼ Metronidazole; T4O¼ Terpinen 4-ol; MBE¼ Microblepharoexfoliation; TTO¼ Tea tree oil; IPL¼ Intense
pulsed light.

pharmacological versus non-pharmacological interventions, there was a 3.5. Cylindrical dandruff

significant difference between subgroups (p = 0.04), favouring phar­
macological interventions (-2.88 Demodex/eyelash (95 % CI -5.74 to Three studies were included [32,35,36], where cylindrical dandruff
-0.02) (Fig. 4). was reported according to a 0–3 score (3 being a high presence of the
clinical finding). An overall decrease with statistical significance was
3.3. Total eradication of mites found: − 0.76 (95 % CI − 1.09 to − 0.44) p ≤ 0.00001. Stratified
meta-analysis did not show significant subgroup differences between
Thirteen studies were included in this analysis [17,23,25,26,28,30, TTO-derived and non-TTO derived treatments (p = 0.94).
33–36,39–41]. An overall RR of 1.84 (95 % CI 1.27–2.66) p = 0.001 was
found, favouring total eradication in the intervention groups. The
3.6. Adverse reactions
highest significant RRs were for 50 % TTO lid scrubs, manuka honey and
topical ivermectin + metronidazole [17,30,41] (Fig. 5). Stratified
To assess the adverse drug reactions, nine studies were included
meta-analysis did not show significant subgroup differences between
[28–30,33,35,36,38,40,41]. The overall effect according to risk differ­
TTO-derived and non-TTO-derived treatments (p = 0.15). However, a
ences was 0.24 (95 % CI 0.08 to 0.41) p = 0.005, with a tendency for
separate evaluation showed a non-statistically significant RR of 1.35 (95
intervention groups to present adverse reactions. The biggest individual
% CI 0.84 to − 2.17) p = 0.21 in TTO-derived treatments, versus a sta­
effects favouring the presentation of adverse reactions were found by
tistically significant RR in non-TTO-derived treatments of 1.68 (95 % CI
Koo et al., evaluating eyelid scrubs with 50 % TTO weekly + 10 % TTO
1.17–2.40) p = 0.007. Comparing pharmacological versus
daily (0.05 (95 % CI 0.02 to 0.08)), and Ngo et al., in the groups eval­
non-pharmacological interventions, there were significant differences
uating TTO and terpinen 4-ol (1.00 (95 % CI 0.93–1.07)) [28,38]. All
that favoured pharmacological interventions (p = 0.04) (Fig. 6).
reported data included mild adverse reactions only, such as discomfort,
ocular irritation, burning sensation and transient ocular stinging. These
3.4. OSDI, TBUT, Schirmer test, osmolarity and clinical improvement
were reported in 2.17 %–100 % of cases, in intervention groups using
the following treatments: Manuka honey, terpinen 4-ol and tea tree oil.
Twelve studies were included in these analyses [25,26,28,31–34,36,
No major adverse events that threatened life or visual function were
37,39–41], but none showed statistically significant results. For clinical
reported in any study.
improvement, the overall RR was 0.96 (CI 95 % 0.90–1.04) p = 0.32, I2
= 0%.
4. Discussion

Blepharitis is a frequent pathology within ophthalmologic and

optometric consultations worldwide [2,42], with Demodex Spp. being

Fig. 4. Forest plot showing the comparison on mites’ count on eyes undergoing pharmacological versus non-pharmacological treatments. CI¼ Confidence Interval;
IV¼ Inverse Variance; SD¼Standard deviation; IVM¼ Ivermectin; MTZ¼ Metronidazole; T4O¼ Terpinen 4-ol; MBE¼ Microblepharoexfoliation; TTO¼ Tea tree oil;
IPL¼ Intense pulsed light.

