Invasive breast cancer and breast cancer mortality after ductal carcinoma in situ in women attending for breast screening in England, 1988-2014: population based observational cohort study

Abstract

Objective: To evaluate the long term risks of invasive breast cancer and death from breast cancer after ductal carcinoma in situ (DCIS) diagnosed through breast screening.

Design: Population based observational cohort study.

Setting: Data from the NHS Breast Screening Programme and the National Cancer Registration and Analysis Service.

Participants: All 35 024 women in England diagnosed as having DCIS by the NHS Breast Screening Programme from its start in 1988 until March 2014.

Main outcome measures: Incident invasive breast cancer and death from breast cancer.

Results: By December 2014, 13 606 women had been followed for up to five years, 10 998 for five to nine years, 6861 for 10-14 years, 2620 for 15-19 years, and 939 for at least 20 years. Among these women, 2076 developed invasive breast cancer, corresponding to an incidence rate of 8.82 (95% confidence interval 8.45 to 9.21) per 1000 women per year and more than double that expected from national cancer incidence rates (ratio of observed rate to expected rate 2.52, 95% confidence interval 2.41 to 2.63). The increase started in the second year after diagnosis of DCIS and continued until the end of follow-up. In the same group of women, 310 died from breast cancer, corresponding to a death rate of 1.26 (1.13 to 1.41) per 1000 women per year and 70% higher than that expected from national breast cancer mortality rates (observed:expected ratio 1.70, 1.52 to 1.90). During the first five years after diagnosis of DCIS, the breast cancer death rate was similar to that expected from national mortality rates (observed:expected ratio 0.87, 0.69 to 1.10), but it then increased, with values of 1.98 (1.65 to 2.37), 2.99 (2.41 to 3.70), and 2.77 (2.01 to 3.80) in years five to nine, 10-14, and 15 or more after DCIS diagnosis. Among 29 044 women with unilateral DCIS undergoing surgery, those who had more intensive treatment (mastectomy, radiotherapy for women who had breast conserving surgery, and endocrine treatment in oestrogen receptor positive disease) and those with larger final surgical margins had lower rates of invasive breast cancer.

Conclusions: To date, women with DCIS detected by screening have, on average, experienced higher long term risks of invasive breast cancer and death from breast cancer than women in the general population during a period of at least two decades after their diagnosis. More intensive treatment and larger final surgical margins were associated with lower risks of invasive breast cancer.

Fig 1

Derivation of study population. BCS=breast conserving surgery; DCIS=ductal carcinoma in situ

Fig 2

Cumulative risk diagnosis of invasive breast cancer (top) and of death from breast cancer (bottom) in 35 024 women with ductal carcinoma in situ (DCIS) detected through screening by year of diagnosis of DCIS. Cumulative risks take into account competing risks from other causes of death. Expected values are based on cancer incidence rates for England and mortality rates for England and Wales. See supplementary text S1 and supplementary tables S7 and S11 for further details. BCS=breast conserving surgery; DCIS=ductal carcinoma in situ. *Number of invasive breast cancers/deaths from breast cancer during interval. †Number of women at risk of invasive breast cancer/death from breast cancer at start of interval

Fig 3

Incidence of ipsilateral invasive breast cancer according to various factors in 24 779 women diagnosed as having unilateral ductal carcinoma in situ (DCIS) as a result of screening during April 2000 to March 2014 and who had surgery. Women with oestrogen receptor positive DCIS and recorded as receiving endocrine treatment were excluded. For each factor, rates are shown relative to the first category shown and adjustment is for all other factors except final margin distance. Final margin distance was not included in adjustment as information on this variable was available only from 2007 onwards. Separate results for “breast conserving surgery with radiotherapy,” for “breast conserving surgery, radiotherapy not recorded,” and for “mastectomy” and results showing final margin distances of 1 mm and 2 mm separately are given in supplementary figures S1 to S3. BCS+RT=breast conserving surgery, radiotherapy recorded; BCS-RT: breast conserving surgery, radiotherapy not recorded. *Tests for trend excluding years 0.5-2.9: crude P=0.40; adjusted P=0.87. †Tests for trend across clear margin categories: crude P=0.001; adjusted P=0.009

Fig 4

Cumulative incidence rates of ipsilateral invasive breast cancer and 95% confidence intervals in 29 044 women with unilateral ductal carcinoma in situ (DCIS) detected as a result of screening during April 2000 to March 2014 and who had surgery. Women with oestrogen receptor positive DCIS and recorded as receiving endocrine treatment were excluded from the top two graphs. BCS+RT=breast conserving surgery, radiotherapy recorded; BCS-RT=breast conserving surgery, radiotherapy not recorded; ER+, endocrine=oestrogen receptor positive DCIS and endocrine treatment recorded; ER+, no endocrine=oestrogen receptor positive DCIS and endocrine treatment not recorded. *Number of ipsilateral invasive breast cancers during interval; †Number of women at risk of ipsilateral invasive breast cancer at start of interval

Fig 5

Cumulative incidence rates of ipsilateral and contralateral invasive breast cancer and 95% confidence intervals in 24 716 women with unilateral ductal carcinoma in situ (DCIS) detected as a result of screening between April 2000 and March 2014 and who had surgery. Women with unknown laterality of subsequent invasive cancer were excluded, as were women with oestrogen receptor positive DCIS and recorded as receiving endocrine treatment. BCS+RT=breast conserving surgery, radiotherapy recorded; BCS-RT=breast conserving surgery, radiotherapy not recorded. *Number of ipsilateral invasive breast cancers during interval. †Number of contralateral invasive breast cancers during interval

Fig 6

Breast cancer mortality among 1273 women registered with invasive breast cancer after previously being diagnosed as having unilateral ductal carcinoma in situ (DCIS) as a result of screening during April 2000 to March 2014 and who had surgery, according to characteristics of original DCIS. For each variable, adjustment was for all other variables shown. BCS+RT=breast conserving surgery, radiotherapy recorded; BCS-RT=breast conserving surgery, radiotherapy not recorded; ER+, endocrine=oestrogen receptor positive DCIS and endocrine treatment recorded; ER+, no endocrine=oestrogen receptor positive DCIS and endocrine treatment not recorded; ER-=oestrogen receptor negative