Breast size impact on adjuvant radiotherapy adverse effects and dose parameters in treatment planning

A comprehensive literature search for clinical and dosimetric findings was carried out using PubMed/ Medline from January 1990, with 30 September 2017 being the last search date. Only English literature was considered, using the following key words: “breast cancer” and “radiotherapy”. Subheadings were searched with “organ size”, “3D-conformal radiotherapy (3D-CRT)”, “intensity modulated radiotherapy (IMRT)”, “hybrid-IMRT”, and “volumetric-modulated arc therapy (VMAT)”, “organs at risk”, “treatment planning”, “Heart/radiation effects”, “Coronary Vessels/radiation effects”, and “dosimetric comparison”. Additional relevant references were found in reference lists published with the articles. Clinical studies were selected independently of the number of the patients included. We also searched for treatment planning studies with at least 2 different treatment modalities (i.e. 3D-CRT, IMRT (multi-beam and tangential), hybrid-IMRT, and volumetric modulated arc therapy (VMAT) for left-sided breast cancer).

All selected articles were reviewed in full-text versions and were further divided into clinical or treatment planning articles. In clinical studies, we searched for acute skin toxicity, heart and lung toxicity, secondary malignancy risk, and for possible strategies to modify the toxicity, again taking into account the different breast size categories. Treatment planning studies were reviewed in detail and selected only if the delineated organs at risk included at least one additional heart substructure, namely coronary arteries or cardiac chambers.

Clinical studies do not define different breast sizes uniformly. Some of the studies differentiate between breast volumes using measures such as clothing and bra size, where a cup size ≥ D categorises woman as having large breasts.7,18 In a study by Pignol et al., breast size was defined as follows: small (USA bra sizes 32A/B, 34A/B, and 36A), medium (USA bra sizes 32C, 34C, 36B/C, and 38A/B/C), or large (all other).7 Some studies use the distance between the edges of the lateral and medial fields, where a breast separation of approximately 18 and 25 cm constitutes medium and larger breast sizes, respectively.19 Modern three-dimensional treatment planning allows for the target volume to be measured, and clinical target volumes (CTV) of ≥ 1.600 cm³, 975–1.600 cm³, and ≤ 500–975 cm³ have been defined as large, medium, and small breasts, respectively.20,21 One study described a standardised and reproducible protocol to measure breast size (the thickness of left and right axillary fat and nipple-to-pectoral muscle distance), finding that anthropometric measurements correlate with the risk of skin toxicity.16

Randomised clinical trials and retrospective clinical data from standard tangential two-dimensional radiotherapy with wedges (2D-RT) vs. IMRT show an improvement in planning target volume homogeneity and conformity with IMRT, which may have a clinically significant benefit in reducing the rates of acute dermatitis, moist desquamation, pruritus, palpable breast fibrosis, and acute and chronic oedema in women with all breast sizes.7,10,22,23,24,25 A detailed investigation about the IMRT technique across the studies revealed different planning approaches. IMRT was partly defined as a manual forward-planned technique (F-IMRT)7,10,22,23,25,26,27 and partly as an inverse algorithm7,23, hybrid IMRT (H-IMRT)28, and typically used physical compensators and step-and-shoot multi-leaf collimator (MLC) fields7,10,22,23,25,26 or enhanced dynamic wedges and dynamic MLCs.29 A systematic review and meta-analysis of side effects associated with the use of IMRT in adjuvant BC treatment can be found elsewhere.30

In 2008, Pignol et al. reported a correlation of increased moist desquamation anywhere in the breast with BMI, increasing breast separation, smaller vs. larger breast sizes, and with a higher relative volume of the breast receiving > 105–115 % of the prescribed dose.7 In a multivariate analysis, IMRT was associated with a decreased risk of moist desquamation (odds ratio, OR, 0.418, p = 0.0034) while breast size (per 100 cm³) (OR 1.23, p < 0.0001) was associated with an increased risk.7 Moist desquamation was also correlated with a reduction in the global health status scale (p = 0.0019), ≥ G2 pain score, and with an increase in the breast symptoms scale (p = 0.0028).7 G2–4 acute pain was not statistically different between IMRT or conventional radiotherapy arms at the end of the radiation treatment, nor was the data on chronic pain in 241 patients available after 9.8 years of follow-up (OR = 0.74, range 0.432–1.271).7,31

