Abstract
Purpose
The COVID-19 pandemic created significant disruptions in the diagnosis and treatment of breast cancer (BC). Several public health measures were taken with limited evidence on their potential impact. In this observational study, we sought to compare the incidence of BC, treatment patterns, and mortality during 2020 versus 2018 and 2019.
Methods
Using the Surveillance, Epidemiology, and End Results program, we identified 37,834 patients with ductal carcinoma in situ (DCIS) and 199,594 with invasive BC between 2018 and 2020. We assessed age-adjusted incidence rates of DCIS and invasive BC as cases per 100,000, treatment patterns, and mortality in 2020 versus 2018 and 2019.
Results
From 2019 to 2020, the incidence of female DCIS decreased from 36.4 to 31.0, and the incidence of female invasive BC decreased from 184.2 to 166.6. Among females, the relative reductions in incidence from 2019 to 2020 were 14.8% for DCIS, 12.1% for stage I, 5.8% for stage II, 2.6% for stage III, and 1.9% for stage IV. Comparing 2020 to 2018–2019 in invasive BC, we observed significant changes in treatment patterns with decreased use of surgery or radiation and increased use of chemotherapy. The 12-month mortality rates were 4.49%, 4.37%, and 4.57% for 2018, 2019 and 2020, respectively. In the Cox model, there were no significant differences in mortality between patients diagnosed in 2020 versus 2018 or 2019.
Conclusions
During 2020, the incidence of BC decreased significantly. There were reductions in surgery and radiation use, but not in chemotherapy. Although vaccines were largely unavailable and COVID-19 treatments were in development, we saw no differences in 12-month mortality in 2020 versus prior years. The impact on BC-specific outcomes requires further follow-up.
Keywords: Breast cancer, COVID, Death, Health policy, Vaccines
Introduction
The coronavirus disease 2019 (COVID-19) pandemic created significant challenges in breast cancer care which were most pronounced during 2020 when screening and treatment access were reduced. Across many locations, screening programs for breast cancer were temporarily suspended or curtailed, and operating rooms were closed for non-emergency surgeries, including cancer-related surgeries [1–7]. In an attempt to balance timely breast cancer treatment versus the need to maintain hospital resources and reduce exposure to COVID-19, several organizations provided guidelines for cancer management during the pandemic [8–10]. These guidelines recommended consideration of neoadjuvant endocrine therapy for early-stage hormone receptor (HR)-positive breast cancer, and neoadjuvant chemotherapy for human epidermal growth factor receptor 2 (HER2)-positive and triple-negative breast cancers, in order to safely delay the time to surgery. Recommendations also included the use of hypofractionated adjuvant radiotherapy schedules. As a result of the COVID-19 measures, several studies have reported reductions in the incidence of breast cancer [11, 12], and potential delays to treatment [13]. In fact, some institutions developed changes to their breast cancer approach in the first part of the pandemic [14–17], with limited evidence on the potential impact of those changes. In addition, some studies reported heterogeneity in adherence to the recommendations between different institutions and even among physicians from the same center [18].
The aims of this study were to evaluate breast cancer incidence, treatments, and short-term mortality in the Surveillance, Epidemiology and End Results (SEER) program, a large, US.-based registry which covers approximately half of the US. Population, during the year 2020 versus the immediate pre-pandemic period (2018 and 2019).
Methods
Data source and study design
We obtained data from the SEER program, using the 17 registry database (November 2022 submission) [19]. Variables collected by the SEER program include patient demographics, primary tumor site, morphology, stage at diagnosis, initial course of therapy, and survival follow-up. SEER collects data on surgery and radiation therapy even when these occurred after neoadjuvant therapy. We included women and men aged 20 years or older, diagnosed with either ductal carcinoma in situ (DCIS) or invasive breast cancer, during 2018, 2019 and 2020. No exclusion criteria were applied. For the breast surgery variable, we grouped breast conservation and mastectomy together as an indicator that patients underwent surgery. Data on chemotherapy were collected from the respective variable in SEER, which categorizes chemotherapy as ‘yes’ or ‘no/unknown’. We categorized the radiation therapy variable in the same manner as the chemotherapy variable, based on the information available from SEER.
We extracted information from the following variables within SEER for this study: age at diagnosis, race and ethnicity, sex, year of diagnosis, tumor grade, tumor histology, tumor subtype, stage at diagnosis, type of surgery, number of lymph nodes examined, number of positive lymph nodes, chemotherapy, radiation therapy, marital status, median income, population size in area of residence, vital status, and cause of death (COD). These variables were defined and categorized as per Table 1.
Table 1.
