Meibomian Gland Dysfunction: Recent Progress Worldwide and in Japan

Abstract

In this review, the importance of Japanese research on meibomian gland dysfunction (MGD) is discussed from the perspective of global academic and clinical research on this topic. Many Japanese physicians and researchers have contributed to recent worldwide progress in various fields of MGD research, including pathophysiology, epidemiology, diagnosis, and therapy. In Japan, recent studies in the field of pathophysiology have provided direct evidence for the hypothesis that lipid composition and reactive oxygen species play a crucial role in the development and worsening of MGD. In the field of diagnosis, slit-lamp examination, in vivo confocal microscopy, and meibography have been widely used in studies from Japan. On the basis of the results of these studies, the MGD working group in Japan has proposed new diagnostic criteria for obstructive MGD. According to these criteria, obstructive MGD is considered present when ocular symptoms, anatomic abnormalities, and meibomian gland obstruction are present. In the field of therapy, devices and drugs newly developed in Japan have been shown to enhance the efficacy of lid hygiene and warm compression. Moreover, diquafosol and vitamin D3 have been shown to be effective for MGD. In conclusion, standardization of the diagnosis and treatment of MGD is necessary to enable all patients with MGD to receive appropriate treatment, and specific diagnostic criteria with cutoff values for each parameter are necessary to standardize the diagnosis of MGD.

The term meibomian gland dysfunction (MGD) was coined by Korb and Henriquez in 1980,¹ although investigators were already studying, and continue to study, MGD. In 2011, the International Workshop on MGD compiled a comprehensive overview of various aspects of MGD, including the definition, classification, prevalence, risk factors, structure of meibomian gland lipids, diagnosis, and treatment. The International Workshop on MGD defined MGD as follows: “A chronic, diffuse abnormality of the meibomian glands, commonly characterized by terminal duct obstruction and/or qualitative/quantitative changes in the glandular secretion. It may result in alteration of the tear film, symptoms of eye irritation, clinically apparent inflammation, and ocular surface disease.”² 

Moving away from this definition of MGD, the present review article describes recent progress and future directions in MGD research. 

Many Japanese researchers have contributed to recent progress in various fields of MGD research, including pathophysiology, epidemiology, diagnosis, and therapy. There is no large disagreement between researchers in Japan and those in other countries with regard to the pathophysiology, epidemiology, diagnosis, and therapy of MGD. Rather, the studies conducted by Japanese physicians and researchers have contributed to the core of knowledge concerning MGD. 

Pathophysiology

A recent important finding in the field of MGD-related pathophysiology was the role of ectodysplasin A from the meibomian glands in the regulation of the corneal epithelial barrier function. Sanming Li's³ group showed that ectodysplasin A is highly expressed in meibomian glands and is detected in human tears. They also showed that ectodysplasin A promotes corneal epithelial cell proliferation through regulation of the epithelial growth factor receptor signaling pathway, thereby playing an essential role in the maintenance of corneal epithelial homeostasis. 

Arita et al.⁴ examined the relationship between the composition of free fatty acids (FFAs) in meibum and clinical parameters in patients with MGD and healthy controls and found that unsaturated FFAs tend to be more abundant in colored meibum. In another study, they reported that the relative amount of linoleic acid in meibum is associated with the severity of telangiectasia and plugging of gland orifices in patients with MGD.⁵ In an animal study, Miyake et al.⁶ reported that MGD developed in hairless mice fed a special diet with limited lipid content. In this mouse model, findings characteristic to MGD, such as plugging of the meibomian gland orifices, telangiectasia, and toothpaste-like meibum, were observed (Fig.). Histopathologically, hyperkeratinization of the ductal epithelium in the meibomian gland was observed. Loss of acini and atrophy of the glands were also observed. These findings were similar to those in patients with MGD. In addition, in this model, azithromycin treatment decreased the number of plugged orifices. It was suggested that the deficiency of n-6 polyunsaturated fatty acids, particularly linoleic acid, plays an important role in the development of MGD. These recent studies provided direct evidence for the hypothesis that alterations in the lipid composition play a crucial role in the development and worsening of MGD. However, the factors affecting the composition of lipids in meibum remain unknown. One strong candidate factor is oxidative stress. Ibrahim et al.⁷ investigated alterations in the meibomian glands in Cu, Zu-superoxide dismutase-1 knockout mice and found an accumulation of large lipid droplets and increased terminal deoxynucleotidyl transferase dUTP nick-end labeling and oxidative stress marker staining in the meibomian gland acinar epithelium, suggesting a pivotal role of reactive oxygen species in the pathogenesis of MGD. 

