Abstract
Background/Aim: This study aimed to assess the time advantages of delivering adjuvant hypofractionated radiotherapy (AH-RT) before third-generation adjuvant chemotherapy (A-CT), compared to the standard sequence (A-CT before AH-RT), in node-positive breast cancer (BC).
Patients and Methods: A total of 45 patients with node-positive BC treated with AH-RT before third-generation A-CT at our institution between 2022 and 2023 (EXP group) were retrospectively enrolled and matched with a control group of 45 patients treated with standard sequencing (CTRL group). The primary endpoints were as follows: gain in time to RT initiation and overall treatment time, RT delay, RT interruptions, which were compared between the two groups. Propensity score matching was performed. Univariate and multivariate Cox-proportional hazards models were generated. Data from the multivariate analysis were confirmed by Pearson’s covariance test, assuming p<0.001.
Results: A significant reduction in the time to AH-RT initiation and overall adjuvant treatment time was recorded in the EXP group. In the EXP group, the mean duration of the entire adjuvant treatment was 35 (29-40) weeks after surgery vs. 42 (39-50) weeks for the CTRL group (p=0.032). Hematological G2-G3 toxicity was responsible for RT delay (p=0.022) in the CTRL group. Multivariate analysis confirmed that acute skin toxicity was significantly associated with RT delay and interruption in the AH-RT15 CTRL arm (p=0.033) in the CTRL group. Pearson’s covariance test confirmed these effects for the CTRL group (p<0.001).
Conclusion: Treatment with AH-RT before third-generation A-CT was found to be safe with a low acute toxicity profile in node-positive breast cancer, providing an advantage in shortening the time from surgery to AH-RT initiation as well as the overall adjuvant treatment time.
Introduction
The current approach for the treatment of node-positive breast cancer (BC) after surgery includes adjuvant chemotherapy (CT), followed by postoperative radiotherapy (PORT), since the publication of the Up-front-Outback trial results in 1996, which were updated in 2005 (1, 2). In this trial, two treatment sequences were compared: radiotherapy (RT) followed by CT versus CT followed by RT. Chemotherapy consisted of 12 weeks of four cycles of cyclophosphamide, adriamycin, and fluorouracil (CAMPF); PORT was delivered in 25-30 fractions (frs) for at least five weeks using standard fractionation of 2 Gy/fr. The median interval between surgery and initiation of RT was 36 days (five weeks) in the RT-first group and 126 days (31 weeks) in the CT-first group. An interval of more than 16 weeks was observed between surgery and RT in 1% of the RT-first group and 84% of the CT-first group. The median interval from the first breast excision to the start of CT was 29 weeks and 13 weeks in the RT-first and CT-first groups, respectively. Regarding acute toxicity, the incidence of skin toxicity was 15% in the RT-first group and 11% in the CT-first group (p=0.43). The difference in the incidence of radiation pneumonitis was not significant between the two groups (p=0.12). Moreover, no differences in event-free status, distant metastases, and overall survival were observed between the two sequencing arms.
The treatment of node-positive BC has changed significantly since then. Several advancements have revolutionized the adjuvant treatment approach in RT and CT. Concerning adjuvant RT, numerous high-quality randomized clinical trials have shown equal outcomes of moderately or accelerated hypofractionated whole or partial-breast PORT over the long-course standard fractionated schedule. Adjuvant hypofractionated RT (AH-RT) has become the new standard of care (3-9). Apropos adjuvant CT (A-CT), the introduction of long-course third-generation CT regimens with new drug combinations have substantially improved survival, achieving a 20% reduction in BC mortality compared with previous regimens (10). However, due to unexpected acute toxicities, these regimens may prolong the entire schedule by more than 12 weeks. In turn, RT initiation could be delayed, prolonging the overall adjuvant treatment time. Although many attempts have been made to optimize this timing, such concurrent or sandwich CT schedules, no definitive conclusions have been reached (11).
