Intensity-modulated adjuvant whole breast radiation delivered with static angle tomotherapy (TomoDirect): a prospective case series

Abstract

Purpose

To report the 2-year outcomes of whole breast intensity-modulated radiotherapy (IMRT) after conserving surgery for early breast cancer (EBC) delivered with static angle tomotherapy (TomoDirect) (TD).

Methods

A prospective cohort of 120 EBC patients underwent whole breast IMRT with TD between 2010 and 2012. Radiation was delivered to a conventionally fractionated whole breast total dose of 50 Gy with TD, followed by a sequential conventionally fractionated tumor bed boost dose of 10–16 Gy with helical tomotherapy (HT). Clinical endpoints include acute and late toxicity, cosmesis, quality of life and local control.

Results

Median follow-up was 24 months (range 12–36 months); maximum detected acute skin toxicity was G0 22 %; G1 63 %; G2 12 % and G3 3 %. Predictors of acute dermatitis were as follows: volume of the whole breast minus boost volume receiving 105, 110 and 115 % of prescription dose, whole breast and boost volume, breast thickness and soft tissue thickness. Late skin toxicity was mild with no >G2 events. Cosmesis was good/excellent in 91.7 % of patients and fair/poor in 8.3 %. Quality of life was preserved over time, but for fatigue, transiently increased.

Conclusion

Adjuvant whole breast IMRT delivered sequentially with both TD and HT provides consistent clinical results. An observed unintended excessive dose outside the tumor bed might increase acute toxicity and eventually affect long-term clinical endpoints. The incorporation of the boost dose within the whole breast phase employing a simultaneous integrated boost (SIB) approach might mitigate this issue.

Introduction

Adjuvant whole breast radiotherapy (WBRT) after conserving surgery (BCS) for early stage breast cancer (EBC) is an integral part of the multimodality treatment strategy for this scenario (Portmans 2007). The most common radiation (RT) approach involves forward-planned conventional wedged tangential beams, with eventual intra-target heterogeneous dose distributions (Ceilley et al. 2005). Intensity-modulated radiotherapy (IMRT) provides evident advantages in terms of dose homogeneity and normal tissue sparing (Harsolia et al. 2007). TomoDirect (TD) is a non-rotational treatment option of the TomoTherapy platform (Accuray Inc., Sunnyvale, CA) allowing for RT planning and delivery with a series of highly modulated linear beam paths, using up to 12 coplanar static beams, with the couch moving through the beam at a constant rate (Franco and Ricardi 2012). The patient is translated along the cranial–caudal axis past the fixed fan beam path during delivery of each field, and beam intensity is modulated by the binary collimator. After each discrete angle, the gantry is rotated to a different position and the patient is again passed through the bore for the delivery of the subsequent field. The application of TD for WBRT is similar to intensity-modulated tangential beams employing a dynamic multileaf collimator sliding window technique except that the patient slides through the fan beam (Reynders et al. 2009). We herein present early results of a prospective case series of WBRT delivered with TD for EBC after BCS.

Methods

Patient cohort

Between May 2010 and May 2012, a total of 120 consecutive patients underwent WBRT with TD after BCS for EBC at the Radiation Oncology Department, AUSL Valle d’Aosta (Aosta, Italy). The Clinical Research and Ethical Review Board of our institution approved the present study. Written informed consent was obtained from all patients. Eligibility criteria were as follows: histologically proven breast adenocarcinoma; prior BCS (quadrantectomy/lumpectomy/wide excision); pTis/pT1/pT2; pN0-N1 stage according to AJCC-UICC staging system (6th edition) and no systemic disease. Exclusion criteria included prior thoracic radiation, synchronous second primary tumor and pregnancy. RT was delivered either immediately after BCS, in patients at low risk of distant failure, or sequentially after adjuvant chemotherapy (CT), for high-risk features.

