MEIBOMIAN GLAND DYSFUNCTION

Meibomian Gland Dysfunction: An Update

A review of research on meibomian gland dysfunction and how it affects ocular comfort and contact lens wear.

By Carolina Kunnen, PhD, & Jason J. Nichols, OD, MPH, PhD, FAAO

The roles of lens material, design, and care solution in contact lens discomfort have been extensively investigated; however, an area in which research is more recently expanding is the influence of contact lens interaction with the adnexa. This article provides an update on meibomian gland dysfunction (MGD) and its relationship with contact lens wear and ocular comfort.

A Closer Look at MGD

It is believed that MGD is one of the leading causes of evaporative dry eye disease (Nichols K et al, 2011). Because there was no uniformity in definition and criteria of MGD, a committee of experts in the field formed the Tear Film and Ocular Surface Society’s (TFOS) International Workshop on Meibomian Gland Dysfunction, with the aim of achieving better understanding and consensus on meibomian gland health and disease (Nichols K et al, 2011; Nelson et al, 2011). The report of this committee, published in 2011, included the following recommended definition: “MGD is 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,” (Nichols K et al, 2011).

It has been hypothesized that the primary cause of obstructive MGD is hyperkeratinization (Korb and Henriquez, 1980; Knop et al, 2011). Although the exact mechanism is not understood, it is hypothesized that when the delivery of meibum is obstructed by, for example, hyperkeratinization, meibum accumulates within the duct. Because of the continuous delivery of meibum, pressure in the ductal system rises, eventually resulting in widening of the duct. After a while, this widening can extend and dilate the ductules up to the acini. This pressure can result in shrinkage and atrophy of the acini, together with loss of the meibocytes, which is visible as meibomian gland dropout (MGDO) during meibography (Knop et al, 2011).

Prevalence and Risk Factors of MGD The prevalence of MGD is reported variously from 3.5% to 70%, with a higher prevalence in Asian populations (Nichols K et al, 2011). Research indicates that aging influences the structure and/or function of meibomian glands (Schaumberg et al, 2011), as MGD is more frequently diagnosed in older subjects. This same research indicates that sex steroid hormones, such as androgen, have an effect on the lipid production of sebaceous glands in the body and appear to exert a similar effect on the meibomian glands; it is hypothesized that androgen deficiency and menopause are correlated with MGD.

Up to 90% of subjects who have ocular rosacea present with eyelid changes, including MGD (Schaumberg et al, 2011). Furthermore, positive correlations are reported between MGD and Sjögren’s syndrome, cholesterol levels, psoriasis, atopy, hypertension, and benign prostatic hyperplasia (Schaumberg et al, 2011). There is a clinical impression that contact lens wear is associated with MGD, and MGD is also often present in subjects who are diagnosed with anterior blepharitis, Demodex folliculorum, and dry eye disease. Lastly, some medications, such as antiandrogens, medication to treat benign prostatic hyperplasia, postmenopausal hormone therapy, antihistamines, antidepressants, and retinoids, are associated with MGD (Nichols K et al, 2011).

Contact Lens Wear and MGD

Contact lens wear has been investigated as a possible risk factor for MGD (Arita, Itoh, Inoue et al, 2009; Arita, Itoh, Maeda et al, 2009; Arita et al, 2010; Nichols J et al, 2005; and others. Full list available at www.clspectrum.com/references.), although this potential association is controversial. Henriquez and Korb (1981) and Korb and Henriquez (1980) reported that obstruction of the meibomian gland orifices was observed more frequently in symptomatic contact lens wearers than in asymptomatic contact lens wearers. It was also more prevalent in contact lens wearers than in non-wearers. This was confirmed by Ong and Larke (1990), who reported that lens wear was a risk factor associated with MGD and that 30% of lens wearers develop some degree of MGD after six months of lens wear, whereas only 20% of non-lens wearers developed this condition after the same period of time. Meibomian gland expressibility was used to assess the meibomian gland function in these studies.

A more recent study from Arita, Itoh, Inoue et al (2009), which used non-contact meibography to assess MGDO, showed that contact lens wearers had a significantly greater degree of MGDO compared to non-lens wearers. It was suggested that contact lens wear may accelerate age-related changes in the meibomian glands. Together with the observation that the difference in MGDO between lens wearers and non-lens wearers was greater for the upper eyelid than for the lower eyelid, it was suggested that mechanical trauma from lens wear may be responsible for the observed meibomian gland changes. Interestingly, GP lens and hydrogel lens wearers showed no statistically significant difference in the average MGDO score, although there was a trend favoring more dropout with GP lens wear. Then again, the duration of contact lens wear was weakly associated with the MGDO score.