Six other clinical studies reported a comparison of the clinically adverse events in regard to the three groups of breast sizes. Four compared 2D-RT vs. IMRT7,10,23,25,27 in the supine position, one study compared 3D-CRT vs. IMRT in the prone position28, and one study compared conventional (CF) vs. HF in the prone or supine position (Table 1).32 The percentages of patients experiencing ≥ G2 acute breast toxicity categorised in groups of small, medium, or large-sized are presented separately in Figure 1.

Figure 1

Table 1

Selected studies evaluating IMRT versus 2D-RT or 3D-CRT. Patients were further stratified by small, medium or large-sized breasts.

Harsolia et al. found a correlation between ≥ G2 dermatitis, the development of chronic hyperpigmentation, and breast oedema with a larger than average breast size.25 Interestingly, no ≥ G3 toxicity was reported in smaller breast volumes in either treatment modality (2D-RT vs. IMRT). In comparison with 2D-RT, IMRT improved the rates of ≥ G2 acute oedema (36 % vs. 0 %, p < 0.001), ≥ G2 chronic oedema (30 % vs. 3 %, p = 0.007), and ≥ G2 hyperpigmentation (41 % vs. 3 %, p = 0.001) regardless of the breast volume. No statistically significant difference was observed in acute or chronic rates of ≥ G2 acute dermatitis, pain, chronic hyperpigmentation, or breast induration.25

Increased rates of acute dermatitis, acute and chronic oedema, and chronic hyperpigmentation, irrespective of the treatment technique (prone vs. supine) or fractionation (CF vs. HF), were noted in large-breasted patients (volume > 1.600 cm³) by Shah et al., when compared to patients with a smaller breast size. In large-breasted patients, IMRT was superior to the 2D-RT technique in reducing the rates of ≥ G2 acute dermatitis (0 % vs. 19 %, p = 0.02) and oedema (7 % vs. 24 %, p = 0.06).23

Acute skin toxicity, especially moist desquamation, is associated with late complications of radiotherapy, namely telangiectasia and late subcutaneous fibrosis, as shown in a ten-year update of a Canadian breast IMRT trial and other studies.31,33 Five-year results of simple IMRT (F-IMRT) support the use of BC adjuvant radiotherapy technique that improves homogeneity: the benefit of IMRT was confirmed in a multivariate analysis for both overall cosmesis (p = 0.038) and skin telangiectasia (p = 0.031), although there was no difference in breast shrinkage, breast oedema, tumour bed induration, or pigmentation.24

Moderate HF schedule is an established adjuvant treatment in lymph node-negative BC after breast conserving surgery with no differences in disease-related outcomes and with a favourable toxicity profile.34,35,36,37 Moreover, shorter treatment schedules are a cost-effective approach for both the patient and healthcare providers.38 The advantages of a hypofractionated schedule over conventional fractionation, e.g. convenience, a less acute pain, fatigue, and dermatitis, were recently confirmed with prospectively collected physician-assessed data and patient-reported outcome measures in a large comparative analysis by Jagsi et al.37 In this multicentre cohort, the mean breast volume, BMI, and separation distance were slightly smaller in the hypofractionated group: 1270 vs. 980 cm³, 23 vs. 21.9 cm, and 30.8 vs. 28.7, respectively.37

