Supine vs Prone Breast Radiotherapy and Acute Toxic Effects of the Skin Among Women With Large Breast Size (2024)

Key Points

Question Can radiotherapy in the prone position reduce acute toxic effects of the skin in women with large breast size being treated for breast cancer?

Findings In this randomized clinical trial of 378 patients treated with adjuvant breast radiotherapy, treatment in the prone position was statistically significantly associated with lower rates of moist desquamation (26.9%) compared with treatment in the standard supine position (39.6%).

Meaning Treatment in the prone position is less toxic than in the supine position in women with large breast size receiving adjuvant breast radiation.

Importance Women with large breast size treated with adjuvant breast radiotherapy (RT) have a high rate of acute toxic effects of the skin. Breast RT in the prone position is one strategy that may decrease these toxic effects.

Objective To determine if breast RT in the prone position reduces acute toxic effects of the skin when compared with treatment in the supine position.

Design, Setting, and Participants This phase 3, multicenter, single-blind randomized clinical trial accrued patients from 5 centers across Canada from April 2013 to March 2018 to compare acute toxic effects of breast RT for women with large breast size (bra band ≥40 in and/or ≥D cup) in the prone vs supine positions. A total of 378 patients were referred for adjuvant RT and underwent randomization. Seven patients randomized to supine position were excluded (5 declined treatment and 2 withdrew consent), and 14 patients randomized to prone position were excluded (4 declined treatment, 3 had unacceptable cardiac dose, and 7 were unable to tolerate being prone). Data were analyzed from April 2019 through September 2020.

Interventions Patients were randomized to RT in the supine or prone position. From April 2013 until June 2016, all patients (n = 167) received 50 Gy in 25 fractions (extended fractionation) with or without boost (range, 10-16 Gy). After trial amendment in June 2016, the majority of patients (177 of 190 [93.2%]) received the hypofractionation regimen of 42.5 Gy in 16 fractions.

Main Outcomes and Measures Main outcome was moist desquamation (desquamation).

Results Of the 357 women (mean [SD] age, 61 [9.9] years) included in the analysis, 182 (51.0%) were treated in the supine position and 175 (49.0%) in prone. There was statistically significantly more desquamation in patients treated in the supine position compared with prone (72 of 182 [39.6%] patients vs 47 of 175 [26.9%] patients; OR, 1.78; 95% CI, 1.24-2.56; P = .002), which was confirmed on multivariable analysis (OR, 1.99; 95% CI, 1.48-2.66; P < .001), along with other independent factors: use of boost (OR, 2.71; 95% CI, 1.95-3.77; P < .001), extended fractionation (OR, 2.85; 95% CI, 1.41-5.79; P = .004), and bra size (OR, 2.56; 95% CI, 1.50-4.37; P < .001).

Conclusions and Relevance This randomized clinical trial confirms that treatment in the prone position decreases desquamation in women with large breast size receiving adjuvant RT. It also shows increased toxic effects using an RT boost and conventional fractionation.

Trial Registration ClinicalTrials.gov Identifier: NCT01815476

Adjuvant breast radiotherapy (RT) continues to be the standard of care for women who choose to have breast-conserving surgery (BCS). Modern data suggest that adjuvant RT improves local control and survival after lumpectomy compared with lumpectomy alone.¹ For many tumor sites, technological improvements have increased the precision of RR, including more conformal delivery and better real-time imaging. These improvements have reduced normal tissue exposure and, hence, toxic effects.² For breast RT, however, the decrease in toxic effects has been relatively modest relative to other cancer sites, particularly for women with large breast size, given their increased risk of radiation-related toxic effects.^3-6 Minimizing acute toxic effects is of considerable importance to improve quality of life (QOL) because it has been correlated with late toxic effects in patients with breast cancer.^7-9

A large randomized clinical trial (RCT) led by our group at Sunnybrook Health Science Centre in Toronto, Ontario, Canada, previously demonstrated that intensity-modulated RT (IMRT), when compared with conventional RT, meaningfully decreased acute toxic effects in women of all breast sizes, but the risk of moist desquamation (desquamation) remained unacceptably high in women with large breast size.¹⁰ For this reason, other options are needed to further decrease toxic effects for women with large breast size.

