ESMO 2022: spotlight on new emerging treatment options in central nervous system tumors

Indoleamine‑2,3‑dioxygenase (IDO) is an enzyme catalyzing the conversion of tryptophan to kynurenine and exerts immune-suppressive properties via a plethora of metabolic and nonmetabolic mechanisms. Indeed, IDO is expressed in the microenvironment of glioblastoma and is associated with infiltration of regulatory T cells which inhibit antitumoral immune responses [5], providing the rationale for the evaluation of IDO inhibitors.

Bevacizumab is frequently used in recurrent glioblastoma for symptomatic perifocal edema, although exerting only little (if any) antitumoral activity [6]. However, inhibition of vascular endothelial growth factor (VEGF) was shown to modulate the tumor microenvironment, providing the rationale for investigating the combination of immune-modulating agents with bevacizumab [7, 8]. In a phase 2 study, the combination of hypofractionated radiotherapy, bevacizumab and the anti-PD‑1 antibody retifanlimab ± the IDO inhibitor epacadostat is being evaluated in patients with recurrent glioblastoma [9]. After confirming the favorable side effect profile of hypofractionated radiotherapy + bevacizumab + retifanlimab in cohort A, the addition of epacadostat is assessed in cohort B. At ESMO 2022, the results of cohort A were presented. In 24 included patients, only one grade 3 toxicity was seen (myositis), and overall response rate was 62.5%, whereas disease control rate reached 87.5%, with durable responses in 37.5% of patients. The median progression-free survival (PFS) and overall survival (OS) times were 7.6 months and 11.1 months, respectively. These results are well in line with other trials evaluating bevacizumab either as control group or as experimental arm in combination with other agents [1, 10]. Data on the addition of epacadostat remain to be awaited as enrollment of patients to cohort B is still ongoing.

Recently, it was shown that intratumoral release of interferon alpha (IFN-alpha) leads to potent antitumoral immune responses within the tumor microenvironment which may even lead to tumor eradication in murine models [11]. Based on this and previous work, autologous hematopoietic stem cells were ex vivo transfected with a lentiviral vector encoding for IFN-alpha under the Tie2 promoter, which is active in tumor-infiltrating monocytes that contribute to a growth-promoting milieu in the glioma microenvironment. At ESMO 2022, early clinical and translational data of a phase 1/2a trial enrolling patients with newly diagnosed, O6-methylguanine methyltransferase (MGMT) promoter-unmethylated glioblastoma (NCT03866109) were shown [12]. The expression of transgenic IFN-alpha was well controlled, as serum IFN-alpha levels were low and no systemic, potentially IFN-alpha-related adverse events were seen. In addition, persistence of transduced cells was observed up to 18 months after administration. Whereas median OS (15 months after first surgery) and PFS (8.3 months) were comparable to historical controls [13], some long-term survivors were observed. Providing mechanistic insights, single cell RNA sequencing of tumor-infiltrating leukocytes at second surgery showed an enrichment of IFN-alpha-related pathways and indicated repolarization of macrophages. Still, superiority over the current standard of care remains to be proven in well-designed randomized controlled trials.

L19TNF is a compound consisting of tumor necrosis factor alpha and an antibody targeting fibronectin (L19). Preclinical and early clinical data have shown that L19TNF causes intratumoral necrosis and accumulation of proinflammatory cytokines within the tumor microenvironment [14]. At ESMO 2022, data on the combination between L19TNF and the alkylating agent lomustine (CCNU) were presented [15]. In murine models, a synergistic effect was observed which seemed to be immune-dependent, as long-term antitumoral immunity was observed. Moreover, the combination was paralleled by an increase of lymphocyte infiltration. Of the first 6 patients enrolled in the phase 1 part of the phase 1/2 trial (NCT04573192), 2 patients achieved objective responses; further data on efficacy remain to be awaited for the phase 2 stage, where patients are currently 1:1 randomized to CCNU alone or CCNU + L19TNF.

Based on data showing that the phosphoinositide‑3 kinase (PI3K) pathway is activated in a large fraction of glioblastomas, the NCT03522298 phase 2 trial explored the use of the PI3K and mammalian target of rapamycin (mTOR) inhibitor paxalisib in newly diagnosed, MGMT promoter unmethylated glioblastoma [16]. After temozolomide-based concurrent radiochemotherapy, patients received adjuvant paxalisib at either 60 mg or 75 mg in a fed and fasted state. Based on safety data, the maximum tolerated dose was 60 mg, and systemic exposure to paxalisib was marginally higher in a fed as compared to fasted state, whereas there were no differences in other pharmacokinetic parameters. Of note, median PFS and OS were 8.6 and 15.9 months, respectively, comparing favorable to historical data on MGMT promoter-unmethylated glioblastoma receiving radiochemotherapy. Further data on efficacy will be generated within GBM AGILE (NCT039070447), an ongoing platform trial assessing the use of various targeted agents in glioblastoma [17].

Regarding radiotherapeutic treatments, data on a nanoliposomal formulation of the radionuclide rhenium-186 (¹⁸⁶RNL) were shown [18]. In this trial, ¹⁸⁶RNL was locally delivered via convection-enhanced delivery in recurrent glioma, utilizing pressure gradients via intratumoral/-cavitary catheters to improve tumor tissue penetration. Local application of radionuclides allows to achieve considerably higher radiation doses which are limited by radiation exposure of healthy tissue in conventional external beam radiotherapy. Indeed, absorbed radiation dose in the trial reached up to 740 Gy which is more than 10-fold compared to conventional radiotherapy, and a significant survival benefit was observed in patients with an effective dose > 100 Gy. Still, there was high heterogeneity in absorbed doses (range 9–740 Gy) and tumor/treated volume ratios (13–100%). Moreover, the number of placed catheters differed considerably, underscoring the need for standardized protocols and optimal patient selection. Although survival data are promising, the superiority towards other methods remains to be proven in randomized controlled trials.

Antibody–drug conjugates are increasingly used in a wide array of solid tumor entities. Recently, an increasing body of evidence showing intracranial activity also in active brain metastases has emerged. One of these trials was the TUXEDO‑1 phase 2 trial, evaluating the intracranial efficacy of trastuzumab deruxtecan (T-DXd) in HER2+ breast cancer brain metastases [19]. Here, an intracranial response rate of 73.3% according to Response Assessment in Neuro-Oncology for Brain Metastases (RANO-BM) criteria was observed as presented at the ESMO Breast Cancer Congress in May 2022. Further data on quality of life followed at ESMO 2022 [20]. Here, maintained quality of life was observed during T‑DXd treatment as determined using the EORTC QLQ-C30 questionnaire. Moreover, specific subdomains such as emotional, social, cognitive and physical functioning remained stable over time, justifying further evaluation of antibody–drug conjugates for active brain metastases.