Accelerated Hypofractionated Radiotherapy With Simultaneous Integrated Boost With Volumetric Modulated Arc Technique in Patients With Breast Cancer: a Phase 2 Study.

Purpose To assess feasibility of accelerated hypofractionated radiotherapy with simultaneous integrated boost (SIB) with volumetric modulated arc technique (VMAT) in patients with breast cancer. Methods Total 27 patients after breast conserving surgery (BCS) were included in this study. Patients were planned on 4-dimensional computerized tomogram (4D-CT) and contouring was done using RTOG guidelines. Dose delivered was 34 Gy/10#/2wk to the breast and 40 Gy/10#/2wk to the tumor bed as SIB with VMAT technique. The primary endpoint was grade 2 acute skin toxicity. Doses to the organs at risk were calculated. Toxicities and cosmesis were assessed using RTOG LENT-SOMA and HARVARD/NSABP/RTOG grading scales, respectively. Disease-free survival (DFS) and overall survival (OS) was calculated with Kaplan Meier curves.


Introduction
Breast cancer is the most common cancer among women globally as well as in our country [1]. Radiotherapy (RT) plays an important role in breast cancer management after breast conserving surgery (BCS) or mastectomy. In patients with BCS, whole breast irradiation (WBI) can be delivered with many techniques. These techniques include 2-dimensional (2D), 3-dimensional conformal RT(3D-CRT) with or without deep inspiration breath hold, eld-in-eld intensity modulated RT(FF IMRT), inverse planning IMRT, tomotherapy, image guided RT(IGRT) and proton therapy. Many BCS patients may bene t from a boost to the primary site to prevent recurrences that occurs within 2cm of the primary tumor location [2].
RT contributes by sterilizing the microscopic disease thus reducing the risk of local recurrence [3,4]. There are many techniques and modalities(photons, electrons and brachytherapy) by which boost can be delivered. The optimal modality, timing, dose fractionation and technique of tumor bed boost have not yet been established, especially with hypofractionated radiotherapy. However, for a patient who may bene t from boost, simultaneous integrated boost (SIB) can be one of the techniques for its delivery. It achieves dose conformity, homogeneity and completes the treatment fast in one plan only. If it is planned with volumetric-modulated arc therapy (VMAT), the treatment delivery is fast, and planning on 4D-CTcan improve its localization and onboard imaging can increase the accuracy of the delivery. Tumor bed boost has been shown to be associated with increased acute and late toxicity [5,6]. However, it depends on the total dose, dose per fraction, volume of the boost, modality and the technique used for boost delivery. In majority of the studies SIB was delivered in 3-5 weeks [7][8][9][10][11]. SIB with accelerated hypofractionation can further reduce treatment duration from 3 weeks to 2 weeks.
In this study, we report dosimetry, acute and late toxicities and the cosmetic outcomes in patients with breast cancer post BCS who were treated with accelerated hypofractionated WBI and SIB with VMAT technique over 2 weeks (10 fractions).

Methods
This prospective phase II study was conducted in the Department of Radiation Oncology, Regional Cancer Centre, XXXX, XXXX. Primary objective was to assess grade 2 acute skin toxicity with hypofractionated WBI with SIB completed in 10 fractions. Secondary objectives of the study were to determine dose distribution, target coverage, dose homogeneity dose conformity of the target volume, late toxicity and cosmetic outcomes.

Patient selection
Patients who had undergone BCS were included in this study. Institutional Ethics Committee approval was taken. Informed consent was taken from all the patients. The trial was registered with clinicaltrials.gov no. XXXX. Inclusion criteria were: primary cancer of breast of any histology, age >18-70 years, post BCS with clear margins, healed scar, Karnofsky performance status (KPS) >70, regional nodal radiation when indicated(depending on risk factors) and no distant metastasis. Neoadjuvant or adjuvant chemotherapy was allowed. Adjuvant endocrine therapy was given to patients with hormone receptor positive tumors. Exclusion criteria were: mastectomy, history of prior primary malignancy, prior irradiation to breast or chest, pregnancy and collagen vascular disease.