Treatment in the prone position has several dosimetric advantages for these patients. It allows for more hom*ogeneous dose distribution owing to the smaller separation when compared with the supine position,¹¹ which decreases deposition of higher doses in the inframammary fold and axilla. One single-center RCT has been published comparing prone and supine treatment for women with large breast size.¹² The prone position resulted in lower rates of desquamation, pain, and other acute toxic effects. Herein, we report the results of a multicenter, phase 3, single-blind RCT in which patients with large breast size were randomly assigned to receive adjuvant breast IMRT in either the supine or prone position.

Patients with large breast size (bra band ≥40 in and/or ≥D cup) and early-stage breast cancer referred for adjuvant RT to 5 comprehensive cancer centers in Canada were enrolled in this phase 3, single-blind RCT. After providing written informed consent, patients were randomized and treated in the prone or supine position using a breast IMRT technique aimed at reducing dose inhom*ogeneity.¹³ Initially, patients were treated with 2 Gy per day conventional fractionation (extended fractionation; 50 Gy in 25 fractions), though the protocol was amended at 3 years to allow hypofractionation (42.5 Gy in 16 fractions), which was almost exclusively used onward. The primary end point was development of desquamation. The study protocol was approved by the institutional research ethics board of each participating center and registered with the National Institutes of Health (Supplement 1).

Women with large breast size with invasive carcinoma or ductal carcinoma in situ who were treated by BCS and referred for adjuvant RT to the breast alone (no nodal coverage) were eligible for inclusion. Patients were ineligible if they had bilateral breast cancer, had unhealed wounds, received previous ipsilateral RT, had active connective tissue disease, had medial tumors where adequate coverage was not feasible, were pregnant, or were unable to tolerate being prone.

Accrued patients were randomly assigned 1:1 to receive breast RT in the supine or prone position. Randomization was performed with blocking on the presence or absence of a boost to the surgical bed using a software-based algorithm. The addition of a boost in the same position was at the discretion of the treating physician and declared prior to randomization.

All patients underwent computed tomographic simulation for RT planning and were treated with an IMRT technique as previously described.⁶ Treatment arms differed only by positioning. Patients randomized to supine were positioned on an angled breast board with the ipsilateral arm abducted over the head. Patients randomized to prone were positioned on a prone board with arms immobilized above the head. Megavoltage portal images were taken on day 1 of treatment to ensure setup accuracy. From study initiation in April 2013 until June 2016, patients (n = 167) received a prescription dose of 50 Gy in 25 fractions (extended fractionation) with or without a boost (range, 10-16 Gy in 4-8 fractions). In June 2016, the trial was amended to allow a hypofractionated regimen of 42.5 Gy in 16 fractions to improve trial accrual. From this date until study completion, the majority of the remaining 190 patients were treated with hypofractionation. Treatment constraints are found in eTable 1 in Supplement 2.

The primary end point of acute skin reactions with desquamation and secondary end points of erythema, edema, and pain were scored using the National Cancer Institute’s Common Terminology Criteria for Adverse Events (CTCAE), version 4.0. As in our previous studies, each CTCAE symptom was captured separately.^6,14 Maximum skin reaction was evaluated and recorded at baseline, weekly during treatment, weekly for up to 6 weeks for patients being followed for resolution of toxic effects, and then again 6 to 8 weeks posttreatment. For desquamation, both any grade and grade 3 experienced at any time point during or after treatment were captured. Maximum breast pain was assessed by CTCAE and a visual analog scale ranging from 0 (none) to 10 (unbearable). Quality of life was assessed using the European Organisation for Research and Treatment of Cancer core QOL questionnaire (QLQ-C30) and breast cancer module (QLQ-BR23). The QOL questionnaires were completed at baseline, the last week of radiation, and at 6 to 8 weeks follow-up. Clinical research assistants (CRAs) involved in consenting, randomizing/assigning arms, and assessing toxic effects were blinded to arm, and assessment was completed independently of the treating physician to ensure nonbiased assessment. Patients and physicians were instructed to not discuss treatment arm with CRAs. All CRAs had an initial training session prior to accrual to standardize toxic effect assessment.

