Less aggressive approaches have been explored globally during the past few decades for breast cancer (BC) treatment (1). Particularly, breast-conserving surgery (BCS) in sequence with whole-breast radiotherapy (WBRT) was able to demonstrate comparable survival outcomes to those earlier radical surgery interventions (2). In the subset of high-risk patients, using a dose-escalated boost targeting the surgical bed has also demonstrated substantial benefits in terms of local control (3, 4), considering that up to 60% of BC patients might have boost indication (5).

Different methods for boost application have been reported, all of them demonstrating outstanding clinical results. Frequently, for external-beam radiotherapy (EBRT), sequential prescriptions of 40–50 Gy in 15 to 25 fractions are employed, summed to a 10–16-Gy boost delivered in 5–8 additional fractions. These schedules correspond to single 1.8- to 2.67-Gy daily doses, yielding in total 4 to 7 treatment weeks (6). Recently published and presented results of simultaneous integrated boost (SIB) techniques have been proposed to shorten treatment times within 15–16 fractions with hypofractionation (HF) (7, 8). Nevertheless, when it comes to EBRT-based boost delivery, this still carries a relatively fair cosmetic toll, plus the uncertainties of postsurgical tumor bed localization, when unmarked (9).

Intraoperative assessment of the surgical cavity under direct view represents the best option to accurately target residual tumor cells. In this regard, interstitial placing of multicatheter for high-dose rate brachytherapy (HDR) is one of the oldest and most studied techniques, with numerous publications and guidelines supporting it (10). One of its potential advantages, in comparison to EBRT, is a remarkably improved patient-reported cosmesis, perhaps as a result of reduced skin exposure (11). Nonetheless, HDR usually requires prolonged hospitalization times in order to fully deliver the required treatment fractions.

Under this same rationale, intraoperative radiotherapy (IORT), first as an electron- and lately as a kilovoltage-based (low-energy X-ray) approach, has recently risen as an additional option for these patients (12, 13). With a growing body of evidence, for both accelerated partial breast irradiation (APBI) and boost modalities, this technique allows a prompt irradiation of the surgical bed as an upfront boost within the same surgical procedure (14, 15). In addition, novel literature has suggested a potential enhancement of immune response, which might positively influence disease control outcomes (16).

Herein we report the combined experiences of four centers from Peru, Spain, and Germany, regarding disease control and survival after IORT upfront boost in BC patients.

Since the seminal studies by Veronesi et al., the paradigm of treating BC with modern conservative approaches was established as the new standard of care (1, 2). The role of WBRT plus boost, for patients with increased risk of IBRT, is currently a key piece within the arsenal to assure good oncologic outcomes (3). Upfront boosting techniques with IORT have been under investigation for over 50 years, with kIORT being the latest option on-scene. Historical data have already proven the advantages and long-term outcomes of electron-based IORT (IOERT) as a boost, leading to modern guidelines recommending its utilization for selected patients (18–20). Nowadays, different aspects sustain the rationale for breast IORT. From an immunological perspective, IORT acts on regulating the surgical cavity environment and pro-inflammatory cytokines, which might be triggered after surgical aggression (21, 22). An immediate downregulation of surgically induced tumor-growth factors could potentially yield improved disease-control outcomes, overcoming a temporal miss factor (23–25). Recent data have suggested an improvement in breast cancer mortality for those patients undergoing IORT; nevertheless, different confusing factors might be associated with these differences and further research is warranted to confirm this statement (13, 26). An interesting feature related to this cohort was the time elapsed between IORT and EBRT start. With a median 74.5 days (>10 weeks), it seems that results might not be impaired by this extended waiting time. It is important to remark that most of this waiting was due to sequencing between chemotherapy and EBRT. Another point lies on logistics and patients’ comfort. Shortening the treatment duration by at least 1 week has direct influence on treatment-related costs for both the patient and healthcare provider, while patient satisfaction increases due to time convenience and improved cosmesis (27–30). Furthermore, reducing the probability of geographical miss by applying radiotherapy directly into the surgical cavity under direct view could be hardly compared to postoperative surgical bed reconstruction, particularly when unmarked and mostly during the era of oncoplastic surgery (31). Despite these advantages, weighing the abovementioned with increased operatory times, longer anesthesia, and surgeon availability should be formally analyzed to define the actual cost-effectivity profile of this intervention.