Patient Characteristics of invasive breast cancer cases
| Year of diagnosis |
||||||||
|---|---|---|---|---|---|---|---|---|
| 2018–2019 |
2020 |
Total |
||||||
| N | % | N | % | N | % | P value | ||
|
| ||||||||
| All patients | 136,604 | 68.4 | 62,990 | 31.6 | 199,594 | 100.0 | ||
| Age at diagnosis, y | < 50 | 24,999 | 18.3 | 12,037 | 19.1 | 37,036 | 18.6 | < 0.001 |
| 50–64 | 47,756 | 35.0 | 21,764 | 34.6 | 69,520 | 34.8 | ||
| > 64 | 63,849 | 46.7 | 29,189 | 46.3 | 93,038 | 46.6 | ||
| Race and Ethnicity | Non-Hispanic White | 88,183 | 64.6 | 40,339 | 64.0 | 128,522 | 64.4 | < 0.001 |
| Non-Hispanic Black | 14,464 | 10.6 | 6876 | 10.9 | 21,340 | 10.7 | ||
| Non-Hispanic American Indian/Alaska Native | 875 | 0.6 | 377 | 0.6 | 1252 | 0.6 | ||
| Non-Hispanic Asian or Pacific Islander | 13,856 | 10.1 | 6283 | 10.0 | 20,139 | 10.1 | ||
| Hispanic (All Races) | 18,059 | 13.2 | 8430 | 13.4 | 26,489 | 13.3 | ||
| Unknowna | 1167 | 0.9 | 685 | 1.1 | 1852 | 0.9 | ||
| Sex | Female | 135,499 | 99.2 | 62,491 | 99.2 | 197,990 | 99.2 | 0.697 |
| Male | 1105 | 0.8 | 499 | 0.8 | 1604 | 0.8 | ||
| Marital status at diagnosis | Single | 21,171 | 15.5 | 10,100 | 16.0 | 31,271 | 15.7 | < 0.001 |
| Married | 73,163 | 53.6 | 33,949 | 53.9 | 107,112 | 53.7 | ||
| Domestic partner | 814 | 0.6 | 455 | 0.7 | 1269 | 0.6 | ||
| Other (Separated/Divorced/Widowed) | 33,622 | 24.6 | 14,952 | 23.7 | 48,574 | 24.3 | ||
| Unknowna | 7834 | 5.7 | 3534 | 5.6 | 11,368 | 5.7 | ||
| Median household income | ≥ $75,000 | 77,544 | 56.8 | 40,501 | 64.3 | 118,045 | 59.1 | < 0.001 |
| $65,000– $74,999 | 27,612 | 20.2 | 7950 | 12.6 | 35,562 | 17.8 | ||
| $55,000– $64,999 | 16,380 | 12.0 | 7652 | 12.1 | 24,032 | 12.0 | ||
| $45,000– $54,999 | 9782 | 7.2 | 4497 | 7.1 | 14,279 | 7.2 | ||
| $35,000—$44,999 | 4402 | 3.2 | 1989 | 3.2 | 6391 | 3.2 | ||
| < $35,000 | 876 | 0.6 | 395 | 0.6 | 1271 | 0.6 | ||
| Unknowna | 8 | 0.0 | 6 | 0.0 | 14 | 0.0 | ||
| Rural-Urban | Metropolitan areas > = 1 million pop | 83,680 | 61.3 | 38,276 | 60.8 | 121,956 | 61.1 | 0.005 |
| Metropolitan areas of 250,000 to 1 million pop | 29,133 | 21.3 | 13,576 | 21.6 | 42,709 | 21.4 | ||
| Metropolitan areas of < 250,000 pop | 9846 | 7.2 | 4644 | 7.4 | 14,490 | 7.3 | ||
| Nonmetropolitan adjacent to a metropolitan area | 7967 | 5.8 | 3854 | 6.1 | 11,821 | 5.9 | ||
| Nonmetropolitan not adjacent to a metropolitan area | 5793 | 4.2 | 2577 | 4.1 | 8370 | 4.2 | ||
| Unknowna | 185 | 0.1 | 63 | 0.1 | 248 | 0.1 | ||
| Grade | I | 29,657 | 21.7 | 12,808 | 20.3 | 42,465 | 21.3 | < 0.001 |
| II | 60,305 | 44.1 | 28,075 | 44.6 | 88,380 | 44.3 | ||
| III | 38,741 | 28.4 | 18,240 | 29.0 | 56,981 | 28.5 | ||
| Unknowna | 7901 | 5.8 | 3867 | 6.1 | 11,768 | 5.9 | ||
| Histology | Ductal | 109,203 | 79.9 | 50,424 | 80.1 | 159,627 | 80.0 | 0.371 |
| Lobular | 14,085 | 10.3 | 6572 | 10.4 | 20,657 | 10.3 | ||
| Ductal and lobular | 4972 | 3.6 | 2228 | 3.5 | 7200 | 3.6 | ||
| Other | 8344 | 6.1 | 3766 | 6.0 | 12,110 | 6.1 | ||
| Tumor subtype | HR + /HER2− | 97,954 | 71.7 | 44,578 | 70.8 | 142,532 | 71.4 | < 0.001 |
| HR + /HER2 + | 12,713 | 9.3 | 5952 | 9.4 | 18,665 | 9.4 | ||
| HR−/HER2 + | 4968 | 3.6 | 2425 | 3.8 | 7393 | 3.7 | ||
| Triple negative | 13,453 | 9.8 | 6451 | 10.2 | 19,904 | 10.0 | ||
| Unknowna | 7516 | 5.5 | 3584 | 5.7 | 11,100 | 5.6 | ||
| Stage | I | 88,717 | 64.9 | 40,414 | 64.2 | 129,131 | 64.7 | < 0.001 |
| II | 16,508 | 12.1 | 7752 | 12.3 | 24,260 | 12.2 | ||
| III | 8267 | 6.1 | 4028 | 6.4 | 12,295 | 6.2 | ||
| IV | 7843 | 5.7 | 3920 | 6.2 | 11,763 | 5.9 | ||
| Unknowna | 15,269 | 11.2 | 6876 | 10.9 | 22,145 | 11.1 | ||
| T | T0/T1 | 76,698 | 56.1 | 34,270 | 54.4 | 110,968 | 55.6 | < 0.001 |
| T2 | 39,187 | 28.7 | 18,377 | 29.2 | 57,564 | 28.8 | ||
| T3 | 7773 | 5.7 | 3829 | 6.1 | 11,602 | 5.8 | ||
| T4 | 4969 | 3.6 | 2418 | 3.8 | 7387 | 3.7 | ||
| TX | 7977 | 5.8 | 4096 | 6.5 | 12,073 | 6.0 | ||
| N | N0 | 92,899 | 68.0 | 42,208 | 67.0 | 135,107 | 67.7 | < 0.001 |
| N1 | 30,488 | 22.3 | 14,606 | 23.2 | 45,094 | 22.6 | ||
| N2 | 4306 | 3.2 | 1936 | 3.1 | 6242 | 3.1 | ||
| N3 | 3806 | 2.8 | 1729 | 2.7 | 5535 | 2.8 | ||
| NX | 5105 | 3.7 | 2511 | 4.0 | 7616 | 3.8 | ||
| M | M0 | 128,343 | 94.0 | 58,871 | 93.5 | 187,214 | 93.8 | < 0.001 |
| M1 | 7843 | 5.7 | 3923 | 6.2 | 11,766 | 5.9 | ||
| Unknowna | 418 | 0.3 | 196 | 0.3 | 614 | 0.3 | ||
| Status | Alive | 125,774 | 92.1 | 61,120 | 97.0 | 186,894 | 93.6 | – |
| Dead | 10,830 | 7.9 | 1870 | 3.0 | 12,700 | 6.4 | ||
| Cause of death | Alive | 125,774 | 92.1 | 61,120 | 97.0 | 186,894 | 93.6 | – |
| Dead from breast cancer | 6936 | 5.1 | 1278 | 2.0 | 8214 | 4.1 | ||
| Dead from other cause | 3690 | 2.7 | 560 | 0.9 | 4250 | 2.1 | ||
| Dead from unknown cause | 204 | 0.1 | 32 | 0.1 | 236 | 0.1 | ||
Unknown patients are excluded from the comparative analysis
- P values for “Status” and “Cause of Death” variables are not shown due to differences in follow-up time
HER2, human epidermal growth factor receptor 2; HR, hormone receptor; pop, population; y, years
Statistical analysis
The three outcomes of interest of this study were incidence, treatment patterns, and short-term mortality. Incidence was estimated as the number of new breast cancer diagnoses in a given year per 100,000 population, and calculated individually for 2018, 2019, and 2020. All incidence calculations were age-adjusted to the 2000 U.S. standard population using the 19 age groups available in SEER*Stat version 8.4.2. The 95% confidence intervals (95% CIs) are reported. Incidence was estimated for DCIS and invasive breast cancer separately. In both cases, estimates were stratified by sex; and in the case of invasive breast cancer, estimates were stratified also by stage and by tumor subtype.
