OBJECTIVES: To conduct a cost-utility analysis comparing first- and second-line drug strategies involving octreotide, lanreotide, pasireotide, and pegvisomant for the treatment of patients with acromegaly who have failed surgery, from a Brazilian public health system perspective.

METHODS: A probabilistic cohort Markov model was developed. One-year cycles were employed. The patients started at 45 years of age and were followed lifelong. Costs, efficacy, and quality of life parameters were retrieved from the literature. A discount rate (5% per year) was applied to both costs and effects. Incremental cost-effectiveness ratios (ICERs) were calculated when applicable, and reported as cost per quality-adjusted life year (QALY). Scenario analyses considered alternative dosages, discount rates, tax exemption, and continued use of treatment despite lack of response. Value of information (VOI) analysis was conducted to explore uncertainty and to estimate the costs to be spent in future research. One American dollar (U$) equals approximately 4 Brazilian reais (R$) in 2020.

RESULTS: Of the 11 strategies compared, four were not dominated. Only lanreotide autogel showed a reasonable ICER for having its use considered in clinical practice (R$ 112,138/QALY compared to “no treatment”). Considering a threshold of 3 times the GDP per capita (R$ 91,645), lanreotide autogel had a 38.19% probability of being cost-effective compared to “no treatment”. Scenario analyses corroborated the base-case result. VOI analysis showed that much uncertainty surrounds the parameters, and future clinical research should cost less than R$ 43,230,000 per year. VOI also showed that almost all uncertainty that precludes an optimal strategy choice involves quality of life.

CONCLUSIONS: With current information, the only strategy that can be considered cost-effective in Brazil is lanreotide autogel. No second-line treatment is recommended. Significant uncertainty of parameters impairs optimal decision-making, and this conclusion can be generalized to other countries. Future research should focus on acquiring utility data.