Our group reports the largest cohort of patients receiving both WBRT and kIORT upfront boost, to date. Based on the previously mentioned, investigating this subject results to be compulsory in an era of shorter treatments and improvement in patients’ quality of life (QOL). Despite its retrospective nature, the cooperation established among centers from three different countries allows a real-world comparison of patients undergoing this treatment. Furthermore, the patients herein included were collected mostly in a prospective fashion, either withdrawn from clinical trials due to screening failures or collected in institutional registries. This feature allowed a quite homogeneous data collection and further interpretation. Notwithstanding, no secondary event assessment was considered due to differences in registration, which largely differed according to local standards. Regarding the latter, previous publications have highlighted a possible equivalence, if not superiority, of combined IORT-EBRT against pure EBRT approaches. An early report by Kraus-Tiefenbacher et al. described similar toxicity profiles, in comparison to historical EBRT data, for a prospectively collected cohort of 73 patients (32). Recently, various reports have reached consistent findings, with median follow-up periods of up to 78 months (14); moreover, data suggested improved cosmetic outcomes when compared to EBRT-boost trials. For example, data pooled from the EORTC 22881/10882 trial demonstrated after physician-based assessment good/excellent cosmetic results in 71% of patients, while kIORT and IOERT data yielded rates over 90% and 86%, respectively (32–35), similar to those obtained after HDR application (36). Additionally, complementary hypofractionated EBRT does not seem to impair cosmetic outcomes, as recently reported by Burgos-Burgos et al. (37)

This patient cohort had rather unfavorable prognostic factors as those described by prospective trials, including ~26% TNBC/HER2, ~34% ≥T2, and ~8% R1, among others. As the main purpose of boost application is increasing local tumor control, the central endpoint to assess in this trial was LR and LFS (6). With a median follow-up period of 55 months, the cumulative recurrence rate herein obtained of 3.4% resembles those reported by the EORTC trial, about 4.3% after 5 years, or the 2.8% reported by the START-B trial. In addition, the former reported a 10-year cumulative recurrence rate of 6.4% compared to 7.9% LR in our study (3, 38, 39). Although this comparison results interesting, no conclusion could be assumed due to methodological and sample size differences (2,661 vs. 653 patients). Nonetheless, interesting hypotheses could be pondered after these findings. Other experiences published along the last decade, assessing kIORT as a boost approach, have reached similar and encouraging outcomes with follow-up times between 28 and 78 months (14, 40–43). In spite of recurrence risks, classical factors, such as patient age <50, higher T and clinical stages, ET, and ypN1 outcome, were related to increased failure rates. In the multivariate analysis, only age was a significant predictor for IBTR. We found with great surprise that other variables, which might be as well associated with failure, such as re-excision after IORT, R1, and in-situ component and its margin, had no relationship with increased recurrence rates. Special attention should be given to re-excision and R1 situations. The former is a relatively common issue at centers practicing IORT, as this has been reportedly performed after IORT application and receiving frozen-section notice. In our study, 151 patients recorded this feature. This is probably due to the surgeon’s preference for reducing surgical time and self-confidence in obtaining an upfront negative margin after primary excision. Depending on the extent of resection, this could potentially impair control outcomes. However, these re-excision interventions are usually focused on the compromised surgical coordinates, while the rest of the cavity remains unmodified after IORT. It is worth mentioning that approximately 30% of the applied dose reaches a 1-cm depth (17), which according to the previously pondered might still be active on the re-excised area. Regarding the latter, R1 has classically pointed the need of dose escalation to achieve comparable control rates to those R0 margins (44). The boost doses applied to these patients, mostly 20 Gy, should suffice to cover the entire clinical area of interest, considering the surface delivery and penetration depth of kIORT. Another item to consider, although not clinically proven, is the relative biological effectiveness (RBE), which might increase the effectivity of IORT at a cellular interaction level. Typically, an average 1.3 factor is employed in addition to 2-Gy per fraction equivalent dose (EQD₂) calculations to obtain equivalent doses to those of EBRT (45). We must remark that high dose-per-fraction EQD₂ calculations carry flaws and cautious interpretation of these data should be done. Despite these physical uncertainties, the growing body of clinical information demonstrates the value of the IORT approach. A further feature was fractionation. We observed no differences between hypo- and normofractionation, in terms of local control. Modern approaches including SIB are currently being investigated, aiming at shortening treatment times to the nowadays 3-week standard timespan for low-risk BC patients (7, 8). Future investigations will be required to assess whether upfront boosting with IORT might confer clinically relevant immune enhancement compared to EBRT.