Changes in treatment patterns between 2018–2019 and 2020 were assessed fitting multivariable logistic regression models for surgery, radiation therapy and chemotherapy. We describe the proportion of patients who received surgery, chemotherapy, and radiation therapy in each year, and stratified by stage. Differences in the proportion distribution over time were evaluated by chi-squared testing.
To account for potential misclassification of COD during the pandemic, we evaluated mortality in the invasive breast cancer cohort, defined as time from breast cancer diagnosis until death from any cause or last follow-up for censored patients through December 31, 2020. We used the Kaplan–Meier failure function to depict mortality probabilities up to 12 months from the time of breast cancer diagnosis to account for the association between follow-up time and year of diagnosis. To assess whether the COVID-19 pandemic led to excess short-term mortality in patients with invasive breast cancer, we compared mortality over 12 months for patients diagnosed with invasive breast cancer in 2020 versus 2018 and 2019. This comparison was tested using a multivariable Cox proportional hazard regression model with diagnosis year 2020 as the reference and report adjusted hazard ratios for 2018 and 2019 with their respective 95% CIs. The Cox model was adjusted for age at diagnosis, race and ethnicity, sex, tumor grade, histology, stage, tumor subtype, surgery, radiation, chemotherapy, marital status, median household income, and rurality. As exploratory analyses, we evaluated breast cancer-specific mortality (BCSM) defined as the interval from breast cancer diagnosis until death from breast cancer or last follow-up for censored patients. The analysis of BCSM was done using cumulative incidence function with other causes of death as competing risk. We also evaluated non-BCSM defined as the interval from breast cancer diagnosis until death from causes other than breast cancer or last follow-up for censored patients. The analysis of non-BCSM was done using cumulative incidence function with BCSM as competing risk. We used Fine and Gray multivariable regression to compare BCSM and non-BCSM over 12 months in patients diagnosed in 2020 as the reference vs 2018 and 2019. All P values were 2-sided, with P < 0.05 considered statistically significant. Statistical analyses were performed using STATA 12.0 (Stata Corporation, College Station, TX) and SPSS 26.0 (IBM Corporation, Armonk, NY).
Ethical declarations
The study was exempted from Institutional Review Board review because we utilized de-identified, previously collected, publicly available data. The study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
Results
Patient characteristics
Between 2018 and 2020, there were 37,834 patients with a diagnosis of DCIS and 199,594 patients with a diagnosis of invasive breast cancer. The distribution of patient characteristics for DCIS cases is shown in Supplementary Table 1 and for the invasive breast cancer cases in Table 1. There were 136,604 cases of invasive breast cancer diagnosed in 2018–2019, and 62,990 cases of invasive breast cancer diagnosed in 2020. The median age and interquartile range (IQR) for invasive cases diagnosed both in 2018–2019 and in 2020 was 63 years (IQR: 53–72 years). At diagnosis, most patients were Non-Hispanic White (64.4%), presented with stage I (64.7%) tumors of intermediate grade (44.3%) and had HR + /HER2− subtype (71.4%).
We observed statistically significant differences of small magnitude between periods of diagnosis. As compared with 2018–2019, invasive cases diagnosed in 2020 were more often in those who were Non-Hispanic Black, had higher income, or resided outside of large metropolitan areas. Cancers had higher stage, higher tumor grade, and were less often HR + /HER2− (Table 1).
Incidence of DCIS and invasive breast cancer
The age-adjusted incidence rates of DCIS and invasive breast cancer during 2018, 2019, and 2020 are shown in Table 2. In general, incidence rates increased from 2018 to 2019 and then declined in 2020. For women, there was a 14.8% relative reduction in incidence of DCIS from 2019 (36.4 cases per 100,000) to 2020 (31.0 cases per 100,000). For invasive breast cancer in women, the incidence rate declined from 184.2 cases per 100,000 in 2019 to 166.6 cases per 100,000 in 2020. In contrast, the incidence of invasive breast cancer in men declined by only 0.1 cases per 100,000 per year during the study period.
Table 2.