Radiotherapy planning
All patients were made to lie supine on a carbon ber breast board or wing board or a T bar with ipsilateral arm abducted to 90 0 and face turned to the opposite side. Radiopaque markers were placed for de ning the superior, inferior, medial and lateral borders and the surgical scar. Three skin markings were placed along with the ducials below the breast folds for the purpose of reproducibility and the location of tumor bed with respect to ducials.
All patients underwent a normal free-breathing scan with virtual computerized tomogram (CT) breast simulation. Axial cuts were taken from the mandible to the upper abdomen with a slice thickness of 3 mm. The 4D-CT images with recording of the respiratory signals were acquired, taking organ motion into account. The delineation of the tumor cavity and contouring of the OARs was done by using RTOG guidelines. Contouring for the target volumes were done on maximum intensity projection (MIP) of the 4D-CT. The OARs contoured were heart, bilateral lungs, contralateral breast, esophagus, spinal cord, left anterior descending artery and thyroid.
The affected breast was contoured as the clinical target volume (CTV) excluding 5mm from the skin. An additional 5mm (0.5cm) margin for setup error and motion was then added to CTV to form the planning target volume (PTV), shredding(removing) it from lungs and body by 5mm to spare the skin. The nodal areas, when indicated according to the risk factors, were also contoured following RTOG contouring guidelines.

Boost RT planning
In each patient, tumor bed was delineated using clinical, radiological (mammography/CT/ultrasound of breast), surgical (intra operative notes, external and internal surgical scar location) ndings, seroma cavity and surgical clips location. ITV was generated by contouring tumor bed in all phases of respiratory cycles on 4D-CT. All delineation was done on MIP images. A margin of 5mm was added to the cavity to form PTV BOOST SIB. A dose of 34 Gy/10#/2wk to the PTV TOTAL and 40 Gy/10#/2wk to the PTV SIB BOOST was delivered with IGRT using the RapidArc® technique. Partial arcs were used for RT planning. Dose distribution and target coverage criteria for PTV TOTAL and PTV SIB BOOST were: 98% of volume should receive >95% of dose and 2% volume should receive < 107% of dose. Conformity and homogeneity indices were also calculated for each plan [12][13][14]. Dose constraints given were; ipsilateral mean lung dose (MLD) ≤10Gy, V16Gy < 20% and contralateral lung V5 <5%. Mean heart dose (MHD) <7Gy, V18 <5% for left side and <1% for the right side. LAD Dmax and Dmean <15Gy and <8Gy, respectively from left breast. Contralateral breast Dmean <3Gy. Thyroid V25 and V30 should be <50% and <25%, respectively. Dmax and Dmean for oesophagus 20Gy and <5Gy, respectively. Dmax for the spinal cord and brachial plexus should be <30Gy and <40Gy, respectively.
Cone beam CT was done on the rst three consecutive days and then orthogonal images were taken daily for set-up veri cation. All patient were treated in free breathing.