In a previous retrospective study in patients with node-positive BC, the authors recorded a substantial gain in the time to RT initiation as well as overall treatment duration using AH-RT up-front to third generation A-CT. Moreover, no differences in acute and subacute toxicities were found upon comparison with studies using AH-RT after third-generation A-CT regimens (12). To confirm these advantages in our experience, we conducted a retrospective study with a propensity score matched experimental (EXP) group comparing 45 patients treated with AH-RT up-front to 45 patients treated with third-generation (long course) A-CT before AH- RT (CTRL group).
Patients and Methods
Objectives. This study was approved by our institutional review board and ethics committee (CEUR23/2024 addendum) and complied with the Declaration of Helsinki. Informed consent was obtained from all patients. This study was designed as a retrospective, propensity score-matched analysis of patients with stage II-III BC who underwent breast conservation surgery (BCS) or mastectomy, followed by adjuvant treatments (RT and CT). Overall, data of 90 patients with node-positive BC treated with AH-RT at our institution between 2022 and 2023 were collected from the clinical records for analysis
The primary endpoint of the study was the gain in the time to RT initiation from surgery and in the overall adjuvant treatment timeline from the date of surgery. The secondary endpoint was the rate of the acute grade (G)2-G3 scores according to the Common Terminology Criteria for Adverse Events, version 5 (CTCAE v5.0) for skin, lung, and hematological toxicities impacting RT completion and days of RT interruption (13).
Study population and treatments. Ninety patients (45 patients in each group) affected by high-risk node positive breast cancer were included. The specific dataset for both groups was developed by the principal investigator. Each group included 45 patients as shown in Table I for the experimental (EXP) and Table II for the control group (CTRL). Each group received adjuvant therapy as follows: the EXP group had up-front RT (RT before CT), while in the CTRL group up-front CT (CT before RT) was delivered. Chemotherapy is described in Table III and Table IV for each group. All patients received adjuvant RT with moderate or ultra-hypofractionated schedules on the breast or chest wall, and nodal areas according to stage and surgery.
Patients and treatment findings [surgery and radiotherapy (RT) volumes] in the experimental (EXP) group.
Patients and treatment findings [surgery and radiotherapy (RT) volumes] in the control (CTRL) group.
Patient characteristics according to pN status and third generation adjuvant chemotherapy in the experimental (EXP) group.
Patient characteristics according to pN status and third generation adjuvant chemotherapy in the control (CTRL) group.
Radiotherapy. The clinical target volume consisted of the remaining breast or chest wall with nodal areas with or without an internal mammary chain, as defined by the updated contouring guidelines devised by the AIRO Breast Cancer Group (14). As shown in Table I and Table II, the moderate hypofractionated RT of 2.67 Gy/fr/15 frs/40.05 Gy was delivered to 57 patients, according to the START B trial protocol. Sequential or simultaneous boost was provided at a dose of 2.5 Gy/fr/4 frs/10 Gy or simultaneous integrated boost (SIB) at a dose of 3.2 Gy/fr/15 frs/48 Gy. Ultra-hypofractionated RT was delivered in five fractions on alternate days at 5.7 Gy/fr/28.5 Gy (SIB 6.4 Gy/fr/32 Gy) and 5.4 Gy/fr/27 Gy on nodal areas to 33 patients. Conformal 3D, intensity-modulated radiotherapy or volumetric modulated arc therapy with 6 MV photons was delivered to all patients using Trilogy® (Varian Medical System, Palo Alto, CA, USA); the treatment plan was calculated on Eclipse®TPs (Varian Medical System).
Chemotherapy. As shown in Table III and Table IV, adjuvant third-generation CT was chosen by medical oncologists according to age, comorbidities, oncotype test results, molecular phenotypes, and nodal involvement, per the current guidelines. Adjuvant third-generation chemotherapy was delivered with anthracyclines alone in 14 patients and with taxanes combined with anthracyclines in 76 patients.
Follow-up. Follow up was conducted after 2 weeks off RT, and 1, 3, 6, 9, 12 months during the first year. Subsequent follow-ups were conducted annually.
Toxicities. Acute and subacute toxicities were recorded by treating physicians one week off RT and during every CT cycle and subsequently scored according to CTCAE v5.0 for adverse skin, pulmonary, and hematologic events.