Set up, simulation and target definition

For set up, patients were positioned on a wingboard with both arms raised alongside the head and radiopaque markers along breast borders. Subsequently, 3–5-mm slice thickness axial images were acquired from the lower mandible aspect to the base of the lungs. The whole-breast clinical target volume (WB-CTV) encompassed breast palpable tissue, with radioopaque wires marking clinically detectable breast borders. The whole-breast planning target volume (WB-PTV) was generated with a 5 mm margin around WB-CTV, but confined to the interior of the patient’s outer contours reduced by 5 mm (excluding heart and lungs when needed). Tumor bed (TB) definition was driven by radioopaque clips placed during surgery. The TB clinical target volume (TB-CTV) was generated with a 5 mm isotropic margin around the TB; the consequent planning target volume (TB-PTV) required a further margin of 5 mm around the TB-CTV. The heart, bilateral lungs and contralateral breast were separately contoured as organs at risk (OARs). The heart was outlined to the pulmonary trunk superiorly, including pericardium and excluding major vessels.

Dose prescription and dose constraints

A prescription dose of 50 Gy/25 fractions was planned to the WB-PTV with TD and a subsequent conventionally fractionated boost dose of 10-16 Gy (according to resection margins) to the TB-PTV using helical tomotherapy (HT). Both RT phases were prescribed to 50 % of respective PTVs. Dose distribution was optimized so that the 95 % of both PTVs received at least 95 % of the prescription dose, minimizing hot spots occurrence (i.e., D _max < 105–107 % of prescribed dose). Dose constraints for OARs were set to V _20Gy < 10 %; V _10Gy < 20 %; V _5Gy < 42 % (ipsilateral lung); V _25Gy < 10 % (heart); maximum dose (D _0.1cc) < 5 Gy (contralateral breast) and V _5Gy < 5 % (contralateral lung). Optimization was addressed to reduce both the mean lung dose (MLD) and mean heart dose (MHD) for ipsilateral lung and heart. Excess irradiation (D _2cc), defined as the percentage of the prescription dose delivered to a volume of 2 cc of normal tissues external to the PTV, was minimized.

Tomotherapy planning

Treatment plans were generated with TomoTherapy Hi-Art (version 4.0.4 or higher) treatment planning software (TPS) (Accuray Inc., Sunnyvale, CA). For each plan, specific field width (FW), pitch (PH) (the TD PH is defined as the distance of couch travel in centimeters per sinogram projection) and modulation factor (MF) are chosen. The dose distribution for each beamlet passing through the target is calculated by a convolution/superposition algorithm (C/S) based on the collapsed cone approach. An iterative least-squares minimization method optimizes the objective function. During the final dose computation, the optimized sinogram is converted into the delivery sinogram, accounting for varying leaf fluence output factors and latency data. A fine calculation 0.196 cm× 0.196 cm dose grid is used (for optimization and calculation). The optimization process is driven by volume- and dose-based objectives, relative penalties and ROI-based weighting factors. For target volumes, minimum and maximum dose values and respective penalties are used in addition to a DVH-based prescription point. OARs objectives are described by maximum dose, DVH-based constraints and respective penalties. For TD planning, 2–4 tangential beams with a 2.5 cm FW were used, representing a suitable compromise between cranial and caudal dose spread (lower than with 5 cm FW) and treatment time (lower than with 1 cm FW). The PH value was set by default to 1/10 of the FW (0.25 cm/projection for the 2.5 cm beam). Beam angles were selected to minimize OARs dose. To account for possible breath-related target movements, 3–5 MLC leaves were opened on the anterior edge of each beam. To calculate the intensity for all leaves required to expand the beam, the average intensity of 2 outermost leaves on the beam edge is used. A 10-mm ring around the WB-PTV was used to help reduce skin overdosage and to improve target dose conformity. Helping structures were created within the body volume outside the WB-PTV where significant hot spots were likely to occur (i.e., at medial/lateral target edges). If necessary, OARs were used as avoidance structures. A planning 2–2.5 MF was used in all plans. The HT plans designed for TB-PTV used a 2.5 cm FW, a PH value set to 0.287 and a maximum MF set to 2. A 10-mm ring structure was generated around the TB-PTV to avoid unnecessary WB-PTV irradiation. TD and HT plans were summed together, for the evaluation of the total dose distribution. Patient-specific quality assurance methods included a 2D dose distribution verification in a coronal or sagittal plane of the Cheese Phantom (Gammex RMI, Middleton, WI) by means of Gafchromic films EBT2 (ISP Inc, NJ) and/or a 3D diode array evaluation with ArcCHECK (Sun Nuclear, Melbourne, FL) (Catuzzo et al. 2012).

