【佳学基因检测】基因检测指导髓母细胞瘤的代谢治疗
根据《个性化肿瘤治疗的方法与途径》，与表观遗传学、转录组学和基因组学相比，髓母细胞瘤代谢组学是一个仍处于起步阶段的领域。因此，针对髓母细胞瘤的代谢和非WNT/non-SHH 髓母细胞瘤的特异性治疗概念很少。一种方法是抑制IDO1。髓母细胞瘤的致病基因鉴定基因解码表明，IDO1是一种主要参与色氨酸分解代谢的酶，已被确定为创造免疫抑制微环境的关键角色[91]。一项小型研究报告了27个髓母细胞瘤样本队列中IDO1的表达，其中包括16名非WNT/非SHH患者[92]。去年完成了一项I期试验，该试验将IDO1抑制剂吲哚昔莫与替莫唑胺联合应用于患有复发性/难治性中枢神经系统恶性肿瘤的儿童群体，并纳入了81名患者。虽然该试验的结果尚未公布，但应用吲哚肟与放化疗联合应用的II期研究目前已开放供参与。除色氨酸外，多胺代谢已被确定为几种癌症的潜在脆弱性，包括SHH髓母细胞瘤，因为多胺参与许多细胞过程的调节。干扰多胺合成的一种可能性是使用二氟甲基鸟氨酸（DFMO/Eflornithine）抑制鸟氨酸脱羧酶（ODC），这是一种有效的ODC抑制剂，已被测试用于各种类型的癌症，包括神经母细胞瘤。DFMO已经被批准用于治疗昏睡病和多毛症，这使其成为一个有趣的旧药新用的创新药物候选者。然而，由于毒性问题和药代动力学方面有许多问题，对其使用仍存在争议。虽然缺乏临床前研究，但二氟甲基鸟氨酸目前正在扩大使用环境（NCT03581240）和探索其作为高危髓母细胞瘤样维持疗法的II期试验中，对非WNT/非SHH 髓母细胞瘤样患者进行试验。经过髓母细胞瘤基因解码，与其他针对非WNT/非SHH 髓母细胞瘤样代谢组学脆弱性的恶性肿瘤一样，这可能是增强治疗支柱的一种有价值的方法。迄今为止，缺乏临床前和临床证据不允许对这一分子亚群做出任何结论或展望。
2.8. Immunotherapy
2.8.1. Immune Checkpoint Inhibitors
Immune checkpoint inhibitors that interfere with the PD1/PDL1- and CTLA4-mediated immunosuppressive crosstalk between malignant and immune effector cells are celebrated as one of the most important developments in oncology during the last decade [114]. However, these strategies rely on the presence of tumor-infiltrating lymphocytes and especially the blockade of the PD1/PDL1-axis is at least partly dependent on the expression of PDL1 on the respective tumor cells. A series of studies used immunohistochemistry and bioinformatic deconvolution to assess immune infiltration and PDL1-expression in MB [115,116,117,118,119,120,121,122,123]. Although most studies only analyzed small case series or cohorts of MB patients, together they suggest that MB is an immunologically “cold” tumor with only sparse immune infiltration. Furthermore, apart from one notable exception [118], all studies showed negligible or no PDL1-expression at all, especially for non-WNT/non-SHH MB [115,116,118,119,121,122]. Additionally, studies investigating intratumoral heterogeneity using single cell RNA-sequencing reported only minor infiltration of tumor-specific lymphocytes and a diverse spectrum of myeloid cells and microglia, which potentially contribute to an immunosuppressive tumor microenvironment [33,124,125]. These findings indicate that it may be challenging to implement immune checkpoint inhibition as part of future treatment strategies for these patients, at least in form of a monotherapy.
