Journal of Neuro-Oncology https://doi.org/10.1007/s11060-019-03309-6 CLINICAL STUDY Intraoperative radiotherapy (IORT) for surgically resected brain metastases: outcome analysis of an international cooperative study Christopher P. Cifarelli1,3 · Stefanie Brehmer5 · John Austin Vargo2 · Joshua D. Hack3 · Klaus Henning Kahl4 · Gustavo Sarria‑Vargas6 · Frank A. Giordano6 Received: 19 August 2019 / Accepted: 5 October 2019 © Springer Science+Business Media, LLC, part of Springer Nature 2019 Abstract Background and objective The ideal delivery of radiation to the surgical cavity of brain metastases (BMs) remains the subject of debate. Risks of local failure (LF) and radiation necrosis (RN) have prompted a reappraisal of the timing and/or modality of this critical component of BM management. IORT delivered at the time of resection for BMs requiring surgery ofers the potential for improved local control (LC) aforded by the elimination of delay in time to initiation of radiation following surgery, decreased uncertainty in target delineation, and the possibility of dose escalation beyond that seen in stereotactic radiosurgery (SRS). This study provides a retrospective analysis with identifcation of potential predictors of outcomes. Methods Retrospective data was collected on patients treated with IORT immediately following surgical resection of BMs at three institutions according to the approval of individual IRBs. All patients were treated with 50kV portable linear accelerator using spherical applicators ranging from 1.5 to 4.0 cm. Statistical analyses were performed using IBM SPSS with endpoints of LC, DBC, incidence of RN, and overall survival (OS) and p < 0.05 considered signifcant. Results 54 patients were treated with IORT with a median age of 64 years. The most common primary diagnosis was non- small cell lung cancer (40%) with the most common location in the frontal lobe (38%). Median follow-up was 7.2 months and 1-year LC, DBC, and OS were 88%, 58%, and 73%, respectively. LMD was identifed in 2 patients (3%) and RN present in 4 patients (7%). The only predictor of LC was extent of resection with 1-year LC of 94% for GTR versus 62% for STR (p = 0.049). Conclusions IORT is a safe and efective means of delivering adjuvant radiation to the BM resection cavities with high rates of LC and low incidence of RN. Further studies are warranted directly comparing LC outcomes to SRS. Keywords Brain metastases · IORT · Intraoperative radiotherapy Introduction * Christopher P. Cifarelli [email protected] The use of adjuvant cranial radiation in patients requiring surgical resection for large brain metastases (BMs) remains 1 Department of Neurosurgery, West Virginia University, 1 a standard of care in most regions. Concerns related to the Medical Center Drive, Suite 4300, Morgantown, WV 26505, potential neurocognitive impact of whole brain radiation USA (WBRT) have resulted in the adoption of stereotactic radio- 2 Department of Radiation Oncology, UPMC, Pittsburgh, PA, surgery (SRS) approaches for use in treatment of metastatic USA lesions with or without surgical resection [4, 19]. Unfor- 3 Department of Radiation Oncology, West Virginia tunately, local control rates and the risks of leptomenin- University, Morgantown, WV, USA geal dissemination (LMD) in post-operative SRS treated 4 Department of Radiation Oncology, University Medical patients have failed to surpass rates achieved via post-oper- Center Augsburg, Augsburg, Germany ative WBRT [3]. Accordingly, proposals have hypothesized 5 Department of Neurosurgery, Medical Faculty Mannheim, that the use of pre-operative SRS could decrease the risks Heidelberg University, Mannheim, Germany of LMD and local failure, while others have developed 6 Department of Radiation Oncology, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany Vol.:(0123456789)1 3 Journal of Neuro-Oncology dedicated SRS targeting guidelines which are still being dose reduction or abandonment of the IORT procedure validated [15, 20]. were reserved for cases in which the dose to critical organs Advances in radiation delivery techniques have led to a exceeded tolerances. BMs in the posterior fossa where re-emergence of intra-operative radiotherapy (IORT) as a excluded from IORT treatment based on proximity to the viable mechanism for delivery of adjuvant treatment fol- brain stem. Follow-up imaging was performed at the dis- lowing tumor resection. The safety of IORT in the setting cretion of each institution’s practice, but generally included of glioblastoma has been established in conjunction with the gadolinium enhanced magnetic resonance imaging (MRI) addition of external beam radiotherapy (EBRT) on the Stupp with perfusion imaging or positron emission tomography protocol and its efcacy is being evaluated in Phase III clini- (PET) for evaluation of radiation necrosis versus disease cal trial [6]. The published data regarding the use of IORT progression. Radiation necrosis was determined by the pres- for surgically resected BMs is limited to single institutional ence of contrast enhancement in the absence of increased studies [9, 24, 