Topical Spironolactone in the Treatment of Ocular Graft-Versus-Host Disease

Abstract

Introduction: This two-part study aimed to investigate the therapeutic potential of topical spironolactone in ocular graft-versus-host disease (oGVHD). While off-label use of topical spironolactone has been described in dry eye, its efficacy in managing signs and symptoms of oGVHD remains unstudied. Preclinically, we tested the hypothesis that spironolactone induces corneal lipid synthesis in a mouse model. Clinically, we assessed patient response to spironolactone with a retrospective observational design.

Methods: Both immortalized and primary human corneal epithelial cells were stained with oil red O after 9 days of treatment with spironolactone. C57BL/6 mice were dosed thrice daily with one drop in each eye for 18 days. Corneal tissue was stained with oil red O and BODIPY™. Twenty eyes with oGVHD, as defined by the International Chronic oGVHD Consensus Group, were studied. Corneal fluorescein staining, lid margin vascularity, meibomian gland obstruction, meibum turbidity, zone A posterior lid margin vascularity, and oGVHD diagnostic criteria severity grading were compared in a pre-post study. Follow-up times ranged from 7 to 21 weeks, with a median time of 12 weeks. Statistical analysis was done with STATA 17 by fitting data to a non-parametric model.

Results: In vitro results showed an increased number and density of oil red O staining granules in the treatment group versus control in both primary and immortalized human corneal epithelium. In vivo, results showed translation to the mouse model with increased corneal epithelial BODIPY™ signal compared to untreated control. oGVHD patients had improved lid margin vascularity (p = 0.046), corneal fluorescein staining (p = 0.021), and International oGVHD Consensus Group severity scores (p = 0.011) after treatment with topical spironolactone. Minimal adverse effects were noted, the most common being mild stinging lasting less than a minute after instillation.

Conclusion: The improved severity scores, lid margin inflammation, and corneal fluorescein staining after weeks of treatment support the rationale that topical spironolactone may benefit oGVHD. The observed lipid production by the corneal epithelium is thought to contribute to this protective effect against ocular surface erosive disease in oGVHD. A mineralocorticoid receptor antagonist, spironolactone may offer therapeutic benefits in oGVHD while avoiding undesirable side effects of topical or systemic glucocorticoids.

Keywords: cornea; dry eye disorder; ocular graft-versus-host disease; ocular keratitis; tear film.

Figure 1. hTCEpi treated with control, 0.003% spironolactone, and 0.015% spironolactone.

(a-c) Immortalized human corneal epithelium without oil red O shows no staining. In panel (c), loss of cell confluency and survival is noted in the group treated with 0.015% spironolactone. (d) Cells treated with control demonstrate baseline oil red O staining. (e) Cells treated with 0.003% spironolactone demonstrate diffusely increased staining in addition to focal, intensely staining lipid granules indicated by arrows. (f) At spironolactone dosage of 0.015%, a toxic effect is noted with decreased cell survival, loss of confluence, abnormal size and morphology. Arrows indicate concentrated granular staining.

Figure 2. Oil red O staining of primary human corneal epithelium treated with spironolactone versus control.

(a, b) Primary human corneal epithelium shows similar size, confluence and morphology in cells treated with control and with 0.0005% spironolactone. (c) Primary HCE stained with oil red O shows background staining level in tissue treated with control. (d) Increased distribution of oil red O stain in treated tissue compared with control is noted in addition to densely staining bodies within cells. Cellular confluence and morphology remains largely similar between both stained control and treated groups.

Figure 3. Mouse cornea stained with oil red O and treated with spironolactone versus control.

(a) Mouse cornea treated with control, cryosectioned, and stained with oil red O demonstrates baseline stain. Epithelium and stroma are well visualized. (b) Mouse cornea treated with 0.001% spironolactone was cryosectioned and stained with oil red O. Corneal anatomy remains grossly similar to panel (a). Diffuse, slightly increased staining is noted in the epithelium, indicated by arrows.

Figure 4. BODIPY staining of mouse corneas treated with spironolactone versus untreated corneas.

BODIPY (493/503) staining of untreated mouse cornea shows minimal foci of fluorescent dye. (b) In a similar orientation, mouse cornea treated with 0.0005% spironolactone shows increased distribution of fluorescence throughout the epithelium. Increase in size and number of fluorescent foci are noted relative to untreated tissue in panel (a).

Figure 5. Schematic of ocular GVHD pathophysiology.

Fibroblast and lacrimal dysfunction contribute to cyclic lipid abnormalities, keratitis, and inflammation. Spironolactone may have therapeutic benefits by competitive inhibition of the mineralocorticoid receptor and subsequent inflammatory fibrotic processes. Adapted with permission from Ogawa [23].