Incidence of ductal carcinoma in situ and invasive breast cancer by year
| 2018 | 2019 | 2020 | Relative incidence change from 2019 to 2020 | |
|---|---|---|---|---|
|
| ||||
| DCIS cases (N) | 12,623 | 13,563 | 11,648 | - |
| Invasive BC cases (N) | 67,308 | 69,296 | 62,990 | - |
| Incidence of DCIS in women (cases per 100,000 and 95% CI) | 34.1 (33.5–34.7) | 36.4 (35.8–37.0) | 31.0 (30.4–31.6) | −14.8% |
| Incidence of invasive BC in women (cases per 100,000 and 95% CI) | 180.8 (179.4–182.2) | 184.2 (182.8–185.7) | 166.6 (165.2–167.9) | − 9.6% |
| Incidence of invasive BC in men (cases per 100,000 and 95% CI) | 1.7 (1.6– 1.9) | 1.6 (1.5– 1.8) | 1.5 (1.3– 1.6) | − 6.3% |
| Incidence of invasive BC by stage in women (cases per 100,000 and 95% CI) | ||||
| Stage I | 114.8 (113.6– 115.9) | 120.4 (119.3– 121.6) | 105.8 (104.7– 106.9) | − 12.1% |
| Stage II | 23.1 (22.6-) 23.6 | 22.5 (22.0– 23.0) | 21.2 (20.7– 21.7) | − 5.8% |
| Stage III | 11.4 (11.1– 11.8) | 11.6 (11.3– 12.0) | 11.3 (10.9– 11.6) | − 2.6% |
| Stage IV | 10.3 (10.0– 10.7) | 10.5 (10.1– 10.8) | 10.3 (9.9– 10.6) | − 1.9% |
| Unknown stage | 21.2 (20.7– 21.7) | 19.3 (18.8– 19.7) | 18.1 (17.6– 18.5) | − 6.2% |
| Incidence of invasive BC by subtype in women (cases per 100,000 and 95% CI) | ||||
| HR + /HER2− | 127.2 (126.1– 128.4) | 131.7 (130.5– 132.9) | 116.5 (115.4– 117.6) | − 11.5% |
| HR + /HER2 + | 18.1 (17.6–18.5) | 17.5 (17.1–18.0) | 16.6 (16.2–17.0) | − 5.1% |
| HR−/HER2 + | 7.0 (6.7– 7.3) | 6.9 (6.6– 7.2) | 6.7 (6.4– 6.9) | − 2.9% |
| Triple negative | 18.3 (17.8– 18.7) | 18.8 (18.4– 19.3) | 17.7 (17.3– 18.1) | − 5.9% |
| Unknown subtype | 10.2 (9.9– 10.6) | 9.3 (9.0– 9.6) | 9.1 (8.8– 9.5) | − 2.2% |
| Incidence of invasive BC by age at diagnosis in women (cases per 100,000 and 95% CI) | ||||
| < 50 years | 75.6 (74.3 – 77.0) | 78.7 (77.3 – 80.1) | 74.6 (73.3 – 76.0) | − 5.2% |
| 50–64 years | 271.9 (268.3 – 275.4) | 279.6 (276.1 – 283.3) | 252.7 (249.3 – 256.2) | − 9.6% |
| > 64 years | 436.1 (431.3 – 441.1) | 435.8 (431.0 – 440.6) | 382.2 (377.8 – 386.7) | − 12.3% |
| Incidence of invasive BC by race and ethnicity in women (cases per 100,000 and 95% CI) | ||||
| Non-Hispanic white | 193.5 (191.6 – 195.5) | 196.6 (194.7 – 198.6) | 179.6 (177.7 – 181.5) | − 8.6% |
| Non-Hispanic black | 181.7 (177.4 – 186.2) | 187.9 (183.5 – 192.3) | 170.8 (166.7 – 175.0) | − 9.1% |
| Non-Hispanic American Indian/Alaska Native | 143.5 (129.3 – 158.7) | 163.9 (148.9 – 180.0) | 127.7 (114.6 – 141.8) | − 22.1% |
| Non-Hispanic Asian or Pacific Islander | 160.8 (156.9 – 164.7) | 160.1 (156.3 – 164.0) | 141.1 (137.6 – 144.7) | − 11.9% |
| Hispanic (all races) | 138.7 (135.7 – 141.7) | 146.5 (143.4 – 149.5) | 127.0 (124.3 – 129.8) | − 13.3% |
-Value not displayed as it is not an incidence calculation
BC, breast cancer; CI, confidence interval; DCIS, ductal carcinoma in situ; HER2, human epidermal growth factor receptor 2; HR, hormone receptor
When comparing 2019 with 2020 among the female population, the incidence of invasive breast cancer by stage had relative reductions of 12.1% for stage I (from 120.4 to 105.8 cases per 100,000, respectively), 5.8% for stage II (from 22.5 to 21.2 cases per 100,000, respectively), 2.6% for stage III (from 11.6 to 11.3 cases per 100,000, respectively), and 1.9% for stage IV (from 10.5 to 10.3 cases per 100,000, respectively) (Table 2). There were reductions in incidence rates of invasive breast cancer in women across all tumor subtypes, which were most pronounced in HR + /HER2− breast cancer where the incidence decreased by 11.5%, from 131.7 cases per 100,000 (95% CI: 130.5–132.9) in 2019 to 116.5 cases per 100,000 (95% CI: 115.4–117.6) in 2020 (Table 2).
We observed that the incidence rates of invasive breast cancer in women decreased in 2020 across all age groups with an increasing trend in relative incidence rate reductions as age at diagnosis increased, ranging from a 5.2% relative incidence rate reduction in women aged < 50 years, to a 12.3% reduction in women aged > 64 years (Table 2). We also evaluated incidence rates of invasive breast cancer in women by race and ethnicity and noted a decreased incidence in 2020 across all groups, with a higher impact in the Non-Hispanic American Indian / Alaska Native population which had a relative incidence rate reduction of 22.1% when comparing 2019 with 2020 (Table 2). An exploratory analysis of incidence across age groups with specific screening recommendations is shown in Supplementary Table 2. In women aged < 40, there was a 3.4% reduction in the incidence rate of invasive breast cancer compared to 5.8% reduction in women aged 40–49, and 10.9–11.2% reduction in women aged > 49. For DCIS, there was a similar trend, although the larger reduction in the incidence rates of DCIS for women aged < 40 may be affected by a potential outlier in incidence in 2019. Analysis of incidence of invasive breast cancer by SEER region showed a reduction in incidence in 2020 across all US territories of different magnitude ranging from 1.3% in Greater Georgia to 52.5% in Alaska, the latter being affected by a potential outlier in incidence in 2019 (Supplementary Table 3).