Follow-up, toxicity, cosmesis and quality of life assessment

During follow-up, patients were examined at 3–6 months and twice a year afterward, with a clinical examination at every time point, plain chest X-ray and mammography once a year; other examinations were performed if needed. Acute skin toxicity, the primary endpoint of the study, was assessed at the end of WBRT and after 3 months; late skin toxicity was scored from 6 months. The maximal detected toxicity was scored according to the Common Terminology Criteria for Adverse Events (version 4.02), for late effects; the RTOG/EORTC toxicity scale was employed for acute effects (NCI-CTEP 2013; Cox et al. 1995). Skin toxicity endpoints were as follows: erythema, edema, desquamation, ulceration, hemorrhage, necrosis, telangiectasia, fibrosis-induration, hyperpigmentation, retraction and atrophy. Cosmetic results were assessed at the end of RT and at every follow-up examination, using the standards set forth by the Harvard criteria, a cosmetic evaluation method based on a physician-rated scale consisting of different categories, comparing treated and untreated breast. An ‘excellent’ score was assigned when the treated breast looked essentially as the contralateral; a ‘good’ score for minimal but identifiable radiation effects; a ‘fair’ score if significant radiation effects were readily observable; and a ‘poor’ score for radiation-induced severe late effects (Rose et al. 1989). Late skin toxicity and cosmesis are referred to the last examination. To investigate clinical and dosimetric factors predictive of acute skin toxicity, we divided observed events into 2 categories (G0–G1 vs. G2–G3) assessing, in multivariate analysis, any dependence with a certain set of variables (grouped in 2 classes: continuous and categorical). Continuous variables were as follows: age, body mass index (BMI), breast and soft tissue thicknesses (perpendicular distances between rib cage and nipple of non-index breast and between sternum and anterior skin surface, respectively), WB- and TB-PTV volume as a set of dosimetric items measuring the amount of dose exceeding 50 Gy delivered to the WB-PTV (i.e., V _52.5Gy, V_55Gy and V_57.5Gy). Categorical variables were as follows: breast side and quadrant, field number, hormonal therapy, adjuvant chemotherapy, trastuzumab-based therapy, diabetes, hypertension, vasculopathy and smoking status. Quality of life (QoL) was assessed with the EORTC QoL-questionnaire QLQ-C30, to measure general cancer QoL, quantifying patient’s capacity to fulfill the activities of daily living. This tool incorporates 30 items exploring global health status/QoL, 5 functioning domains (physical, role, cognition, emotional and social) and 9 symptom scales (fatigue, pain, nausea/vomiting, dyspnea, insomnia, appetite loss, constipation, diarrhea and financial impact). Each item was scored according to the standard scoring rules as in the EORTC QLQ-C30 Scoring Manual (Fayers et al. 1999). We added the EORTC QLQ-BR23, an EORTC QLQ-C30 supplementary module targeted to breast cancer to assess tumor site-related specific symptoms, treatment-related side effects and disease-specific QoL domains. It is composed of 23 items related to 4 functioning domains (body image, sexual functioning, sexual enjoyment and future perspective) and to 4 symptom scales (systemic therapies side effects, breast symptoms, arm symptoms and upset by hair loss). The scoring methods are similar to those of EORTC QLQ-C30. Both EORTC QLQ-C30 and QLQ-BR-23 were assessed at 4 different time points: before and at the end of RT, 6 months and 1 year after WBRT.

