Endothelial keratoplasty in children

Paediatric keratoplasty possesses several distinct challenges in comparison to adults

Dermot McGrath

Posted: Thursday, December 3, 2020

Endothelial keratoplasty may offer several distinct advantages over penetrating keratoplasty for children with visually significant corneal clouding due to endothelial dysfunction, according to Muralidhar Ramappa MD.

“Compared to PK, the advantages of a technique such as Descemet’s stripping endothelial keratoplasty (DSEK) include rapid visual recovery with predictable refractive outcome, minimised suture related events, less stringent follow-up postoperatively and reduced steroid-related issues with the possibility of early tapering or weaning,” Dr Ramappa said at the World Ophthalmology Congress 2020 Virtual.

Dr Ramappa, a consultant ophthalmologist at the LV Prasad Eye Institute, Hyderabad, India, said that children with conditions such as congenital hereditary endothelial dystrophy (CHED) and other endothelial disorders could achieve excellent visual outcomes and realise a significant improvement in corneal transparency by undergoing DSEK or one of its variations.

“Our long-term clinical experience has shown that it is safe and effective and a superior surgical alternative to PK in children with endothelial disorders. However, some further study is justified to establish whether the use of superficial keratectomy with mitomycin C use may offer a substantial benefit in cases with dense anterior stromal opacification due to late presentation,” he said.

Dr Ramappa outlined the results of a retrospective study he carried out of 180 DSEAKs in 167 eyes of 111 children with corneal endothelial dysfunction treated with endothelial keratoplasty.

The median age at surgical intervention was 7.9 years (interquartile range IQR, 5.2 to 11.2 years), and the majority of indications were CHED (70%) followed by failed graft (12%), pseudophakic corneal oedema (7%) and xeroderma pigmentosa (6%). Cox proportional hazards regression analysis showed that the indication for DSAEK (p=0.007; hazards ratio: 2.17 ± 0.62), age at surgery (p= 0.02; hazards ratio: 0.87 ± 0.05) and any subsequent intervention following DSAEK (p=0.003; hazards ratio: 0.11 ± 0.08) were significant risk factors for DSAEK failure. Eyes with CHED had a significantly better graft survival than eyes with a prior failed PK.

Most of the cases (80%) were microkeratome assisted for donor preparation, with a push-through insertion technique used in 91%. “We have found this to be the most consistent approach and has shown the least amount of endothelial cell loss in our hands,” said Dr Ramappa. The ECD loss was 40.1% at six months, 45.4% at year one, 43.9% at year two, 51.6% at year three, 55.2% at year five and 61.9% at seven years from baseline.

The main complications included cataract formation, pupillary block, lenticule detachment and decentration and reversed lenticule. Primary graft failure was encountered in three cases, which Dr Ramappa surmised may have been related to the learning curve, and secondary glaucoma in five patients, which necessitated medical management. Repeat interventions included five for rebubbling, four repeat DSAEK, three PK surgeries, six for cataract, three glaucoma valve surgeries and one glaucoma filtering surgery.

The visual outcomes were excellent overall, said Dr Ramappa. The median BCVA improved from 1.45 ± 0.70 preoperatively to 0.90 ± 0.06 at the last follow-up visit (p<0.0001). The mean spherical equivalent refraction at the last visit had a median of 0.50D (IQR, -2.25 to +3.13).

“We report at a median post-interventional follow-up of 2.5 years (IQR, 0.9-3.4 years) 86.2% (144 of 167 eyes, 95% confidence interval 79.9% to 90.1%) maintained a clear graft.

“Factors affecting vision postoperatively included a late presentation, dense amblyopia, cataract and glaucoma. The cumulative overall long-term graft survival was 92.7%, 86.5%, and 77.7% at one, three and seven years respectively, which is very satisfactory,” he said. These long-term outcomes are much superior to in comparisons to conventional fill thickness corneal grafting.

The risk of allograft rejection and infection rates was significantly lower after an endothelial keratoplasty than conventional keratoplasty. Morphometric and densitometric analysis throughout the follow-up period showed an improvement of the graft-host thickness over time. These children were likely to be a bit hyperopic due to thicker graft host profiles.

Although DSEAK is more advantageous over conventional PK, it comes with specific intra and postoperatic challenges. The small, unstable anterior chamber, an underlying clear lens with poor visibility, makes it challenging to deal with intraoperatively and necessitates several modifications to avoid potential complications. Somewhat difficulty in strip thickened Descemet’s membrane, and there is potential risk of disturbing posterior stromal lamellae, thus results in a sub-optimal interface. Besides inadvertent lens or uveal trauma. Subsequent reinterventions again require general anaesthesia.

Modifications in surgical technique can help to overcome most of these issues, he said.

“For instance, the problem of poor visibility and difficulty in stripping Descemet’s membrane can be taken care of using chandelier illumination, trypan blue, and fine micro forceps,” he said.

Other potential issues to be aware of include insertion difficulties with the unfolding of the donor lenticule in the small anterior chamber and preventing trauma to the crystalline lens.

Postoperatively, supine positioning under sedation for at least 30 minutes helps minimise the lenticule dislocation or decentration. A detailed postoperative evaluation is not technically easier to perform and may need multiple examinations under anaesthesia to ensure the graft is well attached and discern the graft host dynamic well. Postoperative cataract and glaucoma management also require special attention; each of these additional subsequent steps can put the graft under huge risk.

Based on their extensive experience with corneal grafts and keratoplasty over many years, Dr Ramappa and his colleagues have developed a treatment algorithm to help decide each pediatric patient’s best course of action.

“The general rule of thumb is that where the corneal cloudiness is precluding the iris details due to significant anterior stromal alternations, it would be prudent to consider PK, and also to time the intervention as late as possible while caring amblyopia. If the case is mild to moderate and the iris details are seen well, then explore endothelial keratoplasty options and try to expedite to circumvent irreversible stromal changes that typically occur in late interventions. Thus, end up with a superior outcome with a minimal post haze or sub-Bowman’s folds.

Summing up, Dr Ramappa said that DSEK helps eliminate suture-related events and graft-host dehiscence and offers other advantages over PK in younger patients.

“It makes the eye more robust against injuries, delivers early refractive stabilisation as it is more predictable, and significantly, it minimises the risk of cataract formation, secondary glaucoma, and the need for repeat interventions. Furthermore, allograft rejection is at least 10 times less frequent and more reversible.

In conclusion, DSAEK provides superior long-term clinical outcomes with low complication rates suggesting that DSAEK is a safe and effective surgical alternative in children with corneal endothelial disease. Notably, accelerated postoperative visual recovery helps in better long-term functional outcomes by minimising the risk of amblyopia.

Muralidhar Ramappa: