Glioblastoma is the most common and devastating type of primary intracranial tumor (1). According to the 2021 statistical report of the Central Brain Tumor Registry of the United States (CBTRUS), glioblastoma accounts for 49.1% of all malignant brain and central nervous system (CNS) tumors and 58.4% of gliomas (2). Because vascular endothelial growth factor (VEGF) and VEGF receptor (VEGFR) are up-regulated in glioblastoma, antiangiogenic drugs targeting VEGF ligand have been suggested and introduced into the treatment regime with radiotherapy plus concurrent or adjuvant temozolomide (3, 4). These antiangiogenic drugs may inhibit angiogenesis primarily through destruction of existing tumor vasculature, normalization of surviving vessels, and inhibition of new and relapsing tumor blood vessel growth (5–7). Bevacizumab, a humanized monoclonal antibody targeted to the VEGFA ligand, was approved for the treatment of recurrent glioblastoma rather than newly diagnosed glioblastomas by the Food and Drug Administration (8). A series of prospective clinical trials had yielded some results, patients treated with bevacizumab combined with conventional concurrent radiotherapy and temozolomide (CCRT) revealed a median overall survival (OS) ranging from 15.7-19.6 months and median progression-free survival (PFS) 8.4-13.6 months versus those only received CCRT who had a median OS ranging from 14.6-21.1 months and PFS 4.3-9.4 months (9, 10). The PFS rather than OS of patients with newly diagnosed glioblastoma treated with bevacizumab combined with CCRT was significantly improved. However, the OS has been found to be heterogeneous, with a considerable number of patients still benefiting significantly from this combination therapy.

The ability to accurately screen this portion of the population benefiting from PFS and OS could provide a breakthrough in the treatment of newly diagnosed glioblastoma patients. It is well-established that glioblastoma patients with promoter methylation of the DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT) derive more benefit from treatment with temozolomide (11). However, the prediction of the response to antiangiogenic therapy has always been a complicated problem. VEGFA is the major mediator promoting tumor-induced angiogenesis (4). Its predictive value for the response to antiangiogenic therapy has been extensively studied, but inconsistent prediction results have limited its widespread application in clinical practice (12–15). There remains no appropriate parameter for predicting the response of glioblastoma to bevacizumab as precisely as epidermal growth factor receptor (EGFR) mutation predicts the response to tyrosine kinase inhibitor (TKI) treatment (16, 17).

As bevacizumab achieves an anti-tumor effect mainly via changing the density of microvessels and decrease vascular permeability, blood flow, and the degree of oxygenation in the local microenvironment (18), it may be a more rational direction to search for prediction parameters through dynamic imaging examination to solve the current dilemma. Arginine-glycine–aspartic acid (RGD) is the most important integrin for angiogenesis, because it specially binds to the integrin alpha V beta3 (αvβ3), which is highly expressed in new tumor vessels (19, 20). In theory, RGD positron emission tomography (PET) can provide direct imaging of angiogenesis, and ¹⁸F-Galacto-RGD uptake was shown to be significantly correlated with αvβ3 integrin staining intensity in glioblastoma lesions (21). Recently, the novel one-step method for preparing the integrin αvβ3-targeting PET probe ¹⁸F-AlF-NOTA-PRGD2 has made tracer preparation more convenient, which then increases the likelihood that this imaging method can be popularized in clinical practice (22). Quantitative permeability parameters such as the volume transfer constant (K^trans) on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) may also reflect the physiologic characteristics of the microvasculature, permeability, and angiogenesis (23, 24).

Therefore, we registered and performed the present prospective study. ¹⁸F-AlF-NOTA-PRGD2 and DCE-MRI were completed to identify potentially effective biomarkers for the prognosis of newly diagnosed glioblastoma after surgery treated with the combination of bevacizumab with CCRT.

This was an open-label, single-arm, phase IV clinical trial involving patients with newly diagnosed glioblastoma (World Health Organization [WHO], Grade IV) that was confirmed histopathologically after surgery. This trial was approved by the ethics committee of Shandong Cancer Hospital and Institute (number NCT01939574, ID ML28676).

