Interventional procedure overview of nerve graft for corneal denervation
Closed for comments This consultation ended on at Request commenting lead permission
Summary of key evidence on autograft: sural nerve graft with orbital nerves as donor for corneal denervation
Study 1 Capatano J (2019)
Study type | Single arm, single centre, non-randomised, before-and-after study |
|---|---|
Country | Canada |
Recruitment period | 2012 to 2017 |
Study population and number | n=19 eyes in 16 people All people who had corneal neurotisation for neurotrophic keratitis at a hospital for children and young people. |
Age and sex | Mean 12.5 years (SD 8.3 years) with 88% of people aged 18 years or younger; 56% female |
Patient selection criteria | Inclusion criteria: all people who had corneal neurotisation for neurotrophic keratitis at a hospital for children and young people. Further details of patient selection are not described. |
Technique | Sural nerve graft coapted to the ipsilateral or contralateral supratrochlear, supraorbital, or infraorbital nerves. In 14 people, the graft was coapted to the supratrochlear nerve, in 1 person to the supraorbital, and in 1 person to the infraorbital. Coaptation was performed using sutures and/or fibrin glue. In the first 7 eyes (6 people), the fascicles of the sural graft were laid in the perilimbal subconjunctival space. In the following 12 eyes, the ends of nerve fascicles were inserted into the peripheral corneal stroma via a partial-thickness corneoscleral tunnel incision. The authors state that this was to replicate the normal innervation of the cornea and shorten the distance between the donor nerve end and the cornea. After surgery, a temporal permanent tarsorrhaphy was constructed. A central temporary tarsorrhaphy was also placed for 1 week. |
Follow up | Mean 24.0 months (SD 16.1 months) |
Conflict of interest/source of funding | Conflict of interest: the authors declared no conflict of interest. Source of funding: Supported in part by grants from the Plastic Surgery Foundation and American Society of Peripheral Nerve; Moorfields Eye Charity; and HCA International Foundation. |
Analysis
Follow up issues: Reliable testing of corneal sensation was not possible in 1 person due to young age and developmental delay, and this person was excluded from corneal sensory analysis.
Study design issues: This single arm, single centre, non-randomised, before-and-after study evaluated the outcomes of people with neurotrophic keratitis treated with corneal neurotisation. The primary outcome was central corneal sensation, as measured by Cochet-Bonnet aesthesiometry with sensation ranging from 0 mm (absent sensation) to 60 mm (full sensation). Other outcomes included BCVA (reported using LogMAR), episodes of persistent epithelial defect (PED, defined as non-infected corneal ulceration that failed to resolve within 2 weeks of commencing treatment), adjunctive medical treatments and any side effects or complications from MICN were documented.
Differences in continuous variables were assessed with Wilcoxon signed-rank test, while categorical variables were assessed with Fisher's exact test. A p value <0.05 was considered statistically significant. The authors do not report adjustment for multiple comparisons.
Study population issues: Three eyes had Mackie stage 1 neurotrophic keratitis, 7 eyes had grade 2, and 9 eyes had grade 3. The most common aetiology of neurotrophic keratitis was congenital corneal anaesthesia (14 eyes, 74%). Amblyopia was present in 5 eyes (26%), facial nerve palsy in 5 (26%), cataracts in 2 (11%), glaucoma in 1 (5%), and microphthalmia in 1 (5%).
Key efficacy findings
Number of people analysed: 16 people, 19 eyes
Follow up at time of assessment: mean 24.0 months (SD 16.1 months)
Central corneal sensation: Statistically significant increase from a mean of 0.8 mm (SD 2.5 mm) before surgery to 49.7 mm (SD 15.5 mm) after surgery (p<0.0001; Table).
At 3 months after surgery, central corneal sensation in the perilimbal subconjunctival group was 2.5 mm (SD 6.1 mm), and 19.4 mm (SD 23.1 mm) in the corneoscleral tunnel group (p=0.08). Central corneal sensation at other time points was comparable between the groups.
BCVA: No change in BCVA from before surgery to after surgery (Table). The authors note that this is 'due to the presence of corneal stromal scarring in all the eyes in our cohort'.
History of PED: Statistically significant decrease in the number of people with PEDs from 17 (89%) before surgery to 4 (21%) after surgery (p<0.0001; Table).
Neurotrophic keratitis grade: Before surgery, 3 eyes had Mackie stage 1 neurotrophic keratitis, 7 eyes had grade 2, and 9 eyes had grade 3. After surgery 15 eyes had Mackie stage 1 neurotrophic keratitis, 4 eyes had grade 2, and 9 eyes had grade 3 (Table).
