Breast cancer risk in the Swedish haemangioma cohort: Influence of radiation dose, reproductive factors and familial predisposition

*Erik Holmberg, Gothenburg University 

Keywords: Breast cancer, excess relative risk, TSCE, jonizing radiation

Breast cancer risk in the Swedish haemangioma cohort: Influence of radiation dose, reproductive factors and familial predisposition In Sweden a large number of individuals were treated with ionizing radiation for skin haemangioma in childhood during 1920 to 1965. Two sub cohorts, one in Stockholm and one in Gothenburg, with more than 25,000 individuals have been established to study late health effects after radiation therapy in infancy. Most of the individuals (90%) were treated with different Radium-226 applicators, mainly used as ?-irradiators. The rest of the cohort (10%) was exposed to X-rays mainly contact therapy (= 60 kVp, HVL < 2 mm Al). All persons had their first treatment before the age of 18 months. The number of treatments varied between 1 and 37 with a mean of 1.5 treatments. The aim of the treatments was to give the haemangioma an average dose of 800 R at 1–10 mm depth, which corresponds to an absorbed dose of 7–8 Gy.

The pooled cohort has been matched by record linkage with several national population registers using the unique identification number given to all Swedish residents. Information on cancer incidence comes from the Swedish Cancer Register. Using the Swedish Total Population Register, the Emigration Register and the Swedish Cause of Death Register, the follow-up status and person years at risk are estimated. Date of birth and number of children for the women in the cohorts have been collected from the Swedish Multi-Generation Register. This register could also be used to trace parents, siblings and children to the cohort members.

Several publications (1-5) have been done for the cohort on the breast cancer risk after exposure and the latest publication was with follow-up to 2004. For 17,158 women (99% of the total) a complete follow-up including reproductive history was available. Six hundred seventy-eight women were diagnosed with breast cancer. The cohort had a total of 1,515,308 breast years at risk, calculated from 1958 or date of first treatment to date of first breast cancer, date of death, date of emigration or 31 December 2004. The mean and median absorbed doses of the breasts were 0.29 and 0.04 Gy with a range between < 0.01 Gy to 35.8 Gy.

Analyses have been performed with both an empirical excess relative risk (ERR) model and the two-stage clonal expansion (TSCE) model (5). The TSCE model assumes that the key processes necessary to convert a healthy cell to a cancer cell can be reduced to two basic steps. Since the TSCE model is described by biological parameters it is possible to investigate whether potential consequences of genomic instability (GI) are expressed in the data.

Both models agree very well in their risk predictions and a central risk estimate for the cohort is given by the excess risk at the age of 50 years, about the mean age of breast cancer incidence, of ERR = 0.25 Gy-1 (95% CI 0.14; 0.37). This confirms previous result (2) of ERR = 0.35Gy-1 (95% CI 0.18; 0.59) for a mean breast cancer incidence age of 44 years. A linear dependence of risk on dose was found with no indication of a quadratic term. Though the change of risk with age has a significant uncertainty, both models indicate a relatively moderate decrease of ERR with age. In fact, using an ERR model with a linear dependence of excess risk on age, the TSCE model with GI and the empirical ERR model coincide almost perfectly. Both model improved significantly when accounting for number of children in the baseline risk.

Further analysis accounting for familial breast cancer risk factors will be done in the context of the EpiRadBio project (Combining epidemiology and radiobiology to assess cancer risks in the breast, lung, thyroid and digestive tract after exposures to ionizing radiation with total doses in the order of 100 mSv or below) sponsored by European Union Seventh Framework Programme for Research and Technological Development (FP7). This analysis will have a follow-up until 2009 and include information on breast cancer among mothers and sisters to the women in the haemangioma cohort. Preliminary results of this analysis will be presented on the ASA Conference on Radiation & Health in June 2012.

1. M. Lundell, A. Mattsson, P. Karlsson, E. Holmberg, A. Gustafsson, L. Holm, Breast cancer risk after radiotherapy in infancy: a pooled analysis of two Swedish cohorts of 17,202 infants, Radiat. Res. 151 (1999) 626–632.

2. E. Holmberg, L. Holm, M. Lundell, A. Mattsson, A. Wallgren, P. Karlsson, Excess breast cancer risk and the role of parity, age at first childbirth and exposure to radiation in infancy, Br. J. Cancer 85 (2001) 362–366.

3. DL . Preston, A . Mattsson, E. Holmberg, R. Shore, NG. Hildreth, JD. Boice, Radiation effects on breast cancer risk: A pooled analysis of eight cohorts. Radiat Res. 158(2) (2002) 220-35.

4. E. Holmberg, H. Anderson, M. Lundell, P. Karlsson, The impact of reproductive factors on breast cancer risk—the feasibility of using Swedish population-based registers to account for the effect of confounding in cohort studies, Cancer Causes and Control 16 (2005) 235–243.

5. M. Eidemuller, E. Holmberg, P. Jacob, M. Lundell, P. Karlsson, Breast cancer risk among Swedish hemangioma patients and possible consequences of radiation-induced genomic instability. Mutat Res-Fundam Mol Mech Mutagen 669(1-2) (2009) 48-55.