Further variables related to LR, DM, and OS were found accordingly to those published by larger series (46). It drew our attention the large differences in survival among fractionation schemes, favoring hypofractionation. We must note the significant disparity in follow-up times between them, as hypofractionation was initiated later in time. This variability resulted in 19/159 and 264/492 assessable patients after 5 years, respectively. Therefore, a relevant clinical meaning should be disregarded.

Certainly, this investigation burdens a number of limitations. Its retrospective nature and inherent shortcomings are to be mentioned. No toxicity or cosmesis assessment was performed due to the heterogeneity of data collection and classification. Nonetheless, as previously mentioned, different groups have already reported these features. Patient-reported outcomes, added to physician objective assessment, are required to understand the actual toxicity extent and cosmetic implications of both kIORT and IOERT (47). Cancer control results should be interpreted according to the limitations of a retrospective assessment, as under no circumstances should it be assumed that IORT would overcome well-established recurrence risk features. Moreover, these indirect inter-study comparisons against seminal series should be carefully interpreted and taken only for referential purposes. The rather small number of events did not allow a subgroup analysis to adequately identify recurrence-related variables. Despite these drawbacks, major strengths of our study lie on the number of included patients, as well as their multicenter and international character. Careful patient selection and timing between IORT and EBRT start, among others, are important elements to be considered for avoiding undesired toxicity (5, 48). These real-world data, although encouraging, should be assumed preliminary and hypothesis generating, while results from ongoing trials are expected (NCT01343459, NCT01792726).

Upfront boost with IORT might yield similar local control outcomes to those reported by historical trials for high-risk early BC patients. Results from prospective trials, regarding toxicity, cosmesis, and effectivity, are awaited to confirm these findings.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

GS: study conceptualization, data collection, curation, statistical analysis, manuscript drafting and editing. MR: study conceptualization and data collection. AP: data collection and curation. RC: data collection and curation. FC: data collection. AC: data collection. JC: study design and site supervision. AH: site supervision. JG: study design and site supervision. PF-R: study design and data collection. AA: study design and data collection. DM: data collection and site supervision. KC: data curation. GZ: site supervision. LS: manuscript review and editing. LP: manuscript review and site supervision. FW: study conceptualization, manuscript review, and editing. FG: manuscript review and editing. GS: study conceptualization and supervision. ES: study conceptualization, data collection, manuscript review, editing, and supervision. All authors contributed to the article and approved the submitted version.

GS: personal fees and travel costs from Carl Zeiss Meditec AG, personal fees from Roche Pharma AG, personal fees from MedWave Clinical Trials, travel costs from Guerbet SA, not related to this work. FW: reports personal fees from Celgene GmbH, fees Roche Pharma AG, fees Eli Lilly and Company, fees from Ipsen Pharma GmbH and grants and other from Carl Zeiss Meditec AG and Elekta AB, patent by Carl Zeiss Meditec AG, not related to this work, outside the submitted work. FG: grants and personal fees from Carl Zeiss Meditec AG, personal fees from Roche Pharma AG, grants and personal fees from Elekta AB, grants and personal fees from NOXXON Pharma AG, grants and personal fees from Bristol-Myers Squibb, grants and personal fees from MSD Sharp and Dome GmbH, grants and personal fees from AstraZeneca GmbH, non-financial support from ONCARE GmbH, non-financial support from OPASCA GmbH, outside the submitted work. GS: personal fees and travel costs from Carl Zeiss Meditec AG, not related to this work. ES: grants from the Ministry for Science and Arts and others from Carl Zeiss Meditec, outside of the submitted work.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.