Surgical taming of neurotrophic keratopathy: From amniotic membrane grafting to neurotisation

Ophthalmology Times EuropeOphthalmology Times Europe March 2024
Volume 20
Issue 2
Pages: 24 - 25

Clara Chan, MD, shares surgical pearls for neurotrophic keratopathy and explores new pathways through neurotisation

The Helix Nebula, also known as the God's Eye Nebula. Image credit: ©Supernova –

Surgical interventions create many treatment possibilities for patients with neurotrophic keratoapthy. Image credit: ©Supernova –

Neurotrophic keratopathy (NK) is best treated early, but when that is not the case there are several surgical approaches that can be used. Clara Chan, MD, from the University of Toronto, Ontario, Canada, noted the importance of early recognition and aggressive medical treatment of NK to avoid surgery, and maintaining the integrity of the corneal epithelial surface following healing.

A surgical overview

Chan and colleagues evaluated a series of patients with NK from a tertiary care clinic to ascertain the contributory factors to the disease. They found that 54% had had herpes simplex virus (HSV) or herpes zoster virus, 24% had neurosurgical causes and 21% had had retinal surgery;
ocular surface disorders characterised by extreme dry eye disease were a contributing factor among others. Most patients needed a surgical procedure (ie, tarsorrhaphy, amniotic membrane grafts, conjunctival flaps, or tectonic or optical keratoplasty). The visual prognosis was guarded and healing occurred over an extended period (85 days); 20% had a perforation, 2 eyes were eviscerated, and only 60% maintained their best-corrected visual acuity (BCVA).

When medical treatments failed and stage 2 NK was present, aggressive surgery was needed. In Canada, cenegermin, a nerve growth factor, is unavailable. Presentation with or progression to stage 2 requires immediate, aggressive attention.

In patients with stage 2 NK, the ocular surface still must be optimised, infection and inflammation controlled and nutritional support provided. The surface can be optimised with instillation of nonpreserved lubricating drops and discontinuation of toxic agents. Chan reported that 65% of patients achieved healing of the ocular surface without surgery when using cenegermin in US FDA studies. In the presence of stage 3 NK, surgery is critical because melting may have occurred, and the tissue is thinning and possibly perforated.

Surgical options for NK

The surgical pearls for amniotic membrane grafting include debriding the uneven necrotic edges of the epithelial defect and avoiding the use of toothed forceps when handling the amniotic tissue. When positioning the tissue, the stromal side (ie, the sticky side), if placed down, can facilitate integration of the amnion into the corneal stroma. Chan said she prefers to place the basement membrane side (ie, the nonsticky side) down because it acts as a patch graft, dissolves faster and the corneal epithelium grows underneath. Amniotic membrane grafting is generally performed before corneal scarring develops to facilitate rapid healing of the defect.

In a retrospective study covering an 8-year period, 335 patients (354 eyes) out of 305,351 had NK; HSV was the most frequent etiology. Amniotic membrane grafting was used in patients with stage 2 disease with a success rate of 57.2% and a mean healing time of 15 days. In stage 3, the success rate was 63.6% with a mean healing time of 16 days. The risk factors for a worse final corrected distance visual acuity were advanced age, advanced NK stage and decreased corrected distance visual acuity at presentation.

Use of dehydrated amniotic membrane is another option for treating persistent epithelial defects. In a pilot study of this tissue,1 Chan et al found that the amnion was absorbed within 2 weeks in all patients, and the corneal defect resolved in 89%, with a mean time to resolution of 17.8 days. The patient follow-up times ranged from 90 to 265 days and BCVA improved from 10/174 to 20/47, which reached significance (P = .036), Chan reported.

Tarsorrhaphy is another surgical option, though patients are reluctant to accept it. The procedure offers the best rate of defect resolution.

Chan recounted a study2 that followed 77 patients for 5 years. Of these, 24 patients had a temporary tarsorrhaphy and the rest a permanent one. The results showed that after tarsorrhaphy, the defect resolved fully in 91%. Time to healing was a mean of 18 days. Tape tarsorrhaphy may be more acceptable to patients.

The glue patch technique (tectonic patching) using cyanoacrylate glue is another treatment option for corneal perforations or leaks. This procedure requires about 2 to 3 mm of epithelium-free area for the glue to stick to the cornea surface and plug the perforation. Sliding bolster temporary tarsorrhaphy technique resembles a pulley system to allow the eye to be opened for examination and closed afterward. The procedure uses double-armed 6-0 nylon that is passed through the lower and upper eyelid margins with bolsters made of 25-gauge butterfly tubing cut to size.