Sample size calculation for this study assumed a desquamation rate of 50% for women with large breast size treated in the supine position given rates seen in a previous trial.⁶ We hypothesized that treatment in the prone position would result in a relative reduction of toxic effects of the skin by 30% (absolute change from 50% to 35%) given preliminary trial evidence in 36 patients treated while prone. A total of 340 patients, 170 patients treated in each arm, could test the hypothesis with a 2-tailed level of significance of 5% and power of 80%. Accounting for a 10% dropout rate, 378 patients (189 per group) would effectively test the primary hypothesis.

Patient characteristics between treatment arms were compared using χ² tests for categorical variables and Wilcoxon rank sum tests for continuous variables. Comparison between arms of the frequency of grade 2 to 3 toxic effects of the skin at any time during treatment or within 6 weeks of completion was performed using a 2-sample, 2-sided test of proportions providing an odds ratio (OR) and its associated 95% CI. Comparisons were considered significant at the 5% level of significance. Multivariable logistic regression analysis evaluated the OR of toxic effects as the dependent variable with a generalized estimating equation model to adjust for potential within-facility clustering. The various predetermined independent variables tested included position (supine/prone), boost delivery (yes/no), breast size (>40-in bra band vs not), body mass index (comparison of means), chemotherapy (yes/no), and age at RT (comparison of means), with fractionation regimen (extended fractionation/hypofractionation) included post hoc.

The QLQ-C30 and QLQ-BR23 questionnaire data were processed to calculate global health status, pain, and breast symptom scale values at each assessment. Outcomes used for analysis were changes in values from baseline to the end of treatment and posttreatment follow-up. Change over time between arms was evaluated using a linear mixed model, while comparison between individual time points was completed using pairwise comparison. All statistical analyses were performed using SAS software (SAS Institute Inc), version 9.1.

Between April 2013 and March 2018, a total of 378 patients met eligibility criteria and were enrolled. Twenty-one patients were excluded owing to treatment constraint violation, inability to tolerate prone positioning, declining treatment after randomization, or consent withdrawal (Figure). Of the 357 patients eligible for analysis, 182 (51.0%) received RT in the supine position and 175 (49.0%) received treatment in the prone position. Patient groups were well balanced with no statistically significant differences in age, body mass index, bra size, delivery of boost, use of hypofractionation, or use of chemotherapy (eTable 2 in Supplement 2). There was no difference in boost delivery between patients receiving extended fractionation and hypofractionation (52 of 180 [28.8%] patients vs 59 of 177 [33.3%] patients; P = .41).

Patients treated in the supine position had statistically significantly higher rates of desquamation occurring anywhere in the breast compared with those treated in the prone position (72 of 182 [39.6%] patients vs 47 of 175 [26.9%] patients; OR, 1.78; 95% CI, 1.24-2.56; P = .002), as well as more grade 3 desquamation (28 of 182 [15.4%] patients vs 14 of 175 [8.0%] patients; OR, 2.09; 95% CI, 1.62-2.69; P < .001). There was no difference in the secondary end points of erythema, edema, or pain (Table 1).

On univariate analysis, supine position, bra size, boost, extended fractionation, chemotherapy, and older age were statistically significantly associated with increased risks of desquamation (Table 2). On multivariable analysis, supine position, bra size, boost, and extended fractionation remained statistically significantly associated with desquamation (Table 2).