Outcome | Before surgery | After surgery | p value |
CCS (mm) | |||
Mean (± SD) | 0.8 ± 2.5 | 49.7 ± 15.5 | <0.0001 |
Median (range) | 0 (0 to 0) | 60 (40 to 60) | <0.0001 |
BCVA (LogMAR) | |||
Mean (± SD) | 1.01 ± 0.68 | 0.89 ± 0.73 | Not significant |
Median (range) | 0.7 (0.3 to 2.28) | 0.7 (0.0 to 2.28) | Not significant |
History of PED | |||
Total vs. after surgery | 17 (89%) | 4 (21%) | <0.0001 |
1 year prior vs. after surgery | 10 (53%) | 4 (21%) | Not reported |
NK grade (eyes) | |||
Grade 1 | 3 (16%) | 15 (79%) | Not reported |
Grade 2 | 7 (37%) | 4 (21%) | Not reported |
Grade 3 | 9 (47%) | 0 | Not reported |
Abbreviations: BCVA, best-corrected visual acuity; CCS, central corneal sensation; LogMAR, logarithm of the minimum angle of resolution; NK, neurotrophic keratitis; PED, persistent epithelial defect; SD, standard deviation.
Key safety findings
Number of people analysed: 16 people, 19 eyes
Follow up at time of assessment: mean 24.0 months (SD 16.1 months)
There were no intraoperative complications.
After surgery complications:
Persistent epithelial defects, n=5 episodes in 4 eyes of 4 people.
All occurred over 24 months after neurotisation, 3 were following corneal transplant.
All resolved within 4 weeks after presentation with antibiotics with or without a bandage contact lens.
Suture exposure, n=5
Did not occur after switching to absorbable polyglactin sutures.
Study 2 Fogagnolo P (2020)
Study type | Multicentre, non-randomised controlled trial |
|---|---|
Country | Italy |
Recruitment period | 2014 to 2019 |
Study population and number | n=26 eyes in 25 people; 10 eyes by indirect neurotisation (with sural nerve graft), 16 eyes by direct neurotisation Adults with neurotropic keratitis |
Age and sex | Mean 45.4 years; 80% female |
Patient selection criteria | Inclusion criterion: chronic neurotropic keratitis (reduced or abolished corneal sensation measured by Cochet-Bonnet aesthesiometer with a duration time from the onset of >3 months) because of central nervous denervation despite conventional treatment. Exclusion criteria: presence of any active corneal disease other than neurotropic keratitis and diagnosis of polyneuropathy or other types of disorder affecting the peripheral nervous system. |
Technique | Indirect neurotisation or direct neurotisation was chosen according to the person's clinical characteristics and preferences. For indirect neurotisation, a sural nerve graft was coapted end-to-end to the supratrochlear and/or supraorbital nerves. The nerve graft was tunnelled subconjunctivally to the perilimbal area of the cornea. A scleral-corneal tunnel for each fascicle was made into the anterior corneal stroma to help nerve growth toward the centre of the cornea. The nerves were then fixed in the desired position with fibrin glue, and the conjunctiva was repaired with 8-0 Vicryl suture. |
Follow up | Mean 18.76 months; data at 1 year follow up was used for the main analysis |
Conflict of interest/source of funding | Conflict of interest: the authors report 'no proprietary or commercial interest in any materials discussed in the article'. Source of funding: supported by a grant from the Italian Society of Ophthalmology. |
Analysis
Follow up issues: Of people who had indirect neurotisation, 2 were followed up for less than 12 months.
Study design issues: This multicentre, non-randomised controlled trial assessed the comparative efficacy and safety of indirect corneal neurotisation (using a sural nerve graft) versus direct corneal neurotisation in people with neurotrophic keratitis. Consecutive people with neurotropic keratitis who attended the corneal service of the 3 centres were screened for eligibility. Outcomes included neurotrophic keratitis healing, corneal sensation (measured by a Cochet-Bonnet aesthesiometer), BCVA (reported as a decimal), and in vivo confocal microscopy.
The Wilcoxon test was used to compare the continuous variables before surgery and at 1 year follow up in overall people and separately in the 2 groups. The Chi-squared test was used to compare the proportion of people with neurotrophic keratitis in severity stages 1, 2, and 3 in the 2 groups. The Mann-Whitney U test was used to compare the changes in continuous variables between the 2 groups. A p value <0.05 was considered statistically significant. The authors do not report adjustment for multiple comparisons.
