OBJECTIVES

To evaluate the cost-effectiveness of conbercept versus ranibizumab in treating myopic CNV from China payer’s perspective.

METHODS

A Markov model, based on best corrected visual acuity (BCVA) letters, was developed to simulate the progression of myopic CNV and to predict the outcomes of conbercept and ranibizumab treatment over life-time.

Disease progression was characterized by a series of health states referred to as 90, 75, 60, 45, and 30 BCVA letters. A patient’s BCVA was assumed to either increase by 30 letters, increase by 15 letters, remain the same, decrease by 15 letters or decrease by 30 letters each 3-month length cycle.

The first-year transition probabilities of conbercept and ranibizumab were estimated based on clinical trials on Chinese population, i.e. SHINY for conbercept and BRILLIANCE for ranibizumab, respectively. The long-term visual prognosis was assumed the same in both arms based on literatures. The number of injections in two arms were either from the clinical trials (scenario 1) or from real-world studies (scenario 2). Utilities and cost data (including drug cost, costs associated with health states and costs of managing adverse events) were mainly derived from literatures or previous studies.

Both one-way sensitivity analysis and probabilistic sensitivity analysis (PSA) were conducted on selected key parameters.

RESULTS

In scenario 1, the total costs and QALYs of conbercept were ¥180,207 and 9.86 while that of ranibizumab were ¥175,955 and 9.83, respectively. Conbercept was cost-effective under the threshold of 3 times GDP per capita. In scenario 2, the QALYs remained the same and the total costs of conbercept was lower than ranibizumab (¥163,987 vs ¥164,923), making conbercept a dominant choice. PSA indicated that the probability of conbercept being cost-effective in two scenarios were 53.0% and 66.3%, respectively.

CONCLUSIONS

Compared with ranibizumab, conbercept had similar total costs and QALYs while PSA demonstrated a slightly higher level of cost-effectiveness.