Figure

 

The eyelids of mice fed a normal diet (A, B) and a special diet with limited lipid content (the HR-AD diet) (C, D, E) for 11 weeks, from 5 to 16 weeks of age. Rounding (C), plugging and telangiectasia (D, arrowheads), and toothpaste-like meibum (E, arrowheads) can be seen. This figure was extracted from Miyake H, Oda T, Katsuta O, Seno M, Nakamura M. Meibomian gland dysfunction model in hairless mice fed a special diet with limited lipid content. Invest Ophthalmol Vis Sci. 2016;57:3268–3275, with permission from the Association for Research in Vision and Ophthalmology.

Diagnosis

The diagnosis subcommittee of the International Workshop on MGD recommended the following tests for the diagnosis of MGD-related disease in the general clinical setting: symptom questionnaire, blink rate, tear meniscus height, tear osmolality (if available), tear breakup time, fluorescein staining, Schirmer test, slit-lamp findings (lid changes, meibum expressibility/quality), and meibography.⁸ These tests enable physicians to differentiate dry eye from normal eyes and MGD-related evaporative dry eye from aqueous-deficient dry eye. In that report, however, specific diagnostic criteria for MGD were not proposed, which would be necessary to standardize the diagnosis. 

LipiView

More recently, the LipiView interferometer (TearScience, Inc., Morrisville, NC, USA) was introduced as a new tool for MGD diagnosis. LipiView provides values for the tear film lipid layer thickness (LLT). Finis et al.⁹ reported that the sensitivity and specificity for MGD diagnosis were 65.8% and 63.4%, respectively, when the LLT cutoff value was ≤75 nm, and 47.9% and 90.2%, respectively, when the LLT cutoff value was ≤60 nm. These values were not high enough for clinical use, but they could be enhanced when the LLT value was combined with other parameters such as symptom scores, lid margin abnormality score, and meibography score. Jung et al.¹⁰ investigated the factors related to LLT and found that age, female sex, hypersecretory MGD, and lid margin inflammation are associated with increased LLT. 

Slit-Lamp Examination

Findings around the meibomian gland orifices, such as plugging, telangiectasia, and lid margin irregularity, are crucial for the diagnosis of MGD. However, interrater disagreement on the grading of these findings may exist. Arita et al.¹¹ developed new grading scales for MGD signs that were based on many photos of lids in patients with MGD and healthy controls. Vascularity, plugging of gland orifices, and partial glands were assessed on a scale from 0 to 3, and lid margin irregularity, lid margin thickening, and gland dropout were assessed on a scale from 0 to 2. A validation test was performed in another set of patients with MGD to evaluate the efficacy of the proposed grading scales, and the new grading scales showed appropriate inter- and intrarater reliabilities for grading MGD. This suggested the suitability of the grading scales for the standardization of slit-lamp findings in MGD patients. 