Study population issues: There were statistically significant differences in the baseline values of BCVA and neurotrophic keratitis severity stage between the 2 groups. Decimal BCVA was significantly lower in the direct neurotisation group (0.19 [SD 0.23] vs. 0.42 [SD 0.28]; p=0.044). Mackie stage was significantly higher in the direct neurotisation group (0% vs. 40% had stage 1, 31.25% vs 40.00% had stage 2, and 68.75% vs 20.00% had stage 3; p=0.009).
Key efficacy findings
Neurotrophic keratitis healing
Number of people analysed: 10 eyes in 10 people
Follow up at time of assessment: 1 year
Neurotrophic keratitis healed in all people after a mean period of 3.9 (SD 1.5 months; range: 2 to 6 months) (healing rate 100%). There was no statistically significant difference in the healing time of the 2 techniques.
In people treated with indirect corneal neurotisation, there was a statistically significant decrease in the area of the epithelial defect from 12.40 mm2 before surgery to 0.10 mm2 after surgery (p=0.006). There was no statistically significant difference in the change in epithelial defect area between the 2 techniques (p=0.120).
Corneal sensation
Number of people analysed: direct neurotisation = 16 eyes in 16 people; indirect neurotisation = 10 eyes in 10 people
Follow up at time of assessment: 1 year
In people treated with indirect neurotisation, there was a statistically significant increase in corneal sensation from 2.5 mm (SD 5.4 mm) to 22.5 mm (SD 18.3 mm; p=0.002). Note that there are discrepancies in the data provided in the text (2.5 mm to 22.5 mm) and the table below (2.5 mm to 17.5 mm). It is not clear from the publication whether the differences in data represents different analysis sets.
At 3 and 6 months follow up, people treated with direct neurotisation had a statistically significantly higher change from baseline corneal sensation than people treated with indirect neurotisation (p=0.042 and p=0.048, respectively). This difference was not seen at 1 year follow up (p=0.579).
Visit (months) | DCN Group, mean mm ± SD (range) | ICN Group, mean mm ± SD (range) | p-value, DCN vs. ICN as compared to baseline |
Baseline | 4.0 ± 8.9 (0 to 30) | 2.5 ± 5.3 (0 to 15) | 0.867 |
After 1 | 6.5 ±11.6 (0 to 40) | 5.0 ± 10.1 (0 to 20) | 0.785 |
After 3 | 15.2 ± 20.3 (0 to 60) | 6.5 ± 7.5 (0 to 20) | 0.042 |
After 6 | 19.8 ± 17.1 (0 to 60) | 9.3 ± 19.1 (0 to 40) | 0.048 |
After 9 | 23.0 ± 25.1 (0 to 60) | 16.2 ± 22.9 (0 to 45) | 0.432 |
After 12 | 22.3 ± 20.4 (0 to 60) | 17.5 ± 17.3 (0 to 45) | 0.579 |
Abbreviations: DCN, direct corneal neurotisation; ICN, indirect corneal neurotisation; SD, standard deviation.
Visual acuity
Number of people analysed: direct neurotisation = 16 eyes in 16 people; indirect neurotisation = 10 eyes in 10 people
Follow up at time of assessment: 1 year
In people treated with indirect corneal neurotisation, there was a numerical increase in decimal BCVA from 0.42 (SD 0.23) before surgery to 0.48 (SD 0.27) after surgery. This was not statistically significant (p=0.054). There was no statistically significant difference in the improvement of BCVA between the 2 techniques (p=0.089), however, there was a statistically significant difference in the before versus after surgery BCVA of people treated with direct neurotisation (p=0.004).
In vivo confocal microscopy findings
Number of people analysed: 26 eyes in 25 people
Follow up at time of assessment: 1 year
In all but 1 person, corneal sub-basal nerve plexus was not detectable before surgery. After surgery, all people had new nerve fibres appear which gradually formed a regenerated corneal sub-basal nerve plexus that reached near-normal features 1 year after surgery.
Mean corneal nerve fibre length increased from 1.8 mm/mm2 (SD 0.15 mm/mm2) before surgery to 14.67 mm/mm2 (SD 7.92 mm/mm2) after surgery. The change in corneal nerve fibre length from before to after surgery did not differ significantly between the people who had direct or indirect neurotisation (p=0.833).
Key safety findings
Number of people analysed: 10 eyes in 10 people
Follow up at time of assessment: Up to 21 months
In people who had indirect neurotisation, the following safety findings were reported:
Oedema of the upper third of the face, n=10
Partial numbness of the frontal region on the leg where the sural nerve was harvested, n=10
This deficit of sensation gradually reduced in size and intensity within the first year after surgery.
