Left breast irradiation with tangential intensity modulated radiotherapy (t-IMRT) versus tangential volumetric modulated arc therapy (t-VMAT): trade-offs between secondary cancer induction risk and optimal target coverage

Breast cancer is the most common malignancy in the Western World [1]. According to Robert Koch Institute incidence and mortality rate in Germany are reported as being in the range of 114.6/100,000 (incidence) and 23/100,000 (mortality) [2]. Breast cancer as a systemic disease requires a multimodal treatment concept consisting of surgery, chemotherapy and radiotherapy. If necessary, these can be supplemented by hormone and antibody therapy. In case of early breast cancer, breast conserving surgery followed by adjuvant radiotherapy is regarded as standard treatment method. Numerous randomized data including meta-analyses have shown a lower recurrence rate and better long-term survival of 5–10% after adjuvant 3d-conventional irradiation (3d-CRT) in breast cancer patients [3‐5]. However, apart from beneficial effects, irradiation may cause detrimental side effects on normal tissue. Specifically, when irradiating left-sided breast cancer, not only the lungs, and contralateral breast but also the heart with coronary vessels may receive significant dose of ionizing radiation. Recent data from the literature indicated that cardiac dose was associated with an increased rate of coronary events and cardiac mortality [6‐8]. To minimize dose to organs at risk as low as possible, new techniques have been developed which attempt to achieve this aim by increasing homogeneity and conformity of the irradiation and better saving the risk organs [9‐12]. These include the improvement of imaging procedures and further development of treatment planning systems. In addition, new irradiation methods have been introduced, for example deep-inspiration breath-hold (DIBH) in combination with or without tangential intensity modulated radiotherapy (t-IMRT) and tangential volumetric modulated arc therapy (t-VMAT).

The advantage of t-IMRT and t-VMAT technique, compared to former irradiation techniques (3d-CRT), is higher dose conformity and a more uniform dose distribution inside the target volume, especially with atypically shaped target volumes and thus a better protection of normal tissue and organs at risk [13‐15]. In contrast to t-IMRT, in which the irradiation device is positioned at a certain angle around the patient, t-VMAT technology rotates the irradiation device once or several times continuously around the patient. With the t-VMAT technique, it is possible to simultaneously regulate the gantry speed, the treatment field shape using the movement of the MLC leaves and the dose rate during irradiation. Compared to 3D-CRT, additionally to the high conformal dose distribution, improved target volume coverage, t-VMAT can reduce the individual treatment time [16‐18]. One point of concern on t-VMAT and t-IMRT is that at the expense of improved protection of the normal tissue from high radiation doses, the volume of normal tissue receiving low doses increase [19, 20]. This simultaneously increases the risk of developing a second malignancy. In contrast to deterministic side effects, there is no threshold dose value for the development of irradiation-induced malignancies, so that even low doses are significant. This is due to the fact that in the low dose range the changes in the cell cycle are in the foreground compared to cell destruction. Taking into account the excellent outcome of patients with early breast cancer, the risk of irradiation-induced cancer becomes particularly relevant. There are several mathematical formulas for assessing the risk of a second malignancy. One possibility is to estimate the excess absolute risk (EAR) per 10,000 person-years. EAR describes the absolute risk of second cancer of a person exposed to irradiation compared to a control group for specific organs based on the calculation of the organ equivalent dose (OED), which describes the dose-response relationship for cancer caused by irradiation and depends on age at the time of irradiation and age attained [21, 22].

Most of the previous studies compared between 3d-CRT and mainly discussed only the benefits of DIBH in terms of reduction of cardiac dose during left sided breast radiotherapy. Novelty of the present study is that it goes ahead in a further step to compare between two different modern radiation techniques, namely t-IMRT and t-VMAT in DIBH-position, and simultaneously assessed not only the dose reduction to organs at risk but also other parameters such as normal tissue integral dose (NTID), homogeneity- and conformity index as well as the risk of second malignancy.

