Dr. Sohini Mandal

Dr. J. S. TITIYAL, Dr. VINEY GUPTA, Dr. SHIKHA GUPTA

Semi Final

Abstract

Purpose: To compare DM tear morphology in keratoconus (KC) and PCG eyes using ASOCT and histopathology. Methods: ASOCT of PCG eyes with Haab’s striae(15) and KC eyes with hydrops(10) were evaluated and compared with healthy eyes(12) for DM-PDL morphology. These features were assessed on histopathology of corneal button obtained from PCG(13) and KC eyes(14) after keratoplasty and compared with retinoblastoma eyes(6) on light microscopy and immunostaining.

Results: ASOCT showed thicker DM-PDL in PCG(33.1±12.7µ) than that of KC eyes(18.4±2.1µ)(p<0.01). Both groups showed hyper-reflective double layer. On the contrary, histopathology showed thicker DM-PDL in KC(91.7±77.5µ) than that of PCG eyes(62.5±57.3µ)(p-0.01). Despite the increased thickness of DM-PDL, it does not act to seal the break in KC whereas that in PCG acts to seal the break. Conclusion: This differential response of the posterior cornea may be responsible for varied clinical presentation of DM tears between the 2 groups.

Full Text  

Comparative analysis of posterior cornea between hydrops in keratoconus and Haab’s striae in primary congenital glaucoma

Authors:

Sohini Mandal1, MD;Jeewan S Titiyal1, MD; Viney Gupta1, MD; Shikha Gupta1, MD. Dr Rajendra Prasad
Centre for Ophthalmic Sciences, All India Institute of Medical Sciences, New Delhi, India- 110029

Tel: 91 40 68102502
Fax: 91 40 68108339
ORCID ID: 0000-0002-0800-9961

Abstract:

Aim:

To compare the morphological characteristics of Descemet membrane (DM) tears in eyes with primary congenital glaucoma (PCG) and keratoconus (KC) using high-definition anterior segment optical coherence tomography (ASOCT) and histopathology.

Methods

ASOCT of PCG eyes with Haab’s striae (15 eyes of 15 patients) and KC eyes with hydrops (12 eyes of 12 patients) were evaluated prospectively for Descemet’s membrane-Pre-Descemet’s layer (DM-PDL) morphology. These features were further assessed retrospectively on histopathology of corneal buttons obtained from 13 PCG and 14 KC eyes after penetrating keratoplasty and compared with 6 retinoblastoma eyes on light microscopy and immunostaining for Collagen IV.

Results:

ASOCT showed significantly thicker DM-PDL in PCG (77.2 ± 26.2 μm) than that of KC eyes (47.1 ± 25.5 μm) (p<0.01). Both groups showed hyper-reflective double layer. On the contrary, histopathology showed thicker DM-PDL in KC eyes (91.7±77.5 µm) than that of PCG eyes (62.5±57.3 µm) (p-0.01). PCG eyes have generalised thickening of DM and PDL in response to high IOP, more so in areas of Haab’s striae which heal by localised hyper-proliferation. On the contrary, keratoconus eyes with hydrops have higher posterior corneal thickening in areas of rupture which don’t act to seal the break but result in retraction and coiling of DM/PDL.

Conclusion:

This differential response of the posterior cornea may be responsible for the varied clinical presentation of the DM tears between the two groups.

Keywords:

Descemet’s tear, DM-PDL complex, Haab’s striae, corneal hydrops, primary congenital glaucoma, keratoconus.

Introduction

The response of corneal tissue to Descemet’s membrane (DM) tears in two different diseases of primary congenital glaucoma (PCG) and keratoconus is distinctly different. Contrary to PCG, acute corneal hydrops in cases of advanced keratoconus, is associated with stretching and rupture of DM, resulting in transient corneal edema and intrastromal fluid clefts.(1) However, recently Parker et al in their study proved that a defect at the level of DM is not sufficient to elicit an acute corneal hydrops, unless also accompanied by a defect in the posterior corneal stroma.(2) PDL possesses special mechanical properties that includes high tensile strength, impermeability to air, and a distinctive collagen makeup featuring a greater amount of type VI collagen and longer spacing fibrils.