In all people who regained corneal sensation (with either direct or indirect neurotisation), the following safety findings were reported:
Misperception of the corneal tactile stimulation in the contralateral forehead.
This complication occurred in the first 3 to 6 months after surgery, regardless of the technique employed. Then, the sensation shifted from the forehead to the cornea about 6 to 9 months after surgery.
Study 3 Elalfy M (2021)
Study type | Single arm, single centre, non-randomised, prospective, consecutive case series |
|---|---|
Country | UK |
Recruitment period | 2016 to 2018 |
Study population and number | n=11 eyes in 11 people People with neurotrophic keratitis undergoing corneal neurotisation surgery |
Age and sex | Median 43 years; 73% Female |
Patient selection criteria | Inclusion criteria: All people included in the study had episodes of recurrent persistent epithelial defects or corneal ulceration aggressively treated with conventional therapy including temporary tarsorrhaphy, with corneal neurotisation procedure performed once their ocular surface was stabilised. All people had 0 corneal sensation on Cochet–Bonnet aesthesiometer before surgery and full corneal sensation in the contralateral eye. |
Technique | Sural nerve graft, end-to-end coapted to the ipsilateral or contralateral supratrochlear or supraorbital nerve (or both). Four to five sural nerve fascicles were each inserted into corneoscleral lamella tunnels around the entire limbal circumference, with 2 further fascicles laid in the perilimbal area. Coaptation was performed using 10-0 Nylon sutures and fibrin glue (Tisseel). In 3 cases amniotic membrane was also wrapped around the coaptation. |
Follow up | Mean 14.5 months |
Conflict of interest/source of funding | Conflict of interest: The authors declared that no financial or non-financial conflicting interests exist for any author. Source of funding: The authors report no source of funding. |
Analysis
Follow up issues: 6 people did not complete in vivo confocal microscopy. Eight people completed the pre-neurotisation questionnaire, and 6 people completed both pre- and post-neurotisation questionnaire.
Study design issues: This single arm, single centre, non-randomised, prospective, consecutive case series reported the functional, anatomical, and subjective outcomes of people with neurotrophic keratitis who had corneal neurotisation surgery. Functional outcomes included bilateral assessment of visual acuity; slit-lamp examination of corneal and conjunctival staining using fluorescein 1% eye drops (increased staining is indicative of worsening keratitis); tear production (Schirmer's 1 test); tear film break-up time; tear film meniscus height, quality and osmolarity; central corneal thickness; and measurement of corneal sensation using Cochet–Bonnet aesthesiometer. Structural outcomes included changes in corneal nerve density and morphology by in vivo confocal microscopy. Subjective outcomes were collected using the NEI VFQ-25. This questionnaire includes items on general health, vision, difficulties with activities (near, distance, driving) and responses to vision problems (levels of disability, discomfort, and psychosocial impact). Further subjective outcomes included frequency of eye lubricants, visits to the hospital up to 6 months after surgery and the frequency of corneal ulcers 2 years before and up to 6 months after surgery.
No statistical analysis was performed.
Study population issues: Person 4 had corneal neovascularisation, person 2 had corneal scarring and people 7, 8 and 11 had superficial corneal scarring preoperatively. No other ocular comorbidities that could affect the final visual outcome were detected in any of the people.
Key efficacy findings
Functional outcomes
Number of people analysed: 11 eyes in 11 people
Follow up at time of assessment: mean 14.5 months
Vision (Snellen acuity): improved in 6 people, stabilised in 3 people, and deteriorated in 2 people at the last follow up compared to before surgery.
Corneal and conjunctival staining: decreased in 10 people, no change in 1 person.
Tear film breakup time: increased in 10 people, decreased in 1 person.
Tear meniscus height: increased 7 people, no change in 4 people.
Schirmer's test(tear production): increased in 4 people, no change in 7 people.
Tear film osmolarity: decreased 7 people, no change in 2 people, increased in 2 people.
Central corneal thickness: increased in 9 people, no change in 1 person, decreased in 1 person.
Corneal sensation improved in 9 people, no change in 2 people.
Late recovery of corneal sensation was noted in 1 person, from 0 at 6 and 12 months after corneal neurotisation, to 10 mm at 24 months.
Structural outcomes
Number of people analysed: 5 eyes in 5 people
Follow up at time of assessment: 9 to 36 months
Apart from 1 person, nerve parameters including nerve fibre density, nerve branch density, nerve fibre length, and total branch density generally increased over the follow up period.
