Welcome to Part 3 of NBCC’s Key Take-Aways from the 2022 American Society of Clinical Oncology (ASCO) annual conference. In Part 1 we described some of the widely talked-about results from the DESTINY-Breast03 and Breast04 clinical trials. In Part 2 we summarized the long-awaited overall survival results of the PALOMA2 trial.
In this week’s Science Spotlight, we examine two studies that look at changes over the last few decades in breast cancer mortality. We know that mortality rates have decreased. From NBCC’s perspective, the decrease is nowhere near commensurate with the investments to date, including financial, time, and people resources. But the question remains, what contributed to the decrease that did occur? Was it early detection? New treatments? Changes in how data are collected? Or improvements in access to care?
One study, presented in an oral abstract at ASCO, detailed the results of a U.S. population-based modeling study that examined the specific contributions to mortality reductions by subtype, of screening and treatments for early-stage and metastatic breast cancer. We also summarize a study that was published on June 23, 2022, in the European Journal of Public Health (EJPH). The EJPH study, another population-based study, explores how the effectiveness of mammography screening (in terms of reducing breast cancer mortality) is diminishing as more effective treatments for breast cancer emerge, though the harms of screening (e.g., false positives, overdiagnosis) remain unchanged.
U.S. Population-Based Breast Cancer Mortality Reductions by Subtype and Contributions From Metastatic Treatments
Key Takeaway: Breast cancer mortality has declined over the years, variably by subtype, and with variable contributions from screening, early-stage treatment, and metastatic treatment.
Dr. Jennifer Lee Caswell-Jin of Stanford University described the results of a modeling study that explored the contributions of screening, early-stage treatment, and metastatic treatment on breast cancer mortality reduction by breast cancer subtype in U.S. women between the years 2000 and 2019.
The investigators used the CISNET breast cancer models (Cancer Intervention and Surveillance Modeling Network). CISNET is an NCI-sponsored consortium of investigators whose purpose is to measure the effect of cancer-control interventions on the incidence of and risk of death from cancer in the general population. Over the years, the consortium has developed six models, four of which were applied in the current study. Previously published work[i],[ii] used these models to quantify the contributions of “screening” and “treatment in early-stage disease” on breast cancer mortality. The present study added the impact of “treatments for metastatic breast cancer.”
The investigators first used the models to estimate how survival for metastatic breast cancer has changed by breast cancer subtype.
They estimated that median metastatic breast cancer survival in the general population (using SEER data) increased and mortality declined, by the following amounts from the year 2000 to 2019 for each of the respective breast cancer subtypes:
The results across all models suggest that by 2019, breast cancer mortality has declined by approximately 58%, with differences by subtype as shown above in parentheses.
Note: You may be wondering why this number (58%) is much higher than what we read about in the annual statistics that come out every year by the American Cancer Society (~41% across all breast cancer subtypes and ages reported in 2022). The reason for this is that the CISNET investigators compare breast cancer mortality in 2019 to the rate predicted by the models in the year 2000 if we didn’t have any screening or treatments for breast cancer. The most recent American Cancer Society statistics compare the breast cancer mortality data from 2019 (the most recent year for which mortality data are available) to the year when actual breast cancer mortality was at its highest point, in 1989. Read this footnote for more information on how this is calculated. [iii] It is worth noting that breast cancer mortality has been declining by 1-2% every year since 1989, for a total cumulative reduction of about 41% since 1989. Moreover, while the mortality rate was decreasing by about 1.9% annually between 1998 and 2013, those annual declines have slowed to 1.1% per year between 2013 and 2019.
The last set of results that were presented at ASCO from this model-based analysis were the specific contributions over time of the interventions. Notably, the contribution of screening has declined progressively every year from 2000 to 2019 as newer treatments for both early-stage and metastatic breast cancer become available. But it is important to remember that these estimates are all modeled and can vary depending on the assumptions that are made. The accuracy of modeling depends on good-quality data going into the model and reasonable assumptions about unobservable events.
Interestingly, this latter finding is consistent with the results of another recently published study[iv] described below.
Change in Effectiveness of Mammography Screening with Decreasing Breast Cancer Mortality: A Population-Based Study
Christiansen et al., 2022 describe the results of another model-based analysis that looks at how the effectiveness of mammography screening in reducing breast cancer mortality has changed over time as newer and more effective treatments and care management protocols have become available for breast cancer. This study uses data from the Norwegian cancer registry (similar to the US SEER data) that contains breast cancer incidence and mortality data as far back as 1960. The investigators looked at how the “number needed to invite” (NNI) to screening in order to prevent 1 death from breast cancer over the next 10 years has changed over time, along with the balance of benefits and harms (i.e., overdiagnosis).
The main conclusion of this study is that the number of women who need to be invited to screening to prevent 1 death has increased substantially over the years as new more effective treatments are developed. That means mammography screening has become even less effective over time. But it hasn’t become less harmful. The harms of screening are not affected by treatment improvements, so, as the NNI goes up, the balance tilts more in the direction of harm over benefit.
[i] Berry DA, Cronin KA, Plevritis SK, et al. Cancer Intervention and Surveillance Modeling Network (CISNET) Collaborators. Effect of screening and adjuvant therapy on mortality from breast cancer. N Engl J Med. 2005 Oct 27;353(17):1784-92. doi: 10.1056/NEJMoa050518. PMID: 16251534.
[ii] Plevritis SK, Munoz D, Kurian AW et al. Association of Screening and Treatment With Breast Cancer Mortality by Molecular Subtype in US Women, 2000-2012. JAMA. 2018 Jan 9;319(2):154-164. doi: 10.1001/jama.2017.19130. Erratum in: JAMA. 2018 Feb 20;319(7):724. PMID: 29318276; PMCID: PMC5833658.
[iii] How is the annual breast cancer mortality calculated: According to SEER data, in 1989, the age-adjusted breast cancer mortality in the US was reported to be 33.2 deaths per 100,000 women. In 2019, breast cancer mortality was 19.4 deaths per 100,000 women representing a drop in mortality of 41.6%. The 58% reduction predicted by the CISNET models in this study compare breast cancer mortality in 2019 to what the models predict the mortality would be in 2000 (the highest level), in the absence of screening and treatment for early and metastatic breast cancer.
[iv] Christiansen SR, Autier P, Støvring H. Change in effectiveness of mammography screening with decreasing breast cancer mortality: a population-based study. Eur J Public Health. 2022 Jun 23:ckac047. doi: 10.1093/eurpub/ckac047. Epub ahead of print. PMID: 35732293. https://academic.oup.com/eurpub/advance-article/doi/10.1093/eurpub/ckac047/6609838?login=true