Statistical analysis

Bivariate analysis was performed to assess whether the observed skin toxicity was affected by either continuous or categorical variables. Patients were grouped in 2 classes according to skin toxicity (G0–G1 vs. G2–G3). Frequencies of each continuous variable were grouped in an arbitrary number of bins with the same width to generate a (2 × number of bins)-dimensional 2-way table. By definition, categorical variables yielded (2 × 2)-dimensional 2-way table. Assuming the sample as Gaussian, chi-square test was performed for each 2-way table and corresponding variable. The statistical significance to each chi-square test was carried out by calculating the Pearson’s coefficient, a measure of the correlation between the difference in toxicity and the variable studied. A Pearson’s coefficient value close to 1 would suggest strong correlation;values > 0.3 would be interpreted as correlation trend. Changes in QoL over time were analyzed by the Wilcoxon signed-rank-test. A difference among time points was considered clinically relevant if >10 points (as in Osoba et al. 1998) and statistically significant if p < 0.01.

Results

Clinical characteristics

The 120 patients included in the present analysis achieved a minimum follow-up of 12 months. Mean follow-up time was 24 months (range 12–36 months). Baseline characteristics are detailed in Table 1. Patients were aged >50 years (82 %), with few comorbidities (hypertension—22 %), with a mean BMI of 23.5, and a mean breast and soft tissue thickness of 48.4 mm and 15.9 mm. They were affected with left-sided (58 %) outer quadrants (59 %) tumors, with an invasive primary <2 cm (92 %), node negative (81 %), hormone sensitive (84 %), moderately differentiated (73 %) with ductal histology (88 %), low proliferation index (60 %), no c-erb-B2 amplification (86 %) vascular (94 %) and perineural invasion (91 %). Positive to close resection margins (≤2 mm) were found in 12 % of cases. Most of the patients underwent quadrantectomy/lumpectomy and sentinel lymph node biopsy (77 %); 10 % had an axillary dissection. We had 13 % of pNx cases. Up to 81 % received concomitant hormonal therapy, 25 adjuvant CT and 12.5 % trastuzumab. WBRT was always completed without interruptions due to clinical issues.

Table 1.

Cohort characteristics

	N (%)
Pts characteristics
Age
<50 years	22 (18)
>50 years	98 (82)
Mean (years)	60.3
Laterality
Left-sided	68 (57)
Right-sided	52 (43)
Quadrants
Upper inner quadrant	10 (8)
Lower inner quadrant	11 (9)
Upper outer quadrant	44 (37)
Lower outer quadrant	17 (14)
Across inner quadrants	5 (4)
Across outer quadrants	10 (8)
Across upper quadrants	10 (8)
Across lower quadrants	4 (3)
Central quadrant	9 (7)
Axillary tail involvement	2 (2)
Breast thickness
Mean	48.8 mm
Soft tissue thickness
Mean	15.9 mm
Diabetes
Yes	8 (7)
No	112 (93)
Hypertension
Yes	26 (22)
No	94 (78)
Vasculopathy
Yes	6 (5)
No	114 (95)
Smoking status
Yes	21 (17)
No	99 (83)
Regular alcohol intake
Yes	4 (3)
No	116 (97)
BMI
Mean	23.56
Tumor/treatment characteristics
Pathological tumor stage
pTis	16 (13)
pT1a	11 (9)
pT1b	34 (29)
pT1c	49 (41)
pT2	10 (8)
Pathological nodal stage
pN0	81 (68)
pN1	23 (19)
pNx	16 (13)
Histology
Ductal carcinoma	106 (88)
Lobular carcinoma	8 (7)
Mixed ductal/lobular	5 (4)
Mucinous	1 (<1)
Grading
G1	8 (7)
G2	88 (73)
G3	20 (17)
Not available	4 (3)
Estrogen receptor
>80 %	85 (71)
<80 %	16 (13)
0 %	19 (16)
Progesterone receptor
>80 %	51 (42)
<80 %	50 (42)
0 %	19 (16)
c-erb-B2
Amplification	15 (12)
No amplification	93 (78)
Not available	12 (10)
Ki-67
<20 %	65 (54)
20–40 %	30 (25)
>40 %	13 (11)
Not available	12 (10)
Vascular invasion
Positive	16 (13)
Negative	99 (83)
Not available	5 (4)
Perineural invasion
Positive	8 (7)
Negative	102 (85)
Not available	10 (8)
Surgical margins
Positive	14 (12)
Negative	106 (88)
Surgery
Quad/lump	16 (13)
Quad/lump + SLNB	92 (77)
Quad/lump + SLNB + AD	12 (10)
Concomitant hormonal therapy	97/120 (81)
Tamoxifen-based	42 (35)
Aromatasis inhibitor-based	45 (37)
LH-RH an. + tamoxifen	10 (8)
Previous CT	30/120 (25)
TC	12 (10)
FEC + TXT	6 (5)
CMF	12 (10)
Target therapy
Herceptin	15 (13)

