ESTIMATING NATIONAL PATIENT POPULATION SIZES FOR BUDGET IMPACT: A CENSUS-BASED EPIDEMIOLOGICAL MODEL PREDICTING THE NUMBER OF WOMEN WITH BONE METASTASES SUBSEQUENT TO BREAST CANCER IN THE UK

Purpose: Bone metastases (BM) are a common occurrence once cancer progresses: 65% to 75% of advanced breast cancers (BC) progress to BM. Although the incidence of BC is precisely estimated in the UK, the prevalence of BC with BM is not known. This study aimed to estimate the absolute number of patients with BC and BM from 1992 to 2020.  

Methods: A Markov model reflecting the progression of BC was developed to simulate the transitions between health states (silent disease, bone and visceral metastatic disease, death due to BC, other death). The number of incident cases of BC were obtained from the Office of National Statistics. Transition probabilities were obtained from a recent meta-analysis on the effect of chemotherapy. Treatments launched since 1992 were identified and corresponding risk reductions (obtained from the literature) applied to the baseline transition probabilities to reflect the improved prognosis of women with BC. Validation was performed against the number of BC deaths registered by the Office of National Statistics.
Results: Use of constant transition probabilities resulted in an over-estimation of the number of BC deaths by 21% in 2005. After adjustment of the transition probabilities to include recent therapeutic advancements, the model predicted 12,575 BC deaths in 2005, while 12,509 had been observed (0.53% difference).
The model predicted that in 2008, 45,000 new cases of BC will be diagnosed and more than 500,000 women will be living with BC, of whom 39,000 will be suffering from BM. Assuming no major change in treatment efficacy  and that BC incidence by age remains constant, the population of women suffering from BM is predicted to reach 50,000 in 2020.

Conclusion: The number of women suffering from BC and subsequently BM is expected to increase within the next decade. The validation of the model using historical data shows that the benefit of new treatments is directly measurable at the national level.