Published studies illustrating BMH pattern in cancer recurrence

Annual recurrence rates for breast cancer

Kimura et al studied the risk of breast cancer recurrence following surgery in 2209 patients (1). Their Fig. 3 is reproduced here. It displays a bi-modal recurrence pattern (BMH).

Fig. 3 Time-course changes of annual recurrence rates (ARR) according to number of lymph-node metastases

Cumulative incidence of breast cancer recurrence

Khoshnoud et al (2) studied the long-term pattern of disease recurrence among in pre- and post-menopausal patients with primary breast cancer according to estrogen receptor status. The study was based on patients with an operable, invasive breast cancer entered in one of three controlled clinical trials conducted by the Stockholm Breast Cancer Group. We selected those 2,562 patients who had been randomly allocated between adjuvant tamoxifen and no adjuvant systemic therapy. These patients had a known estrogen receptor status. Their Fig.1 is reproduced here.

Fig.1 Cumulative incidence of events among all 2,562 patients according to ER status and randomized treatment allocation (tamoxifen or no tamoxifen)

The cumulative incidence is a manifestation of the bi-modal hazard (BMH). The initial steep slope up to the third year coincides with a rising hazard. When incidence levels off the hazard declines.
-- Hazard rates in localized, regional and distant breast cancers published by Yakovlev et al
-- The link between local recurrence and distant metastases in human breast cancer. Koscielny and Tubiana

Breast cancer recurrence and mortality

The following study illustrates two epidemiological hallmarks of breast cancer:
1. An accelerated mortality and recurrence rate during the first three years
2. An ongoing deceleration thereafter.

The first is a manifestation of the bi-modal hazard rate (BMH) which is explained elsewhere The second hallmark illustrates the improving prognosis of the patient. The longer she lives the better her chances to survive. As cancer progresses patient succeeds living with it better and better.

The objective of the Early Breast Cancer Trialists' Collaborative Group (EBCTCG) (1) study was to examine the relationship between the risk of loco-regional recurrence and long-term breast cancer mortality .Collaborative meta-analyses were undertaken, based on individual patient data, of the relevant randomized trials that began by 1995. Information was available on 42 000 women in 78 randomized treatment comparisons (radiotherapy vs no radiotherapy, 23 500; more vs less surgery, 9300; more surgery vs radiotherapy, 9300). 24 types of local treatment comparison were identified. To help relate the effect on local ( locoregional) recurrence to that on breast cancer mortality, these were grouped according to whether or not the 5-year local recurrence risk exceeded 10% (<10%, 17 000 women; >10%, 25 000 women).

The following figure was reproduced from their WebFigure 2e. The red boxes were added by me. They depict the percent difference between observations made at 5, 10 and 15 years. The curves start with an accelerated phase and gradually decelerate.


1. Kimura, Yasuhiro Yanagita, Tomomi Fujisawa and Tokihiro Koida
Study of time-course changes in annual recurrence rates for breast cancer: data analysis of 2,209 patients for 10 years post-surgery
Breast Cancer Research and Treatment, Volume 106, Number 3 / December, 2007

2. Mahmoud R. Khoshnoud , Tommy Fornander , Hemming Johansson and Lars-Erik Rutqvist
Long-term pattern of disease recurrence among patients with early-stage breast cancer according to estrogen receptor status and use of adjuvant tamoxifen
Breast Cancer Research and Treatment Volume 107, Number 1 / January, 2008

3. Early Breast Cancer Trialists' Collaborative Group (EBCTCG)
Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomized trials
The Lancet Volume 366, Issue 9503, 17 December 2005-6 January 2006, Pages 2087-2106

Back to New Medicine