At low doses (< 2Gy), the dose-response relationship is linear, based on data from atomic bomb attack survivors. At higher doses, as used in radiotherapy, the relationship takes on different forms because there is a different weighting of cell destruction, cell repair and cell regeneration. The dose-response relationship is no longer linear, but decreasing and assumes a plateau. The models for higher doses are based on data from a group of patients with Hodgkin’s lymphoma who received external radiotherapy. Using these data, Schneider and Walsh developed a mathematical model, taking into account various organ-specific and dose-dependent parameters, which allows an exact estimation of the EAR for both low and higher doses [23‐26].The aim of the study was to compare t-IMRT-plans with t-VMAT-plans in deep inspiration breath-hold position (DIBH) for left sided breast cancer after breast-conserving surgery in terms of normal tissue exposure as well as homogeneity and conformity. In addition, the EARs for left and right lung and contralateral breast were calculated and compared for both techniques.

In order to achieve optimal irradiation results for patients with left-sided breast cancer, i. e. best local control, lowest acute and late toxicity and lowest second cancer induction, modern irradiation techniques are used, including DIBH, t-IMRT and t-VMAT. Some studies have shown the superiority of 3D-DIBH over conventional 3D-CRT in regard to improving OAR constraints [9‐11, 18, 32‐34]. Regarding dose conformity and homogeneity there is a clear theoretical advantage for t-IMRT/ t-VMT -plan over 3D (32–33). However, the difference between t-IMRT and t-VMAT is less clear.

In a population of 20 patients with left-sided breast cancer, this planning study compared V5% of total body volume, monitor units (MU), normal tissue integral dose (NTID), as well as homogeneity index (HI), conformity index (CI) and dose to organs at risk for t-IMRT and t-VMAT plans. In addition, the EAR for cancer induction in several organs at risk was estimated for comparison between the two methods.

T-VMAT showed, compared to t-IMRT, a better dose homogeneity and dose conformity. CI and HI are regarded as important quality indicators for irradiation plans and provide recognized assistance in choosing between different irradiation plans. One problem is the inaccurate knowledge of the relationship between these theoretical parameters and the clinical effects of irradiation. In theory, better HI and CI would be expected to be associated with a lower local recurrence rate and fewer radiation-associated side effects. However, there are so far in the literature no studies that provide a clear clinical impact for these parameters. Another factor to consider is the existence of different formulas, both for HI and for CI. Kataria et al [30]. showed a sufficient degree of agreement between different formulae for homogeneity index, so that we limited ourselves to the calculation with one single formula. For both, HI and CI, we chose formulas, which have their optimum value of 1 to make a good comparison between the t-IMRT and t-VMAT.

High conformity and homogeneity, accompanied by a higher number of beams on the target volume, is often at the expense of an increased low-dose exposure for the tumor-surrounding tissue. This may result in a higher risk for the induction of second malignancies, a factor that raised major reluctance in using newer methods compared to conventional 3d-CRT [35]. Our study showed that t-IMRT reduced the V5% of total body, representing the low dose irradiation volume, as well as the normal tissue integral dose significantly, compared to t-VMAT.

Risk of second malignancy formation grows with increasing dose and irradiated volume. Given these facts, the full model for EAR gives a good assessment of the increased secondary cancer risk following irradiation of the respective organs [36, 37]. Mendes et al. demonstrated a relevant risk up to 2.2% of induced-cancer from left breast radiation in a dosimetric planning study, considering the whole thorax organs and Brazilian cancer-incidence [38]. The difference between t-VMAT and t-IMRT with respect to EAR can be observed particularly in the contralateral breast and contralateral lung, both receiving relatively low dose. The difference in exposure of the left lung between the two techniques was much smaller. Nevertheless, the absolute EAR was highest in left lung. Our results compare very favorably with data from the literature. Fogliata et al., as well as Sakthivel et al. showed a lower risk for second malignancy following t-IMRT as compared to t-VMAT [31, 39]. In our study, EAR was calculated exclusively using the full model described by Schneider et al., which best takes into account tissue and dose-dependent cell biological properties such as cell destruction and cell regeneration. Inaccuracies in the estimation could result from the inconsistency and variance in the variable parameters, which are not derived from data of a breast cancer population but from a population of patients surviving atomic bombs and patients irradiated for Hodgkin’s lymphoma. Another uncertainty factor is the heterogeneity in the age of the patients in our study, which has a large influence on the EAR.