25]. Moreover, recent evidence suggests that perfusion or FDG uptake. Imaging was performed at a mini- the dose delivered locally to the resection cavity may be mum interval of every 3 months for the frst two years fol- escalated beyond that safely achievable with traditional SRS lowing IORT dependent on overall survival. Radiographic techniques, eliminating delay in time to initiation of radia- progression was determined based on RANO criteria, with tion treatment following surgery and avoiding the complex- the defnition of distant lesions being those not in continu- ity of target delineation in post-operative period [24]. ity with the post-operative resection cavity not previously The current study represents a retrospective analysis of identifed prior to IORT. multi-institutional data from centers in the United States and Statistical analyses were completed using IBM SPSS ver- Europe designed to provide insight into the ability of IORT sion 24 (Armonk, NY). Local failure (LF, defned as radio- to provide improvement in local control and reduced toxicity graphic progression at the site surgery and IORT), distant compared to historical data from SRS. brain control (DBC, defned as the absence lesions at other intracranial sites independent of surgery and IORT), and overall survival (OS) were calculated using Kaplan–Meier Methods method. Comparisons between groups were made using the log-rank method. Variables examined included: gender, age, Following institutional review board approval, an interna- primary histology, IORT dose, applicator size, lesion loca- tional multi-institutional data set was retrospectively formu- tion, eloquent versus non-eloquent location, and extent of lated between three participating centers. Data was shared in resection based on post-operative imaging. Continuous vari- de-identifed format using a comprehensive data set formu- ables were categorized around the median for comparisons. lated by all study investigators. Inclusion criteria included Due to a low number of events relative to the number of patients treated with IORT in conjunction with resection potential covariates, a valid multivariate model could not of brain metastases. The decision to proceed with surgical be generated. Variables potentially impacting the develop- resection was made by the primary treating neurosurgeon ment of radiation necrosis were compared using a two-sided and followed the guidelines of being a lesion not amenable Chi square test. A p < 0.05 was considered signifcant for all to SRS based on size or requiring surgical removal for mass analyses. efect and/or need for tissue diagnosis. Patients with primary brain tumors were excluded. Patients who underwent surgery and IORT followed by SRS for additional untreated lesions Results were included in the analysis. Although the goal of surgical treatment was to achieve a gross total resection, patients with Baseline patient and treatment characteristics for the post-operative radiographic evidence of a subtotal resection included 54 patients are detailed in Table 1. Briefy the was also included in the analysis. median age was 64 years, the most common histology was All patients were treated with IORT using the Intrabeam non-small cell lung cancer (40%, n = 23), the most common device (Carl Zeiss Meditec AG, Oberkochen, Germany) location was frontal lobe (38%, n = 22), and the majority with spherical applicators ranging from 1.5 to 4.0 cm in were in non-eloquent regions of the brain (72%, n = 42). diameter using a low-energy 50 kV X-ray portable linear Post-operative imaging verifed gross total resection was accelerator. IORT prescription doses were specified to achieved in 81% (n = 44). The median prescription dose for the spherical applicator surface and delivered in a single IORT was 30Gy to the applicator surface [interquartile range fraction following resection of the tumors. Intra-operative (IQR) 20–30Gy]. The median applicator size was 2.0 cm navigation was utilized to determine the distance to criti- (IQR 2.0–2.5), with a median treatment time of 16.8 min cal organs, i.e., brainstem and optic apparatus. Following (IQR 12.1–22.3). from the INTRAGO II trial guidelines for glioblastoma, The median follow-up was 7.2 months (IQR 3.4–15.3). 1 3 Journal of Neuro-Oncology Table 1 Baseline patient and treatment characteristics Baseline characteristics Median (inter- quartile range), or n (%) Gender Male 24 (44%) Female 30 (56%) Age 64 (54–68) Primary Histology Non-small cell lung 23 (43%) Breast 8 (15%) Melanoma 8 (15%) Renal 4 (7%) Gastrointestinal 4 (7%) Gynecologic 2 (4%) Other 5 (9%) Prescription Dose > 18 Gy 4 (7%) 18 Gy 7 (13%) 20 Gy 14 (26%) 30 Gy 29 (54%) IORT applicator size 1.5 cm 9 (17%) 2.0 cm 20 (37%) 2.5 cm 13 (24%) 3.0 cm 8 (15%) 3.5 cm 1 (2%) 4.0 cm 3 (6% IORT treatment time, minutes 16.8 (12.1–22.3) Brain Lobe Frontal 22 (41%) Parietal* 15 (28%) Occipital 6 (11%) Temporal 5 (9%) Posterior Fossa 6 (11%) Eloquent brain Eloquent 12 (22%) Non-eloquent 42 (78%) Extent of resection Gross total 44 (81%) Sub total 10 (19%) Median age of 64 years, diagnosis of non-small cell lung cancer and Fig.