Cardiac dose reduction during tangential breast irradiation using deep inspiration breath hold: a dose comparison study based on deformable image registration

Postoperative radiotherapy (RT) for breast cancer patients has been widely used since it was proven to reduce the risk of local recurrence and improve long-term survival [1, 2]. However, the absolute benefit of RT on long-term overall mortality is less than its benefit on breast cancer-related mortality. The excess mortality in these patients is mainly due to heart disease and lung cancer [1].

RT for left-sided breast cancer often involves some incidental exposure of the heart and ipsilateral lung to radiation. Exposure of the heart to radiation is related to increased rates of coronary events and cardiac mortality [3]. In tangential RT of left-sided breast, part of the anterior heart, including the left anterior descending artery (LAD), usually receives a non-negligible dose. It is not clear whether this results in myocardial damage or coronary artery damage, or both. Generally, the mean cardiac dose causes a proportional increase in the rate of major coronary events per gray. Darby et al. [3] have reported that the rates of major coronary events increased by 7.4 % per gray.

In recent years, considerable effort has been exerted to identify techniques that reduce the dose to the heart in patients receiving postoperative RT for breast cancer. One method involves respiratory gating using the deep inspiration breath hold (DIBH) technique. During deep inspiration, the heart moves posteriorly and inferiorly due to lung expansion and diaphragmatic movements. Thus, the distance between the chest wall and heart is maximized at or near the deep inspiration. Treatment delivery only at deep inspiration reduces the area of the heart that receives a high dose. Exposure of the ipsilateral lung to irradiation is not consistent with DIBH but the relative volume of the irradiated lung usually decreases. Several groups have reported a reduction in the heart and lung dose using DIBH in both dose planning and clinical studies [4, 5].

Our department adopted DIBH in selected left breast cancer patients with a maximum cardiac distance of ≥ 10 mm. Every patient who was treated using DIBH in our department underwent additional computed tomography (CT) with normal breathing (free-breathing CT [FB-CT]) in the same treatment position. We conducted our current CT planning study using two CT sets (DIBH-CT and FB-CT) in the same patient. When comparing the dose to organs at risk, it is essential that the target coverage is as equal as possible. The female breast is usually defined by an external surface structure. Generally, it can extend from mid-axillary line to mid-sternum, and from the second anterior rib to the sixth anterior rib. Because the shape varies depending on the position, it is not reliable to compare target coverage when the position of patient is different. During full inspiration, the shape of the breast tissue changes as the chest wall inflates and moves anterocranially. Thus, we used a deformable image registration tool to achieve the most consistent dose coverage.

Darby et al. [3] have previously conducted a population-based case–control study that suggested a reliable prediction model of major coronary events based on individual cardiac risk factors and the irradiated cardiac volume. Based on this model, we tried in our present study to predict the absolute risk reduction in women who were treated using RT after breast cancer surgery.

Assuming that the same dose coverage has been achieved, the heart and lung dose parameters were significantly reduced in the present study patients by using respiratory gating. The average heart distance in the tangential field was reduced from 2.1 cm to 0.7 cm, and the mean cardiac dose was reduced from 724 cGy to 279 cGy (-61 %). The reduction in the mean cardiac dose was similar to that of earlier studies but the absolute cardiac dose and range of dose reduction was higher in our current study. Swanson et al. [12] reported a mean heart dose reduction of 4.2 Gy with FB vs. 2.5 Gy (-40 %) with DIBH. All other heart parameters evaluated favored the delivery of DIBH over FB plans. Vikstrom et al. [5] showed an average mean dose of 3.7 Gy with FB for the heart and 1.7 Gy (-54 %) with DIBH. Stranzl et al. [13] reported a mean heart dose reduction of 2.3 Gy with FB vs. 1.3 Gy (-56 %) with DIBH. DIBH also shows benefit when combined to modern technique RT. In a study by Hayden et al. [14], DIBH resulted in a significant reduction in radiation dose to the heart and LAD compared with an FB, utilizing tangential intensity modulated RT (IMRT) plans with simultaneous integral boost. Nissen et al. [4] found that the mean heart dose reduced from 5.2 Gy to 2.7 Gy (-48 %) with a DIBH and IMRT plan. As the prescribed dose in all studies, including our current report, was 50 Gy/25 fx, there are two possible explanations for the discrepancy. First, not all left breast cancer patients were treated using DIBH at our center. We specifically included selected women with a maximum cardiac distance ≥ 10 mm in CT simulation. Second, due to the smaller lung volume of Asian women compared with Caucasians, the distance between the left ventricle and chest wall might be shorter [15, 16]. This is supported by the higher D_max of the heart in the DIBH plan (25–27 Gy vs. 49 Gy) and the lower proportion of patients (59 vs. 6 %) in our current series whose heart was completely outside the fields. Thus, we suggest that DIBH is particularly helpful for patients with a long maximal heart distance and small lung volume, such as Asian women.

