Association of Breast Cancer Screening Behaviors With Stage at Breast Cancer Diagnosis and Potential for Additive Multi-Cancer Detection via Liquid Biopsy Screening: A Claims-Based Study

Abstract

Purpose: To evaluate mammography uptake and subsequent breast cancer diagnoses, as well as the prospect of additive cancer detection via a liquid biopsy multi-cancer early detection (MCED) screening test during a routine preventive care exam (PCE).

Methods: Patients with incident breast cancer were identified from five years of longitudinal Blue Health Intelligence^® (BHI^®) claims data (2014-19) and their screening mammogram and PCE utilization were characterized. Ordinal logistic regression analyses were performed to identify the association of a biennial screening mammogram with stage at diagnosis. Additional screening opportunities for breast cancer during a PCE within two years before diagnosis were identified, and the method extrapolated to all cancers, including those without recommended screening modalities.

Results: Claims for biennial screening mammograms and the time from screening to diagnosis were found to be predictors of breast cancer stage at diagnosis. When compared to women who received a screening mammogram proximal to their breast cancer diagnosis (0-4 months), women who were adherent to guidelines but had a longer time window from their screening mammogram to diagnosis (4-24 months) had a 87% increased odds of a later-stage (stages III or IV) breast cancer diagnosis (p-value <0.001), while women with no biennial screening mammogram had a 155% increased odds of a later-stage breast cancer diagnosis (p-value <0.001). This highlights the importance of screening in the earlier detection of breast cancer. Of incident breast cancer cases, 23% had no evidence of a screening mammogram in the two years before diagnosis. However, 49% of these women had a PCE within that time. Thus, an additional 11% of breast cancer cases could have been screened if a MCED test had been available during a PCE. Additionally, MCED tests have the potential to target up to 58% of the top 5 cancers that are the leading causes of cancer death currently without a USPSTF recommended screening modality (prostate, pancreatic, liver, lymphoma, and ovarian cancer).

Conclusion: The study used claims data to demonstrate the association of cancer screening with cancer stage at diagnosis and demonstrates the unmet potential for a MCED screening test which could be ordered during a PCE.

Keywords: breast cancer; cancer screening; claims data analysis; earlier detection; liquid biopsy; mammography; multi-cancer early detection; preventive care.

Figure 1

(A) Breakdown of the study into three subpopulations: women, all cancer, and breast cancer-specific populations. Inclusion criteria were ages 50-64 and at least 2 years of continuous enrollment. Screening mammogram and preventive care exam (PCE) utilization were characterized for all women and breast cancer populations, while only PCE utilization was analyzed for the all cancer (men and women) population. (B) Breakdown of the study into main aims and specific research questions, along with their associated figures. Note that analysis on the frequency of screening mammogram history required at least four years of continuous enrollment. A description of the methods for this analysis can be found in the Supplementary Material .

Figure 2

(A) Breast cancer stage distribution by presence of a biennial screening mammogram (yes vs. no) and (B) Breast cancer stage distribution by time from screening mammogram grouped into R-MAM (screening mammogram <4 months from diagnosis) and D-MAM (screening mammogram 4-24 months from cancer diagnosis). (C) The cumulative proportion of each stage of incident breast cancers by months from the screening mammogram to breast cancer diagnosis. (D) Distribution of months from screening mammogram to breast cancer diagnosis is bimodally distributed and modeled by a two-component Gaussian mixture. Note that percentages may not add to 100% due to rounding.

Figure 3

(A) Combinations of preventive care exam (PCE) and screening mammogram utilization in the all women aged 50-64 and breast cancer cohorts, and (B) PCE utilization in the all member aged 50-64, incident cancer, and metastatic cancer cohorts. Note that percentages may not add to 100% due to rounding.

Figure 4

Current screening rates for breast cancer and the top 5 cancers that are both the leading causes of cancer death and lack a USPSTF recommended screening modality were compared with the potential screening rates possible with an additive multi-cancer screening blood test performed at a preventive care exam (PCE). The potential screening rate for breast cancer was the proportion of breast cancer patients who had either a screening mammogram or PCE in the two years before their diagnosis date, while the potential screening rate for other cancer types was the proportion of patients who had a PCE in the two years preceding diagnosis. The top 5 cancers that are leading causes of cancer death were limited to cancer types that do not currently have a screening modality and that form solid tumors, which include prostate, pancreatic, liver, lymphoma, and ovarian cancers (24). Conclusions could not be made for colorectal, lung, and cervical cancers because their screening adherence rates were not explored in this study.