William D. Mathers

Meibomian gland dysfunction in chronic blepharitis

We examined 57 patients with symptoms of chronic blepharitis using meibomian gland expression, meibography, tear osmolarity, and the Schirmers test. We also performed meibography on 20 normal patients free of chronic blepharitis. We found that 42 blepharitis patients (74%) had evidence of meibomian gland loss, whereas only four of 20 normal patients (20%) had any gland dropout. We performed cluster analysis on the data from the patients with blepharitis and found that these patients tended to fall into distinct groups with clinically relevant characteristics. We also found that tear osmolarity correlated positively with gland dropout (+0.413) and negatively with excreta volume (−0.499). This study demonstrates that an objective analysis of meibomian gland function may be used to assess chronic blepharitis and define subsets of blepharitis with measurable differences. It also supports the significance of meibomian gland dysfunction on tear osmolarity and the evaporative state of the eye.

Meibomian gland morphology and tear osmolarity: changes with Accutane therapy.

We evaluated the meibomian gland function of 11 patients before and during treatment with isotretinoin (Accutane) by assessing tear osmolarity, meibomian gland morphology, tear production, rose bengal staining, and meibomian gland excreta. We found, during Accutane use, that meibomian glands appeared significantly less dense and atrophic by meibography. Excreta thickness increased from 1.7 ± 0.9 to 3.1 ± 1.2 (p < 0.005), and expressible excreta volume decreased from 1.52 ± 0.68 to 1.10 ± 0.3 (p < 0.05) (scale 1-4). We also found a significant increase in tear osmolarity from 304.9 ± 11 to 316.3 ± 10 mosmol/L (p < 0.005). There was no significant change in the Schirmer test during treatment. We suggest that the clinical symptoms of blepharitis during Accutane therapy are related to decreased meibomian gland function and consequent increased tear evaporation and tear osmolarity.

Immunopathology and Electron Microscopy of Acanthamoeba Keratitis

In order to evaluate the local cellular immune response to Acanthamoeba infection we performed immunohistochemical examinations of the corneal buttons of two patients with Acanthamoeba keratitis. We found that the corneal stroma was infiltrated with polymorphonuclear leukocytes and HLA-DR positive macrophages that appeared to be stromal keratocytes by light microscopy. Despite the presence of chronic inflammation in both patients, no stromal lymphocytes were seen in one patient and a sparse lymphocytic infiltrate was seen in the other patient. Electron microscopy confirmed the presence of macrophages, neutrophils, and Acanthamoeba organisms in these two patients.

Corneal rim cultures.

We performed a retrospective analysis of 291 consecutive scleral rim cultures, taken at the time of corneal transplantation, from 1981 to 1986. There were 82 positive cultures (39%), and the most common pathogen was Staphylococcus epidermidis, with 24 positive cultures (30%). Ten of these (42%) were resistant to gentamycin. There were no cases of endophthalmitis or other infections in the immediate postoperative period. We also found 15 positive cultures of Streptococcus viridans. Of these, 13 (87%) were resistant to gentamycin but sensitive to most other antibiotics. Even though endophthalmitis is rare following penetrating keratoplasty, the emergence of resistant strains of Streptococcus viridans and Staphylococcus epidermidis suggests that eyebanks need to consider the addition of a second antibiotic to their MK medium.

Vortex Keratopathy of the Corneal Graft

We examined the surface epithelium of 30 eyes with color specular microscopy at 3-month intervals 1–18 months following penetrating keratoplasty. At least three examinations were performed on most eyes. We found a vortex keratopathy to be present in 70% of these eyes. In nine eyes, this vortex pattern was present at every examination, and in 12 eyes, it was present at some examinations and not at others. There was no vortex pattern seen at any examination in nine eyes. We also found alterations in the normal epithelial morphology around sutures and at the wound margin, with palisading parallel to the sutures and perpendicular to the wound edge. This altered morphology was also demonstrated by digitization of cells for shape and size, which showed significant differences compared with normals. We conclude that the morphology of the epithelial cells on the surface of the cornea remains abnormal for up to 18 months following penetrating keratoplasty.

Corneal steepening with final suture removal after penetrating keratoplasty.

We reviewed 53 patients after they underwent penetrating keratoplasty, comparing the change in average corneal curvature before and after final suture removal 15-18 months postsurgery. In 29 eyes with keratoconus, there was a significant increase in average curvature of 3.48 diopters ± 3.76 (p < 0.0005). In 24 eyes with other diagnoses, there was also a significant increase in the average curvature of 3.26 diopters ± 3.98, (p < 0.0005). The difference between these groups was not significant. The amount of steepening correlated negatively ( - .6494) with the average corneal curvature before suture removal (p < 0.005). We found no change in the average astigmatism after suture removal; however, large changes in astigmatism occurred unpredictably with some patients. This study shows that corneal curvature after penetrating keratoplasty steepens after suture removal, particularly in flat corneas, and astigmatism may shift unpredictably.

Correlation of full-thickness corneal wound length with endothelial cell loss

We studied 12 patients who had sustained penetrating corneal lacerations with corneal wound lengths ranging from 1 to 9 mm to determine whether wound size was directly proportional to endothelial cell loss 3 or more months after injury. Endothelial cell counts in the uninjured eye averaged 2973 +/- 330 cells/mm2 (range, 2500 to 3700). In the injured eye, near the wound (less than 2 mm from the wound), the cell counts averaged 1260 +/- 590 cells/mm2 (range, 575 to 2500); away from the wound, where cells appeared healthiest, counts averaged 1619 +/- 544 cells/mm2 (range, 850 to 2750). The average total cell loss near the wound was 1713 +/- 622 cells/mm2 (range, 500 to 2575), vs 1354 +/- 582 cells/mm2 (range, 250 to 2150) away from the wound, compared with the uninjured eye. Cell loss near the wound was not significantly different from that away from the wound. There was a positive correlation between wound length and total cell loss measured both near the wound (r = 0.830) and away from the wound (r = 0.755). Pars plana lensectomy was not associated with detectable additional cell loss. Our results suggest that patients with corneal lacerations sustain significant endothelial cell loss, which correlates closely with wound length, and that patients with larger wounds may be at greater risk for developing corneal decompensation with additional procedures or trauma.