A higher daily fraction size (> 1.8–2.0 Gy) and hot spots (> V105 %) may contribute to so-called ‘triple trouble’ or an unequal distribution of the biological effective dose (BED), although the risk is probably insignificant.39,40 To avoid any of the possible complications (greater fibrosis or late normal tissue effects) with HF, it is generally recommended to limit the volume of hot spots and not to exceed 107 % of the prescribed dose.40 Some authors suggest that patients with a large breast size that precludes achieving the maximum dose of > 107 % should be offered a dose/fractionation that is biologically less intense, for example 45 Gy in 25 daily fractions at 1.8 Gy daily with an addition of a boost dose.41

Similarly as with the CF schedule, high BMI, an increasing PTV volume with a cut off value as small as 500 cm³32,35,42, and excessive radiation dose in the target volume (i.e. V107–110 %) contribute to increased acute skin toxicity.17,42 When CF was compared to HF, the CF schedule was a predictive factor for increased ≥ G2 toxicity.17,32,35 One study reported no differences in acute skin toxicity when HF was compared to CF in similar groups of largebreasted women (volume > 1.500 cm³), BMI > 30, or breast separation > 25 cm. Breast volume was the only patient factor significantly associated with moist desquamation in a multivariable analysis (p = 0.01). A very large breast volume (> 2.500 cm³) had a higher rate of focal moist desquamation (40.7 %) compared to breast volume < 2.500 cm³ (11.1 %) (p = 0.002).11 In a randomised clinical trial by Shaitelman et al., where three-quarters of patients were overweight or obese and half of the patients had a Dmax of 107 % or higher, the authors have confirmed the administration of the HF schedule in regard to acute toxic effects to be safe.36

The dose-volume predictors for acute and late radiation-induced toxicities are established for the lung and heart as a whole structure.43,44,45 Recent studies have evaluated the dose to the whole heart and the proportional increase in cardiac events after BC radiotherapy. An estimated linear increase of 7.4 % and 4.1 % was found per every Gy mean dose to the whole heart for major coronary events and cardiac mortality, respectively.9,45 A systematic literature review on modern radiation doses to heart and lung in BC radiotherapy showed an estimated absolute 30-year risk for cardiac mortality of 1 % for smokers and 0.3 % for non-smokers.9 Patient-related factors (age and smoking), systemic therapy, the fractionation schedule (total dose and daily fraction), and dose-volume parameters of radiation treatment plan such as mean dose to the whole lung and V20, all constitute risk factors for pneumonitis and lung fibrosis.15,43

In a recent retrospective clinical study of 4688 WBRT-treated BC patients, it was reported that larger breast separation (> 22 cm) was one of the factors significantly increasing the mean heart dose (MHD) for CF by 1.5 % per 1 cm and in HF by 1.7 % per every 1 cm increase, respectively.46 It has been demonstrated that the dose to the heart can be significantly reduced in both CF and HF by means of breathing adaptation and prone positioning.13,46

Hannan et al. found that increasing breast size results in increased mean and maximum point heart doses.13 PTV, BMI, or age were generally unrelated to ipsilateral lung dose. The lung dose decreased markedly in the prone compared to supine position for the whole group of patients. Largebreasted prone-treated patients received a higher ipsilateral lung D5 (greatest dose delivered to 5 cm³) compared to small-breasted prone-treated patients (PTV < 1000 cm³), but without significant differences in V5 or V20.13 Breast shape (i.e. pendulous breasts) can contribute to a higher maximum heart dose as well.47 One small dosimetric study in free-breathing supine-position radiation therapy did not find a correlation of increasing breast separation with higher heart doses, but instead found a correlation with an increased dose to the ipsilateral lung (parameters V5 Gy, V10 Gy and V20 Gy), with the greatest increase noted in breast separation between 25 and 27 cm.48 By contrast, Hardee et al. found an opposite association between breast size and lung dosimetric parameters in the prone position (in-field lung volume, V5, and Dmax). All parameters decreased as breast size increased.28