Treatment patterns
In multivariable logistic regression adjusted for important covariates including age, stage, tumor subtype, and others, we saw significant changes in treatment patterns of invasive breast cancer. Specifically, patients diagnosed in 2020 had a relative reduction of 18% in the odds of undergoing surgery as compared with 2018–2019 (P < 0.001), and a relative reduction of 11% in the odds of undergoing radiation therapy as compared with 2018–2019 (P < 0.001) (Table 3). In contrast, there was a relative increase of 7% in the odds of chemotherapy use in 2020 versus 2018–2019 (P < 0.001) (Table 3). Multivariable logistic regression models evaluating treatment patterns among patients with stage I, II and III showed similar results to those from the entire invasive cohort (Supplementary Table 4). The treatments administered for each stage of breast cancer by year of diagnosis are listed in Supplementary Table 5.
Table 3.
Multivariable logistic regression analyses for individual treatments in the invasive breast cancer cohort
| Surgerya |
Radiation therapyb |
Chemotherapyc |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | P | Odds ratio | 95.0% CI for odds ratio |
P | Odds ratio | 95.0% CI for odds ratio |
P | Odds ratio | 95.0% CI for odds ratio |
|||
| Lower | Upper | Lower | Upper | Lower | Upper | |||||||
|
| ||||||||||||
| Year of diagnosis | ||||||||||||
| 2018–2019 | Reference | Reference | Reference | |||||||||
| 2020 | < 0.001 | 0.822 | 0.790 | 0.855 | < 0.001 | 0.893 | 0.875 | 0.911 | < 0.001 | 1.069 | 1.042 | 1.096 |
Model for surgery adjusted for: age at diagnosis, race and ethnicity, sex, tumor grade, histology, tumor stage, tumor subtype, radiation, chemotherapy, marital status, income and rurality
Model for radiation therapy adjusted for: age at diagnosis, race and ethnicity, sex, tumor grade, histology, tumor stage, tumor subtype, surgery, chemotherapy, marital status, income and rurality
Model for chemotherapy adjusted for: age at diagnosis, race and ethnicity, sex, tumor grade, histology, tumor stage, tumor subtype, surgery, radiation, marital status, income and rurality
CI, confidence interval
Mortality outcomes
To evaluate the impact of COVID-19 in the breast cancer population, we evaluated mortality over 12 months from initial breast cancer diagnosis for each diagnosis year. Within 12 months from diagnosis, there were 2995 deaths for patients diagnosed in 2018, 2992 deaths for patients diagnosed in 2019, and 1870 deaths for patients diagnosed in 2020. The distribution of the causes of death for each year overall and by stage are shown in Supplementary Tables 6, 7, 8, 9, 10 and 11. Breast cancer was the most frequent COD in each year. There were 43 deaths from COVID-19 in patients diagnosed in 2019 and 77 deaths from COVID-19 in those diagnosed in 2020.
The overall mortality rate at 12 months for patients diagnosed in 2018, 2019, and 2020 were 4.49% (95% CI: 4.33%–4.65%), 4.37% (95% CI: 4.22%–4.53%), and 4.57% (95% CI: 4.33%–4.82%), respectively (Fig. 1A). In the adjusted Cox model, there were no significant differences in the risk of death within 1 year between patients diagnosed in 2018 and in 2020 (adjusted Hazard Ratio: 0.95; 95% CI: 0.90–1.01; P = 0.13) or between those diagnosed in 2019 and in 2020 (adjusted Hazard Ratio: 0.95; 95% CI: 0.89–1.01; P = 0.08) (Table 4). The cumulative incidence of non-BCSM at 12 months for patients diagnosed in 2018, 2019, and 2020 were 1.39% (95% CI: 1.30%–1.48%), 1.39% (95% CI: 1.30%–1.48%), and 1.67% (95% CI: 1.51%–1.85%), respectively (Fig. 1B). In the adjusted Fine and Gray model, there were no significant differences in the risk of non-BCSM within 1 year between patients diagnosed in 2018 and in 2020 (adjusted Subdistribution Hazard Ratio: 0.91; 95% CI: 0.82–1.02; P = 0.10) or between those diagnosed in 2019 and in 2020 (adjusted Subdistribution Hazard Ratio: 0.91; 95% CI: 0.82–1.01; P = 0.08) (Table 4). We evaluated mortality from any cause and BCSM by individual stages and observed that outcomes during 2020 are not significantly different than prior years (Supplementary Table 12).
Fig. 1.
A Kaplan–Meier failure curve for mortality from any cause over 12 months from initial diagnosis by year of diagnosis and B cumulative incidence curve for non-breast cancer-specific mortality (non-BCSM) over 12 months from initial diagnosis by year of diagnosis
Table 4.
Multivariable Cox regression for mortality from any cause and multivariable Fine and Gray regression for non-breast cancer-specific mortality within 12 months from breast cancer diagnosis
| Variable | Mortality |
Non-breast cancer-specific mortality |
||||||
|---|---|---|---|---|---|---|---|---|
| P | Hazard ratioa | 95.0% CI for hazard rati o |
P | Subdistribution hazard ratioa | 95.0% CI for subdistribution hazard ratio |
|||
| Lower | Upper | Lower | Upper | |||||
|
| ||||||||
| Year of diagnosis | ||||||||
| 2020 | Reference | Reference | ||||||
| 2018 | 0.126 | 0.954 | 0.899 | 1.013 | 0.095 | 0.911 | 0.816 | 1.016 |
| 2019 | 0.083 | 0.949 | 0.895 | 1.007 | 0.075 | 0.908 | 0.816 | 1.010 |
Both models are adjusted for age at diagnosis, race and ethnicity, sex, tumor grade, histology, tumor stage, tumor subtype, surgery, radiation, chemotherapy, marital status, income and rurality
CI, confidence interval
Discussion
Our study showed that in the U.S. there was a significant reduction in the incidence of DCIS and invasive breast cancer during the first year of the pandemic. At the same time, we observed no significant increase in the short-term mortality risk during 2020.
