The rapid expansion of immuno-oncology treatment options has led to concerns around their long-term affordability. Evidence on the potential budget and health impact of these new treatment options is required to inform public health policy and ensure adequate allocation of budget for the future.

The Health Impact Projection model was developed to compare the economic impact and health outcomes observed with and without PD-1/PD-L1 inhibitors using traditional budget impact analysis. Seven types of high-incidence cancers were included: melanoma, first- and second-level non-small cell lung, bladder, head and neck, renal cell carcinoma, and triple negative breast. Inputs were based on publicly available data and literature, and over 10 key experts (oncologists, health economists) were involved in the model development. The model draws on five-year budget impact analysis.

Using the experience of Belgium, Slovenia, Switzerland, and Italy, the model estimates budget and health impact of the PD-1/PD-L1 inhibitor class. It shows that for 2018-2022, the class will provide additional life years and avoid high-grade adverse events (AEs) with a manageable budget impact per year compared to the standard of care. The model also enables policy-makers to assess the adequacy of their budget for the near future and explore the implications of different policy decisions. Results for Belgium show that over the five-year period the PD-1/PD-L1 inhibitors will save 10,635 additional life years, avoid 7,597 AEs and have a budget impact of approximately EUR 260 million. Results for Slovenia show 1,468 additional life years gained and 869 AEs avoided with a budget impact of approximately EUR 116 million; for Switzerland, 6,775 life years gained, 6,953 AEs avoided, and EUR 106 million budget impact; and for Italy, 5,019 life years gained, 2,040 AEs avoided, and EUR 627 million budget impact.

Although limitations exist, the model informs planning by helping quantify the potential impact of immune-oncology treatments on health and budget in different scenarios.