Analysis of data and statistical methods. Data were retrieved from the medical records by the radiation oncologist and medical oncologist. To avoid bias in patient selection, propensity score matching was used to ensure that age, surgery, stage, tumor grades and molecular subtypes, systemic therapy, hypofractionated radiation schedule, and duration of follow-up of the study groups matched as closely as possible. The duration of the entire adjuvant treatment, time to RT and CT initiation, days of RT interruption, and acute/subacute toxicities were compared and analyzed. The overall adjuvant treatment time was defined as the time elapsed from surgery to the last day of treatment (RT fraction or CT cycle). The times to RT and CT initiation were calculated from the day of surgery. Nearest-neighbor matching without replacement within a caliper was used. The caliper width was 0.2 standard deviations (SD) of the logit-transformed propensity score. Case matching was performed in a 1:1 ratio using a propensity score incorporating many factors that may have an effect on acute toxicity, RT delay, and RT interruption. Univariate analysis with paired t-tests and multivariate Cox-proportional hazards models were used to evaluate the correlation between G2-G3 skin, lung, and hematological toxicities; hazard ratios with 95% confidence intervals are reported. Statistical significance was set at p<0.05. Pearson covariance was used to confirm data in the multivariate analysis, with significance set at p<0.001. For statistical analyses, SPSS (version 28; SPSS Inc., IBM Statistics, Chicago, IL, USA) was used.
Results
Data of 90 patients (45 patients in each group) affected by high-risk node positive breast cancer were collected. Outcomes in terms of gap timing, RT interruptions, overall treatment time, acute and subacute toxicity were compared between the two groups and analyzed as shown in Table V.
Gap timing between the experimental (EXP) group and control (CTRL) groups.
Gap timing outcomes. In the EXP group, RT was initiated at a mean 35 days (five weeks) after surgery (30-45 days). In the CTRL group, RT was initiated at a mean 196 days after surgery, corresponding to 28 (25-35) weeks [p=0.02, odds ratio (OR)=0.12, 95% confidence interval (CI)=0.01-0.99)]. Adjuvant-CT was started after a mean 10 days (8-15 days) off RT, corresponding to a median eight weeks from surgery (6-12 weeks). In the CTRL group, A-CT was started after a mean 7.2 (6-12) weeks from surgery [p=0.572; OR=1.15 (95%CI=0.90-1.55)].
RT interruptions. No RT interruptions occurred in the EXP group, irrespective of the RT schedule. In the CTRL group, a mean of eight days of interruption (5-12 days) were recorded. RT interruptions occurred in the moderately hypofractionated arm within the second week at a mean of eight days (6-10) of RT.
Overall treatment time. In the EXP group, the mean adjuvant treatment timeframe was 35 weeks (26-42) after surgery, while that in the CTRL group was 42 weeks (38-56) [p=0.032; OR=1.75 (95%CI=1.53-1.93)].
Acute and subacute toxicity. As indicated in Table VI, at a mean of eight months follow-up (6-12), G2-G3 skin toxicity occurred in both groups. Fourteen of 45 patients in the EXP group (35%) versus 21/45 (46%) in the CTRL group developed G2-G3 skin toxicity [p=0.45, OR=1.88 (95%CI=0.99-1.74)].
Acute G2-G3 toxicity.
In the CTRL group, multivariate analysis revealed that acute hematological G2-G3 toxicity bore significant associations with RT delay [p=0.0219; OR=0.11 (95%CI=0.01-0.95)] and RT interruption [p=0.043; OR=0.40 (95%CI=0.023-0.97)].
In this group, acute G3 skin toxicity was significantly associated with RT interruption in the moderate hypofractionation RT group [p=0.0319; OR=1.71 (95%CI=1.51-1.95)].
According to multivariate Pearson’s covariance, a significant relationship (2-tailed p<0.001) was evident between skin and hematological toxicities and RT interruption, RT delay, overall treatment time and taxane-based chemotherapy in the CTRL group as shown in Table VII.
Pearson’s covariance multivariate analysis.
At six months, the cumulative incidence of G2-G3 acute-subacute skin toxicity in the EXP and CTRL groups was 3% and 4.5% [p=0.472, OR=0.58 (95%CI=0.30-1.64)], respectively, and that of lung toxicity was 3.2% and 3.5% [p=0.836; OR=0.97 (95%CI=0.78-1.09)], respectively as depicted in Figure 1.