4
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

Fig. 5. Forest plot demostrating the RR of total eradication of mites in eyes randomized to intervention or comparative groups. CI¼ Confidence Interval; M-H¼
Mantel-Haenszel; RR¼Risk Ratio; SD¼Standard deviation; IVM¼ Ivermectin; MTZ¼ Metronidazole; T4O¼ Terpinen 4-ol; HA¼ Hyaluronic acid; TTO¼ Tea tree oil;
IPL¼ Intense pulsed light.

Fig. 6. Forest plot demostrating the RR of total eradication on eyes undergoing pharmacological versus non-pharmacological treatments. CI¼ Confidence Interval;
M-H¼ Mantel-Haenszel; RR¼Risk Ratio; SD¼Standard deviation; IVM¼ Ivermectin; MTZ¼ Metronidazole; T4O¼ Terpinen 4-ol; HA¼ Hyaluronic acid; TTO¼ Tea
tree oil; IPL¼ Intense pulsed light.

responsible in some cases. Dozens of different treatments have been favouring the interventions, except for symptomatic improvement. Due
explored with variable responses [6,18,19,22]. to the different methods used to calculate the global effects sizes be­
Navel et al. [43] performed the first meta-analysis reported in the tween this study and Navel et al., it is not possible to make a completely
world literature about the effectiveness of these treatments, which objective comparison of the results.
included both randomized controlled clinical trials and observational
studies. In order to achieve a more objective assessment of the evaluated 4.1. Treatment considerations for Demodex blepharitis
interventions, the present study limited the included articles exclusively
to randomized and non-randomized clinical trials. Navel et al. [43] Systemic and topical ivermectin + metronidazole [23,30] were
calculated the global effect sizes using the Hedge method (> 0.2 small shown to be the most effective interventions in reducing mite counts.
effect, > 0.5 moderate effect and > 0.8 large effect) and found results of This finding is explained by the excellent acaricidal effect of ivermectin,
1.68 (95 % CI 1.25–2.12) for mite counts, 0.45 (0.26 to 0.64) for erad­ as described by Holzchuh et al. [44] and the broad anti-inflammatory
ication rates and 0.76 (0.59 to 0.90) for symptomatic improvement. effect provided by metronidazole, through the T-lymphocytes and
Comparatively, the present study showed overall effects for the same neutrophil mediated reduction of reactive oxygen species [45].
variables (calculated as RR and mean differences through random ef­ Although adverse reactions were not reported with these two treat­
fects models) of -2.07 (95 % CI − 3.99 to − 0.15), 1.84 (95 % CI ments, caution is recommended due to possible hypersensitivity re­
1.27–2.66) and 0.96 (95 % CI 0.90–1.04), with statistical significance actions and drug interactions, especially in systemic treatments.