This prospective study examined the potential value of imaging biomarkers for assessing the prognosis of patients with postoperative newly diagnosed glioblastoma treated with bevacizumab combined with CCRT. A large decrease in the tumor SUV_mean on RGD-PET for PFS as well as small decrease in K^trans on DCE-MRI from week 3 to week 10 were favorable factors for improved survival, indicating that these imaging modalities offer early screening benefits for patients treated with combination therapy to better guide treatment planning.

Our exciting findings of the value of the new imaging markers for predicting survival following treatment with bevacizumab and CCRT were in accordance with the prior literature. Provost et al. reported that imaging via ⁶⁸Ga-RGD PET/CT rather than ¹⁸F-FDG PET/CT revealed changes in the SUV of tumors following bevacizumab and/or temozolomide treatment in mice models bearing a U87MG tumor (30). In another study employing an bevacizumab-containing therapy in six cases of ovarian and cervical cancers, a large decrease in the RGD uptake was noted in patients with an early objective response (31). Eight glioblastoma patients received bevacizumab treatment and underwent pretreatment ¹⁸F-FPPRGD2 PET, and in these patients, a small decrease in ¹⁸F-FPPRGD2 uptake tended to indicate a poor prognosis (32). Xia et al. showed that a lower decrease in K^trans was associated with a better response to antiangiogenic treatment followed by chemoradiotherapy in 11 soft tissue sarcoma patients (33). However, these studies were limited by the use of only animal models, a small sample size, or a non-standard treatment mode. They mainly focused on monitoring rather than comparing survival to identify clinical characteristics associated with statistically significant survival benefits.

The newly diagnosed glioblastoma patients enrolled in this study postoperatively were receiving the standard combination treatment of bevacizumab with CCRT and underwent dynamic combined imaging examination of RGD PET and DCE-MRI. As described in previous studies, RGD PET served as an excellent imaging method for glioblastoma patients with a clear background, according to the passage of the tracer through the damaged blood-brain barrier within the tumor (22, 34). Patients with a larger decrease in the RGD SUV_mean from before to after bevacizumab treatment significantly showed a better PFS. Bevacizumab is a humanized monoclonal antibody targeting VEGFA. It limits tumor growth mainly by blocking the blood supply within the existing tumor vasculature, inhibiting the function of the tumor microvasculature as well as endothelial cell migration, thereby preventing regrowth over time (18, 35). The complexes formed by αvβ3 integrin and VEGFRs are considered to be the most important integrin interactions in angiogenesis, and the αvβ3-targeting RGD uptake values may reflect VEGF pathway activity (36–38). Therefore, the greater decrease in SUV_mean may represent stronger inhibition of tumor neovascularization and migration by bevacizumab to achieve superior antiangiogenic efficacy.

In addition, as shown in our study, K^trans showed a significant decrease during the period of combination therapy, and a small decrease in K^trans from week 3 to week 10 on DCE-MRI was predictive of improved OS. It is well known that the parameter of K^trans characterized by wash-in rates is positively correlated with the capillary permeability in the tissue of interest (23). Because bevacizumab acts as an anti-tumor agent primarily by reducing the vascular permeability and blood flow, the K^trans values could be decreased along with the reduction in permeability (18). However, the steep decrease may aggravate local hypoxia, which may further lead to treatment resistance (39, 40). Meanwhile, bevacizumab may cause cell death and lysis and, thereby, reduce the cell density and the associated restrictive barriers of cell membranes. The destruction of the cell barrier perhaps leads to an increase in permeability, corresponding to an increase in K^trans. On the other hand, it has been suggested that bevacizumab may transiently normalize tumor vasculature by pruning the immature and inefficient vessels and remodeling the remaining vessels, which would reduce the leakage of contrast agent (41–46). Therefore, the relationship between K^trans and OS may be related to the initial vascular status at week 3. With more normal blood vessels present, there may be fewer targets for vascular normalization during antiangiogenic therapy. Accordingly, the decrease in K^trans value would be smaller. Furthermore, the relationship between maintenance of the normal tumor vasculature and prognosis also has been reported in multiple studies (47, 48). Hence, the ultimate trend indicates that a small decrease in k^trans may be sufficient to identify high-risk individuals who would likely achieve improved OS with antiangiogenic therapy.

