Neurotrophic keratitis / keratopathy ( NK ), a rare degenerative corneal disease, lacks effective pharmacologic therapies.
Because NK pathology involves trigeminal nerve damage and loss of corneal innervation, nerve growth factor ( NGF ) is surmised to promote healing of neurotrophic keratitis.
Preliminary studies with murine NGF demonstrated efficacy for treating corneal neurotrophic ulcers; however, the complex tertiary structure of NGF has complicated the production of recombinant human NGF ( rhNGF ) suitable for clinical development.
To this end, researchers have developed an Escherichia coli–derived rhNGF formulation that demonstrated to be well tolerated and safe for topical ophthalmic use in a phase I study in healthy volunteers.
Phase I results of topical rhNGF ( Cenegermin; Oxervate ) for patients with moderate-to-severe neurotrophic keratitis were reported.
NGF0212/REPARO was a phase I/II randomized, double-masked, multicenter, vehicle-controlled, parallel group study that evaluated the safety and efficacy of rhNGF eye drops ( 10 or 20 microg/ml, 6 drops/day for 8 weeks ) in patients with moderate ( stage 2 ) or severe ( stage 3 ) neurotrophic keratitis.
Patients aged 18 years or more with stage 2 or 3 NK were enrolled according to published diagnostic criteria and inclusion / exclusion criteria described in the REPARO phase II report.
Eighteen phase I patients ( 2 cohorts of 9 consecutively enrolled patients each ) with stage 2 or 3 NK gave informed consent and were randomized 7:2 to rhNGF 10 microg/ml versus vehicle ( cohort A ) or rhNGF 20 microg/ml versus vehicle ( cohort B ).
Sample size was based on clinical feasibility ( i.e., no formal power calculation was performed ), because phase I aimed primarily to assess the safety and systemic absorption of topical rhNGF to support proceeding with phase II, which was conducted, analyzed, and reported separately.
Patients, researchers, and site/sponsor staff were masked to primary randomized treatment.
The study included an 8-week controlled treatment period and a 48- or 56-week follow-up ( duration determined by treatment allocation and corneal healing status during controlled treatment ).
In the event of treatment failure during the 8-week controlled treatment period ( predefined as failure to achieve corneal healing, recurrence of neurotrophic keratitis after healing, or deterioration as described in the phase II report ), vehicle-treated patients were eligible to receive 8 weeks of uncontrolled rhNGF treatment ( cohort A: 10 microg/ml; cohort B: 20 microg/ml ) before continuing follow-up ( total follow-up: 56 weeks ).
However, no phase I patients entered the 56-week follow-up period.
The primary safety variable was incidence of adverse events, defined per Good Clinical Practice guidelines as any untoward medical occurrences in patients who received study treatment, regardless of causal or temporal association.
Other safety parameters included visual analogue scale for ocular tolerability ( described in the phase II report ), best-corrected distance visual acuity measured in Early Treatment Diabetic Retinopathy Study letters, intraocular pressure, dilated fundus ophthalmoscopy, vital signs, hematology, and clinical chemistry.
Eye pain and headache were the most frequently reported adverse events during controlled treatment, each occurring in 2 patients ( 28.6% ) in the rhNGF 20 microg/ml group.
Treatment-related adverse events ( TAEs ) reported during controlled treatment occurred in 1 of 18 patients each.
No TAEs were reported during the 48-week follow-up. No deaths occurred during controlled treatment or follow-up, nor were there any notable trends or clinically significant differences over time or between treatment groups in laboratory parameters, vital signs, or other ocular safety assessments.
Pharmacokinetics profiling was performed. Only 2 patients had detectable serum NGF at any time point.
Of note, the patient in the rhNGF 10 microg/ml group had only 1 positive NGF measurement during the study, and the patient in the rhNGF 20 microg/ml group had detectable serum NGF levels at all time points, even before initiating study treatment.
Taken together, the pharmacokinetics results suggest individual fluctuations of endogenous NGF independent of study treatment.
Although the phase I study was not designed or powered for efficacy outcomes, corneal healing ( less than 0.5 mm fluorescein staining in the lesion area ) was assessed in clinical pictures by central readers ( masked to treatment assignment and duration ) at week 4 ( primary end point ) and week 8 ( key secondary end point ).
At week 4, based on postbaseline last-observation-carried-forward analysis, corneal healing was achieved by 1 of 4 patients ( 25.0% ) receiving vehicle, 3 of 7 patients ( 42.9% ) receiving rhNGF 10 microg/ml, and 3 of 7 patients ( 42.9% ) receiving rhNGF 20 microg/ml.
Of patients with responses available at week 8, corneal healing was achieved by 1 of 2 patients ( 50% ) receiving vehicle, 4 of 6 patients ( 66.7% ) receiving rhNGF 10 microg/ml, and 6 of 7 patients ( 85.7% ) receiving rhNGF 20 microg/ml.
No phase I patients discontinued because of a lack of efficacy or inadequate control of neurotrophic keratitis.
Before week 8, no patients in any treatment group experienced deterioration.
At week 8, 1 patient who received rhNGF 20 microg/ml experienced a decrease in best-corrected distance visual acuity score of more than 5 Early Treatment Diabetic Retinopathy Study letters.
The REPARO phase I study has demonstrated that topical ophthalmic rhNGF ( 10 or 20 microg/ml ), administered 6 drops/day for 8 weeks, was well tolerated in patients with stage 2 or 3 neurotrophic keratitis.
No safety concerns arose; most adverse effects were ocular, mild, and transient, and did not require discontinuing or corrective treatments.
Most patients had undetectable serum NGF, and systemic adverse effects were infrequent and mild.
Favorable trends in corneal healing suggest that topical ophthalmic rhNGF may be effective for treating patients with moderate-to-severe neurotrophic keratitis. ( Xagena )
Bonini S et al, Ophthalmology 2018; 125: 1468-1471