Villani et al (2011) investigated the relationship between contact lens wear and morphological changes in the meibomian glands by using an in vivo laser scanning confocal microscope (Figure 1), which is a relatively new method employed to image meibomian glands. The duration of lens wear was associated with changes in the diameter of the meibomian gland acini. According to the authors, morphological changes observed using this method were indicative of MGD.

Figure 1. In vivo laser scanning confocal microscopy image, central duct surrounded by acini.
COURTESY OF BRIEN HOLDEN VISION INSTITUTE

Conversely, a large study by Hom et al (1990), which compared the frequency of MGD within contact lens wearers and non-lens wearers, found no significant difference in meibomian gland expressibility between both groups. Another study by Marren (1994), in which gentle expression was used to assess the meibomian gland function, was also unable to find significant differences between contact lens wearers and non-lens wearers. A more recent study from Nichols J and Sinnott (2011) used meibography to assess MGDO, similarly to Arita, Itoh, Inoue et al (2009), and found no significant differences in MGDO between symptomatic and asymptomatic lens wearers.

It is evident from these various studies that the true influence of contact lens wear on the meibomian glands is still unknown, although there is mounting evidence of structural changes to the glands themselves. One possible explanation for the discrepancies in these results might lie in the different definitions of MGD used in each study. A greater consensus in the definition of MGD has been achieved by the International Workshop on MGD. However, the studies described above were conducted prior to this report.

MGD Assessment and Its Relationship with Ocular Comfort

Currently, MGD and MGD-related disorders are diagnosed based on abnormal appearance or performance of the glands. In this update, the key clinical signs of MGD—meibomian gland expression, eyelid morphology, and MGDO as listed by the International Workshop on MGD (Tomlinson et al, 2011)—are highlighted, as is the relationship of these signs with ocular comfort.

Meibomian Gland Expression and Quality It is believed that MGD has an asymptomatic preclinical phase. Asymptomatic MGD may be diagnosed by meibomian gland expression (Tomlinson et al, 2011). It is believed that normal meibum is clear, but in MGD, it becomes cloudy or yellow/white or cannot be expressed (Knop et al, 2011). With progression, MGD becomes more symptomatic (Tomlinson et al, 2011).

Korb and Henriquez (1980) observed cloudy/creamy meibum expression in asymptomatic subjects who had normal eyelid appearance, which indicated MGD according to authors. It is therefore recommended to perform meibomian gland expression to detect the presence of asymptomatic MGD (Tomlinson et al, 2011; Meadows et al, 2012).

In a later report, Blackie et al (2010) described this phenomenon as non-obvious obstructive meibomian gland dysfunction (NOMGD). It is possible that NOMGD is a precursor for obstructive MGD. The authors reported that patients who have NOMGD often present with severe symptoms, which suggests that meibum changes without lid margin abnormalities can present both with and without symptoms. This may indicate that there is a fine line between symptomatic and asymptomatic MGD.

Furthermore, an absence of meibum expression does not necessarily indicate overt MGD. Norn (1985) showed that only 45% of meibomian glands are active at a given time. Korb and Blackie (2008) confirmed this finding and reported that, on average, only six to 10 glands of the lower eyelid are actively producing meibum at a given time. Studies also show that the nasal glands are the most active, followed by the central glands and, finally, the temporal glands (Knop et al, 2011; Tomlinson et al, 2011; Korb and Blackie, 2008). After expression of the gland, the recovery time from the same gland was approximately two hours (range: 1.5 to 2.5 hours) (Blackie and Korb, 2009).

According to the International Workshop on MGD, it is possible that meibomian glands work in a cycle of activity that differs from gland to gland over time. This theory suggests that each gland has periods of activity when meibum is released (Tomlinson et al, 2011). However, research has shown that if meibum is yielded in the morning, there is a high likelihood that the gland will still yield meibum after nine hours and vice versa (Blackie and Korb, 2010). Thus, it is believed that the cycle of meibum activity is a slow one (Tomlinson et al, 2011). Therefore, if only some of the glands are actively secreting meibum at any one time, it may be difficult to differentiate between healthy glands and those not expressible for physiological or pathological reasons (Tomlinson et al, 2011).