It is also suggested that the risk of a second malignancy increases with a higher number of monitor units [37]. In our study, the number of MUs following t-IMRT was significantly lower compared to t-VMAT (31%, p = 0.0000004). Sakthivel et al. evaluated the relationship between number of MUs and EAR. According to their results, however, no strong relationship between the two parameters was seen [39].

Dumane et al. showed significant dosimetric gains with a lower dose to the heart, lungs and contralateral breast following DIBH irradiation of the left reconstructed chest wall and regional nodes [40]. Sakka et al. reported a significant additional decrease in heart and LADCA dose by t-IMRT in DIBH technique compared with t-VMAT [41]. In another dosimetric analysis of left breast irradiation in DIBH position Yu et al. found no difference between t-VMAT and t-IMRT [42].

In left-sided breast cancer irradiation, the most important organs at risk in terms of deterministic late effects are the heart, the coronary arteries as well as the ipsilateral lung. Given the well-known toxicity of numerous agents like doxorubicine and trastuzumab, long-term effects, especially coronary events, may be of crucial importance in the case of left-sided breast irradiation. Cardiovascular complications may progress over time and occur years after initial exposure. Various studies have shown that the risk for subsequent ischemic events increased linearly with the mean heart dose [7, 8, 43]. For older women, cardiovascular damage is a higher mortality risk than breast cancer itself [1]. In our study t-IMRT as compared to t-VMAT reduced the mean heart dose by 26% (p = 0.002) and the mean dose to LADCA by 20% (p = 0.03).

An example for an early late toxic effect is radiation-induced pneumonia, which is well correlated with the dose at the ipsilateral lung. The higher the mean lung dose when irradiating left-sided breast cancer, the higher the risk for developing radiation-induced pneumonia [44‐46]. In contrast to the heart, in our study t-VMAT showed better sparing effects compared to t-IMRT and reduced the mean radiation dose to the ipsilateral lung by 5% (p = 0.006), which may be statistically different albeit not clinically relevant. Other planning studies comparing t-IMRT and t-VMAT have shown contradictory results. Ekambaram et al. demonstrated better homogeneity and conformity for the t-VMAT technique in left-sided breast cancer irradiation by using t-IMRT with 7 beams and t-VMAT with 2 semi arcs. However, in their study t-VMAT also showed better sparing effects, with a lower dose to heart and ipsilateral lung, as well as lower NTID and less MU. In line with these data, Popescu et al. reported superior results for t-VMAT in terms of sparing effects to OAR, while Zhao et al. and Sakka et al. demonstrated better sparing effects for the t-IMRT technique [41, 47‐49].

However, comparisons between the different studies remain difficult because of differences in patient population, dose prescription, delineation and definition of target volumes and radiation techniques; e.g. number of fields and arcs used in t-IMRT and t-VMAT. This may explain the partially contradictory results with regard to HI, CI, MU and OAR sparing.

One interesting question remains: How much of the detected difference in dose distribution was due to the different irradiation techniques and how much was mainly a consequence of different optimization algorithms? This question must remain open since the individual performance, endurance and expertise may play a significant role in medical physics.

Compared to t-VMAT, t-IMRT in left-sided breast irradiation significantly and considerably reduced dose to organs at risk as well as normal tissue integral dose, V5% total body and total number of monitor units. Excess absolute risk (EAR) for the induction of secondary tumours was significantly lower following t-IMRT than t-VMAT, especially in the contralateral lung and breast (2–5/10,000 person years) but not for ipsilateral lung. This risk appears small compared to the survival benefit achieved with adjuvant radiotherapy after breast conserving therapy in this group of early breast cancer patients. However, t-VMAT-plans achieved better homogeneity and conformity, based on index calculation. This may be relevant in individual cases where sufficient coverage of the parasternal nodes is warranted.