Assessments
Toxicities: Baseline assessment was done for all the patients. The physicians examined patients for any toxicity every week during treatment, at the treatment completion and during the follow-up visits. First follow up was at 1 month of completion of. Patients were followed every 3 months in the 1st year, every 4 months in the 2 nd year, every 6 months thereafter. Toxicities were scored according to Radiation Therapy Oncology Group (RTOG) and Late Effects on Normal Tissues (LENT)/Subjective, Objective, Management and Analytic (SOMA) grading scale. Acute toxicities are reported at 1 and 3 month of completion of radiotherapy. Late effects are reported at 6 months and 4 years follow up.
Cosmesis: Cosmetic effects were assessed in the treated breast and compared with the opposite breast and also with the baseline photographic evaluation. Both objective and subjective parameters were used. The Harvard/NSABP/RTOG scale proposed by Harris et al. was used to evaluate the cosmetic parameters [15]. Variability in both objective and subjective assessment was evaluated. Changes in terms of colour, shape, size, any swelling, symmetry, texture and position of nipple were noted in the treated breast. The assessment was done at baseline (before the start of radiation treatment), at the time of completion of treatment, at 1 month, 3 months, 6 months, 1 year and 3 years after completion of treatment. The long-term cosmetic effects were reported at 4 years. For subjective evaluation, a standard scale for assessment of cosmetic effect due to RT after BCS was used. For objective quali cation, digital photography of the patient was used, before and after the treatment. Digital photo, in a front view of the patient including the sternal notch and both the breast with a light background with adequate light were taken. Two views with hands by side and hands raised above the head were taken for all the patients. A picture of the scar was also taken by the same person to avoid variability of clicked photos.
For cosmesis, both subjective (hyperpigmentation, change in shape, change in size, nipple changes, heaviness, pain) and objective (skin reaction, overall grade, edema, induration, subcutaneous brosis, tenderness, scar changes and any other skin changes/ulceration) response was considered. All the parameters were noticed for any change with time and graded upon accordingly.
Clinical Outcomes: Disease-free survival (DFS) and overall survival(OS) were summarized by Kaplan-Meier curves. Local recurrence was de ned as recurrence in the in the involved breast, axilla, supraclavicular fossa and internal mammary nodes. Distant metastases were de ned as disease occurring in the other sites. Local recurrence and distant metastases were used to calculate DFS. Time was calculated from the date of completion of RT. OS was de ned from the date of diagnosis till the last follow-up or death due to breast cancer.

Statistical analysis
The purpose of the trial was to reject the experimental treatment from further study if it is too toxic, and to accept it for further study if the toxicity is acceptable. The primary endpoint was grade 2 acute skin toxicity, and other toxicities were considered secondary endpoints. The study was designed as a phase II trial with the following assumptions: Grade 2 skin toxicity ≥ 36% was considered unacceptable, and grade 2 skin toxicity ≤ 11% was considered acceptable. Hence the hypotheses of interest were H 0 : r ³ 36% against H A : r £ 11%, where r is the proportion of patients with grade 2 skin toxicity The type I error rate (a, probability of accepting an overly toxic treatment, a false positive outcome) was set to 5% The type II error rate (b, probability of rejecting an acceptably toxic treatment, a false negative outcome) was set to 10% -i.e., the power is equal to 90% Under these assumptions, using a one-sided Fisher's exact test, the design consists of treating 27 evaluable patients, and if at most 5 patients have grade 2 skin toxicity, the treatment was considered acceptable (5/27=19%) if at least 6 patients have grade 2 skin toxicity, the treatment was considered too toxic (6/27=22%)

Acute Toxicity
Grade 1 and 2 acute skin toxicity was observed in 9 (33%) and 5 (18.5%) patients, respectively (Table 4). Acute grade 2 skin toxicity in patients with and without nodal radiotherapy was 1(14.2%) and 4(20%), respectively. There was no grade 3 acute skin toxicity. This rate of grade 2 acute skin toxicity met the prede ned criteria of ≤ 5/27 for acceptable toxicity.
All of the secondary toxicities at 1 month also met the prede ned criteria for acceptable toxicity. Grade 2 hyperpigmentation, edema and induration were observed in 1 (3.7%), 2 (7%) and 4(14.8%) patients, respectively. At 1 month, patient reported acute toxicities were mild swelling, heaviness and pain in 1(3.7%), 4(14.8%) and 8(29%) patients, respectively. Mild di culty in swallowing was reported by 1(3.7%) patient in whom internal mammary nodes were also treated. None of the patients developed acute radiation pneumonitis. Dmax to the oesophagus in this patient was 32.8Gy.