5
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

Stratified meta-analysis did not show significant subgroup differ­ neomycin-dexamethasone, Manuka honey and micro­
ences between TTO-derived and non-TTO derived treatments in terms of blepharoexfoliation devices, should be clarified. The first example is a
mite counts, total eradication and cylindrical dandruff. When total diol, used mainly in the treatment of pediculosis, whose mechanism of
eradication was evaluated separately in non-TTO treatments however, action is based on the disruption of the parasite’s lipid cover, allowing a
the overall RR was statistically significant (RR of 2.42 (95 % CI dehydration effect, which has been extrapolated to Demodex Spp. The
1.27–4.59) p = 0.007). This can be explained by the better tolerability study by Murphy et al. [25] is the only included study that evaluates this
and fewer adverse reactions of these treatments [17,28,38,46]. Strati­ treatment, reaching total eradication rate of 36.7 % at the end of
fied analysis comparing pharmacological versus non-pharmacological follow-up versus a total eradication rate of 32 % for 38 % TTO. Hygiene
measures found a statistically significant subgroup difference in terms with neutral shampoo, with or without metronidazole or
of the reduction of mite counts and total eradication (p = 0.04), polymyxin-neomycin-dexamethasone, is another therapeutic option
favouring pharmacological interventions for total eradication (p = studied for Demodex blepharitis. Worldwide, this is one of the first-line
0.003) and leading to a hypothesis of greater mite susceptibility to these therapies for blepharitis, independent of etiology and gives some de­
measures. Currently, there are no published studies that have evaluated gree of symptomatic improvement in 100 % of patients [1,32]. The
comparisons of this type. analyses conducted during the present work showed that its use as a
In last few years, TTO and its derivatives (especially terpinen 4-ol) monotherapy or combined with metronidazole/polymyxin
have become more visible due to their excellent acaricidal effects [17, -neomycin-dexamethasone has led to a global reduction in cylindrical
28,47,48]. According to the studies included in this meta-analysis, these dandruff score of -0.80 (95 % CI V1.28 to − 0.32). Micro­
preparations attained total eradication rates of between 23.6 % and 77.8 blepharoexfoliation devices allow an aggressive initial cleaning effect,
% at different concentrations. However, other, less well recognized as well as a reduction of the bacterial load on the eyelid edge. Murphy
therapeutic options, such as IPL and systemic or topical ivermectin + et al. [25] showed that the total eradication rates of micro­
metronidazole, can achieve total eradication rates of up to 100 % [23, blepharoexfoliation versus 38 % TTO and 1, 2 octanediol were very
26,30], perhaps with better tolerability and fewer adverse reactions similar (32.1 % versus 36.7 % versus 32.1 %). Manuka honey, evaluated
than have been reported for TTO and its derivatives [17,28,38,46]. by Craig et al. [41], has a high concentration of methylglyoxal; com­
Despite this, it is currently unclear whether a total eradication of mites is plexed with α-cyclodextrin, it has demonstrated antibacterial and anti­
required for clinical improvement, or whether it is only necessary to demodectic efficacy with total eradication rates of up to 60 %.
reduce the number of mites, given that Demodex originated and diver­
sified with early humans and are ubiquitous inhabitants of human skin 4.3. In-office versus at home treatment
[49,50].
There was an additional probability for adverse reaction presenta­ An issue that has not been studied so far is the recommendation
tion of 24 % in the intervention groups (p = 0.005). Both Koo et al. and about in-office versus at home treatment. Some of the included studies
Ngo et al. [28,38] found a significantly greater probability of adverse administered the therapy in-office, at home or both. However, there is
reaction in interventions derived from TTO, similar to that described in no exact recommendation in the current literature on when to use one or
the literature [17,28,38,46]. Ngo et al. affirm that compounds based on the other. It is possible to make a recommendation for the use of in-office
TTO, specifically Cliradex®, OUST Demodex Swabstix® and TheraLid®, protocols in two cases, based only on clinical experience: (1) when
are associated with sudden or progressive irritation and discomfort after mechanical therapies that require medical supervision and special
application to the eyelids, with a delay in comfortable eye opening of 60, equipment, such as microbleharoexfoliation, are used; (2) when the use
600 and 38 s, respectively [38]. The studies of Karakurt et al. and Ergun of pharmacological treatments that have a high concentration is pro­
et al. evaluated the use of lower concentrations of TTO (7.5 % and 3% posed (i.e., 50 % TTO), making it difficult for the patient to ensure good
respectively), which allowed a sufficient acaricidal effect but with adherence due to the high probability of an adverse reaction leading to
reduced adverse effects and improved tolerability [33,35]. the interruption of treatment.
All adverse reactions presented in the studies include for the present
analysis were classified as mild, and no major events were described. 4.4. Study strengths and limitations
The relationship of these reactions and compliance with the different
treatments requires further study. The strengths of this work include the inclusion of controlled trials
only and the implementation of stratified subanalyses and analysis of
4.2. Therapeutic alternatives adverse reactions, neither of which have previously been described in
the literature.
The existence of therapeutic choices different to TTO-derived treat­ However, there are limitations that must be mentioned, such as the
ments and systemic or topical ivermectin/metronidazole must be general lack of standardization and comparability among existing
considered, since it represents a promising area in the research of studies, which limits the performance of all analysis. While heteroge­
Demodex blepharitis. neity between studies could alter the interpretability of the results, this
In the case of IPL, which was evaluated by Zhang et al. [26] although possibility was considered from the conception of this work and all the
not statistically significant, improvements in TBUT (3.15 s, 95 % CI analyses were therefore made under random effects models. Studies
2.24–4.06) and OSDI (− 13.06, 95 % CI − 25.95 to − 0.17) were found such as Wong et al., [40] had to be excluded from some analyses because
compared to other interventions included in the analysis. Zhang et al. they did not report standard deviations in their measurement of vari­
did not find significant effects in terms of mite counts in their study but ables but used interquartile ranges, preventing the normality of the data
the close relationship of demodicosis and MGD with the lid inflamma­ being defined. Other limitations include the inability to perform strati­
tory process has led to IPL being proposed as an adjunctive treatment in fied sub-analysis of local versus systemic treatments, because only one
demodicosis. The proposed beneficial effect of pulsed light on the MGD relevant article was found. There were very few studies with adequate
that frequently accompanies demodicosis is through heat transfer, methodologies oriented exclusively to evaluating adverse reactions,
softening the meibum, decreasing the inflammatory mediators due to which made it necessary to extract information from the included
vasculature thrombosis around the meibomian glands, decreasing the effectiveness studies, where this was available. Another limitation re­
bacterial load of the eyelid margin and decreasing the reactive oxygen lates to the need to obtain information from studies that were catalogued
species [51,52]. as completed in clinical trial platforms, but whose complete data were
The applicability of other options, such as 1, 2 octanediol, hygiene not always available. Finally, one of the most important limitations is
with neutral shampoo with/without metronidazole or polymyxin- the total absence of information in the current literature about the rates

