Cost Utility Analysis of Corneal Neurotization and Recombinant Nerve Growth Factor for Neurotrophic Keratopathy

To evaluate the cost-utility of surgical corneal neurotization (CN) versus topical recombinant nerve growth factor (cenegermin), using a Markov model from a health care system perspective, in the management of neurotrophic keratopathy (NK).

Three Markov models were utilized: treatment with cenegermin, treatment with CN, and combined treatment with cenegermin and CN.

At the per-patient level, CN resulted in total costs of $36,866 and generated 3.66 Quality of Life Years (QALYs), whereas cenegermin resulted in costs of $161,926 and 3.32 QALYs, respectively. The combined strategy of cenegermin followed by CN resulted in costs of $137,048 and 3.82 QALYs. Probability sensitivity analysis using 1000 Monte Carlo simulations suggested that CN was associated with lower mean total costs ($15,808 vs $161,352) and higher mean QALYs (3.89 vs 3.71) when compared with cenegermin. In the comparison between CN and the combined strategy, CN remained less costly ($15,808 vs $124,962) with nearly equivalent effectiveness (3.895 vs 3.905 QALYs). At a willingness to pay (WTP) threshold of $100,000 per QALY, CN was cost-effective in 100% of simulations against cenegermin or combined CN/cenegermin strategies. By deterministic sensitivity analysis (DSA) between cenegermin and CN, CN revealed a positive net monetary benefit (NMB) across all parameter variations, most influenced by the probability of healing following cenegermin treatment, utility associated with a healed cornea and the cost of cenegermin. Incremental cost analysis suggested that CN remained the preferred strategy across the modeled parameter ranges.

Within this five-year cost-utility model, corneal neurotization provided greater economic value and quality-of-life outcomes than cenegermin. These findings suggest that corneal neurotization may offer meaningful long-term economic and patient benefits for moderate-to-severe neurotrophic keratopathy, although results remain dependent on the clinical and economic assumptions incorporated into the model.