Breast Cancer Radiotherapy and Coronary Artery Stenosis: Location, Location, Location

Radiation therapy (RT) plays an integral role in the treatment of breast cancer. Lumpectomy followed by whole breast RT provides equivalent outcomes to mastectomy. In a meta-analysis of nearly 42,000 women who were treated within clinical trials, the use of RT after mastectomy or lumpectomy improved local control, breast cancer–specific survival, and overall survival. Unfortunately, the use of RT also has a dark side. In that same Oxford meta-analysis, the hazard ratio for death secondary to heart disease, presumably radiation related, was 1.27. Although the incidence of cardiac events was low in the first 5 years of follow-up, it increased over time and persisted after year 15. In the article that accompanies this editorial, Nilsson et al provide the oncology community with another useful and elegantly performed study that addresses radiation-associated heart disease. Previous studies from this group and others have suggested that RT for breast cancer can clearly increase the risk of cardiovascular disease, including pericarditis, coronary artery disease (CAD), conduction abnormalities, congestive heart failure, and valvular disease. In addition, essentially all of the increased risk of clinically meaningful cardiac events is not manifest until more than 10 years after RT. Given this long latency, prospective studies with clinical end points have been difficult; therefore, much of the current data is derived from population-based studies and a few clinical trials with long follow-up. Additionally clouding the issue is that RT techniques have been improved during the last several decades and much of the cardiac toxicity data come from older series. RT is also associated with reductions in regional perfusion as assessed by single photon emission computed tomography (SPECT) scans, in a manner consistent with microvascular injury, relatively soon after RT (eg, 6 months to 5 years). These perfusion defects seem to largely persist with longer follow-up, but their clinical relevance is not yet known. Previous studies have not clearly defined the anatomic distribution of RT-associated CAD. One would reasonably hypothesize that the increasedriskofCADwouldbedosedependentandwouldmanifest largely in the coronary arteries that are directly within the radiation portal. In the few studies that have considered a dose response, doses are typically described as those received by the entire heart, left ventricle, and/or left anterior descending artery (LAD). Doses to the individual branches/portions of the coronary arteries are not typically considered. Nilsson et al took a meaningful step to address this lack of data in a systematic and logical manner. They examined a Swedish cohort of patients with breast cancer who were identified through their national registry and who had been treated between 1970 and 2003. They then cross referenced this with their coronary angiography registry and found 199 women who were treated for breast cancer and who went on to have coronary angiography at some point after their treatment. A comparison group of patients who were not treated for breast cancer was also identified. A radiologist who was blinded to the radiation information reviewed the angiographies and scored the degree of coronary artery stenosis within 18 segments of the major coronary arteries. Because detailed, three-dimensional dose data were not available for the patients, the authors defined coronary subsegments that were most likely to be directly included within the RT beams for the patients with leftversus right-sided disease, and for the different RT techniques (eg, with or without a separate anterior internal mammary nodal [IMN] field). Among all of the women who received RT, those with left-sided breast cancer had a statistically significant increased rate of stenosis in the coronaryarterybranchesontheleft-anteriorsurfaceoftheheart(themid, distal, and distal diagonal branch of the LAD) when compared with those with right-sided cancer. This makes perfect sense given the location of typical RT fields. Interestingly, there was not an increased risk of stenosis in the left main coronary artery or in the proximal LAD, likely because of their relatively posterior locations. This is illustrated nicely in Figure 2A of the article by Nilsson et al. The rate of stenosis in the proximal right coronary artery is (nonstatistically significantly) slightly higher in the patients with rightversus left-sided disease—again, a logical finding given thelocationoftheproximalrightcoronaryarteryrelativetothetypicalRT fields. The lack of a more dramatic difference may be a result of the frequent use of an anterior IMN field in patients for whom the right and left-sided arteries are both at risk. The patients who received RT were additionally subdivided into those who received high-risk (as termed by the authors) versus low-risk RTbeams,onthebasisofthelocationofthebeamsrelativetothecoronary arteries at risk. High-risk beams included any anterior IMN field or leftsidedtangents.Amongthepatientswithcoronarylesions, thedistribution of lesions was similar in the reference subjects (no breast cancer and no RT) to the patients with breast cancer who either received no RT or received RT using low-risk beams. Conversely, in the patients who were treated with high-risk beams, the distribution of coronary lesions differed from that of the reference subjects. There was a marked increase in rightmain disease in those patients who were treated with a right IMN field, andtherewasasomewhatmoresubtle increase inmidanddistal leftmain and distal diagonal disease in the patients with left-sided disease who received RT using high-risk beams. Again, the RT seems to alter the distribution of coronary lesions. There are several shortcomings that might limit the interpretation of the data. For example, there may have been biases with respect to which patients were sent for an angiogram, on the basis of their Editorials