6
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

of reinfestation, an issue that is crucial to understanding the effective­ [15] Gao Y-Y, Di Pascuale M, Li W, Liu DT, Baradaran-Rafii A, Elizondo A, et al. High
prevalence of Demodex in eyelashes with cylindrical dandruff. Invest Ophthalmol
ness of a specific therapeutic measure. The available studies only eval­
Vis Sci 2005;46(9):3089–94.
uated their treatments with a follow-up between 28–90 days, after [16] Randon M, Liang H, El Hamdaoui M, Tahiri R, Batellier L, Denoyer A, et al. In vivo
which there were no data to establish reinfestation rates. It is essential confocal microscopy as a novel and reliable tool for the diagnosis of Demodex
for future studies to employ much longer follow-up times. eyelid infestation. Br J Ophthalmol 2015;99(3):336–41.
[17] Gao Y-Y, Di Pascuale M, Li W, Baradaran-Rafii A, Elizondo A, Kuo CL, et al. In vitro
and in vivo killing of ocular Demodex by tea tree oil. Br J Ophthalmol 2005;89(11):
4.5. Conclusions 1468–73.
[18] Fulk G, Clifford C. A case report of demodicosis. J Am Optom Assoc 1990;61(8):
637–9.
The effectiveness of topical and systemic ivermectin + metronida­ [19] Fulk G, Murphy B, Robins M. Pilocarpine gel for the treatment of demodicosis-a
zole and 50 % TTO weekly + 10 % TTO daily for Demodex count and case series. Optom Vis Sci 1996;73(12):742–5.
total eradication, 5% TTO + 0.1 % betamethasone + hydrox­ [20] Cheung IMY, Xue AL, Kim A, Ammundsen K, Wang MTM, Craig JP. In vitro anti-
demodectic effects and terpinen-4-ol content of commercial eyelid cleansers. Cont
ypropylmethylcellulose for Demodex count, and IPL for total eradication Lens Anterior Eye 2018;41(6):513–7.
is clear. The superiority of systemic and topical ivermectin + metroni­ [21] Hecht I, Melzer-Golik A, Sadi Szyper N, Kaiserman I. Permethrin cream for the
dazole therapy, over any other therapeutic measure, in reducing treatment of demodex blepharitis. Cornea 2019;38(12):1513–8.
[22] Jones L, Downie L, Korb D, et al. TFOS DEWS II management and therapy report.
Demodex count must be highlighted, which makes them interesting Ocul Surf 2017;15(3):575–628.
molecules for further research. Pharmacological interventions are su­ [23] Salem D, El-Shazly A, Nabih N, Benitez-Del-Castillo JM, Dana R, Deng SX, et al.
perior to non-pharmacological interventions in terms of total eradica­ Evaluation of the efficacy of oral ivermectin in comparison with ivermectin-
metronidazole combined therapy in the treatment of ocular and skin lesions of
tion rate. Adverse events are more likely to occur in the intervention Demodex folliculorum. Int J Infect Dis 2013;17(5):e343–7.
groups, especially in TTO-derived treatments. It is recommended that [24] Kane NS, Hirschberg B, Qian S, Hunt D, Thomas B, Brochu R, et al. Drug-resistant
comparable randomized, masked, standardized, controlled studies are Drosophila indicate glutamate-gated chloride channels are targets for the
antiparasitics nodulisporic acid and ivermectin. Proc Natl Acad Sci U S A. 2000;97
implemented to strengthen future integrative studies. It is necessary to
(25):13949–54.
execute long follow-up, prospective studies to determine reinfestation [25] Murphy O, O’Dwyer V, Lloyd-McKernan A. The efficacy of tea tree face wash, 1, 2-