Late toxicity
Late toxicities were either grade 1 or 2 (Table 5). In comparison to the baseline, toxicities increased till 6 months then decreased after that. Late grade 1 and grade 2 breast induration at 4 years was observed in 4(14.8%) and 1(3.7%) patient, respectively. These were present at baseline also. Breast edema was seen in 2(7.4%) patients at baseline, which reduced at 4 years to 1(3.7%) only. Grade 1 breast brosis was observed in 1(3.7%) patient at 4 years. Grade 1 arm edema was seen in 2(7.4%) patients at baseline, which persisted in 1(3.7%) patient till 4 th year.
Patient reported outcomes were mild to moderate only. At baseline mild to moderate breast pain was reported by 2(7.4%) patients, which became mild at 4 years. Breast heaviness was reported by 2(7.4%) patients at baseline, which persisted till 4 th year. Mild breast shrinkage was reported by 1(3.7%) and 2(7.4%) patients at baseline and 4 years, respectively. Mild arm/shoulder discomfort was reported by 1(3.7%) patient only. Arm swelling at 4 years was reported by only 1(3.7%) patient. There were no grade 3 late toxicities. There was no brachial plexopathy or rib fracture with this schedule. We did not observe any late cardiac or pulmonary toxicity (Table 5).

Discussion
In this study we reported the doses to the target organ, the OARs, acute and late toxicities and the cosmesis in breast cancer patients post BCS who were treated with accelerated hypofractionated locoregional RT schedule of 34Gy/10#/2week (3.4Gy/fraction) to the whole breast and 40Gy (4Gy/fraction) to the tumor area with SIB with VMAT technique in 12 days. Dose constraints were achieved in the majority of patients with low rates of acute and late toxicities. There was no adverse cosmesis. Local control and survival were good with this schedule. Since grade 2 skin toxicity occurred in 5 (18.5%) patients, this treatment is acceptable according to the assumption in null hypothesis for this study.
WBI dose fractionation has changed over the years. We modi ed dosimetric constraints for the lung to V16 and heart to V18, which would be biologically equivalent to V20 and V25 of the conventional schedule, respectively. MHD dose in the current study was 7.25Gy, which may be because of the partial arcs, which were used for planning. This MHD may not be acceptable currently because of the risk of late-term cardiac complications. In a study by Darby et al. they reported that the rate of major coronary events increased by 7.4% for each 1Gy increase in MHD. They also demonstrated a threshold MHD of 3 Gy, implying an attributable absolute increased cardiac mortality of 0.5 to 0.7% for women <50 years depending on number of cardiac risk factors. As per their observations MHD was a better predictor of coronary events than the mean LAD dose and these events started within 5-years of treatment [16].
However, their study was from 2D era based on average anatomy and lacked individual dosimetric information hence its rami cations remains unresolved. Recently we published our results at 5-year with this schedule with 2D technique. We did not encounter excess late arm/shoulder and cardiac toxicity, although 5-year may not be adequate to report cardiac toxicities [17]. MHD of 7.25Gy in the current study is higher so there might be a risk of coronary events in the future. Earlier studies have also reported that VMAT increases MLD, MHD and dose to the opposite breast [18]. Considering this risk 3D-CRT with deep inspiration breath hold, inverse planned xed eld IMRT, treatment in prone position, hybrid techniques of combining tangential IMRT with VMAT and proton therapy may be more appropriate in achieving lower MHD and doses to other OARs [19][20][21]. IMRT has been shown to improve target coverage and reduce dose to the OARs [19]. Taylor et al. in another population-based study calculated the absolute risk of mortality from lung cancer at 5Gy MLD and ischemic heart disease at 4Gy MHD after breast RT for smokers and non-smokers to be 0.3% and 4.4%; and 0.3% and 1.2%, respectively. However, these doses were estimated retrospectively [22]. In a recent study Merzenich et al. reported that average MHD of 4.6Gy for left-sided breast RT and only pre-exiting cardiac disease was associated with risk of cardiac death[23]. While another study reported V25 and V30 to be detrimental to the heart[24].