BMI body mass index, Quad quadrantectomy, Lump lumpectomy, SLNB sentinel lymphnode biopsy, AD axillary dissection, LH-RH an. LH-RH analogue, CT chemotherapy, TC docetaxel-cyclophosphamide, FEC fluorouracil-epirubicin-cyclophosphamide, TXT docetaxel, CMF cyclophosphamide-methotrexate-fluorouracil

Dosimetric results

Dosimetric parameters (Table 2) are reported as mean values and corresponding standard deviations. For target coverage, the boost (TB-PTV) dose inhomogeneity was in average ±1 % dose variation, while the whole breast (WB-PTV) dose inhomogeneity was in average ±3 % dose variation (considering only the first RT phase of 50 Gy/25 fr); the WB-PTV minimum significant dose (D _98%) and the maximum significant dose (D _2%) were in average above 94 and 119 %, respectively. Those high doses were located in the proximity of the lumpectomy cavity and related to the dose contribution of the HT-delivered boost phase to the WB-PTV; hence, the mean volumes of the WB-PTV receiving 105, 110 and 115 % of the prescribed dose were 32.6, 19.9 and 11.5 %. Dose to lungs was kept within tolerance levels; ipsilateral lung V ₂₀ and MLD were around 10 % and 6 Gy regardless of boost dose; maximum dose to contralateral lung was around 2 Gy. Heart did not receive high doses (V ₂₅ below 3 % and MHD around 1.5 Gy for left-sided tumors). Contralateral breast was adequately spared (D _max < 3 Gy).

Table 2.

Dosimetric results

	Boost dose
	10 Gy		16 Gy
	Mean	SD	Mean	SD
PTV
WB
D 98 (Gy)	47.2	3.1	48.6	2.7
D 2 (Gy)	59.9	0.5	66.8	2.1
V 95 (%)	95.6	15.2	99.9	5.1
V 105 (%)	38.8	12.6	54.0	10.4
Boost
D 98 (Gy)	58.6	0.7	65.0	2.6
D 2 (Gy)	59.5	7.6	67.6	2.0
V 95 (%)	99.7	1.1	101.0	4.8
V 105 (%)	0.0	0.0	0.0	0.0
WB (excluding boost)
Receiving 52.5 Gy (%)	32.6	9.9	46.0	9.0
Receiving 55 Gy (%)	19.9	10.1	35.2	11.6
Receiving 57.5 Gy (%)	11.5	6.1	22.7	8.3
OARs
Ipsilateral lung
V 5 (%)	20.2	4.4	22.4	6.5
V 10 (%)	14.1	3.7	15.4	4.5
V 20 (%)	9.7	2.9	10.8	3.1
D max (Gy)	53.8	2.3	57.6	4.2
MLD (Gy)	6.0	1.3	6.5	1.3
Contralateral lung
D max (Gy)	1.9	0.9	2.3	1.4
Heart left-sided tumors
V 5 (%)	11.7	9.6	12.3	4.5
V 10 (%)	5.6	3.9	6.6	1.3
V 20 (%)	3.5	2.8	4.2	8.6
V 25 (%)	2.9	2.5	2.4	8.5
MHD (Gy)	1.5	1.5	1.8	2.5
D max (Gy)	29.6	23.7	19.1	23.8
Contralateral breast
D max (Gy)	2.6	1.1	3.0	1.2

SD standard deviation, PTV planning target volume, WB whole breast, OARs organs at risk, MLD mean lung dose, MHD mean heart dose