rates for different therapeutic interventions, allowing profiling of long- Octanediol and microblepharoexfoliation in treating Demodex folliculorum
blepharitis. Cont Lens Anterior Eye 2017;41(1):77–82.
term effects and refining therapeutic schemes in treating pathology of
[26] Zhang X, Song N, Gong L. Therapeutic effect of intense pulsed light on ocular
high frequency, such as Demodex blepharitis. Studies focusing on demodicosis. Curr Eye Res 2019;44(3):250–6.
adverse drug reactions should also be carried out, in order to define [27] Deeks E. Ivermectin: a review in rosacea. Am J Clin Dermatol 2015;16(5):447–52.
tolerability and adherence rates. [28] Koo H, Kim TH, Kim KW, Wee SW, Chun YS, Kim JC. Ocular surface discomfort and
Demodex: effect of tea tree oil eyelid scrub in Demodex blepharitis. J Korean Med
Sci 2012;27(12):1574–9.
Funding [29] Tseng S. Demodex Blepharitis Treatment Study (DBTS), NCT 01647217.
ClinicalTrials.gov, Accessed July 2019. 2017. https://clinicaltrials.gov/ct2/sho
w/study/NCT01647217.
None. [30] Ávila MY, Martínez-Pulgarín DF, Rizo C. Topical ivermectin-metronidazole gel
therapy in the treatment of blepharitis caused by Demodex spp.: a randomized
Declaration of Competing Interest clinical trial. Cont Lens Anterior Eye 2020;(May). https://doi.org/10.1016/j.
clae.2020.04.011. S1367-0484(20)30084-30089.
[31] Epstein IJ, Rosenberg E, Stuber R, Choi MB, Donnenfeld ED, Perry HD. Double-
The authors do not have any financial interest in any of the products masked and unmasked prospective study of Terpinen-4-ol lid scrubs with
mentioned in this study. Two of the authors (DM and MA) conducted one microblepharoexfoliation for the treatment of demodex blepharitis. Cornea 2020;
39(4).
of the studies included in the meta-analysis. [32] Arrua M, Samudio M, Fariña N, Cibils D, Laspina F, Sanabria R, et al. Comparative
study of the efficacy of different treatment options in patients with chronic
References blepharitis. Arch Soc Esp Oftalmol 2015;90(3):112–8.
[33] Ergun S, Saribas G, Yarayici S, Elmazoglu Z, Cardak A, Ozogul C, et al. Comparison
of efficacy and safety of two tea tree oil based formulations in patients with chronic
[1] Lindsley K, Matsumura S, Hatef E, Akpek EK. Interventions for chronic blepharitis.
blepharitis: a double-blinded randomized clinical trial. Ocul Immunol Inflamm
Cochrane Database Syst Rev 2012;(5). CD005556.
2019;0:1–10.
[2] Stapleton F, Alves M, Bunya VY, Jalbert I, Lekhanont K, Malet F, et al. TFOS DEWS
[34] Inceboz T, Yaman A, Over L, Ozturk AT, Akisu C. Diagnosis and treatment of
II epidemiology report. Ocul Surf 2017;15(3):334–65.
demodectic blepharitis. Turkiye Parazitol Derg 2009;33(1):32–6.
[3] Rivera N, Molina P, Torres A. Determinación de índice de infestación por Demodex
[35] Karakurt Y, Zeytun E. Evaluation of the efficacy of tea tree oil on the density of
spp, en pacientes con blefaritis crónica y en pacientes sin otra patología ocular. Rev
demodex mites (Acari: Demodicidae) and ocular symptoms in patients with
Chilena Infectol 2013;30(5):494–501.
demodectic blepharitis. J Parasitol 2018;104(5):473–8.
[4] Post CF, Juhlin E. Demodex folliculorum and blepharitis. Arch Dermatol 1963;88
[36] Messaoud R, El Fekih L, Mahmoud A, Ben Amor H, Bannour R, Doan S, et al.