In BC radiotherapy, differences in body habitus may influence doses to organs at risk, but it is not known if small differences in radiation exposure at the time of the first radiation course significantly influence the risk of a secondary non-breast cancer. Compared to the general population, BC patients have an increased risk of secondary non-breast cancers, five or more years after BC diagnosis with and without radiation therapy (RR 1.12; 95 % confidence interval [CI] 1.06–1.19).49 But probably less than 3.5 % of secondary malignancies in BC survivors are attributable to radiation therapy.50 The total dose of radiation, premenopausal age (< 40–45 years) and the irradiated volume of normal tissue all increase the risk for secondary lung, oesophageal, or thyroid cancer, and secondary sarcomas.49,51 The risk of lung cancer increases with the mean dose to the whole lung.52 A systematic review of modern radiation doses to the lung in BC radiotherapy showed an estimated absolute 30-year risk for lung cancer of 4 % for long-term continuing smokers and 0.3 % for non-smokers.9

The radiation dose to the lung increases with lymph node irradiation and the use of IMRT techniques, and decreases with breathing adaptations and prone/lateral decubitus positioning.52 Younger women (< 40 years of age) with an absorbed radiation dose > 1.0 Gy to the contralateral breast have an elevated long-term risk of developing a second primary contralateral BC.52

A study by Bhatnagar et al. suggests that the size of the primary irradiated breast significantly affects the scatter dose to the contralateral breast but not the ipsilateral lung or heart when using IMRT for breast irradiation. The mean volume of the primary irradiated breast in the study was 1167.9 cm³.53 However, Jin et al. found that in the population of women with smaller breasts (360.8 ± 149.1 cm³), the size of the treated breast does not significantly increase the dose to the contra-lateral breast with 2D-RT, 3D-CRT with 3–5 subfields, or tangential IMRT techniques.54

Different strategies in BC radiotherapy exist to lower the dose to organs at risk. Approaches could be further divided according to patient or breast positioning modification, breathing adaptation, and treatment volume reduction.

The prone positioning setup demonstrated to be an excellent strategy to spare the ipsilateral lung in 100 % and heart dose in 85–87 % of the patients, independently of their BMI or breast size12,55,56,57,58, but a particular benefit was observed in large-breasted women (CTV > 1000 cm³).12,13,57 Similar findings were confirmed in a study by Formenti et al. but, in this study, the prone treatment position did not necessarily spare the heart in patients with breast volumes smaller than 750 cm³.57 On the other hand, one study reported having achieved similar heart doses for prone and supine 3D-CRT WBRT in women with large breast volumes (the average treated volume was 1804 cm³ in right-sided breasts and 1500 cm³ in left-sided breasts). They also noted a significantly higher incidental dose to the LADCA in the prone position with left-sided BC.59

3D-CRT lateral isocentric decubitus position was recently described as a treatment planning solution. Long-term toxicity results were published by the Institut Curie group.14,60 Women with a median BMI of 26.3 were treated with different types of fractionation. Acute dermatitis of any grade was present in 93 % of the patients and G3 dermatitis in only 2.8 %. In a 1-year follow-up, 94.1 % of cases had no skin reaction, making this technique feasible with excellent toxicity rates, but the results need to be confirmed with a longer follow-up. In a multivariate analysis, the cup size and the fractionation had a significant influence on acute dermatitis.14

The deep inspiration breath-hold (DIBH) technique helps to minimise the “trade-off” between the target and OAR, a compromise often required, and is less resource-intensive than the IMRT technique. It reduces the low-dose irradiation to the heart, left anterior descending coronary artery (LADCA), and lung, ultimately benefiting women of all breast sizes.61 DIBH can be accurately clinically implemented with an acceptable reproducibility and stability in both supine and prone position.61,62 In a group of women with a volume of the treated breast > 750 cm³, supine voluntary DIBH enabled a cardiac sparing and reproducibility superior to that of free-breathing prone position.63

