The aim of CN is to transfer sensory innervation from a donor source nerve to the recipient denervated cornea, enabling donor axons and Schwann cells to populate and protect the ocular surface.

Although earlier trigeminal nerve reconstruction concepts had been previously described, the modern era of CN began in 2009 when Terzis et al described her approach. In their series of patients with combined facial nerve palsy and ipsilateral trigeminal nerve dysfunction causing corneal denervation, they employed direct surgical CN using the contralateral supraorbital (SON) and supratrochlear (STN) nerves as donor sources. The procedure involved a bicoronal incision to isolate the healthy donor nerves, which were then mobilized and tunneled subcutaneously to a small lid crease incision of the contralateral anesthetic eye. The nerve branches were then passed through the superior fornix and within the sub-Tenon space via small limbal conjunctival incisions. The nerve fascicles were secured with sutures at the perilimbal region. Although the technique required a cosmetically unappealing bicoronal scar and could only be used in unilateral disease with intact contralateral SON or STN sensation, Terzis established the foundation for subsequent technical refinements.

The introduction of minimally invasive CN (MICN) by Elbaz et al addressed two significant constraints of the Terzis technique: the need for intact ipsilateral SON or STN sensation, and limitation to unilateral disease. By interposing a sural nerve graft (SNG) between the STN nerve and the anesthetic cornea—accessed through a transverse incision over the medial upper eyelid—MICN offered improved cosmesis through much smaller incisions and applicability to a broader range of patients, including those with bilateral corneal anesthesia. Further modifications have included sutureless techniques using fibrin glue to secure nerve fascicles within sclerocorneal tunnel incisions at the limbus.

The selection of an appropriate donor nerve is fundamental to surgical success. Surgeons should perform nerve transfers using donor sensory nerves with robust axon counts that ideally carry protective pain sensation. Understanding the sensory distribution of potential donor nerves guides proper selection. The STN nerve provides sensation to the medial forehead, nasal bridge, and upper eyelid. The SON nerve supplies the lateral forehead, upper eyelid, and anterior scalp. The infraorbital (ION) nerve innervates the lateral nose, cheek, upper lip, and upper dental arch. When none of these trigeminal branches are available, sensation supplied by the greater auricular nerve (GAN), arising from cervical nerve roots C2 and C3, offers an extratrigeminal alternative. Preoperative sensation testing for this nerve can be confirmed by earlobe pinch testing, particularly in infants.

The algorithm for donor nerve selection follows a hierarchical approach based on the pattern of sensory loss ( Figure ). In patients with intact ipsilateral trigeminal sensation, the ophthalmic nerve of trigeminal (V1) or maxillary branch of trigeminal (V2) branches are an option for direct nerve transfer. When possible, ipsilateral donor nerves are favored due to their proximity to the target cornea, with contralateral nerves considered when the ipsilateral option is unsuitable or unavailable. In cases where ipsilateral sensation is absent but contralateral trigeminal sensation remains intact, cross-face V1 nerve transfers are used, extended with nerve grafts. GAN is usually available for patients lacking trigeminal nerve sensation. Contraindications to CN may include active ocular infections and active epithelial defects or corneal melt, all of which should be treated prior to CN. Abnormal sensation in the donor nerve dermatome is an absolute contraindication. Patients with prior strabismus, vitreoretinal, glaucoma, or other procedures involving the conjunctiva may have extensive conjunctival scarring. This may create a technical challenge to achieving a reliable full-thickness vascularized coverage of the nerve grafts and warrants careful consideration. Similarly, future surgery involving the conjunctiva may affect the donor nerves.

Corneal neurotization treatment algorithm.

Direct CN (DCN) involves the transfer of a donor sensory nerve to the perilimbal region without the use of an interpositional nerve graft. For patients with isolated injury to the long ciliary nerves, such as herpes simplex keratitis (HSK), multiple ocular surgeries or pan-retinal photocoagulation, the ipsilateral STN and/or SON nerves can be transferred. The shorter distance in ipsilateral DCN, compared to the contralateral approach (which requires a 7-8 cm nerve segment) facilitates faster reinnervation. Operative time is shorter at 60 to 120 minutes vs 3 to 4 hours for nerve grafting, and eliminates sural nerve donor site morbidity. When available, the ipsilateral ION represents another option for either direct or indirect CN (ICN).

ICN uses a nerve graft to conduct axon growth from the donor nerve to the cornea. Coaptation of donor graft to the donor sensory nerve can be performed end-to-end, maximizing axonal load at the cost of donor dermatome sensation, or end-to-side, preserving more donor sensation but diverting fewer axons to the cornea. Elbaz et al described coapting an SNG to the source contralateral STN in two cases and a bilateral case using both STN nerves. A transverse incision over the medial eyelid left a cosmetically acceptable scar. Sensation in the forehead was preserved, and bilateral cases were treated. ICN, via nerve grafting, uses more proximal donor nerve segments, which exhibit a higher axon density and may therefore support greater corneal reinnervation and enhanced sensory recovery. The sensory deficit after SNG is limited to the dorsolateral foot and the GAN to the earlobe, which are clinically acceptable to most patients. However, autografts incur longer operating time, additional surgical expertise beyond ophthalmology, donor site morbidity and longer time to recovery of innervation.