Use of a mattress suture facilitates creation of a temporary lateral tarsorrhaphy that can be performed at the bedside. It requires a spatulated double-armed 5-0 Prolene or nylon suture passed through the tarsus at the grey line and will stay tight for about 6 months to 1 year. Chan believes this is the most efficient way to perform a tarsorrhaphy. Injection of about 15 units of botulinum toxin (Botox) into the superior orbital rim is a simple surgical method of achieving a continuously drooping eyelid that lasts up to 3 months. A downside is that the Bell reflex can be lost if the superior rectus muscle is inadvertently hit.

Conjunctival flaps can be created that are total (Gundersen flaps) or partial. In a study of shingles and HSV that caused NK, the main issue was that the flap can retract; this can be addressed by regluing/resuturing the flap. When creating Gundersen flaps, all corneal epithelium must be removed. The limbus is cauterised to achieve a smooth surface for the conjunctiva. The flap should be a few millimeters larger than needed, as tissue tends to shrink. The flap is secured with 9-0 Vicryl at the inferior margin and glue at the base. Chan advised removing the Tenon capsule, which causes retraction.

Surgical pearls for NK

Chan listed the key takeaways:

  • Surgical treatments are used for stage 2 and 3 NK to avoid melting/perforation.
  • Almost everything can be glued.
  • Tape tarsorrhaphy gives the patient control.
  • Amniotic grafts are best before development of stromal scarring (~60% success).
  • If grafting fails, tarsorrhaphy works well (~90% success).
  • Tarsorrhaphy and tectonic grafting “go together like a horse and carriage.”
  • Send the recipient button for HSV testing to rule out subclinical active disease.
  • Gundersen flaps are the least intensive option for eyes with poor vision potential.

Neurotisation: A new approach

Neurotisation is defined as restoration of corneal sensation with regional nerve transfers and nerve grafts. The beauty of neurotisation is that it restores corneal nerve sensation. The sural nerve in the calf can be harvested and connected from the patient’s contralateral eye to the eye with NK. Once the nerve is embedded into the cornea in a few insertion sites in each corneal quadrant, a corneal transplantation procedure can be performed, Chan explained. The procedure was developed at The Hospital for Sick Children, Toronto.

“This procedure prevents scarring, allows corneal transplant grafting to be performed, and amblyopia is minimised,” she said. In addition to the absence of corneal sensation, patients cannot blink and this leads to substantial corneal exposure and risk of perforation. One neurotisation option is using the contralateral supratrochlear and supraorbital nerve, which was reported by Elbaz et al.3 Roberto Pineda, MD, and Rong Guo, MS, from Massachusetts Eye and Ear and Harvard Medical School, both in Boston, Massachusetts, also developed a neurotisation procedure using the great auricular nerve. The advantages of using this nerve are the availability of more axons, ipsilateral sensation with minimal donor site morbidity and avoidance of facial incisions.


1. Mimouni M, Trinh T, Sorkin N, et al. Sutureless dehydrated amniotic membrane for persistent epithelial defects. Eur J Ophthalmol. 2022;32(2):875-879.doi:10.1177/11206721211011354
2. Cosar CB, Cohen EJ, Rapuano CJ, et al.Tarsorrhaphy: clinical experience from a cornea practice. Cornea. 2001;20(8):787-791. doi:10.1097/00003226-200111000-00002
A photo of Clara Chan, MD
3. Elbaz U, Bains R, Zuker RM, Borschel GH, Ali A. Restoration of corneal sensation with regional nerve transfers and nerve grafts: a new approach to a difficult problem. JAMA Ophthalmol. 2014;132(11):1289-1295. doi:10.1001/jamaophthalmol.2014.2316

Clara Chan, MD | E:

Chan is with the University of Toronto and has no financial interest in this subject matter.

Related Videos
Josefina Botta, MD, MSc, at ASCRS 2024
Dr Nir Shoham Hazon, Director, Miramichi EyeNB Centre of Excellence, New Brunswick, Canada
J. Morgan Micheletti, MD, speaks at the 2024 ASCRS meeting
Dr William Wiley of Cleveland Eye Clinic, Northeast Ohio
© 2024 MJH Life Sciences

All rights reserved.