Breast cancer mortality in 500 000 women with early invasive breast cancer diagnosed in England, 1993-2015: population based observational cohort study

Abstract

Objectives: To describe long term breast cancer mortality among women with a diagnosis of breast cancer in the past and estimate absolute breast cancer mortality risks for groups of patients with a recent diagnosis.

Design: Population based observational cohort study.

Setting: Routinely collected data from the National Cancer Registration and Analysis Service.

Participants: All 512 447 women registered with early invasive breast cancer (involving only breast and possibly axillary nodes) in England during January 1993 to December 2015, with follow-up to December 2020.

Main outcome measures: Annual breast cancer mortality rates and cumulative risks by time since diagnosis, calendar period of diagnosis, and nine characteristics of patients and tumours.

Results: For women with a diagnosis made within each of the calendar periods 1993-99, 2000-04, 2005-09, and 2010-15, the crude annual breast cancer mortality rate was highest during the five years after diagnosis and then declined. For any given time since diagnosis, crude annual breast cancer mortality rates and risks decreased with increasing calendar period. Crude five year breast cancer mortality risk was 14.4% (95% confidence interval 14.2% to 14.6%) for women with a diagnosis made during 1993-99 and 4.9% (4.8% to 5.0%) for women with a diagnosis made during 2010-15. Adjusted annual breast cancer mortality rates also decreased with increasing calendar period in nearly every patient group, by a factor of about three in oestrogen receptor positive disease and about two in oestrogen receptor negative disease. Considering just the women with a diagnosis made during 2010-15, cumulative five year breast cancer mortality risk varied substantially between women with different characteristics: it was <3% for 62.8% (96 085/153 006) of women but ≥20% for 4.6% (6962/153 006) of women.

Conclusions: These five year breast cancer mortality risks for patients with a recent diagnosis may be used to estimate breast cancer mortality risks for patients today. The prognosis for women with early invasive breast cancer has improved substantially since the 1990s. Most can expect to become long term cancer survivors, although for a few the risk remains appreciable.

Fig 1

Crude annual breast cancer mortality rates and cumulative breast cancer mortality risks in 512 447 women with early breast cancer by time since diagnosis according to calendar period of diagnosis: for all women (top panel), for women aged 50-64 years (who would all have been eligible for screening) according to whether their cancer was screen detected (middle panels), and according to oestrogen receptor (ER) status (bottom panel). Vertical lines are 95% confidence intervals. Numbers of women at risk, numbers of deaths from breast cancer, annual rates, and cumulative mortality risks are given in tables S4 to S6. Further analyses of women by ER status are in figures S4 and S5. Analyses of non-breast cancer mortality and all cause mortality are in figures S23 and S24. Analyses of women with metastatic as well as early breast cancer, and including women who received neoadjuvant therapy, are in figure S25

Fig 2

Adjusted annual breast cancer mortality rates in 512 447 women with early breast cancer with oestrogen receptor (ER) positive or ER negative disease, by various characteristics. For each characteristic, rates are adjusted for all other characteristics shown and also for time since diagnosis, breast cancer laterality, index of multiple deprivation, and region of residence. Vertical lines are 95% confidence intervals. Results for breast cancer laterality, index of multiple deprivation, and region of residence are in figure S6. Further details are in figure S7, and separate rates for years 0-5, 5-15, and ≥15 since diagnosis are in figures S8-S10. Analyses of all cause mortality are in figures S26-S28

Fig 3

Adjusted annual breast cancer mortality rates in 512 447 women with early breast cancer with oestrogen receptor (ER) positive or ER negative disease by calendar period of diagnosis, according to age at diagnosis. Rates are adjusted for whether cancer was screen detected, tumour size, number of positive nodes, tumour grade, time since diagnosis, breast cancer laterality, index of multiple deprivation, and region of residence. Vertical lines are 95% confidence intervals. Further details are in figure S13. Results including only five years of follow-up (so that all age groups have the same length of follow-up) are in figure S11. Results for breast cancer laterality, index of multiple deprivation and region of residence are in figure S12. Results for all cause mortality are in figures S29 and S30

Fig 4

Adjusted annual breast cancer mortality rates in 512 447 women with early breast cancer with oestrogen receptor (ER) positive or ER negative disease by calendar period of diagnosis, according to whether their cancer was screen detected. Rates are adjusted for age at diagnosis, tumour size, number of positive nodes, tumour grade, time since diagnosis, breast cancer laterality, index of multiple deprivation, and region of residence. Vertical lines are 95% confidence intervals. Further details are in figure S14

Fig 5

Adjusted annual breast cancer mortality rates in 512 447 women with early breast cancer with oestrogen receptor (ER) positive or ER negative disease by calendar period of diagnosis, according to tumour size. Rates are adjusted for age at diagnosis, whether cancer was screen detected, number of positive nodes, tumour grade, time since diagnosis, breast cancer laterality, index of multiple deprivation, and region of residence. Vertical lines are 95% confidence intervals. Further details are in figure S15

Fig 6

Adjusted annual breast cancer mortality rates in 512 447 women with early breast cancer with oestrogen receptor (ER) positive or ER negative disease by calendar period of diagnosis, according to number of positive nodes. Rates are adjusted for age at diagnosis, whether cancer was screen detected, tumour size, tumour grade, time since diagnosis, breast cancer laterality, index of multiple deprivation, and region of residence. Vertical lines are 95% confidence intervals. Further details are in figure S16

Fig 7

Adjusted annual breast cancer mortality rates in 512 447 women with early breast cancer with oestrogen receptor (ER) positive or ER negative disease by calendar period of diagnosis, according to tumour grade. Rates are adjusted for age at diagnosis, whether cancer was screen detected, tumour size, number of positive nodes, time since diagnosis, breast cancer laterality, index of multiple deprivation, and region of residence. Vertical lines are 95% confidence intervals. Further details are in figure S17

Fig 8

Cumulative five year breast cancer mortality risks in 156 338 women with early breast cancer diagnosed during 2010-15 by categories of tumour grade, tumour size and number of positive nodes in women with oestrogen receptor (ER) positive or ER negative disease. Figures are split by human epidermal growth factor receptor 2 (HER2) status, age and screening status. Vertical lines are 95% confidence intervals. Points are plotted for groups of women with data on ≥40 women and include 153 006/156 338 (97.9%) of women. Further details, including values of plotted points, are given in tables S9 and S10. Results on a square root scale are in figure S22. Results for all cause mortality are in figure S31