When the ION is available, it can be used in ICN by unroofing the ION canal. V2 hypoesthesia following ION harvest is generally less well-tolerated compared to the forehead numbness associated with V1 transfers. To minimize this potential morbidity, the authors advocate performing an epineural window in the side of ION, dividing only a minority of ION fibers during the nerve repair.

Fogagnolo et al compared DCN and ICN in a nonrandomized multicenter interventional study of 26 eyes. Both produced comparable epithelial healing, corneal sensitivity, and subbasal nerve regeneration at 1 year, suggesting that donor anatomy and surgical indication may be more important than intrinsic superiority of either technique.

The SNG has been the preferred autograft due to its ease of harvest, long length, and minimal donor site morbidity, with sensory deficit limited to the dorsolateral foot. The GAN was reported as a nerve graft in one case, with the advantage of a single operating field and a mean diameter of 2.6 mm—only slightly larger than the STN—compared to 3.6 mm for the sural nerve. Confocal microscopy demonstrated nerve regeneration at 3 months, with Cochet–Bonnet esthesiometry (CBA) of 10 mm first recorded at 9 months; vision and corneal appearance remained unchanged at 1 year. Albavera et al used the superficial peroneal nerve in 4 cases, but this nerve carries a potential motor deficit of foot eversion, so is not recommended. Bourcier et al reported lateral antebrachial cutaneous nerve use in a patient with HSV-related NK, achieving improvement in corneal sensation from <5 to 40 mm and best-corrected visual acuity (BCVA) from 20/200 to 20/80 at 12 months. The lateral antebrachial cutaneous nerve offers a caliber comparable to the SON nerve, with donor site numbness typically resolving by 6 months.

Use of an acellular nerve allograft (ANA) in CN eliminates donor site morbidity and reduces operative time. The ANA measures 70 mm × 1 to 3 mm, which may limit its use in patients with widely-spaced orbits or in those with more distal donor nerves as undesirable tension on the coaptation site may limit nerve regeneration and corneal sensory outcomes. Practical limitations include cost (particularly in outpatient surgical centers), freezer storage requirements, and restricted availability. Avance processed nerve graft (Axogen Inc) is produced by processing human nerve tissue with a combination of detergent processing, gamma irradiation, and enzymatic removal of chondroitin sulfate proteoglycan to minimize structural damage while permitting axonal regeneration.

One limitation of ANAs is their lack of Schwann cells, which are essential for axon regeneration. Repopulation depends on host Schwann cell migration into the graft, a process subject to progressive quiescence that may restrict effective graft length. One cohort of 17 ANA patients demonstrated improvement in CBA from 3.6 to 44.2 mm ( P <.01), with a mean time to first sensory recovery of 3.7 months. All patients with chronic epithelial defects achieved resolution at 6 months. No randomized controlled trials comparing autograft to ANA exist.

After CN, younger age and congenital etiology are associated with more favorable corneal sensation outcomes compared to adults. Sensation outcomes may be attributed to younger eyes having greater baseline corneal sensation and higher subbasal nerve fiber density. ^, However, visual outcomes in children may be poorer compared to adults and tend to deteriorate over time due to corneal scarring, neovascularization, recurrent microbial keratitis, and amblyopia. This underscores the severity of the disease trajectory in this population and highlights that while children may demonstrate greater relative improvements following CN than adults, visual outcomes may remain guarded, with further procedures such as corneal transplantation required.

NK occurs in approximately 6% of HSK cases, with rates up to 27% reported, making it the predominant cause of NK. CN yields favorable outcomes in this population, with visual acuity improvement in the majority of cases, alongside significant improvements in Mackie staging and epithelial defect resolution rates of up to 78%. ^, Corneal sensation improvements of 16 to 36 mm ( P =.028) have been demonstrated, with earlier intervention associated with greater sensory recovery. Ocular hypertension was a notable postoperative complication, reported in up to 41.9% of cases, and likely attributable to prolonged steroid use for keratouveitis management. Recurrence of HSK has been reported in 2 cases post-CN manifesting as disciform keratitis and PED respectively, both resolving with antiviral and anti-inflammatory treatment. The authors feel that antiviral prophylaxis is vital to long term surgical success pre- and postoperatively with the literature quoting valacyclovir dosing 3 g/d weaning down to 0.5 g/d postoperatively and in one study additional serial PCRs postoperatively. ^, We are unaware of evidence of transmission of HSV to the contralateral ganglion post-CN.