Tumor control, toxicity, cosmesis and QoL

No local relapse was observed. The maximum acute skin toxicity (Table 3) was Grade 0 in 22 %, Grade 1 in 63 %, Grade 2 in 12 %, Grade 3 in 3 %. No categorical variable was found to be predictor of acute toxicity, while several continuous variables (volume of WB-PTV minus TB-PTV receiving 105, 110, 115 % of prescription dose, whole breast and boost volume, breast thickness and soft tissue thickness) demonstrated some trend toward being predictive of G2–G3 skin acute events versus G0–G1 (Fig. 1). Late skin and subcutaneous toxicity was generally mild (Table 4): no events >Grade 2 were observed. A Grade 1 score was assessed for fibrosis/induration in 2.5 % of patients, for atrophy in 1.6 %, telangiectasia in 0.8 %, hyperpigmentation in 20 % and striae in 3 %. A Grade 2 score was observed only for fibrosis (2 %) and hyperpigmentation (6 %). Cosmetic results (Table 4) were excellent in 68.3 % of patients, good in 23.4 %, fair in 5.8 % and poor in 2.5 %. QoL was generally preserved over time (Fig. 2). The only difference among time points was found for fatigue (between pre-RT and 1-year after RT vs. the end of RT) with a >10 points decrease (p = 0.002).

Table 3.

Acute toxicity

Skin toxicity	Grade	Patients	%
No change over baseline	0	26	22
Follicular, faint or dull erythema/epilation/dry desquamation/decreased sweating	1	75	63
Tender or bright erythema, patchy moist desquamation/moderate edema	2	15	12
Confluent, moist desquamation other than skin folds, pitting edema	3	4	3
Ulceration, hemorrhage, necrosis	4	0	0

Fig. 1.

Predictors of acute toxicity. BMI body mass index, thck thickness, Vol volume (cm³), WB-Vol _105% whole breast volume receiving 105 % of the prescribed dose, V _52.5Gy–V _55Gy–V _57.5Gy whole breast volume minus boost volume receiving 52.5 Gy–55 Gy–57.5 Gy

Table 4.

Late toxicity and cosmesis

Parameter	Grade (%)
	G1	G2	G3	G4
Induration-Fibrosis	3 (2.5)	2 (1.7)	0	–
Atrophy	2 (1.6)	0	–	–
Telangiectasia	1 (0.8)	0	0	–
Hyperpigmentation	24 (20)	6 (5)	–	–
Striae	3 (2.5)	0	–	–
Ulceration	–	0	0	0

Cosmesis
Definition	Poor	Fair	Good	Excellent
	3 (2.5)	7 (5.8)	28 (23.4)	82 (68.3)

Fig. 2.

