Long-term side effects of radiotherapy in breast cancer : studies in ischemic heart disease and lung cancer

Abstract: Breast cancer (BC) is the most common cancer in women worldwide. Due to early detection and advances in adjuvant therapies, most women diagnosed with early BC will be cured of their disease, and issues of survivorship are of great importance. Adjuvant radiotherapy (RT) in BC is well established and significantly reduces local recurrences and BC mortality. Still, it usually involves some accidental irradiation to the heart and lungs, which may lead to long-term side effects, mainly ischemic heart disease (IHD) and lung cancer (LC). The overall aim with this thesis was to study IHD and radiation-induced LC in women receiving RT for BC from the early 1990s until recently.In paper I and paper II a cohort of women (n=182) receiving computed tomography (CT)-based RT (3DCRT) for BC during 1992 to 2012, who subsequently were referred to a coronary angiography and treated for coronary stenosis, was studied. Paper I was a reproducibility study with the aim to examine the inter-observer variation in delineation of the coronary arteries (CAs) in CT scans used for 3DCRT planning. All patients treated at one of the participating RT departments (n=32), were selected from the larger cohort, and the CAs were delineated in the patients’ CT-scans by three oncologists independently, with a validated CT-based heart atlas as guideline. Spatial difference between the different delineations, and variance in radiation dose was calculated. The median distance between the centers of the arteries was 2-8 mm for the right coronary artery (RCA), and 1-4 mm for the left main coronary artery (LMCA) and the left anterior descending artery (LAD). The intraclass correlation coefficient (ICC) was derived to quantify the variance in estimated doses. The ICC for mean doses varied from 0.76 to 0.98 for LMCA-LAD, and from 0.73 to 0.92 for RCA, indicating that variation in radiation doses was mainly due to interpatient variation. In conclusion, the study showed high consistency in contouring the CAs in the patients’ planning CTs, in particular the LMCA-LAD. In paper II, the aim was to examine the relationship between radiation dose to the CAs and subsequent coronary stenosis that required a coronary intervention at this location. The CAs were delineated and divided into segments in the 182 patients’ planning-CTs and doses were recalculated based on the dose distribution of the original RT plans. The location of the CA stenosis was identified from the Swedish Coronary Angiography and Angioplasty Register (SCAAR). Mean doses to the heart and the LAD were substantially higher in women receiving left-sided RT compared to right-sided RT. Segment-wise analyses were performed to assess the risk of developing a coronary stenosis that required an intervention at a certain radiation dose. Segments receiving radiation doses < 1 Gray (Gy) were used as reference. The main finding was a five-fold increase in risk of a clinically relevant coronary stenosis in the mid LAD at mean doses over 20 Gy, compared to doses of 0-1 Gy (odds ratio 5.23; 95 % CI (confidence interval) 2.01-13.6). There were iv too few events to calculate increase in risk per Gy. Still, the result of this study supports that the radiation dose to the LAD should be considered at RT planning and kept as low as possible.In paper III and IV, the BcBaSe cohort was used to examine risk of IHD, and radiation-induced LC after adjuvant RT for BC. The BCBaSe consists of 68089 women diagnosed with BC during 1992 to 2012, and 340352 age-matched women without BC diagnosis. In paper III, Cox regression analyses were performed to estimate risk of IHD, by comparing women with BC to women without BC diagnosis, and by comparing left-sided BC to right-sided BC. Kaplan-Meier analysis was performed to assess cumulative incidence of IHD. Women with BC had a lower risk of IHD compared to women without BC diagnosis at follow-up (hazard ratio (HR) 0.91; 95 % CI 0.88-0.95). Women irradiated for left-sided BC had a higher risk of IHD compared to women irradiated for right-sided BC (HR 1.18; 95 % CI 1.06-1.31). The HRs increased with more extensive lymph node involvement and with addition of systemic therapy. The cumulative IHD incidence was increased in women receiving left-sided RT compared to rightsided RT, starting from the first years after RT and sustained with longer followup. In paper IV, Kaplan-Meier analyses were performed to assess cumulative incidence of LC and LC-specific survival. Cox regression analyses were performed to estimate risk of LC after adjuvant RT for BC, comparing women with BC to women without BC diagnosis. Women with BC receiving RT had a higher cumulative incidence of LC compared both to women with BC not receiving RT and women without BC. This became apparent 5 years after RT and increased with longer follow-up. Women with BC receiving RT had a higher risk of LC compared to women without BC diagnosis (HR 2.35; 95 % CI 1.54-3.59). LCspecific survival was significantly higher in women with a prior BC compared to women without a prior BC diagnosis. In paper III and paper IV information on individual dosimetry data was not available. Most women likely received 3DCRT given with tangential fields and were treated before breathing adaption techniques were implemented in Sweden. The results of these studies emphasize the importance of further development and implementing of RT techniques and regimens that lower the cardiac and lung doses.In conclusion, we found that radiation doses to the LAD remained high in women receiving 3DCRT for BC between 1992 and 2012, and were associated with an increased risk of clinically relevant CA stenosis. Delineating the LAD was feasible and the results of these studies support that the LAD radiation dose should be considered in RT treatment planning. The register-based studies confirmed that the risk of IHD was significantly increased in women receiving left-sided RT and that the risk of LC after BC RT was significantly increased in this large cohort of women with BC.