That the glands are only intermittently expressible may explain the discrepancies in results reported in various studies. A study conducted by Ibrahim et al (2010) showed significantly higher grades of expressibility in MGD patients compared to the control group, suggesting that it was harder to express meibum in the MGD patients. However, Arita, Itoh, Maeda et al (2009) showed that the meibum score was poor at differentiating patients who have obstructive MGD from normal subjects. According to the authors, a possible explanation for this lack of differentiation was that pressure in this particular study was exerted on the center of the eyelid, while most distinguishing changes are observed at the nasal and temporal regions. However, it may be possible that subjects who were categorized as normal were, in fact, subjects who had NOMGD or were asymptomatic subjects in a preclinical stage of MGD.

Contradicting findings between meibomian gland expressibility and ocular comfort have been reported in the literature (Cuevas et al, 2102; Korb and Blackie, 2008; Srinivasan et al, 2012). While there is a lack of clinical signs in NOMGD, lid margin changes are evident in obstructive MGD; several changes in the eyelid have been observed in MGD.

Meibomian Gland Orifices and the Eyelid Margin Changes such as plugging or pouting of the orifices are an early sign of MGD. Plugging may eventually lead to obliteration of the orifices, with atrophy of the meibomian gland duct (Tomlinson et al, 2011). In MGD, the number of meibomian glands may also be reduplicated, or reduced. Orifices may be capped by a ring of meibum with a tough surface, and the cap may be epithelialized (Foulks and Bron, 2003). Capping has been reported mainly on sporadic orifices. Antero-posterior position of the meibomian glands in relation to the mucocutaneous junction is also a sign of MGD (Bron et al, 2011), such that more posterior location of the glands in relation to the mucocutaneous junction was associated with greater MGD. Loss of orifice architecture occurs with age and early in MGD. Other orifice changes are characterized by dimpling, telangiectasia, epithelial ridging between meibomian gland orifices, cystoid changes in the glands, and formation of concretions within the acini (Tomlinson et al, 2011). Several classification schemes have been developed to assess lid margin abnormalities; although lid margin abnormalities have been associated with MGD, no associations with ocular comfort were observed (Nichols K et al, 2004; Tong et al, 2010; Golebiowski et al, 2012; Sullivan et al, 2014).

Meibography As mentioned earlier in this update, MGDO refers to the loss of acinar tissue and can be detected by meibography (Figure 2) (Tomlinson et al, 2011). There are two different types of meibography: transillumination of the everted eyelid and non-contact meibography.

Figure 2. An example of an eyelid without MGDO (left) and with MGDO (right) using non-contact meibography.
COURTESY OF BRIEN HOLDEN VISION INSTITUTE

In transilluminating meibography, the eyelid is everted over the light source, and the meibomian glands are observed in silhouette (Tomlinson et al, 2011; Pult and Nichols J, 2012). Transilluminating meibography was first performed with white light; this was later performed with infrared light (Pult and Nichols J, 2012) because meibomian glands are infrared hyper-reflective (Arita, Itoh, Inoue et al, 2009; Arita, Itoh, Maeda et al, 2009). Non-contact meibography, introduced by Arita, Itoh, Inoue et al (2009) and Arita, Itoh, Maeda et al (2009), increased the scope of the technique and made it more patient-friendly (Tomlinson et al, 2011). Partial or complete loss of the meibomian glands could be monitored with a non-contact meibographer that uses an infrared filter and an infrared charge-coupled device video camera.

MGDO increases with age in normal subjects (Arita et al, 2008; Feng et al, 2014) and develops earlier in men than in women (Arita et al, 2008). The first meibomian gland changes are observed in women after the age of 30 years and in men after the age of 20 years (Arita et al, 2008), although such changes are not necessarily in response to the presence of obstructive MGD (Tomlinson et al, 2011). The changes may also occur as part of an age-related atrophic process (Obata, 2002). However, it is believed that MGDO is an aspect of MGD and that more MGDO represents more severe MGD. MGDO may present proximally (at the orifices’ side), centrally, or distally (at the eyelid fold), or it may involve the entire gland (Tomlinson et al, 2011). MGDO most frequently presents at the distal side, especially in contact lens wearers (Arita, Itoh, Inoue et al, 2009).

There is, as yet, no gold standard in assessing meibography images (Pult and Nichols J, 2012). Several grading scales have been developed for use with meibography assessment since 1998 (Pflugfelder et al, 1998). These scales grade the proportion of partial glands or MGDO and range from three- to seven-point scales. Successful use of grading scales is highly dependent upon clinician ability and accuracy (Nichols J et al, 2005; Pult and Riede-Pult, 2013). To minimize this variability, automated systems and manual image measure methods for meibography have been developed (Srinivasan et al, 2012; Pult and Riede-Pult, 2013; Ban et al, 2013; Arita et al, 2014; and others). Most of these systems have the same focus as the grading scales and calculate the area of MGDO. Arita, Itoh, Maeda et al (2009) observed various meibomian gland changes in subjects who had obstructive MGD, such as gland shortening, distortion, dilation, and hairpin-loop-like changes. Furthermore, these features differed from those seen with advancing age.

