Cost-Effectiveness of Breast Cancer Screening
Editor’s Note: The 2023 China Congress of Integrative Oncology (2023 CCHIO) will be held in Tianjin from November 16th to 19th, 2023. Dr. Alexander Mundinger, who is on board of directors of the International Society of Breast Diseases (SIS), will deliver a speech about “Cost-effectiveness of breast cancer screening” at this congress. He was interviewed by Oncology Frontier on the eve of 2023 CCHIO.
Oncology Frontier: What are the differences in prevalence of breast cancer globally? How might it affect breast cancer screening in different places?
Prof. Alexander Mundinger: The prevalence and incidence of breast cancer increases in parallel globally. Breast cancer specific prevalence refers to the number of people currently diagnosed with breast cancer, , and incidence is measured as the new breast cancer cases being diagnosed over a specific period of time. Globocan estimates of breast cancer show the highest values for example in Belgium and other western countries (age standardized incidence rate about 85 to 95 per 100 0000), intermediate estimates in the Caribbean, South America, North and South Africa (about 50 to 56 per 100 000) and lower rates for most Asian territories (about 41-47 per 100 000). The mortality is highest in Barbados and West Africa. To date breast cancer has become the most frequent cancer type worldwide in women, also for China. Interestingly, the incidence rate in rural areas of China has been shown to be about half of the incidence rate of the urban regions. On the other hand, the survival of breast cancer patients is poor in these rural areas because cancers are detected at a later stage. Globally women living in transitioning countries have 17% higher mortality rates compared to women in transitioned countries (Globocan 15.0 vs. 12.8 per 100 000 respectively). The elevated incidence rates in higher Human Development Index (HDI) countries reflect a longstanding higher prevalence of reproductive and hormonal risk factors, lifestyle risk factors, increased detection by mammography screening and mutations in high-penetrance genes, such as BRCA1 and BRCA2. amongst small subpopulations.
The question arises facing the breast cancer inequities in outcome, different prevalence and unequally distributed financial resources worldwide: Can state of the art mammography screening be established worldwide to improve breast cancer specific mortality? Large trials have shown that in western countries of a very high/high income mammography screening programs have been able to reduce breast cancer specific mortality up to 40% since 1980s. But the standards of early diagnosis, adequate heath service and financial coverage cannot be transferred to most countries with a low/medium income. The answer given by WHO and other international stakeholders is very clear. If a country has low financial resources and other social priorities, it cannot spend enough financial resources on detecting, diagnosing and treating breast cancer. Thus, the alternative concept in these countries is focusing on symptomatic breast cancer patients and to improve the individual diagnosis and the treatment of a minor collective of symptomatic cancer patients. We know how to make it better. Low/middle income countries need cancer specialists and specialized breast centers for diagnosis and treatment. In low resource settings, where populations tend to be younger, ultrasound is an effective low-cost technology to amend clinical examination and guide minimal invasive biopsies. Palpation and ultrasound for detection, fine-needle aspiration or better core biopsy for diagnosis are probably more important than a mammography screening for very low-prevalence regions that would be prone to associated false positive screening cases.
At the same time, it is very difficult to lower the mortality of symptomatic advanced cancers, whatever you do. Several trials have shown that palpation is not able to reduce the breast cancer specific mortality. As a consequence, the surgical and the antihormonal treatment and chemotherapy of these symptomatic cancer patients are in the focus. The rationale behind this is to downstage these symptomatic cancers. This is a different approach compared to the Western countries that are hunting small cancers in asymptomatic women and provide a networks of certified breast cancer centers. Western breast cancer patients in the stage one present with the same mortality following to adequate treatment as healthy women. Unfortunately, no alternative compromise low-cost policy for low/mid income countries can fully compensate the harm of a late-stage diagnosis of breast cancer.
As a summary, different prevalence and inequities of financial resources have an impact on the decision to screen or not to screen asymptomatic women. Stage of breast cancer at diagnosis and poverty rate explains the geographic variations in breast cancer survival best even in developed economies such as USA and Europe.
Oncology Frontier: What do you think is cost-effective for breast cancer screening in different regions?
Prof. Alexander Mundinger: First, you have to be familiar with the principles and definitions of cost-effectiveness. It does not mean that breast cancer screening is less expensive than no screening. It only means the costs of screening are less expensive than a threshold that needs to be defined by the government or responsible health care providers. This threshold of cost effectiveness refers to the “willingness to pay” (WTP) of a health system. Usually, it is set between one to three times the gross domestic product per capita (per person). In two Chinese cost-effectiveness studies (on screening in rural areas and comparing high-risk and low-risk patients) the willingness to pay was set three times the GDP per capita. I appreciate this approach for standardization: three times the GDP per capita or per person. Almost all estimates in European economic screening models fall far below 30 000 € (median below 10 000 €) per life-years or quality‐adjusted life years gained. However, some countries show reluctance to give information about implicit or explicit thresholds to avoid misinterpretations about rationing. Finally, the politicians have to make up the decision. Unfortunately, the willingness to pay and ability to pay is only moderately correlated in comparative studies. Poor economic and disastrous social-political situations counteract with the visions and missions of fighting breast cancer at appropriate costs. It may be necessary for a country to first focus on other fields of health politics where addressing starvation, perinatal child death, infections or war injuries is the highest priority. Thus, the local situation has to be considered and to be wisely integrated in socio-economic decisions including screening of breast cancer. But the vision of breast health federations such as Senologic International Society (SIS) is to argue for the women who have cancer. And therefore, we cannot consider all the other conditions. We have to focus on our mission and visions that are in line with the WHO`s Global Breast Cancer Initiative that started 2021.
Oncology Frontier: How do you think of the application of artificial intelligence in breast cancer screening?
Prof. Alexander Mundinger: The application of artificial intelligence in breast cancer screening is a great future horizon. And it is going to transform the old policies. The older approaches using computer assisted detection (CAD) systems were disappointing. But today artificial intelligence makes use of convolutional neural networks (CNN), deep learning algorithms and includes transformer technology. This enables transfer learning and use of bilateral and prior images for detecting asymmetries and growth of lesions. Recently the authors of the first prospective randomized trial, the MASAI Trial, concluded that the AI support was safe. AI-supported reading found more invasive cancers (184 vs. 165) and more in-situ cancers (60 vs. 38) than state of the art double reading. Further, screen-reading workload was reduced by 44.3% using AI.
Currently three different applications of AI are examined by studies: (a) AI-supported reading of digital mammography and DBT, (b) AI as a stand-alone system for decision support, (c) AI pre-selection of normal cases. The results are encouraging. In my opinion, AI will become a useful tool in the near future and most importantly it will help reducing the costs and an initial shortage of specialized radiologists. But use of AI in screening will be a privilege for countries with a digitalized structure and a lot of money. By the way, China is a country that has the most publications of artificial intelligence according to OECD website and is well digitalized to my knowledge. But any overinflated hype on AI could disappoint. The initial investment costs are high, and the return on investment will come late. Finally, we need evidence by prospective randomized trials whether or not AI algorithms can adopt to women with predominantly dense breasts and early onset of breast cancer. This pattern is typically seen in Asian, African and Latin American women.