Subjective outcomes
Number of people analysed: 11 eyes in 11 people
Follow up at time of assessment: not reported
Use of lubricants: reduction in lubricant use after surgery in 3 people; no change in 8 people.
Frequency of hospital visits: reduced after surgery in 5 people; no change in 5 people; increased after surgery in 1 person.
Corneal ulcers: No person developed any further corneal ulcers following the procedure.
Number of people analysed: 8 (pre-neurotisation questionnaire); 6 (pre- and post-neurotisation questionnaire).
Follow up at time of assessment: not reported
NEI VFQ-25results
Overall vision: 1 person reported that vision remained poor after surgery, 5 people reported fair or good vision.
Ocular discomfort: all had moderate ocular discomfort before surgery. Four people improved to mild discomfort, 3 people continued to experience moderate discomfort. The source of this data is unclear, as only 6 people were reported to have completed both before and after surgery questionnaires.
Reading: improved from extremely or moderately difficult to no or mild difficulty in 4 people, no change in 2 people.
Distance vision: 4 people reported no or mild difficulty with distance vision after surgery. The before surgery outcomes are not reported.
Limitation of activities: all people reported limitation of activity most of the time before surgery, all reported none or some limitation of activity after surgery.
Psychosocial impact of the disease: all people reported psychosocial impact of the disease most of the time before surgery, all reported none or some psychosocial impact of the disease after surgery.
Key safety findings
No during or after surgery complications were observed in any people at either donor or recipient sites.
Study 4 Weis E (2018)
Study type | Single arm, single centre, non-randomised, prospective, consecutive case series |
|---|---|
Country | Canada |
Recruitment period | Not reported |
Study population and number | n=6 eyes in 6 people Adults with neurotrophic keratitis |
Age and sex | Mean 57 years (SD 19 years); 83.3% male |
Patient selection criteria | Inclusion criteria: consecutive people with neurotrophic keratitis who had corneal neurotisation. |
Technique | Sural nerve graft coapted to the ipsilateral or contralateral supratrochlear or supraorbital nerves. Coaptation was performed either end-to-end or end-to-side, using 10-0 Nylon sutures and fibrin sealant. Each nerve fascicle was sutured to the sclera at the limbus with 9-0 Vicryl sutures. After surgery, a temporary suture tarsorrhaphy was performed on the affected anaesthetic eye. |
Follow up | Mean 12 months |
Conflict of interest/source of funding | Conflict of interest: the authors declared 'no proprietary or commercial interest in any materials discussed in this article.' Source of funding: not reported. |
Analysis
Follow up issues: follow up ranged from 6 to 17 months.
Study design issues: This single arm, single centre, non-randomised, prospective, consecutive case series reported this centre's early experience with corneal neurotisation for neurotrophic keratitis. Outcomes included corneal sensation, visual acuity, and epithelial condition. Corneal sensation was measured by gently pressing a pieces of tissue paper against the cornea in 4 peripheral quadrants of the eye as well as centrally. If light touch was present, this was recorded as positive. If no light touch was noted, then harder pressure was evaluated with a Schirmer strip. This method of evaluation of corneal sensation is likely less accurate than use of an aesthesiometer. Visual acuity was measured with an Early Treatment Diabetic Retinopathy Study (ETDRS) chart.
No statistical analysis was described.
Key efficacy findings
Number of people analysed: 6 eyes in 6 people
Follow up at time of assessment: 6 to 17 months
All people regained corneal sensation at the last follow up and 5 of 6 people had improved visual acuity.
Person | Follow up (months) | Visual acuity | Epithelial status | Regained corneal sensation | ||
Before surgery | At follow up | Before surgery | At follow up | |||
1 | 6 | 20/60 | 20/40 | Intact, central dense scarring | Intact, central dense scarring | Yes |
2 | 9 | 20/800 | 20/160 | Ulcer | Intact | Yes |
3 | 13 | Hand motion | 20/25 | Ulcer | Intact | Yes |
4 | 16 | 20/300 | 20/70 | Intact, central scarring | Intact | Yes |
5 | 11 | 20/50 | 20/70 | Dense corneal scar and peripheral punctate epithelial defects | Dense corneal scar and inferior Punctate epithelial defects | Yes |
6 | 17 | 20/100 | 20/80 | Ulcer | Corneal scarring, intact epithelium | Yes |
Key safety findings
No complications were reported.
How are you taking part in this consultation?
You will not be able to change how you comment later.
You must be signed in to answer questions