In Vivo Confocal Microscopy

The usefulness of both microscopic (in vivo confocal microscopy) and macroscopic (meibography) methods for MGD diagnosis has been reported in Japan. The details of research using in vivo confocal microscopy in Japan are described in another article in this issue. Briefly, Ibrahim et al.¹² measured four confocal microscopic parameters in patients with MGD and healthy controls: meibomian gland acinar longest diameter, meibomian gland acinar shortest diameter, inflammatory cell density, and meibomian gland acinar unit density. These parameters showed a strong and significant correlation with tear function, ocular surface vital staining, meibomian gland expressibility, and meibomian gland dropout grades. All parameters showed high sensitivity and specificity for MGD diagnosis, suggesting the potential of in vivo confocal microscopy. Moreover, Matsumoto et al.¹³ used in vivo confocal microscopy to evaluate the anti-inflammatory treatment response in patients with MGD. Inflammatory cell density was significantly reduced, suggesting that inflammatory cell density is a new parameter for evaluation of the treatment response. Ibrahim et al.¹⁴ also used in vivo confocal microscopy to evaluate meibomian gland alterations in patients with atopic keratoconjunctivitis, and the averages of the four parameters were significantly worse for patients with atopic keratoconjunctivitis than for patients with obstructive MGD and normal controls. This result shows that changes in meibomian glands are more severe in patients with atopic keratoconjunctivitis than in patients with obstructive MGD. 

Meibography

In addition to the microscopic method (confocal microscopy), a macroscopic method known as noncontact infrared meibography has been developed in Japan for the observation of changes in the meibomian glands. The details of studies performed using noncontact infrared meibography in Japan are described in another article in this issue. Briefly, meibography can be used to obtain information about the morphologic characteristics of meibomian glands. Although meibography was developed three decades ago, it is not widely used because of patient discomfort and a long examination time. In contrast, the newly developed meibography technique is a noncontact and less time-consuming method that uses an infrared filter and infrared charge-coupled device video camera. Using noncontact meibography, Arita et al. revealed meibomian gland loss associated with aging,¹⁵ contact lens wear,¹⁶ and long-term topical antiglaucoma medications.^17,18 

Diagnostic Criteria

Diagnostic criteria are essential for MGD detection. Thus far, various groups worldwide have used different diagnostic criteria, which makes comparisons of MGD studies difficult. Arita et al.¹⁹ compared the scores for clinical findings, including ocular symptoms, lid margin abnormalities, meibography findings, meibum, superficial punctate keratopathy, and tear film production, between patients with obstructive MGD and normal controls to propose diagnostic criteria for obstructive MGD. The area under the receiver operating characteristic curves indicated that the ocular symptom score had the highest diagnostic power, followed by the lid margin abnormality score, meiboscore, and tear breakup time. On the basis of these findings and other results, the Japanese working group on MGD proposed diagnostic criteria for obstructive MGD (Table).^4,20,21 According to these diagnostic criteria, obstructive MGD is considered present when ocular symptoms, anatomic abnormalities, and obstruction of meibomian glands are present. 

Table

 

Diagnostic Criteria for Obstructive Meibomian Gland Dysfunction Proposed by the Japanese MGD Working Group

 

Prevalence

The prevalence of MGD ranges from 3.6% to 68.0%.^22–28 The prevalence of MGD is higher in Asians (46.2%–68.0%)^24–28 than in Caucasians (3.6%–30.5%).^22,23 However, these studies used different diagnostic parameters, namely, telangiectasia and plugging,^22–27 gland dropout,²⁸ and meibum expressibility.²⁸ Therefore, the prevalence of MGD in these studies cannot be compared directly. Amano and Inoue²⁹ investigated the prevalence of MGD in Japanese patients using the same diagnostic criteria used in a previous population-based study conducted in Spain.²³ MGD was diagnosed when one or more of the following signs were present in at least one eye: viscous or waxy white secretion upon digital expression, presence of two or more lid margin telangiectases, and plugging of two or more gland orifices. The estimated prevalence of symptomatic and total (symptomatic + asymptomatic) MGD was 11.2% and 74.5%, respectively, in Japanese patients and 8.6% and 30.5%, respectively, in Spanish patients; this suggested that the prevalence of MGD is higher in Asians than in Caucasians. 

Using the diagnostic criteria proposed by the Japanese working group on MGD, Amano and Inoue²¹ estimated the prevalence of MGD in the Japanese population. The study included 510 participants. The prevalence of symptomatic MGD and total (symptomatic + asymptomatic) MGD was 18.0% and 47.5%, respectively. The prevalence of total MGD increased significantly with increasing patient age (P < 0.0001). Slit-lamp findings (lid margin abnormalities, shift of the mucocutaneous junction, telangiectasia, and plugging) were more frequent, the tear film breakup time was shorter, and the meiboscore was higher in the total MGD group than in participants without MGD. 