In our previous study with 3D-CRT in patients with left-sided breast cancer postmastectomy; MHD, LAD, proximal LAD, and distal LAD doses were 3.364 Gy, 16.06 Gy, 2.7 Gy, and 27.5 Gy, respectively. Left MLD, V10, and V20 were 5.96 Gy, 14%, and 12.4%, respectively. Mean dose to the right lung and the opposite breast was 0.29 Gy and 0.54 Gy, respectively. V25 for heart was 4.25% [25]. In another study with 3D-CRT in left-sided patients with BCS; MHD in the supine and prone positions was 4.55 Gy and 2.06 Gy (p= 0.02), respectively. MLD in the supine and prone positions was 6.58 Gy and 0.85 Gy (p= 0.001), respectively [26]. All these doses are quite low as compared to the current study. DIBH reduces cardiac volume in the RT eld, hence it lead to reduction in all dose parameters(mean, maximum and volume based) of the heart[27,28]. It has been shown to reduce cardiac mortality by 4.7% compared to free breathing with median cardiac mortality normal tissue complication probability of 0.1% in patients leftsided breast cancer [29]. Because of changes in dose fractionation (from conventional to hypofractionation) and techniques of RT for breast cancer(from 2D to 3D-CRT/FIF IMRT), it still remains unclear what dose constraints to be placed for heart, LAD and lungs; and how is it going affect the cardiac related mortality? Although MHD has been the gold standard for prediction of late cardiac effects in the past studies but recent studies have suggested that reporting doses to the heart substructures such as apical part of left ventricle and LAD may be of relevance[30,31] Hypofractionation may result in lower equivalent doses (EQD2) to the heart as compared to conventional fractionation and comparable late effects[32,33]. However, till data comes clear on these aspects, patients with left-sided breast cancer should be offered techniques, which reduces dose to the heart and lungs.
Second cancers after breast radiation are also a possibility with VMAT because of low dose to larger volume of OARs. In the present study 50% of the contralateral lung received 2 Gy, so it may put this OAR for second cancer. Hall et al. in their study estimated 1% to 1.75% increase in the incidence of second cancers after 3D-CRT and IMRT at 10 years [34]. VMAT technique was also reported to increase this risk in one study [19] where as it was comparable in another for the contralateral breast and lung, but less risk in the ipsilateral lung because of reduced MLD with VMAT [35]. In a recent review, it was observed that VMAT increases contralateral lung V5 by 25% as compared to other techniques [36]. In our study contralateral lung V5 of 3.74% is still lower as compared to other studies. This reduction in V5 is associated with reduction in secondary cancer [35,36] Since, ipsilateral MLD, contralateral lung V5 and breast mean doses in our study are comparable to those observed by Zhang et al., we may expect similar risk of second cancers in our patients. However, this risk may vary with distance of the organ from the sternum, patient anatomy, dose optimization, set up errors, organ motion [37] and smoking [22]. In our past series we have reported second cancers in the contralateral breast, oesophagus and lung cancers in 3.3%, 0.22% and 0.05% patients, respectively [25].
Many dosimetric studies have explored the potential bene ts of integrating boost with WBI, but the majority of them are with conventional fractionation [38][39][40][41][42]. There are few studies where boost has been integrated with moderate hypofractionation and treatment completed in 3-4 weeks [7][8][9][10][11][43][44][45]. and 5-6 weeks in others [47,50]. A multicentric study of 151 patients by Dellas et al. from Germany with RT dose of 40Gy in 16 fractions for WBI and a SIB with 0.5Gy/fraction to the primary area, reported that dose constraints could be achieved and SIB was feasible with hypofractionation [42].
In the present study we integrated boost with accelerated hypofractionation and completed treatment in 2 weeks only. With changes in hypofractionation schedules in breast cancer we have to look for OARs constraints, which can be achieved with a particular dose fractionation schedule. In our study we achieved dose constraints to the OARs such as lungs, heart(high dose volume), contralateral breast and oesophagus in >80% of patients, except for the MLD, MHD, mean dose to contralateral breast (<3Gy) and V30 (<25%) to the thyroid that could be achieved in 70%, 59%, 77% and 70% of patients, respectively ( Table 2). MLD was slightly higher with SCF treatment (10.08Gy vs 9.22Gy). There was no impact of SCF treatment on the MHD. One of the observations of our study was that dose to the thyroid could be reduced signi cantly with head rotation. Dose to the thyroid was 7.84Gy and 34.73Gy(p<0.0001) with and without head rotation, respectively (Table 3).