In patients treated with extended fractionation (n = 180), supine position was associated with increased toxic effects compared with prone position when evaluating desquamation occurring anywhere in the breast (47 of 92 [51.1%] patients vs 31 of 88 [35.2%] patients; OR, 1.92; 95% CI, 1.62-2.72; P < .001), grade 3 desquamation (22 of 92 [23.9%] patients vs 9 of 88 [10.2%] patients; OR, 2.76; 95% CI, 2.45-3.10; P < .001), and pain (12 of 92 [13.0%] patients vs 5 of 88 [5.7%] patients; OR, 2.49; 95% CI, 1.48-4.19; P < .001). In patients treated with hypofractionation (n = 177), overall rates of toxic effects were lower in both arms, with no statistically significant difference in any outcome (Table 3).

For all patients, extended fractionation was associated with more toxic effects than hypofractionation when evaluating desquamation occurring anywhere in the breast (78 of 180 [43.3%] patients vs 41 of 177 [23.2%] patients; OR, 2.56; 95% CI, 1.50-4.37; P < .001), grade 3 desquamation (31 of 180 [17.2%] patients vs 11 of 177 [6.3%] patients; OR, 3.14; 95% CI, 1.62-6.11; P = .001), and pain (17 of 180 [9.4%] patients vs 6 of 177 [3.4%] patients; OR, 2.97; 95% CI, 1.06-8.31; P = .04). These differences were primarily driven by the rates of toxic effects in patients treated in the supine position. In patients treated in the supine position, extended fractionation was associated with increased desquamation compared with hypofractionation when evaluating desquamation occurring anywhere in the breast (47 of 92 [51.1%] patients vs 25 of 90 [27.8%] patients; OR, 2.72; 95% CI, 1.18-6.24; P = .02) and grade 3 desquamation (22 of 92 [23.9%] patients vs 6 of 90 [6.7%] patients; OR, 4.40; 95% CI, 1.83-10.57; P < .001). Patients receiving extended fractionation also had a higher risk of a pain score of 2 per the CTCAE that did not reach statistical significance (12 of 92 [13.0%] patients vs 2 of 90 [2.2%] patients; OR, 6.60; 95% CI, 0.91-48.08; P = .06) (Table 4). In patients treated in the prone position, extended fractionation was also associated with increased desquamation occurring anywhere in the breast (31 of 88 [35.2%] patients vs 16 of 87 [18.4%] patients; OR, 2.41; 95% CI, 1.65-3.52; P < .001) and grade 3 desquamation (9 of 88 [10.2%] patients vs 5 of 87 [5.7%] patients; OR, 1.87; 95% CI, 1.05-3.32; P = .03), while pain scores were not statistically significant (Table 4).

No differences in QOL as measured by the global health status, breast symptom, or pain scales were observed between arms. A statistically significant change in QOL using the global health status scale was seen when comparing baseline with end of treatment, which then statistically significantly improved by 6 to 8 weeks follow-up within the supine arm (eTable 3 in Supplement 2). Change in QOL was temporary given that there was no difference in the global health status scale when comparing baseline to 6 to 8 weeks. This temporary decrease in QOL was not seen in the patients treated in the prone position.

Quality of life as measured by the breast symptom scale decreased from baseline to end of treatment in both arms, then improved at 6 to 8 weeks follow-up but did not go back to baseline, remaining decreased in both arms. Although the pain as measured by the pain scale did increase by the end of treatment compared with baseline in both the supine and prone arms, this returned back to the baseline in both arms, with no statistical difference when comparing baseline with posttreatment pain scale score (eTable 4 in Supplement 2). When comparing pain using the visual analog scale, patients treated in the supine position had a greater increase from baseline to the timing of maximum toxic effects and baseline to 6- to 8–week toxic effects than those treated in the prone position.