As described earlier in this update, in obstructed meibomian glands, meibum accumulates within the duct, and constant production of meibum can cause dilatation of the gland before it results in atrophy (Knop et al, 2011). Nevertheless, the current systems (Arita et al, 2014) used to assess meibomian gland changes primarily consider the area of MGDO or only a relatively limited range of gland morphology characteristics (Pult and Riede-Pult, 2013; Ban et al, 2013). It may be useful to automatically determine other features of the meibomian gland morphology, such as meibomian gland width and the number of glands. Those parameters may provide more detailed information and may serve as precursors for a more severe stage in meibomian gland morphology, namely MGDO. Research in this area is currently being conducted.

Arita, Itoh, Maeda et al (2009) showed a significantly higher level of MGDO in obstructive MGD subjects compared to controls. This finding was confirmed by Ibrahim et al (2010). The level of MGDO was also able to differentiate MGD subjects from those who had aqueous-deficient dry eye (Arita et al, 2010). Furthermore, MGDO has been linked with noninvasive breakup time (NIBUT) (Pult and Riede-Pult, 2012), lipid layer thickness (Pult et al, 2012), lipid layer pattern (Pult and Riede-Pult, 2012; Eom et al, 2013), lid margin abnormalities (Arita, Itoh, Maeda et al, 2009; Arita et al, 2010), and meibum quality (Arita, Itoh, Maeda et al, 2009).

Conversely, a study from Murakami et al (2014) showed that there was no relationship between the level of MGDO and the number of functional meibomian glands and/or meibomian gland volume. The authors concluded that the current assessment method (Pult and Riede-Pult, 2013) (grading scale) for meibography cannot be used to predict meibomian gland function unless there is a complete absence of meibomian glands.

Despite the balance of evidence suggesting that MGDO can differentiate between MGD and controls or aqueous-deficient dry eye, contradicting findings between meibomian gland morphology and ocular comfort have been reported in the literature (Srinivasan et al, 2012; Feng et al, 2014; Pult and Riede-Pult, 2012; Pult et al, 2012).

Conclusions

Although it is believed that MGD is the leading cause of evaporative dry eye (Nichols K et al, 2011), contradicting findings in associations between ocular symptoms and meibomian gland morphology and function have been reported in the literature (Korb and Blackie, 2008; Srinivasan et al, 2012; Nichols K et al, 2004; Sullivan et al, 2014; and others). The theory in the opening sections of this update indicates that meibomian gland obstruction leads to alterations in the lipid layer thickness and tear film stability (Arita, Itoh, Inoue et al, 2009; Craig et al, 1995). It is expected that these tear film changes would result in ocular discomfort, especially when the tear film is compartmentalized by a contact lens (Craig et al, 2013).

However, in practice there is no clear connection between clinical signs and symptoms. The few statistically significant correlations found between signs and symptoms were weak to moderate (Johnson, 2009). This does not indicate that there are no correlations; however, poor study designs, poor repeatability of tests, different ocular comfort questionnaires, and subject selection are possible causes of discrepancy between these findings.

This update shows that diagnosing MGD remains challenging; however, new diagnostic tools will extend the knowledge of meibomian gland function, morphology, and associated symptoms, which will aid our understanding and contribute to better management of MGD, with the ultimate goal of improving care for MGD and dry eye patients worldwide. This will, in turn, benefit eyecare practitioners when managing dry eye, resulting in improved quality of life for dry eye patients and reduced contact lens dropout. CLS

For references, please visit www.clspectrum.com/references and click on document #236.

Dr. Kunnen is a postdoctoral research fellow at the University of Alabama-Birmingham. Her research interests are dry eye and meibomian gland dysfunction. Dr. Kunnen is an optometrist and orthoptist from the Netherlands. She completed her PhD at the Brien Holden Vision Institute and the University of New South Wales in Australia. She is also the chairperson of a non-government organization, The Optical Foundation.

Dr. Nichols is an assistant vice president for industry research development and professor at the University of Alabama-Birmingham as well as editor-in-chief of Contact Lens Spectrum and editor of the weekly email newsletter Contact Lenses Today. He has received research funding or honoraria from Vistakon, Alcon, and Allergan.