Cherif and associates demonstrated that compression sutures placed in the deep posterior stroma may be an effective treatment for corneal hydrops and theorized that posterior stromal breaks were the root cause of the condition.(3) Hence, both these studies point towards the fact that the DM tears alone may not be sufficient to induce hydrops in keratoconus eyes and PDL tears may also be a necessary pre-requisite for inducing the breaks. Although the prevalence of an episode of corneal hydrops in clinic-based patients with keratoconus has been estimated to be between 2.6 and 2.8%.(4,5)

The differential responses of PCG and keratoconus to the corneal tears, is the subject of intrigue for us. While corneal tears in PCG induce Haab’s striae with resultant corneal edema which often resolves on lowering of IOP if done in time, tears in keratoconus induces corneal hydrops with poor resolution rates. The hypothesis leading to the study is that whether the rapid endothelial regeneration in children eyes in tandem with a thickened PDL act concurrently to prevent stromal edema to ensue and to last to some extent in these cases. This led us to envisage this study to assess the morphological characteristics of DM and PDL both in vivo (ASOCT) and on histopathology and to explore possible differences between the PCG and keratoconus eyes, especially in areas with Haab’s striae.

Methods

ASOCT Study

The prospective study included two groups, one of which had confirmed cases of advanced keratoconus with corneal hydrops (acute and healed) and the other had confirmed cases of primary congenital glaucoma (PCG) who had undergone a single surgery (trabeculectomy and trabeculotomy). All the patients were old enough (>8 years) to cooperate for a slit-lamp mounted ASOCT examination and gonioscopy. Cases of corneal hydrops with vision of at-least finger counting at 1 meter, were compared with those of unilateral/ bilateral PCG with enlarged corneal diameter (>12 mm).

All patients underwent a complete ocular examination including uncorrected distance visual acuity (logMAR), corrected distance visual acuity (logMAR), intraocular pressure (mmHg) and refractive error (spherical equivalent). At slit-lamp examination, corneal diameter (mm), corneal clarity (clear/ mild haze with visible iris details/ moderate haze with slightly visible iris details/ severe haze with absence of iris details), location of hydrops/ Haab’s striae (central/ peripheral/ both) and extent of Descemet’s membrane tear (clock hours) were noted.

Any other form of primary corneal ectatic disorder such as pellucid marginal degeneration, keratoglobus, posterior keratoconus, and previous history of corneal collagen crosslinking or keratoplasty procedure were excluded in the keratoconus group. Similarly, PCG eyes with isolated trabeculodysgenesis were included, whereas those having evidence of any other ocular comorbidity like prominent Schwalbe line, angle recession, angle pigmentation ≥ grade 3, peripheral anterior synechiae, nystagmus, poor fixation, microcornea, corectopia, ectropion uveae, cataract, Peter syndrome, aniridia, or a history of steroid use were excluded.

The study was conducted after approval from Institutional Ethics Committee and was carried out in accordance with the tenets of Declaration of Helsinki. Written and informed consent was obtained from all patients/parents before inclusion in the study. All study patients underwent clinical photography of the anterior segment with additional goniophotography for PCG patients on Eye Cap (Haag-Streit International, Koniz, Switzerland).

High resolution in vivo ASOCT (Spectralis, Software version 6.5; Heidelberg Engineering GmbH, Heidelberg, Germany) was performed twice for all study patients, 2 weeks apart by two different examiners using an anterior segment lens to assess the cornea. This was done to ensure reproducibility and an average of the 2 readings was taken for analysis. Corneal scans from central, midperipheral, and peripheral cornea were taken on “high resolution anterior segment module” for scanning DM and PDL. For the purpose of this study, horizontal cross-sectional scans through 3- and 9- o’clock meridians were used for analysis.

