Medical Safety & Global Health
Open Access

Abstract

Mitigating the Paucity-of-Data Problem for Target Population Sizing: Exploring a Model-Based Approach for Advanced Gastroenteropancreatic Neuroendocrine Tumours

Aurore Bergamasco, Gabrielle Nayroles, Anne-Marie Castilloux, Jérôme Dinet, Anthony Berthon, Sylvie Gabriel and Yola Moride

Background: Gastroenteropancreatic Neuroendocrine Tumors (GEP-NETs) are rare neoplasms. For innovative treatments, payer recommendations frequently involve sub-populations more restricted than approved indications. Paucity of epidemiologic data specific to sub-populations is a challenge for reimbursement strategies. Objectives: To estimate the population size by site and type of GEP-NETs in the US, EU, and Australia, over a five-year period.

Methods: Two GEP-NET sub-populations, respectively approved and restricted indication for reimbursement, were considered: i) Stable/slow progressing well-differentiated, functioning and non-functioning GEP-NETs and unresectable locally advanced/metastatic disease; and ii) Stable/slow progressing well-differentiated, nonfunctioning GEP-NETs and unresectable locally advanced/metastatic disease. For both, tumours originating from the hindgut were excluded. Following identification in the literature of crude prevalence and incidence rates for a broader GEP-NET, estimates were obtained for each sub-population using proportions of GEP-NETs by site and type derived from clinical studies. Then, these figures were further refined using clinical expert opinions. A 5-year target population growth model was developed.

Results: Over 5 years, respectively for the first and second sub-population, number of patients is expected to increase from 7,473 to 9,393 and 5,231 to 6,575 in selected European countries; from 8,051 to 10,119 and 5,636 to 7,083 in the US; and from 593 to 746 and 415 to 522 in Australia. Because the second sub-population is a subgroup of the first, lower estimates were obtained.

Conclusion: In the absence of epidemiologic data on specific sub-populations, the development of a population growth model can be used to estimate trends in population size under varying labelling hypotheses.