Cumulative incidence at six months of G2-G3 acute-subacute skin toxicity (blue line): 3% and 4.5% (p=0.472); G2-G3 acute lung toxicity (red line): 3.2% and 3.5% (p=0.836). EXP S: Skin toxicity in the EXP group; CTRL S: skin toxicity in the CTRL group; EXP L: lung toxicity in the EXP group; CTRL L: lung toxicity in the CTRL group; EXP: experimental (dotted lines); CTRL: control (solid lines); G: grade.
Discussion
The optimal sequence of adjuvant RT, i.e., before or after chemotherapy, remains an unresolved conflict between medical and radiation oncologists. Retrospective studies and randomized trials have reported the advantages and disadvantages of different sequences for over 30 years.
Doubts voiced by Bucholtz in 1993 were rooted from a retrospective study conducted in 295 patients with BC receiving non-random cyclophosphamide, methotrexate, and fluorouracil (CMF) or anthracycline-based CT and RT. In that study, patients with BC who received RT within 16 weeks of surgery showed an actuarial 5-year local failure rate of 5% versus 35% in patients who received RT more than 16 weeks after surgery (15).
Thereafter, in 1996, two randomized trials–the IBCSG trial VI and trial VII–which evaluated concurrent CMF in node-positive pre/perimenopausal and postmenopausal patients, respectively. In a combined cohort of 718 patients, the studies concluded that a 4 to 7 months delay in RT following breast-conserving surgery and CMF was acceptable, despite an increased risk of local recurrence (16). Later, the standard sequencing for clinical application was established by the Up-front-Outback randomized trial, as described before (1, 2).
Avoiding prolonged delays in adjuvant BC treatments is an understandable concern. The optimal timing has not been established yet, although an anecdotal interval of at least 12 weeks for the initiation of CT and RT after surgery is the most accepted timeframe (17); however, it is still under investigation. Recently, Yung et al. conducted a study assessing the associations of delays across surgery, CT and RT with survival among 3,368 women with stage I-III BC, where delays were defined as intervals of >8 weeks between each modality and diagnosis or prior treatment. They found that a delay in CT was associated with a higher risk of BC-specific mortality (BC-SM) [hazard ratio (HR)=1.71; 95%CI=1.07-2.75] and all-cause mortality (A-CM) (HR=1.39; 95%CI=1.02-1.90). A delay in radiation therapy was associated with an elevated BC-SM risk (HR=1.49; 95%CI=1.00-2.21) but not AC-M risk (HR=1.19; 95%CI=0.99-1.42). However, delays across multiple treatment modalities were associated with a threefold increase in BC-SM risk (95%CI=1.51-6.12) and 2.3-fold increase in A-CM risk (95%CI=1.50-3.50) comparable to the impact of adverse cancer characteristics, such as nodal involvement (18). When considering the timing of RT initiation, the impact of RT delay on survival is summarized in the “ASARA” principle expressed by Mackillop, which states that delays in RT should be “as short as reasonably achievable” (19). However, the permissible delay due to long course CT is a matter of debate. Yock et al. conducted a retrospective study and showed that delaying RT for more than 7 months, in order to administer longer adjuvant CT, did not compromise local control (20). On the contrary, in a large study of 7,800 patients, mainly comprising node-positive patients with BC treated with systemic therapies, Mikeljevic et al. showed a trend towards an elevated relative risk of death when RT was started after >9 weeks (21). In another large population cohort, van Maaren found no relationship between delayed RT and CT in terms of distant metastases-free survival (DMFS) in a group of patients who underwent BCS-chemotherapy-RT and BCS-RT-chemotherapy. Surprisingly, in the BCS-RT-chemotherapy group, the DMFS was higher when the interval exceeded 55 days than when it was less than 42 days (22).