The decrease in incidence observed in our study was most notable in DCIS and early-stage invasive breast cancer. This is likely due to the combination of the suspension of screening programs [20], stay-at-home mandates [21], and public fear to seek care during 2020 [22]. Further supporting this hypothesis are the findings that incidence decreased more in older women and did not decrease in men. Our study showed that overall, there was a 9.6% reduction in breast cancer incidence from 2019 to 2020, which is particularly concerning to observe during a period when the incidence of breast cancer has otherwise been increasing year over year. The incidence observed in our study matches that of approximately 37 years ago, when around 1983 the reported incidence was approximately 165 cases per 100,000 [23]. Recent studies have also documented a significant decrease in the incidence of breast cancer during 2020 in the U.S. and around the world [11, 24, 25]. With a report from Caswell-Jin et al. describing that the largest incidence reduction occurred within the first 2 months of the pandemic [11]. A recent update in the reporting of U.S. breast cancer statistics showed a rebound in incidence of DCIS and invasive breast cancer in 2021. [26].
Despite the screening interruption that occurred during 2020, our study shows a very small ‘stage shift’ between 2018–2019 (the pre-pandemic period) and 2020. This may be due to several reasons, including that the pause in screening was relatively short, patients may have presented for screening once operations resumed, and tumors may not have grown significantly during that time.
Our study showed that there were treatment pattern changes during the first year of the COVID-19 pandemic. Specifically, we observed relative reductions in surgery and radiation, and an increase in chemotherapy use. The reduction in surgeries is most likely due to operating room closures that occurred during the first part of the pandemic, as many institutions followed Centers for Medicare & Medicaid Services recommendations to delay non-essential surgical procedures to preserve personal protective equipment, beds and ventilators [27]. In line with this and given that radiation therapy, when necessary, typically occurs after surgery in the management of breast cancer, it is not surprising to see a reduction in the administration of radiation therapy. The increase of chemotherapy use observed during 2020 occurs during a period where there was a nationwide increase in the trend of chemotherapy administration [28, 29]. Nonetheless, chemotherapy use increased in 2020 despite potential fears of immunosuppression during the pandemic and concerns about the potential COVID-19 disease severity in this setting.
Given the impact of the pandemic and its threat to the population, we expected to see an increase in short-term mortality during 2020 because of COVID-19-related deaths. In fact, our study showed that the risk of death within 12 months from diagnosis was not statistically different between 2020 and both pre-pandemic years. It is remarkable that 2020 was the first year of the pandemic and treatments for COVID disease were mostly in development. In addition, vaccines were approved for public use in December of 2020 but were only available for health care workers at that time, or for participants in the clinical trials that led to their development and approval. Therefore, mortality outcomes observed in this study are reflective of a most likely unvaccinated population. It is also noteworthy that the mortality risk was not increased in 2020 despite 37.1% of patients in 2020 undergoing chemotherapy with its potential immunosuppressive effects. Our study suggests that for patients with an established diagnosis of breast cancer, the first year of the pandemic did not result in an increased risk of mortality. While this is reassuring, the reduction in screening and the treatment disruptions we observed in 2020 are potentially concerning considering the long-term mortality risks associated with a breast cancer diagnosis. [30, 31].
We acknowledge that our study has several limitations. Our mortality findings come from a population of patients with a diagnosis of breast cancer and caution should be taken when extrapolating these findings to other populations. The reasons for focusing on only one cancer type were to conduct an in-depth comprehensive analysis of the incidence, treatment patterns and mortality that could be interpreted appropriately within the context of a homogeneous population; and to be able to adjust the mortality results to specific tumor characteristics (stage of disease, hormone receptor status, HER2 status, tumor histology, tumor grade, among others). These are potential confounders that are unique to breast cancer, and the adjusted analyses allow for a more robust interpretation of the results. At present, only data from the first year of the pandemic (year 2020) are available in SEER and further follow-up is needed to assess the impact during subsequent years. Unfortunately, SEER does not provide incidence data stratified by month and results are available by calendar year, SEER obtains cause-specific death data from death certificates, therefore our cause-specific mortality results such as death from COVID-19, breast cancer, or other specific causes, could be affected by misclassification bias, particularly at a time of crisis such as the pandemic. It is possible that the number of COVID-19 deaths could have been under- or over-reported. To account for this, we chose mortality from any cause as the primary endpoint for the study. Unfortunately, SEER does not provide information on the number of COVID-19 cases and the severity of the course of the disease for those who contracted it. Similarly, detailed information on specific treatments is unavailable in SEER (e.g., chemotherapy regimens, hypofractionated radiation), which prevented analyses on how utilization of those could have changed. Our study includes data from the U.S. and outcomes may differ in other countries, although the reduction in incidence and changes in treatment patterns have been reported worldwide. [7, 24, 25, 32, 33] Lastly, due to the short follow-up in our study, BCSM is immature. Given this, we advise caution when interpreting BCSM results, and further follow-up is required.
Despite the limitations, our study has considerable strengths. The population-based nature of the design allows us to report on the outcomes of the pandemic in the U.S. at the population level. While mortality over 12 months is a short follow-up for BCSM, it is sufficient to evaluate acute mortality from COVID-19. Our study has the strength that the outcomes reported are reflective of a period largely without COVID-19 vaccines, therefore removing possible biases from access to vaccination and differences in COVID-19 severity between vaccinated and unvaccinated populations. Lastly, given that COVID-19 started in 2019, a strength of our study is the inclusion of 2018 as a comparison year given that some deaths during 2019 could have resulted from the early spread of the virus and in fact, 43 COVID-19 deaths were reported in SEER during 2019.