Recently, immunocheckpoint inhibitors represented by PD⁃1 have been widely studied in glioblastoma. Series of phase III clinical trials displayed disappointing results of PD⁃1 in glioblastoma, including the OS of recurrent glioblastoma in CheckMate-143, OS of newly diagnosed MGMT⁃unmethylated glioblastoma, PFS of newly diagnosed MGMT⁃methylated glioblastoma (49–51). In addition, it’s also worth to note that the efficacy of neoadjuvant was better than the adjuvant PD⁃1 therapy in recurrent glioblastoma (52). Although another study of the combination of neoadjuvant and adjuvant PD⁃1 therapy showed no survival benefit from neoadjuvant therapy in patients with resectable glioblastoma (53), both studies revealed that the neoadjuvant therapy was capable of regulating tumor immune microenvironment. These results indicated the importance of appropriate administration timing, combination therapy and tumor microenvironment status during immunotherapy. At the same time, potential biomarkers to predict the efficacy of immunotherapy maybe more conducive to individualized treatment. Researchers revealed that PTEN mutations, MAPK pathway alterations (PTPN11, BRAF) was correlated with anti⁃PD⁃1 immunotherapy (54). Since the activation of immune cells promotes normalization of tumor vessels, and RGD-based PET and DCE-MRI was potential to monitor the changes of tumor microenvironment by characterizing the intracranial neovascularization during immunotherapy (55).

Overall, the independent predictive roles of SUV_mean and K^trans may be related to their ability to represent the status of the tumor vasculature within the whole tumor, avoiding the influence of local heterogeneity. We measured mean values of permeability parameters (K^trans, Kep, Ve, and Vp) instead of maximal or minimal values, which may also be useful to overcome heterogeneity. Based on their different predictive values for prognosis, the two imaging modalities evaluated may provide complementary rather than similar information regarding the prognosis of patients treated with antiangiogenic therapy in this study.

To the best of our knowledge, this is the comprehensive prospective registered clinical trial in which possible prognostic imaging biomarkers were detected. These encouraging results indicate that accurate, image-guided individual therapy is possible, and further research is warranted. Our clinical trial was a small sample exploratory investigation, and thus, larger studies with data for more molecular markers are needed to validate our findings.

The present study shows that ¹⁸F-RGD PET/CT and DCE-MRI examination maybe useful for identifying glioblastoma patients likely to benefit from treatment with the combination of bevacizumab and CCRT, with a greater decrease in SUV_mean predicting better PFS as well as a smaller decrease in K^trans predicting improved OS.

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

The studies involving human participants were reviewed and approved by Shandong Cancer Hospital and Institute. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Study design: SY and JY; Data acquisition and analysis: LL, NL, LX, and YL; Patient sample collection: HZ, RT, SZ, ZC, and ZF. All authors have read and agreed to the published version of the manuscript.

This study was partially funded by Natural Science Foundation of China (NSFC81872475, and NSFC82073345), Shandong Key Research and Development Plan (2017CXGC1209) to SY as well as Research Unit of Radiation Oncology, Chinese Academy of Medical Sciences (2019RU071) and the Academic Promotion Program of Shandong First Medical University (2019ZL002), the foundation of National Natural Science Foundation of China (81627901, 81972863 and 82030082), the foundation of Natural Science Foundation of Shandong ZR201911040452). to JY Tasly Diyi Ltd. provided the drug Temozolomide (DIQING). Roche Ltd. provided the drug Bevacizumab and part of funding. Tasly Diyi Ltd. and Roche Ltd. played no role in study design, data collection, data analysis and data interpretation.

The 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.