Further, an Ontario population-based study on 1028 women with stage I-II BC treated with adjuvant RT-CT focused on the impact of delayed initiation of A-RT. A delay in RT of over 12 weeks from surgery or six weeks from the end of CT was related to worse survival probability. In the RT-only group, a waiting time of 12 weeks or more from surgery to the start of radiation was associated with worse event-free survival after a median follow-up of 7.2 years (HR=1.44; 95%CI=0.98-2.11; p=0.07). In the group that received intervening adjuvant CT before RT, a waiting period of more than six weeks from the completion of CT led to worse event-free survival after a median follow-up of 7.4 years (HR=1.50; 95%CI=1.00-2.22; p=0.047) (23).
In these two recent decades, several therapeutic advances in BC have significantly revolutionized the entire course of the adjuvant timing duration. Concerning CT, a web-based decision tool called Adjuvant! has classified adjuvant chemotherapy regimens as first-, second-, and third-generation based on the introduction of anthracyclines (doxorubicin, epirubicin) and/or taxanes (paclitaxel, docetaxel) alone or in combination and added targeted therapies (24, 25). The adoption of CT by this classification is related to a reduction in BC mortality from 35% by the first-generation CT to more than 20% by the third-generation A-CT (10), although with greater accompanying toxicity. Further, the A-CT prescription process has become better customized due to the development of gene array technology (Oncotype®, Mammaprint®, etc.), which provides a more individualized and tailored use of chemotherapy (26). However, these tests waste precious time to CT initiation, which in turn delays RT initiation. Regarding postoperative RT, a better understanding of the α/β ratio for tumor and normal breast tissue has led to the introduction of hypofractionated RT as a standard modality worldwide. High-quality randomized clinical trials enrolling more than 7,000 patients with BC treated with moderate or fast hypofractionated RT schedules within these last 20 years have demonstrated equal outcomes in terms of local control and survival compared with standard fractionated RT (3-9). In addition to other benefits such as better RT compliance, quality of life, reduced RT waiting list and costs, hypofractionated RT has shortened the RT treatment time (27). Thus, moderate AH-RT with 16-15 frs or fast AH-RT with five frs has been adopted as the standard regimen worldwide. Given this background, it is reasonable to rethink the timing and sequencing in order to reduce the RT delay and shorten the overall adjuvant treatment time (28). Many efforts in the past have focused on the use of concurrent schedules mainly with CMF and standard RT (29). Concurrent standard RT with taxane-containing CT has been also evaluated, showing an increased incidence of toxicity. Several studies conducted in the 2000s have reported G2-3 skin toxicity in 50% of patients treated concurrently with RT and taxanes (30-32). Taghian reported that the risk of severe pneumonitis requiring steroids was 15-30% in patients treated with concurrent RT and paclitaxel (33).
Furthermore, studies have also investigated concurrent hypofractionated RT and taxane CT. Chen et al. conducted a phase II trial on 44 patients with stage II or III BC treated with hypofractionated RT of 39.6 Gy in 22 fractions and a tumor bed boost of 14 Gy in seven fractions with concurrent paclitaxel, finding no significant pulmonary deficit in diffusing capacity of the lung for carbon monoxide (DLCO), while only two patients developed acute G3 skin toxicity (34). In the CONCERT study based on taxane-based CT and hypofractionated RT of 40 Gy in 15 fractions, an acceptable toxicity profile was recorded (35). However, the increased acute toxicity incidence remains the main feared effect influencing the delay and continuity of adjuvant RT. In addition, the intermediate time between chemotherapy and hypofractionated radiotherapy (ITCR) may delay RT initiation to minimize the risk of radiation-induced acute toxicity. Zygogianni et al. showed significantly higher (p<0.05) acute toxicity when the ITCR was less than 20 days, suggesting a resting period before starting RT after CT (36). Our comparative study confirmed a substantial reduction in the interval between surgery, RT initiation, and CT initiation using the “hypofractionated RT-first” approach, mirroring the timing outcomes reported in RT-first arm of the Up-front-Outback trial in 1995 (1).
The high rate of skin and hematological acute toxicity induced by the addition of taxanes to anthracyclines, as shown by the most representative phase III trials conducted in the past and confirmed by our study, is well acknowledged (10).