In conclusion, we observed a major decrease in the incidence of breast cancer during the first year of the pandemic, largely driven by a decrease in DCIS and early-stage diagnoses. It will be critical to evaluate future trends in cancer incidence beyond 2020, as it is possible that the reduction in early-stage diagnoses represent missed diagnoses that will be identified at a later stage in subsequent years. The year 2020 was associated with reductions in the use of surgery and radiation; however, despite potential concerns about immunosuppression, chemotherapy use did not decrease. Although vaccines were largely unavailable and COVID-19 treatments were in development, we saw no differences in mortality among patients with a breast cancer diagnosis during the study years. The impact on breast cancer-specific outcomes will require further follow-up.
Supplementary Material
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10549-024-07562-w.
Acknowledgements
We would like to thank Kaitlyn Bifolck for editorial and submission assistance with this manuscript. She is a full-time employee of Dana-Farber Cancer Institute.
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript. This study was presented in part at the 2023 San Antonio Breast Cancer Symposium.
Footnotes
Competing interests JPL received research funding from Kazia Therapeutics and Lilly; and consulting honoraria from Minerva Biotechnologies. SMT reports consulting or advisory role for Novartis, Pfizer (SeaGen), Merck, Eli Lilly, AstraZeneca, Genentech/Roche, Eisai, Sanofi, Bristol Myers Squibb, CytomX Therapeutics, Daiichi Sankyo, Gilead, Zymeworks, Zentalis, Blueprint Medicines, Reveal Genomics, Sumitovant Biopharma, Umoja Biopharma, Artios Pharma, Menarini/Stemline, Aadi Bio, Bayer, Incyte Corp, Jazz Pharmaceuticals, Natera, Tango Therapeutics, Systimmune, eFFECTOR, Hengrui USA, Cullinan Oncology, Circle Pharma, Arvinas, BioNTech, Johnson&Johnson/Ambrx, Launch Therapeutics, Zuellig Pharma, and Bicycle Therapeutics; institutional research support from Genentech/Roche, Merck, Exelixis, Pfizer, Lilly, Novartis, Bristol Myers Squibb, Eisai, AstraZeneca, Gilead, NanoString Technologies, Seattle Genetics, OncoPep, Daiichi Sankyo, and Menarini/Stemline; and travel support from Eli Lilly, Sanofi, Gilead, Jazz Pharmaceuticals, Pfizer, and Arvinas. NUL reports consulting honoraria from Puma, Seattle Genetics, Daiichi-Sankyo, AstraZeneca, Olema Pharmaceuticals, Janssen, Blueprint Medicines, Stemline/Menarini, Artera Inc., and Eisai; institutional research support from Genentech (and Zion Pharmaceutical as part of GNE), Pfizer, Merck, Seattle Genetics (now Pfizer), Olema Pharmaceuticals, and AstraZeneca; royalties from Up to date (book); and travel support from Olema Pharmaceuticals and AstraZeneca.
Ethics approval The study was exempted from Institutional Review Board review because we utilized de-identified, previously collected, publicly available data. The study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
Data availability
The data for this manuscript is publicly available at SEER.
References
- 1.Yin K, Singh P, Drohan B, Hughes KS (2020) Breast imaging, breast surgery, and cancer genetics in the age of COVID-19. Cancer 126(20):4466–4472. 10.1002/cncr.33113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Patt D, Gordan L, Diaz M et al. (2020) Impact of COVID-19 on Cancer Care: How the Pandemic Is Delaying Cancer Diagnosis and Treatment for American Seniors. JCO Clin Cancer Inform 4:1059–1071. 10.1200/CCI.20.00134 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bakouny Z, Paciotti M, Schmidt AL, Lipsitz SR, Choueiri TK, Trinh QD (2021) Cancer Screening Tests and Cancer Diagnoses During the COVID-19 Pandemic. JAMA Oncol 7(3):458–460. 10.1001/jamaoncol.2020.7600 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen RC, Haynes K, Du S, Barron J, Katz AJ (2021) Association of Cancer Screening Deficit in the United States With the COVID-19 Pandemic. JAMA Oncol 7(6):878–884. 10.1001/jamaoncol.2021.0884 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Nyante SJ, Benefield TS, Kuzmiak CM, Earnhardt K, Pritchard M, Henderson LM (2021) Population-level impact of coronavirus disease 2019 on breast cancer screening and diagnostic procedures. Cancer 127(12):2111–2121. 10.1002/cncr.33460 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sprague BL, Lowry KP, Miglioretti DL et al. (2021) Changes in Mammography Use by Women’s Characteristics During the First 5 Months of the COVID-19 Pandemic. J Natl Cancer Inst 113(9):1161–1167. 10.1093/jnci/djab045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rocha A, Freitas-Junior R, Ferreira GLR, Rodrigues DCN, Rahal RMS (2023) COVID-19 and Breast Cancer in Brazil. Int J Public Health 68:1605485. 10.3389/ijph.2023.1605485 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bartlett DL, Howe JR, Chang G et al. (2020) Management of Cancer Surgery Cases During the COVID-19 Pandemic: Considerations. Ann Surg Oncol 27(6):1717–1720. 10.1245/s10434-020-08461-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Dietz JR, Moran MS, Isakoff SJ et al. (2020) Recommendations for prioritization, treatment, and triage of breast cancer patients during the COVID-19 pandemic. the COVID-19 pandemic breast cancer consortium. Breast Cancer Res Treat 181(3):487–497. 10.1007/s10549-020-05644-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bhangu A, Lawani I, Ng-Kamstra JS et al. (2020) Global guidance for surgical care during the COVID-19 pandemic. Br J Surg 107(9):1097–1103. 10.1002/bjs.11646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Caswell-Jin JL, Shafaee MN, Xiao L et al. (2022) Breast cancer diagnosis and treatment during the COVID-19 pandemic in a nationwide, insured population. Breast Cancer Res Treat 194(2):475–482. 