Subjects
Patients attending the glaucoma clinic in the Department of Ophthalmology (Chonnam National University Hospital) with features of NTG were consecutively enrolled in this prospective study between September 2018 and February 2019. This study adhered to the Declaration of Helsinki and was approved by the Chonnam National University Hospital Institutional Review Board. Written informed consents were obtained after the subjects were fully informed of the purposes of this study.
NTG was diagnosed using the following criteria: untreated maximum IOP lower than 21 mmHg during the repeated Goldmann applanation tonometry measurements taken in a routine period of daytime on different days, a normal open angle observed on gonioscopy, and the occurrence of glaucomatous optic neuropathy identified in fundus photographs (Kowa Nonmyd 7 fundus camera; Kowa Co., Ltd, Tokyo, Japan) and optical coherence tomography (OCT) (Cirrus HD-OCT; Carl Zeiss Meditec Inc., Dublin, CA, USA) with the corresponding visual field (VF) defects assessed using automated perimetry (Humphrey Field Analyzer; Carl Zeiss Meditec Inc., Dublin, CA, USA). A glaucomatous VF defect was defined as a cluster of three or more contiguous points in the pattern deviation plot with P < 0.05, at least one of which must have been P < 0.01; a pattern standard deviation with P < 0.05; or a glaucoma hemifield test result outside of normal limits. VF defects had to be repeatable on at least 2 subsequent tests with reliable analyses (that is, false-positive rate of ≤ 15%, false-negative rate of ≤ 15%, and fixation loss rate of < 20%).
To qualify for inclusion, the NTG patients were required to satisfy the following criteria: 18 years or older, starting PGA treatment within two weeks, and no MGD-related signs, including abnormal lid margin, altered gland secretions, or conjunctival and corneal staining [21]. Patients were excluded if they had blepharitis, wore contact lens, used other topical eye drops except for PGA or DQS, had undergone ocular surgery, had androgen deficiency, rosacea, or Stevens-Johnson syndrome, or were undergoing systemic medication treatment (such as isotretinoin, antiandrogens, antidepressants, antihistamines or postmenopausal hormone therapy). These are all factors that may affect the meibomian gland [22].
According to the topical medications used by patients, they were classified into the PC-PGA monotherapy group, PF-PGA monotherapy group, and PC-PGA with 3% DQS (PC-PGA + DQS) combination therapy group. PC-PGA included latanoprost 0.005% with benzalkonium chloride (BAK) 0.02% (Xalatan®, Pfizer Inc., New York, USA) and tafluprost 0.0015% with BAK 0.001% (Taflotan®, Santen Pharmaceutical Co, Ltd., Osaka, Japan). PF-PGA included latanoprost 0.005% without BAK (Monoprost®, Thea, Clermont-Ferrand, France) and tafluprost 0.0015% without BAK (Taflotan-S®, Santen Pharmaceutical Co, Ltd., Osaka, Japan). Three percent DQS ophthalmic solution without BAK (Diquas-S®; Santen, Osaka, Japan) was originally intended for those who were worried about dry eye symptoms. All patients were instructed by the glaucoma specialists to apply PGA once per night and DQS four times per day (if appropriate) before enrollment.
Apart from the routine glaucoma examinations which involve the assessment of the IOP, optic nerve head, retinal nerve fiber layer thickness, and VF, the meibomian gland morphology evaluation was the main focus of this study. All examinations were performed at baseline and at each follow-up (1, 3, 6, 9, and 12 months).
To evaluate the meibomian gland morphology, non-contact infrared meibography was performed for the upper eyelid of each eye using Keratograph® 5M (OCULUS, Wetzlar, Germany) [23]. Photographs from the meibography were then analyzed by using ImageJ 1.52a (National Institute of Health, USA) to calculate the ratio of the meibomian gland dropout area to the total tarsal area. This ratio was named the meibomian gland loss (MGL; %), which was found to be reliable for grading the meibomian gland morphology [24, 25].
More specifically, we first manually outlined the total tarsal area using the polygon selection tool. Next, the original color photograph was converted into an 8-bit type image. We then applied automatic threshold identification to discriminate the meibomian gland area from the non-meibomian gland area [26]. The non-meibomian gland area was regarded as the meibomian gland dropout area. In case the automatic threshold identification was not performed properly, the threshold was manually adjusted and the misidentified area was modified using the paintbrush tool. Finally, the number of pixels within the meibomian gland dropout area was counted and its relation to the pixels in the total tarsal area was calculated as a fraction (0-100%) (Fig. 1).
MGL of each meibography photograph was evaluated by two experienced observers (Y.G. and J.Y.H.) who were masked to the treatment of each patient and the results of prior observations. The mean MGL value was analyzed statistically.