Quality of life

Discussion

WBRT is considered a standard option for EBC after BCS (Portmans 2007). 3D-conformal techniques (3DCRT) (2 opposed tangential fields) are generally employed, with wedges to compensate heterogeneous breast thickness and blocks to provide adequate target coverage and OARs shielding (Cante et al. 2011). Some concerns might be raised against 3DCRT in breast cancer in terms of intra-target and surrounding soft tissues dose inhomogeneity (large size breasts) and normal tissue irradiation (lung–heart) with potential impact on clinical outcomes (Hurkmans et al. 2002). IMRT has been shown to improve target coverage and dose homogeneity and to spare normal tissue over conventional approaches within dosimetric comparison studies (Hong et al. 1999; Caudell et al. 2007). Clinical endpoints, collected within prospective IMRT studies, demonstrated consistent long-term results, with a low rate of local relapse, mild late effects and good/excellent cosmesis (Keller et al. 2012). Regarding acute toxicity, researchers at FCCC demonstrated a statistically significant reduction with IMRT in the incidence and duration of acute Grade 2/3 dermatitis, compared to conventional radiation, in a retrospective series of 804 consecutive breast cancer patients (maximum detected toxicity Grade 0/1:48 %, Grade 2/3: 52 % for IMRT; 25 and 75 % for conventional RT; p < 0.0001) (Freedman et al. 2009). The study also analyzed the incidence of acute dermatitis during each treatment week, showing a decrease in time spent with acute skin reactions with IMRT (82 % of RT weeks: Grade 0/1, and 18 %: Grade 2/3 dermatitis in the IMRT group vs. 29 and 71 % for conventional; p < 0.001). This has been confirmed within a randomized phase III trial, where Pignol et al. (Pignol et al. 2008) showed a significant reduction (31.2 vs. 47.8 %; p = 0.002) with IMRT over conventional RT (due to increased dose homogeneity) in moist desquamation, an event strongly correlated with Grades 2–3 breast pain, global health status scale reduction and breast symptoms scale increase. Regarding late toxicity and cosmesis, Donovan et al. (2007), within a randomized phase III trial (2D-RT vs. IMRT, designed with change in breast appearance as primary endpoint) reported a reduction in palpable breast induration/negative changes in breast appearance in the IMRT arm (58 vs. 44 %). The 2-year interim results of the Cambridge randomized trial (patients having inhomogeneous plans with standard tangentials were randomized to forward-planned IMRT or standard RT) showed a reduction in the telangiectasia rate with IMRT (Barnett et al. 2012). TD allows image-guided IMRT delivery at discrete angles with tomotherapy using a fixed gantry and provides a suitable solution for clinical situation where beam arrangement is constrained to a limited number of restricted directions (Franco et al. 2011). TD had been investigated in planning and clinical studies in breast RT, but also in other oncological settings (Franco et al. 2011; Gonzalez et al. 2006; Schubert et al. 2011; Borca et al. 2012; Murai et al. 2013; Fiandra et al. 2012). TD provides adequate target coverage of the intact breast, with reduction in high doses to target and OARs over conventional techniques and a decrease in low doses to normal tissue over HT (Schubert et al. 2011). Our results confirm robust dose consistency on both whole breast and boost coverage and lungs–heart–contralateral breast avoidance. Toxicity profile was generally mild both in the acute and in the chronic phase (with short-term follow-up). Cosmesis seems consistent even if assessed with the Harvard criteria, a physician-rated scale comparing the index breast with the contralateral, and not with photographic assessment (more objective as it includes postsurgical/pre-radiotherapy baseline documentation). Quality of life was essentially unaffected by WBRT, apart from transient fatigue increase. Thus, clinical results seem promising, but the crude acute skin toxicity rate (G2–G3 15 %) deserves attention. We investigated clinical and dosimetric factors correlated with acute events: breast/soft tissue thickness and breast/boost volume are well known and related to intra-target dose inhomogeneity (as V _105 % from the whole breast phase) (Pignol et al. 2008). Conversely, adjunctive dose received by the WB-PTV minus TB-PTV (V _52.5Gy, V _55Gy, V _57.5Gy) due to the boost phase needs consideration. The approach we used, with WBRT employing TD and a sequential boost employing HT, lead to the delivery of adjunctive dose to the whole breast volume from the boost phase so that more than 1/3 of the WB-PTV receives 105 % of the prescribed dose; almost 1/5 receives 110 % and more than 1/10 gets 115 %. We ascribed the skin acute toxicity rate to these dosimetric issues. The incorporation of the boost dose within the WBRT phase leads to a decrease in unintended excessive dose outside the TB with a favorable toxicity profile and cosmetic outcome (van der Laan et al. 2007; Bantema-Joppe et al. 2012). A LINAC-based concomitant boost approach has been tested at our institution over the last 8 years with reliable results (Cante et al. 2013). Thus, we are presently running a phase II prospective trial of hypofractionated WBRT delivered with TD using a ‘simultaneous integrated boost’ (SIB) approach to improve dosimetric results and to investigate whether it might affect toxicity.

Conclusion

Adjuvant WBRT after BCS for EBC delivered with a sequential approach employing both TD and HT provides consistent clinical results (mild toxicity, promising cosmesis and QoL). The delivery of unintended excessive dose outside TB might increase acute toxicity and eventually affect other clinical endpoints on a long-term basis. The incorporation of the boost dose within the whole breast phase employing a SIB approach might mitigate this issue.