Therapy

The management and treatment subcommittee of the International Workshop on MGD proposed a treatment algorithm in which treatment is added depending on the severity of MGD.³⁰ The sequence of treatment addition is eyelid hygiene, eyelid warming and massage, artificial lubricants, topical azithromycin, topical emollient lubricant, oral tetracycline derivatives, lubricant ointment, and anti-inflammatory therapy. 

Lid Hygiene and Warm Compression

Eyelid warming and massage have been the basic treatments for MGD. Recently, a thermopulsation treatment system (Lipiflow, TearScience, Inc.) was introduced as an automated alternative for eyelid warming and massage. In three randomized clinical trials,^31–33 a single application of Lipiflow improved meibomian gland secretion, the tear breakup time, and ocular symptoms up to 1 or 3 months after treatment, and the results were better than or at least similar to those achieved with the daily application of conventional warm compression methods. In nonrandomized comparative studies^34,35 and noncomparative studies,^36,37 Lipiflow improved meibomian gland secretion, the tear breakup time, and ocular symptoms for 3 to 12 months. Yeo et al.³⁸ showed that the tear evaporation rate was significantly decreased from baseline up to 3 months after a single application of Lipiflow in patients with MGD. The decrease in the tear evaporation rate is larger after Lipiflow application than after twice-daily lid warming with a hot towel. These results suggest that Lipiflow is a useful treatment for MGD that does not depend on patient compliance. 

Kobayashi et al.³⁹ evaluated the effects of a newly developed eye-cleansing formulation for MGD. The product contains components that are anti-inflammatory and moisturizing and promote hair growth. In patients with MGD, significant improvements were observed in parameters such as subjective symptoms and meibum secretion. Kaido et al.⁴⁰ investigated the efficacy of eyelid margin cleansing with ofloxacin ointment in patients with obstructive MGD and found that slit-lamp parameters and fluorescein staining scores significantly decreased, while the tear breakup time significantly increased, after the treatment. In addition, the symptoms alleviated after treatment in most cases. These results suggested that cleansing with ofloxacin ointment is effective for patients with obstructive MGD. In another study, Arita et al.⁴¹ evaluated the effect of commercially available eyelid warming devices on ocular temperatures, tear film function, and meibomian glands in normal subjects and patients with MGD. Two nonwet and three wet devices were compared. A single treatment session with each of the five warming devices used alone improved the ocular symptoms score, the tear breakup time, and ocular temperatures. In the repeated warming session, Azuki no Chikara, a nonwet warming device, induced a stable improvement in the tear breakup time and increased the conjunctival temperature. It also improved the meibum grade. These results suggest that repeated eyelid warming with a nonwet device improves tear film and meibomian gland function in patients with MGD. Lid hygiene and warm compression have been the treatments of choice for MGD. The devices and drugs newly developed in Japan may enhance the efficacy of lid hygiene and warm compression in patients with MGD. 

Intense Pulsed Light (IPL)

Another new therapeutic modality is IPL therapy. IPL therapy has been an effective treatment for a range of dermatologic diseases, and it results in the alleviation of telangiectasia and facial erythema.⁴² Therefore, IPL therapy has been evaluated as a treatment for MGD. Craig et al.⁴² evaluated the effects of IPL therapy for MGD in a prospective, double-masked, placebo-controlled, paired-eye study and found that the lipid layer grade, tear breakup time, and symptom score improved from baseline up to day 45. Yin et al.⁴³ compared the results of IPL therapy and eyelid hygiene and found that IPL therapy and eyelid hygiene were equally effective in improving symptoms and meibomian gland function. Jiang et al.⁴⁴ evaluated the efficacy of IPL therapy in a prospective noncomparative study and found that IPL improved symptoms, the tear breakup time, and the meibomian gland secretion quality and expressibility. In two prospective, noncomparative studies,^45,46 serial IPL therapy sessions combined with meibomian gland expression improved symptoms and signs, including the meibomian gland secretion quality and expressibility. Proposed mechanisms by which IPL improves MGD signs and symptoms include heat transfer, thrombosis of the vasculature around the meibomian glands, and reduced inflammatory cytokines in tears.^42,47 However, the mechanism needs to be elucidated in future studies. 