In our study, we recorded a mild significant rate of acute skin toxicity in the CTRL group occurring in patients treated with moderate hypofractionated RT. Moreover, adverse skin and hematological events were responsible for increased RT treatment time due to interruptions during the second part of RT. Thus, it is important to remember that cytotoxicity tends to decrease with increased delivery time, as shown by Elkind et al. 50 years ago (37). Further, RT interruptions are detrimental to local control because of the well-known loss of biological effectiveness due to the rapid repopulation rate of cancer cells when interruptions occur in the second part of RT (38). The impact of RT interruptions on breast cancer has been showed by Bese et al. (39). They analyzed the impact of unplanned interruptions during PORT in 853 patients with stage I-III BC after adjuvant systemic treatment. Patients who experienced interruptions of eight days had significantly lower locoregional control and overall survival compared with patients without any interruptions or interruptions of 1 week. A gap of one week also was an independent prognostic factor in the multivariate analysis, producing an average 5% decrease in the 5-year local control rate. Another consideration is that RT interruptions are not easy to compensate radiobiologically.
An emerging strategy involves sequencing hypofractionated radiotherapy prior to the administration of third-generation chemotherapy. Previously, a retrospective study in 45 high-risk nodal positive patients with BC treated with adjuvant hypofractionated RT administered up-front to third-generation CT was conducted. The authors found a reduction in the time to PORT initiation and overall adjuvant treatment time. PORT was initiated at a median of five weeks after surgery, and A-CT was performed at a median of nine weeks after surgery. The median duration of the entire adjuvant treatment was 35 weeks post-surgery. These results were compared with times spent in the Up-front-Out ack trial and the gain difference was impressive. The rates of acute skin and lung toxicity were similar to data reported in the Up-front-Outback trial and recent studies using hypofractionated RT following adjuvant long-course chemotherapy (12). These advantages have been confirmed by the CTRL group in this study. Due to the short follow-up time, we do not have data on survivals outcomes, but if timing matters on local control and survival, promising outcomes are expected in subsequent analyses. Radiobiological considerations support the rationale for reversing the sequencing of treatment, as demonstrated by the experimental studies of Formenti et al. and Belletti et al. (40, 41). Since the current study is a retrospective trial, a new randomized trial may be useful.
Conclusion
In accordance with the ASARA concept, in this comparative study, we confirmed a significant gain in RT initiation in up-front hypofractionated RT over up-front CT; meanwhile, the adjuvant CT timing has been set within the anecdotal period of 12 weeks. In turn, a reduction in the adjuvant overall treatment time was recorded. In the CTRL group treated with moderate hypofractionated RT, skin and hematologic acute toxicity was the main cause of RT interruptions, delayed RT initiation, and overall adjuvant treatment time prolongation. These findings suggest that initiating treatment with fast hypofractionated radiotherapy may offer advantages and could be particularly beneficial in emerging treatment settings such as the KEYNOTE-522 protocol for triple-negative breast cancer involving adjuvant pembrolizumab (42), where immune-related adverse events during RT can lead to permanent treatment interruptions (43). This is especially relevant given that the duration of RT interruptions in triple-negative breast cancer has been associated with poorer overall survival outcomes (44).
Although the data does not prove the superiority of any modality administered first and the novelties in therapies, it is common practice that third-generation CT is administered before adjuvant hypofractionated RT, while research on concurrent administration is still ongoing in order to reduce the RT delay and overall treatment time within new long-course CT combinations. The short-term analysis in this retrospective paired matched analysis showed that up-front hypofractionated RT followed by third-generation adjuvant CT appears to be a feasible option with a low acute toxicity rate, yielding a gain in time for the entire RT course and overall adjuvant treatment time.
Acknowledgements
The Authors thank Dr. Annalisa Digennaro for logistic help.
Footnotes
Authors’ Contributions
Conceptualization: GL, AB; Investigation and methodology: VM, RT; Project administration: GL, AZ, IS, TV; Resources & Writing of the original draft: GL, AM, IB; Supervision & Writing and Editing: GL, VM, BD; Data curation: all Authors. All the Authors have proofread and approved the final version of the paper.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
Artificial Intelligence (AI) Disclosure
No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning-based image enhancement tools.
- Received June 19, 2025.
- Revision received July 13, 2025.
- Accepted July 17, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
Jump to section
Related Articles
Cited By...
- No citing articles found.