(3):298–302.
Improvement in ocular symptoms and signs in patients with Demodex anterior
[5] Kabataş N, Doğan AS, Kabataş EU, Acar M, Bicer T, Gurdal C. The effect of
blepharitis using a novel terpinen-4-ol (2.5%) and hyaluronic acid (0.2%)
demodex infestation on blepharitis and the ocular symptoms. Eye Contact Lens
cleansing wipe. Clin Ophthalmol 2019;13:1043–54.
2017;43(1):64–7.
[37] Mohammadpour M, Maleki S, Khorrami-Nejad M. The effect of tea tree oil on dry
[6] Luo X, Li J, Chen C, Tseng S, Liang L. Ocular demodicosis as a potential cause of
eye treatment after phacoemulsification cataract surgery: a randomized clinical
ocular surface inflammation. Cornea 2017;36:S9–14.
trial. Eur J Ophthalmol 2019:1–6.
[7] Liu J, Sheha H, Tseng SC. Pathogenic role of Demodex mites in blepharitis. Curr
[38] Ngo W, Jones L, Bitton E. Short-term comfort responses associated with the use of
Opin Allergy Clin Immunol 2010;10(5):505–10.
eyelid cleansing products to manage demodex folliculorum. Eye Contact Lens
[8] Yu J, Asche CV, Fairchild CJ. The economic burden of dry eye disease in the United
2018;44:S87–92.
States: a decision tree analysis. Cornea 2011;30(4):379–87.
[39] Murphy O, O’ Dwyer V, Lloyd-Mckernan A. The efficacy of warm compresses in the
[9] Clegg JP, Guest JF, Lehman A, Smith AF. The annual cost of dry eye syndrome in
treatment of meibomian gland dysfunction and demodex folliculorum blepharitis.
France, Germany, Italy, Spain, Sweden and the United Kingdom among patients
Curr Eye Res 2020;45(5):563–75.
managed by ophthalmologists. Ophthalmic Epidemiol 2006;13(4):263–74.
[40] Wong K, Flanagan J, Jalbert I, Tan J. The effect of Blephadex Eyelid Wipes on
[10] Uchino M, Uchino Y, Dogru M, Kawashima M, Yokoi N, Komuro A, et al. Dry eye
Demodex mites, ocular microbiota, bacterial lipase and comfort: a pilot study. Cont
disease and work productivity loss in visual display users: the Osaka study. Am J
Lens Anterior Eye 2019;42(6):652–7.
Ophthalmol 2014;157(2):294–300.
[41] Craig JP, Cruzat A, Cheung IMY, Watters GA, Wang MTM. Randomized masked
[11] Waduthantri S, Yong SS, Tan CH, Shen L, Lee MX, Nagarajan S, et al. Cost of dry
trial of the clinical efficacy of MGO Manuka Honey microemulsion eye cream for
eye treatment in an Asian clinic setting. PLoS One 2012;7(6):e37711.
the treatment of blepharitis. Ocul Surf 2020;18(1):170–7.
[12] Amescua G, Akpek EK, Farid M, Garcia-Ferrer FJ, Lin A, Rhee MK, et al. American
[42] Galvis-Ramírez V, Tello-Hernandez A, Álvarez-Osorio L, Rey-Serrano JJ. The
Academy of Ophthalmology preferred practice pattern cornea and external disease
prevalence of Demodex folliculorum infection in patients attending a general
panel. Blepharitis preferred practice pattern®. Ophthalmology 2019;126(1):
ophthalmological consultation. Rev Salud Pública 2011;13(5):990–7.
P56–93.
[43] Navel V, Mulliez A, Benoist d’Azy C, Baker JS, Malecaze J, Chiambaretta F, et al.
[13] Cheng A, Sheha H, Tseng S. Recent advances on ocular Demodex infestation. Curr
Efficacy of treatments for Demodex blepharitis: a systematic review and meta-
Opin Ophthalmol 2015;26(4):295–300.
analysis. Ocul Surf 2019;17(4):655–69.
[14] Fromstein S, Harthan J, Patel J, Opitz DL. Demodex blepharitis: clinical
perspectives. Clin Optom 2018;10:57–63.