Anti-Inflammatory Treatment

Anti-inflammatory treatment is important for MGD. Zhang et al.⁴⁸ used 1% azithromycin eye drops for 4 weeks in 16 patients with MGD and examined changes in the expression of proinflammatory and anti-inflammatory mediators in the eyelid margin and conjunctiva. The expression of proinflammatory mediators (interleukin [IL]-1β, IL-8, and matrix metalloproteinase 9) and transforming growth factor beta-1 (TGF-β1) was higher and lower, respectively, in patients with MGD than in healthy individuals. The expression of proinflammatory mediators decreased during treatment, although the levels returned to pretreatment values at 4 weeks after azithromycin withdrawal. Expression of TGF-β1 increased during treatment and remained at levels similar to those in healthy controls after drug withdrawal. These results suggest that the topical application of azithromycin is likely to reduce the inflammation in the eyelid and ocular surface of patients with MGD by changing the expression of pro- and anti-inflammatory mediators. Perry et al.⁴⁹ investigated the efficacy of topical cyclosporine A 0.05% (tCsA) for the treatment of MGD in a prospective study. Thirty-three patients with symptomatic MGD were randomized to receive either tCsA or placebo twice daily for 3 months. At the 3-month visit, several objective parameters, including the number of lid margin vascular injections, tarsal telangiectasis, and fluorescein staining scores, showed a significant improvement in the tCsA group, but not in the placebo group. The most significant finding was the greater decrease in the number of meibomian gland inclusions in the tCsA group than in the placebo group. These results suggest that topical CsA may be helpful in the treatment of MGD. 

Diquafosol

Topical diquafosol solution has been used to treat dry eye because it increases fluid secretion from conjunctival epithelial cells and mucin secretion from conjunctival goblet cells via the P2Y₂ receptor.⁵⁰ Because P2Y₂ receptor expression is observed in sebaceous cells and ductal cells in the meibomian gland,^51,52 diquafosol is expected to have some effects on meibomian glands. 

Arita et al.⁵³ used 3% diquafosol ophthalmic solution in patients with obstructive MGD for more than 4 months. Ocular symptoms, lid margin abnormalities, the superficial punctate keratopathy score, and the meibum grade were decreased, while the tear breakup time, tear film meniscus area, and meibomian gland area were increased. These results suggest that topical diquafosol therapy is effective for patients with obstructive MGD. In addition, Amano and Inoue⁵⁴ administered topical 3% diquafosol ophthalmic solution for 3 months in patients with MGD. The number of telangiectases and plugged gland orifices significantly decreased after 1 month of treatment, while the meibum score and meiboscore significantly decreased at 3 months. The increase in the tear LLT at 20 minutes after diquafosol administration became statistically significant at 3 months. Altogether, diquafosol ophthalmic solution has the potential to improve signs and symptoms in patients with MGD. 

As described in Amano and Inoue's article,⁵⁴ a recent study by Kam et al.⁵⁵ reported that UTP, a P2Y₂ receptor agonist, had no effect on immortalized human meibomian gland epithelial cells. In contrast, Wen et al. found that P2Y₂ receptor agonists enhance lipid production in primary cultured rabbit meibocytes (Wen Q, et al. IOVS 2002;43:ARVO E-Abstract 3146). UTP easily degrades in aqueous solutions, which necessitates its use immediately after dissolution in water.^56,57 In contrast, diquafosol is stable when dissolved in water.^56,57 Because Kam et al.⁵⁵ did not mention whether UTP solutions were immediately used, it is not known for certain whether their results are applicable to the actions of diquafosol on meibomian gland acinar cells. Either way, our results suggest that topical diquafosol is beneficial to meibomian glands. 