De Rose et al. reported a phase II trial of hypofractionated RT with VMAT in 787 patients with early breast cancer with a dose of 40.5Gy to whole breast and 48Gy to the tumor bed in 15 fractions over 3 weeks with VMAT. Grade 1 and 2 acute toxicity was observed in only 51% and 9.7% patients, respectively. At a median follow-up of 45 months, grade 1 toxicity was 13.5% and 4 patients had distant relapse. Cosmetic outcomes were excellent/good in 100% patients [8]. In our study, grade 1 acute toxicity was less than those reported by De Rose et al., perhaps because of the lower total dose used in our study. We also observed only one local recurrence and two distant metastases at a median follow-up of 48 months. Both patients with distant metastases had triple negative disease. Our results are quite consistent with the studies published in the literature ( Table 6) in terms of acute toxicity, cosmetic outcomes, local control, DFS and OS. Acute grade 2 toxicity in these studies ranged from 4%-43%, and upper limit of 95% CI of our study 35%, lie well within this range. Higher grade 2 toxicity in the study by Freedman et al. could be because of delivery of higher total dose (56Gy) [9]. However, local control was comparable to our study.
We did not observe any late grade 3 toxicities. At 4-years, loco-regional control and excellent/good cosmesis of 96.5% and 100% in our study are also comparable to those reported in literature (97%-100%) and (77%-100%), respectively (Table 6). Joppe et al. reported cosmetic outcomes with 8.5% grade 2 brosis in the boost area, chest wall pain in 6.7% patients, and telangiectasia grade ≥ 2 in 3.7% patients at a median follow-up of 30 months. All-grade brosis outside the tumor bed was observed in 50% of patients. Higher brosis, chest wall pain and telangiectasia rate could be because of a high total dose delivered (64.4-67.2 Gy) in their study [47][48][49][50]. We did not observe any telangiectasia or chest wall pain in our study. So the present schedule may be better in terms of toxicities and cosmetic outcomes. In our study the treatment was completed in 12 days only with a similar toxicity pro le, cosmetic outcomes and comparable local control, DFS and OS.
There are a few limitations to our study. The number of patients enrolled was less, because of the study design. Median follow-up of 48 months is modest; therefore; late toxicities and prolonged clinical outcomes need to be further assessed since we have delivered accelerated hypofractionation regimen with a dose of 3.4Gy/fraction to the breast and 4Gy/fraction to the tumor bed which may lead to late radiobiological consequences, although the likely risk is less because of the total delivered dose (40Gy) with one of the optimal techniques of RT. Low doses to lungs and contralateral breast may also not favor VMAT implementation but these can be further reduced by using tangential VMAT or hybrid VMAT. Lastly, it is an expensive technique and one of ASTRO Choosing Wisely Campaign initiatives is "Don't routinely use IMRT to deliver whole breast radiotherapy as part of breast conservation therapy"[51].
This study has shown comparable results with the previous studies (Table 6). Since the higher dose per fraction was used, the overall treatment time was reduced to only 12 days. It helped in increasing treatment compliance of the patients because of less acute toxicity. It also helped in reducing cost to the patient with increased convenience by reducing the number of hospital visits and has potential to reduce risk of local recurrence with acceptable toxicities in the breast because of its low α/β ratio. Therefore, the implication of this study is, reduction of total treatment time from 4 weeks to 2 weeks and reduction in the waiting time for the other patients.
To conclude, this study demonstrated that accelerated hypofractionated radiotherapy with SIB boost is feasible and safe in terms of acute and late breast toxicities. Radiation induced heart disease and stochastic effects might be a concern with higher MHD and low dose bath with this technique. VMAT plans may be used when conformal techniques are not able to achieve the desired dosimetric constraints. A phase III randomized controlled trial with same fractionation schedule with 2D/3D-CRT and DIBH techniques (XXXX; NCT XXXX) is on going and has completed patient accrual.