10.1007/s10549-022-06634-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Eijkelboom AH, de Munck L, Vrancken Peeters M et al. (2021) Impact of the COVID-19 pandemic on diagnosis, stage, and initial treatment of breast cancer in the Netherlands: a population-based study. J Hematol Oncol 14(1):64. 10.1186/s13045-021-01073-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ilgun AS, Ozmen V (2022) The impact of the COVID-19 pandemic on breast cancer patients. Eur J Breast Health 18(1):85–90. 10.4274/ejbh.galenos.2021.2021-11-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Soran A, Gimbel M, Diego E (2020) Breast cancer diagnosis, treatment and follow-up During COVID-19 pandemic. Eur J Breast Health 16(2):86–88. 10.5152/ejbh.2020.240320 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Al-Shamsi HO, Alhazzani W, Alhuraiji A et al. (2020) A practical approach to the management of cancer patients during the novel coronavirus disease 2019 (COVID-19) pandemic: an international collaborative group. Oncologist 25(6):e936–e945. 10.1634/theoncologist.2020-0213 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zarzaur BL, Stahl CC, Greenberg JA, Savage SA, Minter RM (2020) Blueprint for restructuring a department of surgery in concert with the health care system during a pandemic: the university of Wisconsin experience. JAMA Surg 155(7):628–635. 10.1001/jamasurg.2020.1386 [DOI] [PubMed] [Google Scholar]
- 17.Ribeiro R, Wainstein AJA, de Castro Ribeiro HS, Pinheiro RN, Oliveira AF (2021) Perioperative cancer care in the context of limited resources during the COVID-19 pandemic: Brazilian society of surgical oncology recommendations. Ann Surg Oncol 28(3):1289–1297. 10.1245/s10434-020-09098-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Poggio F, Tagliamento M, Di Maio M et al. (2020) Assessing the impact of the COVID-19 outbreak on the attitudes and practice of Italian oncologists toward breast cancer care and related research activities. JCO Oncol Pract 16(11):e1304–e1314. 10.1200/OP.20.00297 [DOI] [PubMed] [Google Scholar]
- 19.Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER Research Plus Data, 17 Registries, Nov 2022 Sub (2000–2020) - Linked To County Attributes - Time Dependent (1990–2021) Income/Rurality, 1969–2021 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, released April 2023, based on the November 2022 submission. [Google Scholar]
- 20.The American Society of Breast Surgeons. Joint Statement on Breast Screening Exams During the COVID-19 Pandemic. https://www.breastsurgeons.org/news/?id=45
- 21.Moreland A, Herlihy C, Tynan MA et al. (2020) Timing of state and territorial COVID-19 stay-at-home orders and changes in population movement—United States, March 1–May 31, 2020. Morb Mortal Wkly Rep 69(35):1198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Czeisler MÉ, Marynak K, Clarke KE et al. (2020) Delay or avoidance of medical care because of COVID-19–related concerns—United States, June 2020. Morb Mortal Wkly Rep 69(36):1250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Giaquinto AN, Sung H, Miller KD et al. (2022) Breast cancer statistics, 2022. CA Cancer J Clin 72(6):524–541. 10.3322/caac.21754 [DOI] [PubMed] [Google Scholar]
- 24.Eijkelboom AH, de Munck L, Larsen M et al. (2023) Impact of the COVID-19 pandemic on breast cancer incidence and tumor stage in the netherlands and norway: a population-based study. Cancer Epidemiol 87:102481. 10.1016/j.canep.2023.102481 [DOI] [PubMed] [Google Scholar]
- 25.Gathani T, Dodwell D, Horgan K (2023) The impact of the first 2 years of the COVID-19 pandemic on breast cancer diagnoses: a population-based study in England. Br J Cancer 128(3):481–483. 10.1038/s41416-022-02054-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Giaquinto AN, Sung H, Newman LA et al. (2024) Breast cancer statistics 2024. CA Cancer J Clin. 10.3322/caac.21863 [DOI] [PubMed] [Google Scholar]
- 27.Centers for Medicare and Medicaid Services. CMS Releases Recommendations on Adult Elective Surgeries, Non-Essential Medical, Surgical, and Dental Procedures During COVID-19 Response. https://www.cms.gov/newsroom/press-releases/cms-releases-recommendations-adult-elective-surgeries-non-essential-medical-surgical-and-dental
- 28.Tarantino P, Leone J, Vallejo CT et al. (2023) Prognosis and trends in chemotherapy use for patients with stage IA triple-negative breast cancer (TNBC): a population-based study. J Clin Oncol 41(16_suppl):510–510. 10.1200/JCO.2023.41.16_suppl.510 [DOI] [Google Scholar]
- 29.Waks AG, Tarantino P, Freedman RA et al. (2023) Outcomes according to treatment received for small node-negative HER2+ breast tumors in the surveillance, epidemiology, and end results (SEER) database, 2010–2019. J Clin Oncol 41(16_suppl):517–517. 10.1200/JCO.2023.41.16_suppl.51736162037 [DOI] [Google Scholar]
- 30.Caswell-Jin JL, Sun LP, Munoz D et al. (2024) Analysis of breast cancer mortality in the US-1975 to 2019. JAMA 331(3):233–241. 10.1001/jama.2023.25881 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Pan H, Gray R, Braybrooke J et al. (2017) 20-year risks of breast-cancer recurrence after stopping endocrine therapy at 5 years. N Engl J Med 377(19):1836–1846 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Haribhai S, Bhatia K, Shahmanesh M (2023) Global elective breast- and colorectal cancer surgery performance backlogs, attributable mortality and implemented health system responses during the COVID-19 pandemic: a scoping review. PLOS Glob Public Health 3(4):e0001413. 10.1371/journal.pgph.0001413 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Fujita M, Hashimoto H, Nagashima K et al. (2023) Impact of coronavirus disease 2019 pandemic on breast cancer surgery using the national database of Japan. Sci Rep 13(1):4977. 10.1038/s41598-023-32317-w [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data for this manuscript is publicly available at SEER.