Vitamin D3

Another promising drug for MGD is vitamin D3. Because hyperkeratinization is a major cause of obstructive MGD, maxacalcitol (a noncalcemic analogue of the active form of vitamin D3), which is effective for psoriasis and ichthyosis, is expected to be effective in patients with MGD.⁵⁸ Arita et al.⁵⁸ applied maxacalcitol ointment to the lid margins of patients with obstructive MGD for 2 months and observed a significant improvement in clinical parameters (scores for plugging of meibomian gland orifices and lid margin vascularity, the tear breakup time, the meibum grade, and the meibomian gland area). An analogue of vitamin D3 was found to improve the condition of patients with obstructive MGD. 

Association of MGD With Other Diseases and Conditions

A few recent studies from Japan have described the association of MGD with other diseases or conditions. Arita et al.⁵⁹ examined tear film and meibomian gland parameters in patients with or without MGD and aqueous-deficient dry eye. Interestingly, meibomian gland losses had a positive correlation with tear secretion values only in patients with MGD. Therefore, the authors thought that the loss of meibomian glands is compensated for by an increase in tear secretion. As the authors discussed in the article, however, the mechanisms for this compensation need to be clarified in future studies. 

Matsumoto et al.⁶⁰ examined changes in the ocular surface and meibomian glands of 10 patients receiving an anticancer drug combination of tegafur, gimeracil, and oteracil potassium (S-1; Tahiho Pharmaceutical, Tokyo, Japan). Occluded orifices of meibomian glands and marked loss of meibomian glands on meibography were observed. In vivo confocal microscopy showed glandular atrophy and periglandular inflammation. The tegafur concentration in tears was high during the period of drug intake. These results suggest that S-1 damages the meibomian glands, with the most likely mechanism being the toxicity of the drug in tears. 

Ban et al.⁶¹ examined the lipid layer of the tear film in patients with chronic graft versus host disease (GVHD) after hematopoietic stem cell transplantation (HST) using tear film lipid layer interferometry (DR-1; Kowa, Tokyo, Japan). The DR-1 score for severity was higher in patients with dry eye and GVHD after HST than in patients without dry eye. Moreover, the DR-1 score for severity was positively correlated with the dry eye severity. These results suggest that MGD is one of the mechanisms for dry eye due to GVHD. 

Future Directions

To enable all patients with MGD to receive appropriate treatment, standardization of the diagnosis and treatment is necessary. To standardize the diagnosis of MGD, specific diagnostic criteria with cutoff values for each parameter are necessary. With the application of emerging novel technologies, new useful tools for MGD have been developed, such as noncontact meibography, in vivo confocal microscopy, and LipiView. Although these diagnostic devices are useful for visualizing subclinical changes related to MGD, they are expensive and not available in all clinics. Parameters that can be measured with devices available in most clinics should be used. Considering these requirements, the Japanese working group for MGD proposed a novel list of diagnostic criteria for obstructive MGD (Table). The validity of these new diagnostic criteria should be evaluated in future studies. 

The International Workshop on MGD proposed a treatment algorithm in which treatment is added depending on the severity of MGD. The sequence of treatment addition is eyelid hygiene, eyelid warming and massage, artificial lubricants, topical azithromycin, topical emollient lubricant, oral tetracycline derivatives, lubricant ointment, and anti-inflammatory therapy. However, the efficacy of this treatment algorithm is yet to be evaluated. In addition to these treatments, new therapeutic modalities such as LipiView and IPL have emerged. However, more clinical evidence is necessary for these new therapies to become a standard in the treatment of MGD. 

Acknowledgments

Funding of the publication fee and administration was provided by the Dry Eye Society, Tokyo, Japan. The Dry Eye Society had no role in the contents or writing of the manuscript. 

Disclosure: S. Amano, Alcon (R), AMO (R), Otsuka (R), Pfizer (R), Santen (F, R), Senju (R)