If these were marred by the presence of the Haab striae or hydrops, scans from adjacent horizontal meridians were considered. The ASOCT manual caliper tool was used to measure the DM thickness at 3 points in central 6-mm cornea (avoiding the areas with Haab striae and hydrops), zoomed at 400% for ease in identification, and an average of the 3 readings was taken for analysis. The 3 random points chosen were such that 2 were at the periphery of the chosen section and one at the centre while making sure that the caliper was kept perpendicular to the endothelium.

Initial medical management of acute hydrops was done with topical 5% sodium chloride 6 times, 1% prednisolone phosphate 6 times, 0.5% moxifloxacin 3 times, 0.5% timolol 2 times, homatropine 2% 2 times and lubricants 4 times a day. Intracameral perfluoropropane (14% C3F8) injection with or without supra-descemetic fluid drainage was performed preferably on the same day of presentation.

Statistical Analysis:

Data were entered in Microsoft Excel and analysed using Stata 12.0 software (StataCorp LP, College Station, TX). Data were presented as number (%) or mean ± SD as appropriate. A 2-sample t test was calculated to derive the difference between the two ASOCT study groups. A Mann-Whitney U test was used for data that did not have a normal distribution. A p value <0.05 was considered statistically significant.

Results:

ASOCT Study:

Twelve eyes of 12 keratoconus patients and fifteen eyes of 15 PCG patients were analysed on ASOCT. The mean age of the patients with keratoconus at the time of OCT examination was 21.2 ± 6.4 years (range 14-36 years) whereas that of PCG was 16.1 ± 4.9 years (9-25 years) (p<0.001). The disease duration was significantly higher in KC eyes (21.2 ± 6.4 years) than that of the PCG eyes (14.8 ± 8.1 years) The mean highest baseline IOP documented was 15.43 ± 2.59 mm Hg and 31.81 ± 6.11 mm Hg in the keratoconus and PCG group respectively. The mean spherical equivalent was -5.50±4.81 DS (-14.00 to -2.25 DS) and -3.25±4.78 DS (-10.00 to +3.00 DS) in keratoconus and PCG group respectively.

The mean keratometry and corneal astigmatism was significantly higher in the KC group (56.1 ± 11.5 D and 6.00 ± 5.50 DC). The mean horizontal corneal diameter was 11.93 ± 0.39 mm and 13.61 ± 0.86 mm in keratoconus and PCG eyes respectively. 66.7% of PCG eyes and 41.7% of KC eyes undergoing ASOCT presented with clear cornea. Mild to moderate to severe corneal haze was present in the remaining eyes. ASOCT revealed significantly thinner corneas in the KC group (394.8 ± 70.1 μm) in comparison to the PCG group (526.7 ± 56.5 μm). Additionally, ASOCT also showed significantly thinner DM-PDL complex inside (47.1 ± 25.5 μm) the areas of hydrops in keratoconus group when compared to the PCG group (77.2 ± 26.2 μm) (p<0.01).

Similar findings were noted outside the areas of hydrops (15.6 ± 4.3 μm) in keratoconus group in comparison to the PCG group (20.4 ± 5.5 μm) (p<0.01). Both the groups showed the DM-PDL complex as two separate hyper-reflective lines with translucent space in between, giving it a “double membrane” appearance. The DM was found to be globally thickened in all scans of PCG eyes with intracameral protuberances of varying sizes and shapes in the areas of Haab’s striae.

