To our knowledge, this is the largest published mono institutional series of patients treated with IORT after resection of BM. Since 2020 more patients were treated with IORT than with SRT after microsurgical resection of BM. This is due to patients´ choices, who find the “one stop shop” character of this treatment very appealing. Nevertheless all patients of the UKA, who are planned for BM resection after decision of the MTB, are offered the option of a postoperative cavity SRT as well as the option of IORT.
The observed LC rate of 90.4% with an estimated LC of 84.2% at 1 year (95% CI: 71.3–91.6%) is as well in line with the smaller IORT series of De Castro et al. and Cifarelli et al. [8, 9] as with postoperative cavity SRT series [3, 14–17] and it stayed unchanged compared to our last publication 2021 [10] after the first 40 treated patients. Remarkably, the high rate of suspected incomplete resections on postoperative MRI of 34.6% did not transfer into worse local control (84.7% (95% CI: 66.5–93.4%) R0 vs. 82.1% (95% CI: 58.9–93.0%) R1; p = 0.606) in univariate analysis. This stands in contrast to the findings of Cifarelli et al. [8], who found this to be the only predictor of LC in their data set. We hypothesize that contrast enhancements on the cavity rim as a reaction of the brain tissue after IORT might blur the distinction between residual disease and IORT related tissue changes.
Patients´ median OS of 18.2 months (range 0.5–79.0 months) stresses the relevance of LC without relevant treatment associated neurotoxicity in this population. Most patients live long enough to experience the impairments either by local recurrence or by therapy associated neurotoxicity. Taking in account the estimated DC of 47.9% at 1 year (95% CI: 34.7–60.0%) close MRI FU is of uttermost importance in this setting to detect the frequent distant relapses before they become symptomatic and deteriorate patients´ quality of life. With the applied UKA FU policy of three-monthly MRIs we were able to detect all distant brain relapses in an asymptomatic state. Most of them could be salvaged with SRT. Only 16% of our patients required WBI due to multiple BMs and/ or LMD during their whole course of cancer treatment. This is in line with data from BM patients treated with SRT [7, 14, 18]. LMD among our patients with an estimated one- year LMD rate of 10.4% (95% CI: 4.7–22.1%) was not observed more frequently than reported in patient series treated with postoperative SRT [14, 19–21]. Hence, there is no evidence that the IORT procedure affects the risk of tumor cell spillage to the cerebrospinal fluid by neurosurgery.
IORT did not increase the perioperative complications rate of brain surgery. Although our series contains cases with huge metastases up to 7cm diameter, 12% of the treated patients fitted to RPA class 3 and 13% of the metastases were located in the posterior fossa, the 30-days- mortality and complication rate were within the range described in recent publications of patient cohorts after neurosurgical BM resection with or without IORT [3, 9, 22–26]. The complete IORT procedure prolonged OR time for 25 minutes in mean including less than 16 minutes radiation time. This and the limited radiation protection measures, comparable to the protection measures required for the use of fluoroscopy, made it possible for our center to integrate the IORT procedure in our routine workflow for microsurgical BM resections easily.
Median time from surgery to discharge from hospital was 6 days (range 2–41 days) for our patients, which is as long as it is for patients after BM resection without IORT. Dejonckheere et al.[27] compared the mean time to next treatment of patients after BM resection with IORT to patients treated with postoperative SRT to the cavity in a retrospective single center study. They found a significant difference in favor of patients treated with IORT (36 (9 − 94) days) versus patients treated with postoperative SRT (52 (11 − 126) days). We observed a mean time to systemic treatment (TTST) of 31days (median TTST: 24 days (range 1 -136 days)) in our patient set, which is in line with the results of the quoted study.
Within this retrospective single center study we observed an extremely low rate of RN during FU. With an observed RN rate 2.6% and an estimated RN rate of 3.9% at one year, we report a markedly lower RN rate compared to other published data on focal BM cavity radiotherapy. In the current literature the rates of RN rage from 5 to 25% for patients treated with SRT after resection of BM [3, 14, 28–30]. Our comparatively low RN rate could possibly be explained by the relatively small volume of surrounding brain tissue receiving 10 Gy (V10), due to the steep dose gradient of 50 kV x-rays [10]. V10 is an established risk factor for RN in SRT. In our series the mean applicator size was 2.0 cm (range 1.5–4.0 cm), corresponding to a nominal mean V10 of 6.12 cm3 (range 3.08–35.95 cm3). Taking into account relative biological efficacy (RBE) of low energy X-rays, the corresponding mean V10 (RBE) is 12.97 cm3 (range 4.6–48.94 cm3). This data supports the clinical benefit gained from the steep dose gradient of 50kV x-ray IORT around the resection cavity already described by other groups [31, 32].
Nevertheless there are some restrictions to the interpretation of our data. As a retrospective study it is subjected to bias due to patient selection and reporting. We treated the majority of our patients within the last two and a half years, which puts a limit to the FU time of this data. Some patients were lost to FU, which is a source of further uncertainty. These effects may be partly balanced by the large number of procedures reported. Thus the results represent the clinical status of IORT after microsurgical resection of BM, which can be achieved as institutional standard treatment in an university center.
Conclusion After117 procedures IORT as an institutional standard treatment appears to be a safe and appealing way to perform cavity RT after microsurgical resection of BM with low toxicity and excellent LC. For patients with additional systemic tumor burden IORT in this situation holds the chance for an early start of adjacent systemic therapy. Three-monthly FU with MRI is paramount to detect the frequent distant brain failure (DBF) early. In this setting WBI could be avoided for over 80% of the patients in the further course of disease using SRT as effective salvage therapy for DBF.