To our knowledge, this study is the first multicenter, phase 3, single-blind RCT that demonstrates decreased acute toxic effects for women with large breast size treated in the prone position compared with the supine position, thus validating the existing literature. In the previously published single-center RCT,¹² the rate of desquamation was lower than we observed, being 6% with prone and 20% with supine, as compared with 26.9% and 39.6%, respectively. This difference is likely owing to the inclusion of women with larger breast size (≥D cup vs ≥C cup), and 50.4% of the patients in this study being treated with an extended fractionation schedule, which previous studies show confers more toxic effects than hypofractionation.^15,16 In addition, patients in the study by Mulliez et al¹² were evaluated only up to 2 weeks posttreatment, even though a previous breast IMRT RCT demonstrated that the cumulative occurrence of desquamation can continue to increase up to 4 weeks posttreatment.⁶

This study’s protocol was amended 3 years after initiation to allow for hypofractionation. Given that greater than 90% of the patients accrued after this amendment were treated with hypofractionation, this allowed for a presumably nonbiased unplanned comparison of the effect of position between patients treated with extended fractionation and hypofractionation. Extended fractionation was statistically significantly associated with increased toxic effects compared with hypofractionation, an observation consistent with existing literature.^15,16 Fractionation schedule was a stronger factor than position in predicting desquamation (Table 2). Although toxic effects were worse in all arms that received extended fractionation, the difference was more marked among those treated in the supine position.

Among patients treated with extended fractionation, the supine position was still associated with increased toxic effects in desquamation occurring anywhere, grade 3 desquamation, and pain scores, though for those patients treated with the hypofractionation regimen, these associations were not statistically significant. This difference is likely because the study was powered to detect differences in rates of toxic effects expected for extended fractionation, and the rates seen in those treated with hypofractionation were not high enough to reach statistical significance, though the relative reduction in toxic effects seen was similar in both regimens. Therefore, though unintended, this study supports the literature that hypofractionation RT is associated with a decrease in toxic effects of the skin even in women with large breast size, independent of position.

It is unclear if the prone position further decreases this risk in patients treated with hypofractionation because this study did not have sufficient power to detect such an association. This question is important and timely as the UK FAST-Forward dose-fractionation scheme is being increasingly used to limit the number of treatments patients must attend during the COVID-19 pandemic.¹⁷ Using a lower total dose of 26 Gy makes it less likely to develop desquamation; however, using a higher dose of 5.2 Gy per fractions may increase the risk of late adverse effects. In the study by Veldeman et al,¹⁸ which used daily doses of 2.67 Gy per fraction up to 40.05 Gy followed by a boost of 10 Gy in 4 fractions, there was a trend toward increased permanent edema (26% vs 11%) with the supine position, and also degradation of the cosmetic result on clinical photograph assessment. Because late adverse effects have been found to be highly correlated with the occurrence of desquamation,^8,9 and because the prone position statistically significantly reduces desquamation, there are arguments to recommend using the prone technique for women with large breast size even with ultrahypofractionated regimens, which may provide further benefit given the decreased lung and heart exposure. In particular, this may help decrease the risk of radiation-induced secondary lung cancer.¹⁹

We observed that not all patients can be treated in the prone position because 7 patients were unable to tolerate the position and 3 were considered ineligible given an unacceptable heart dose (mean heart dose, >3 Gy or volume of heart receiving 50% of the prescribed dose [V50] >10%). Efforts to help improve comfort in the prone position and increasing use of image guidance, such as daily cone beam computed tomography to ensure cardiac sparing, may allow for more women to be treated safely while in the prone position.

Despite the considerable difference in desquamation between arms, differences in QOL in the subacute time frame (6-8 weeks post-RT) were not seen. Quality of life was found to initially decrease from baseline to end of treatment as measured by the global health status and pain scales, but then returned to a level that was not statistically significantly different from baseline in both arms. Quality of life as measured by the breast symptom scale decreased from baseline to end of treatment, and though both improved at 6- to 8–weeks follow-up, they remained statistically significantly worse compared with baseline in both arms. These observations suggest that QOL in general decreases throughout treatment and in most cases returns to normal, and although there is a clear difference in the incidence of desquamation between the arms, these do not translate into an appreciable difference in QOL using these instruments.