7
D.F. Martínez-Pulgarín et al. Contact Lens and Anterior Eye xxx (xxxx) xxx

[44] Holzchuh F, Hida R, Moscovici B, Villa Albers MB, Santo RM, Kara-José N, et al. associations between host ancestry and mite lineages. Proc Natl Acad Sci U S A
Clinical treatment of ocular Demodex folliculorum by systemic ivermectin. Am J 2015;112(52):15958–63.
Ophthalmol 2011;151(6):1030–4. [50] Thoemmes MS, Fergus DJ, Urban J, Trautwein M, Dunn RR. Ubiquity and diversity
[45] Persi A, Rebora A. Metronidazole in the treatment of rosacea. Arch Dermatol 1985; of human-associated Demodex mites. PLoS One 2014;9(8):e106265.
121:307–8. [51] Craig J, Chen Y, Turnbull P. Prospective trial of intense pulsed light for the
[46] Gao Y-Y, Xu D, Huang L, Wang R, Tseng CG. Treatment of ocular itching associated treatment of meibomian gland dysfunction. Invest Ophthalmol Vis Sci 2015;56(3):
with ocular demodicosis by 5% tea tree oil ointment. Cornea 2012;31(1):14–7. 1965–70.
[47] Gao Y-Y, Di Pascuale M, Elizondo A, Tseng SC. Clinical treatment of ocular [52] Huang X, Qin Q, Wang L, Zheng J, Lin L, Jin X. Clinical results of Intraductal
demodecosis by lid scrub with tea tree oil. Cornea 2007;26(2):136–43. Meibomian gland probing combined with intense pulsed light in treating patients
[48] Tighe S, Gao Y-Y, Tseng S. Terpinen-4-ol is the most active ingredient of tea tree oil with refractory obstructive Meibomian gland dysfunction: a randomized controlled
to kill demodex mites. Transl Vis Sci Technol 2013;2(7):1–8. trial. BMC Ophthalmol 2019;19(1):211.
[49] Palopoli MF, Fergus DJ, Minot S, Pei DT, Simison WB, Fernandez-Silva I, et al.
Global divergence of the human follicle mite Demodex folliculorum: persistent