In this pooled analysis from two large tertiary centers, we conducted a comparative assessment of patient outcomes undergoing either early adjuvant Re-RT or no adjuvant Re-RT for patients with rGBM after MRI approved GTR. The results did not reveal a significant difference in OS between adjuvant Re-RT and no adjuvant Re-RT. Incorporating re-RT into a multivariate model did not reveal a significant correlation between the use of adjuvant Re-RT and OS. These findings align with recent randomized trials investigating Re-RT for recurrent GBM (rGBM), which have shown an improved PFS but no significant difference in OS [9, 10]. Furthermore, in a retrospective study comparing second surgery with re-RT (repeated radiation therapy) to second surgery without Re-RT, no significant difference in overall survival (OS) was observed, too [8].

Reasons for the lack of OS benefits from adjuvant Re-RT are speculative. Possible explanations could be related to treatment factors, namely inadequate dose or to small target volume. Alternatively, toxic effects could consume the oncologic treatment effect, hence, a PFS benefit would not result in an OS benefit because of lifetime limiting toxicity. One option to enhance the therapeutic ration in Re-RT is the addition of Bevacizumab, a drug known to reduce the incidence of radiation necrosis after Re-RT by 83% [13]. Notably, due to restrictions in re-imbursement in Germany, only one patient received Bevacizumab in the present cohort.

Most of the reported patients received some kind of chemotherapy. However, patients in the Re-RT cohort were slightly more likely to receive chemotherapy as compared to patients in the no Re-RT-cohort. As all patients were deemed to be in a good condition after second surgery, as defined within the inclusion criteria, one would expect an at least equal proportion of patients receiving chemotherapy despite of not receiving adjuvant Re-RT. Hence, the Re-RT group was treated more intensive also in the view of systemic treatments. A recent meta-analysis from Marwah et al. showed that combination therapy increased OS as compared to mono-RT, but it did not increase OS when compared to mono-CTx [13]. Notably, the underlying data are mostly from trials reporting treatment of macroscopic disease. The potential added efficacy of Re-RT might therefore be limited in the context of a former GTR. Consequently, the patient number in our cohort is not large enough to reach significant differences from no adjuvant treatment or mono therapy to Re-RT in combination with chemotherapy.

The data presented in this study offer further insights into prognostic factors for patients with rGBM. Our results highlight the relevance of patient-specific factors, including post-surgery performance status and MGMT methylation status. Furthermore, the time interval between the first RT and the second surgery, which are surrogates for the individual aggressiveness of the disease, has emerged as a crucial contributing factor, resulting in a strong tendency with borderline significance in a multivariate model. These factors, including post-surgery performance status, MGMT methylation status, and the time interval between the first RT and the second surgery, have not only been identified by our research team but also by other researchers [7, 8, 14,15,16,17]. Moreover, these factors have been integrated into prognostic scores specifically designed for rGBM [7, 17].

For patients with rGBM, a second surgery is a well-established salvage strategy that should be considered in selected cases. The selection criteria typically include patients with a good performance status, small tumor volume, and a location distant from eloquent areas [18]. Pooled data from multiple large-scale centers in Germany have demonstrated the safety and feasibility of this approach, particularly benefiting patients who have undergone GTR [4]. Additionally, a secondary analysis of the DIRECTOR-Trial, which evaluated the efficacy of two different temozolomide regimens in rGBM, indicated a significant advantage for patients based on the extent of resection. Notably, patients who underwent GTR experienced substantial benefits in this prospective cohort [3]. However, conflicting data exist regarding subtotal resection (STR), with some reports indicating no or even detrimental effects of STR in the DIRECTOR analysis. Conversely, Yang et al. reported a progressively increasing benefit with the relative extent of resection, with even small remnants after the second surgery showing some remaining, albeit diminished, benefit for patients [19]. Furthermore, a meta-analysis of 1906 cases demonstrated a significant improvement in patient outcomes when a second surgery was performed (hazard ratio 0.72, p < 0.001) [20]. Therefore, a second surgery for rGBM is– despite a lack of prospective comparative evidence– emerging as a potential treatment approach, particularly in carefully selected patients based on good performance status, small tumor volume, and location distant from eloquent areas [2]. Noteworthy, given the previous selection criteria, our data do not support the use of age as a general exclusion criteria. This also aligns to our previous experience with Re-RT in GBM [21].

The strengths of this article are the homogeneous patient selection in accordance with a published trial protocol as well as the bicentric inclusion of patients, which allows a better generalizability of the data. Limitations are mostly due to retrospective nature of the assessment [22]. Given that positive predictive factors were skewed in favor of patients who received Re-RT, it is unlikely that an effect of Re-RT was missed in this cohort. Exact matching of the cases was not possible due to the low number of patients fulfilling the rigorous inclusion criteria. Additionally, the low patient number reduces the statistical power, a true existing therapeutic effect thus could be overseen. Furthermore, the Re-RT dose regimens were not standardized which introduced some uncertainty, and imaging analysis after Re-RT was not included, hence information on PFS after Re-RT are not available. Additionally, a substantial number of patients had an unknown IDH status. It is known that IDHmt is associated with longer OS after Re-RT [23]. However, since there was no statistical difference in OS between IDHwt and IDH unknown groups, we assume that the IDH unknown cohort had none or very few IDHmt patients which should not influence the overall conclusion of this report. Lastly, analysis of safety was beyond the scope of this article, as comparative data from the control group were not available.

The ongoing GlioCave trial will focus on the impact of Re-RT on PFS. The early results of this randomized trial are eagerly awaited as they might clarify whether Re-RT is safe and can lead to a clinically meaningful improvement in PFS. Importantly, within a palliative setting, a safe treatment that results in a significant improvement in PFS could be an important treatment modality, even in the absence of an OS benefit.