In this study, the primary end point of desquamation was captured using the CTCAE, which compounds various symptoms that are collectively not specific nor sensitive. Desquamation itself, though, is a parameter that is specifically due to radiation, is easily recognized, and is most clinically relevant because it is associated with pain, acute and delayed decreased QOL, and long-term cosmetic effects.^6,8,14

Although this was a multicenter trial, a large number of patients were accrued in the largest participating institution. To control any potential bias this may have caused, the multivariable analysis incorporated a generalized estimating equation model. Because almost all patients accrued after the protocol amendment were treated with hypofractionation, we presume that there is limited bias in the unplanned comparison between the 2 regimens. Because no other aspect of the treatment protocol changed, it is not expected that there would be any other confounding factors. We acknowledge, though, that the study was not designed to compare these regimens and should be considered hypothesis generating rather than a firm causality finding. Bra size was self-reported by study participants, and given the lack of standardization of band/cup sizes, variability exists in the analysis on bra size as a predictor of toxic effects of the skin. In addition, the results are relevant to those women receiving whole-breast RT. It is unclear if the association of lower rates of toxic effects of the skin in the prone position would be seen with partial-breast approaches.

In this RCT, treatment in the prone position was associated with reduced moist desquamation in women with large breast size receiving postoperative breast RT. There were fewer toxic effects of the skin in women treated with hypofractionated RT compared with extended fractionation, though it is unclear if this effect can be further reduced when women are treated in the prone position. Treatment in the prone position should be considered for women with large breast size requiring adjuvant RT.

Accepted for Publication: March 28, 2022.

Published Online: May 26, 2022. doi:10.1001/jamaoncol.2022.1479

Corresponding Author: Danny Vesprini, MD, MSc, Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, 2075 Bayview Ave, Toronto, Ontario, Canada M4N3M5 (danny.vesprini@sunnybrook.ca).

Author Contributions: Dr Vesprini had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Vesprini, Bosnic, Fenkell, El-Mallah, Garcia, Rakovitch, Pignol.

Acquisition, analysis, or interpretation of data: Vesprini, Davidson, Truong, Vallieres, Fenkell, Comsa, Garcia, Stevens, Nakonechny, Tran, Kiss, Rakovitch, Pignol.

Drafting of the manuscript: Vesprini, Truong, El-Mallah, Tran, Kiss, Pignol.

Critical revision of the manuscript for important intellectual content: Vesprini, Davidson, Bosnic, Truong, Vallieres, Fenkell, Comsa, Garcia, Stevens, Nakonechny, Tran, Rakovitch, Pignol.

Statistical analysis: Vesprini, Kiss, Rakovitch.

Obtained funding: Vesprini, Tran, Pignol.

Administrative, technical, or material support: Vesprini, Davidson, Bosnic, Truong, Fenkell, Garcia, Stevens, Nakonechny, Tran, Rakovitch.

Supervision: Vesprini, Truong, Fenkell, El-Mallah, Stevens, Pignol.

Other—technique development and implementation, including equipment commissioning, treatment planning, and verification: Garcia.

Conflict of Interest Disclosures: Dr Rakovitch reported grants from Genomic Health Inc and personal fees from AstraZeneca outside the submitted work. Dr Pignol reported employment at Accuray Inc outside the submitted work. No other disclosures were reported.

Funding/Support: This research was supported by a Canadian Cancer Society (formerly Canadian Breast Cancer Foundation) grant (#307954).

Role of the Funder/Sponsor: The Canadian Cancer Society had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Meeting Presentation: This work was presented at the European Society for Radiotherapy and Oncology Conference #38; April 29, 2019; Milan, Italy.

Data Sharing Statement: See Supplement 3.