Another strategy to lower the absorbed radiation dose to the heart is partial breast irradiation. Patients may benefit in terms of lower mean whole heart doses with moderate HF using 3D-CRT and the accelerated partial breast irradiation technique (APBI) with an external beam or interstitial brahcytherapy.64,65,66,67 Meszaros et al. demonstrated the reproducibility of image-guidance intensity modulated APBI and feasibility in terms of acute toxicity and the cosmetic outcome with a median follow up of 3.2 years. In a study, 55 % of the patients had cup size C and 21 % cup size ≥ D.67 Investigators emphasised the necessity of image guidance prior to each radiation fraction to reduce the CTV to PTV margins.67

A thermoplastic bra which helps to raise the lateral breast border is also an option to lower the dose to thoracic organs in BC radiotherapy. Piroth et al. demonstrated an excellent reduction in radiation exposure to the heart (mean dose reduction by ≈ 23 %) and ipsilateral lung (mean dose reduction by ≈ 30 % and V20 by 39.5 %) without additional skin toxicity in women of all breast sizes (clinical target volume ranging from 283.1 to 1581.6 cm³).68

Many different treatment planning options are available in the modern treatment era: 3D-CRT with or without wedged filters, forward-planned IMRT (F-IMRT), inverse-planned IMRT (INV-IMRT), Helical Tomotherapy (HT), VMAT, and hybrid techniques (H-IMRT). The recommended first choice for WBRT varies across numerous treatment planning studies comparing different modalities (i.e. 3D-CRT vs. IMRT vs. VMAT) and usually depends on the available equipment, technical innovations, irradiated volume, treatment planning system with dose calculation algorithm, and skills of the planner.

There are numerous publications comparing the dosimetric parameters of radiotherapy plans, mostly for patients with left-sided early BC. Some of them have been summarised in a review of treatment planning options by Balaji et al. where most authors favoured hybrid planning techniques (3D CRT + IMRT/VMAT) while weighing the target coverage vs. dose to the organs at risk.69 While INV-IMRT is not routinely recommended after breast-only radiation, the use of advanced techniques is increasing in challenging anatomy cases (up to 9.4 % of all BC patients treated with radiotherapy).70 The majority of techniques can be combined with DIBH and/or the prone position. The role of the VMAT technique in clinical practice is still not known precisely and the technique itself is not routinely recommended. The dosimetric data, although promising, need to be validated from a clinical point of view.71 The VMAT technique is sometimes the first choice in complex anatomy cases (including large breasts, bilateral BC adjuvant radiotherapy, pectus excavatum, etc.).71 HT (TomoDirect) can be delivered in a 3D-CRT or IMRT modality in whole-breast adjuvant BC radiotherapy. Both HT modalities have a good PTV coverage and dose homogeneity, but some caution is needed as the dose to ipsilateral lung and heart can be significantly high with the 3D-CRT modality in specific patient anatomic situations. Authors have proposed that simple anatomic measures like maximal heart distance can be helpful in selecting the appropriate treatment strategy.72

For the purpose of this review, we have evaluated the treatment planning studies which included at least coronary arteries or cardiac chambers as organs at risk. Overall, we have found eight treatment planning studies comparing radiotherapy plans in free-breathing CTs.54,73,74,75,76,77,78,79 The studies indicate an improved dose homogeneity with the IMRT54,73,75,78,79 or VMAT73,74 techniques regardless of the PTV volume, but the number of CT study sets compared was relatively low (10–20). The sizes of the target volumes reported by investigators comparing treatment planning approaches were dissimilar, ranging from 296 cm³ (mean value) to 1160 cm³ (median value).74,79 A study by Tan et al. found additional heart subvolumes (left ventricle or LV and anterior myocardial territory or AMT) to be helpful in the IMRT plan optimization process, although there have been no reports available so far for the dose–volume constraints in these two OARs.75

Besides the heart as a whole structure, authors typically delineated LADCA73,74,76,77,78, LV75,76,78,79, right ventricle (RV)76,78, left atrium (LA)76,78, right atrium (RA)76,78, great vessels78, and AMT.54,74,75 The delineation of heart substructures was not uniformly defined or was rarely guided by written instructions, making it difficult to compare the presented studies.54,74,75