Discussion:

Because PCG and keratoconus needs lifelong therapy, follow-up and visual rehabilitative measures, understanding the corneal pathology is equally important for the clinicians, which needs to be accounted for. This comparative study showed thickening of DM in PCG eyes in vivo on ASOCT. DM is instrumental in maintaining the morphology and function of corneal endothelial cells under physiological conditions. Increased DM thickness in PCG eyes when compared to KC eyes, as seen in our study, may be attributed to the fibrotic proliferative response that ensues following DM tears during corneal stretching as has been postulated in previous study by Gupta et al.(4)

DM tears may present in association with transient corneal edema, corneal decompensation, or even with a clear cornea once the gap is sealed by regenerating endothelium. Tears in DM stimulate the endothelial cells to migrate and seal the gap through normal wound healing process. Chen et al, in a rabbit model of endothelial cell injuring vivo, showed that there was intracorneal fibrous tissue formation in the group that underwent DM stripping along with injury to the endothelial cells; this fibrotic response was absent in eyes with endothelial cell injury alone but with intact DM.

Most of the cells in the posterior fibrotic tissue did not originate from endothelium.(5,6) As DM ruptures, its edges have been shown to curl inside and entrap keratocytes in contact with the corneal stroma and advancing endothelial cells, stimulating a proliferative response eventually sealing the DM break with deposition of new material.(7) This DM thickening response to injury, as seen in PCG eyes, is not observed in eyes with keratoconus with DM tears. In our PCG cases, the DM was not wavy or irregular except in areas overlying the Haab striae where there were focal protuberances on ASOCT. However, the PDL could be appreciated as a thickened layer on both ASOCT and H&E staining.

Cherif et al had demonstrated that compression sutures placed in the deep posterior stroma, apparently to repair the PDL breaks in eyes with hydrops with keratoconus, may be an effective treatment for resolution of corneal edema. This is further supported by observations of Parker et al, who showed that in eyes with keratoconus, contrary to what is widely believed, DM breaks may not alone suffice for the development of acute hydrops and that additional discontinuity in PDL is imperative for this complication to set in. They showed that intraoperatively those eyes which underwent Bowman layer transplant in cases with advanced keratoconus developed corneal edema because of injury to PDL, unlike in eyes which underwent endothelial keratoplasty alone with intact PDL.

One of the limitations of our study was that our patient cohort had been operated in the past. Moreover, eyes in the histopathological arm of the study having undergone keratoplasty might represent a more severe form of corneal involvement compared with those which retained a clear cornea at the time of imaging. This was apparent from the greater DM thickness that we observed from the histopathological samples of keratoconus eyes in comparison to in vivo imaging. One of the strengths of our study is the comprisal of a large cohort of well characterized PCG eyes where children were old enough to undergo an ASOCT evaluation. Owing to the retinoblastoma controls being younger in age at the time of surgery, the age matching of cases and controls in the histopathological arm was not possible.

References:

1. Feder RS, Wilhelmus KR, Vold SD, O’Grady RB. Intrastromal clefts in keratoconus patients with hydrops. Am J Ophthalmol 1998;126(1):9–16.

2. Parker JS, Birbal RS, van Dijk K, et al. Are descemet membrane ruptures the root cause of corneal hydrops in keratoconic eyes? Am J Ophthalmol. 2019;205:147–152.
3. Yahia Cherif H, Gueudry J, Afriat M, et al. Efficacy and safety of pre-Descemet’s membrane sutures for the management of acute corneal hydrops in keratoconus. Br J Ophthalmol. 2015;6:773–777.
4. Gupta S, Chaurasia AK, Sen S, Bhardwaj M, Mandal S, Titiyal JS, Gupta V. The Descemet Membrane in Primary Congenital Glaucoma. Cornea. 2021 Feb 1;40(2):172-178.
5. Chen J, Li Z, Zhang L, et al. Descemet’s membrane supports corneal endothelial cell regeneration in rabbits. Sci Rep. 2017;1:6983.
6. Haab O. Atlas of Internal Diseases of the Eye, Along With a Layout of Your Pathology and Therapy. Munich, Germany: JF Lehmann; 1899:210.
7. Cibis GW, Tripathi RC. The differential diagnosis of Descemet’s tears (Haab’s striae) and posterior polymorpous dystrophy bands. A clinicopathologic study. Ophthalmology. 1982;6:614–620.