With DIBH, we found that the relative reductions in cardiac doses were similar between low- and high-dose regions. For postoperative RT of breast cancer, the aspect that is most responsible for ischemic heart disease, for example, the mean dose or maximum dose of the heart or LAD artery, is still debated [17, 18]. In addition, the part of the heart anatomy most affected by radiation is still inconclusive. In a study by Correa et al. [19], a statistically significant higher prevalence of cardiac stress test abnormalities was seen among left-side irradiated patients vs. those of the right side (8 vs. 59 %). Moreover, most left-sided abnormalities in that study (70 %) were in the LAD artery territory. Mark et al. [20] found that RT causes volume-dependent perfusion defects in approximately 40 % of patients within 2 years of RT. The perfusion defects were significantly associated with the left ventricular volume included within the RT field. With these results, the target of the deleterious radiation in breast cancer treatment is thus the coronary artery. In a study by Nilsson et al. [21], a direct link between the RT field and location of coronary artery stenosis was observed. However, microvascular (fibrotic) damage is also possible after a longer latency period [22].

Although the mechanism is unclear, the relationship between the mean cardiac dose and cardiac disease risk is well established. Furthermore, the LAD artery dose closely correlates with the mean cardiac dose [23]. Thus, estimation of the risk of the radiation-related ischemic heart disease might be most appropriately derived from the cardiac mean dose. In our present study, the average reduction in the cardiac mean dose was 444.8 cGy. The estimated 10-year risks of coronary events were 4.03 and 2.55 % for our patients treated with FB and DIBH, respectively. Although the absolute risk reduction was relatively small (1.48 %), our current findings may be clinically significant given the high incidence of breast cancer and the high prevalence of long-term survivors. The estimated clinical benefit of DIBH was also studied previously by Eldredge-Hindy et al. [10]. The risk of ischemic heart disease was reduced with DIBH in that study, regardless of the baseline cardiac risk, although the largest benefit was observed in the high-risk group. Korreman et al. [24] used the relative seriality model to calculate the expected reduction in cardiac mortality from the use of DIBH in 16 cases. They found that the cardiac mortality probability was reduced from 4.8 % in FB to 0.1 % for DIBH.

Another strategy to minimize radiation exposure is the treatment position. In the prone position, the breast is elongated and falls away from the trunk. Studies comparing supine and prone whole-breast irradiation have shown that the prone position can reduce lung volume exposure [25]. A major disadvantage of the prone setup is the gravity-induced anterior displacement of the heart toward the field edge. To address the problem of a higher heart dose in the prone position, prone DIBH has been attempted. In a study by Mulliez et al. [26], four treatment plans—supine shallow breathing, supine DIBH, prone shallow breathing, and prone DIBH—were compared. The authors found that DIBH was able to reduce the heart dose in both positions, with the results of prone DIBH at least as favorable as those of supine DIBH. While preserving the lung-sparing ability of prone positioning, prone DIBH was able to reduce the heart dose compared with prone shallow breathing. To maximize normal tissue sparing, DIBH in the prone position could be considered, especially for young small-breasted patients who are able to perform prone DIBH.

There were some limitations to our present study. First, our analysis was limited to 32 nonrandom cases. However, it is likely that larger numbers of patients would show similar results. The strength of our current study was that exactly the same patient cohort was included in the FB and DIBH plans. Using the deformable image registration technique, the equality of treated mammary tissue was maintained. Also, both breast-conserving surgery and mastectomies were included in the analysis.

DIBH allows tangential treatment of left-sided breast cancer patients and considerably reduced radiation doses to the heart, lung, and LAD artery without compromising target coverage. For our study patients, the probability of a major coronary event within 10 years was reduced from 4.03–2.55 %.