However, limited preclinical findings still warrant further validation of immune checkpoint blockade as a therapeutic concept for non-WNT/non-SHH MB (Figure 2), e.g., as part of a combination therapy that induces a “hotter” tumor microenvironment such as the abovementioned INFORM2 NivEnt trial [74,117,122]. Currently, several clinical trials that offer immune checkpoint inhibitors as a monotherapy are also open to patients with recurrent or relapsed MB (NCT02359565: Pembrolizumab and NCT03173950: Nivolumab) (Table 1). However, for the latter trial only patients >18 years of age are eligible, and these patients most often do not harbor non-WNT/non-SHH MB. Another phase II trial that tested either Nivolumab as a monotherapy or in combination with Ipilimumab is in the stage of finalization (NCT03130959). Lastly, a currently running industrial trial tests the combination of Nivolumab with the immunostimulant Bempegaldesleukin, a recombinant form of IL-2, in children and young adults with treatment-resistant cancer (NCT04730349). It should be noted though that none of these trials is exclusively recruiting non-WNT-/non-SHH or even generally MB-patients, and it remains to be seen if enough patients will be enrolled to allow for subgroup-specific analysis.
2.8.2. Cellular Immunotherapy
The success of cellular immunotherapy as a new treatment for hematologic malignancies has sparked intensive research activities that aim at translating these novel therapeutic concepts into the clinic for CNS- and other solid tumors. Currently, two main concepts of cellular immunotherapy are tested in MB patients: chimeric antigen receptor (CAR) T-cells and natural killer (NK) cell therapy [126]. CAR T-cells are produced by isolating the patient’s own cytotoxic T-cells, which are then equipped with a CAR that can recognize any form of surface marker and subsequently activates the T-cell to mount an immunological attack against the target cell [127]. It is of crucial importance to choose target antigens that are highly expressed on cancer cells, but not or only negligibly on normal tissue to avoid severe on target/off tumor-toxicity. Currently, several promising targets are investigated for MB-directed CAR T-cell therapy: HER2/Neu, B7-H3 (also called CD276), EPHA2, IL-13Rα2, and PRAME [128,129,130,131,132,133]. HER2/Neu plays an important role as an oncogenic antigen in a number of solid tumors, including breast cancer and glioblastoma, and has also been identified as a possible target for MB-directed cellular therapy both in vitro and in vivo [129,130,134]. Currently, one phase I trial testing HER2/Neu-specific CAR T-cells is recruiting and open to MB patients (NCT03500991), and a recently published interim analysis of the first three enrolled patients (anaplastic astrocytoma, ependymoma) showed no dose-limiting toxicity and presented evidence of immune activation (Table 1) [135]. B7-H3/CD276 is a pancancer antigen that is strongly expressed by MB [128]. Similar to HER2/Neu, B7-H3-specific CAR T-cells have shown preclinical activity against non-WNT/non-SHH MB-models both in vitro and in vivo as well as in a number of other (pediatric) cancers, including atypical teratoid/rhabdoid tumor (AT/RT), another aggressive CNS-tumor of early childhood [128,136,137,138]. These preclinical findings provide a strong rationale to test B7-H3 targeting CAR T-cells in the clinic, which is currently undertaken in one phase I study that enrolls children with B7-H3 positive CNS-malignancies, including MB (NCT04185038). While HER2/Neu and B7-H3 targeting CARs have already been translated into the clinic, several other targets may be of interest based on preclinical data: one study showed strong expression of EPHA2 and IL-13α2 in human Gr.3 MB and subsequently tested both monovalent EPHA2- and trivalent EPHA2-/HER2/Neu-/IL-13α2-targeting CAR T-cells in a Gr. 3 MB-mouse model, with promising results [132]. An IL-13α2 directed CAR T-cell trial is recruiting adult patients with leptomeningeal metastases, including MB, and the results may be of interest to inform future trials in pediatric populations (NCT04661384). Furthermore, another study provided evidence in vitro that PRAME might represent a promising target for CAR T-cell therapy in MB [133]. Additionally, two more phase I CAR-T cell trials are currently open to pediatric patients with CNS-malignancies, including MB. However, the respective target antigens EGFR806 and GD2 have not been tested preclinically in MB patients (NCT03638167 and NCT04099797). Lastly, it should be noted that the delivery of CAR T-cells to the brain poses challenges due to the role of the blood–brain barrier, which could lower the effectiveness of intravenously applied cellular immunotherapies [129]. To date, the best application route for CAR T-cell in neurooncology has not yet been determined. However, the majority (4/5) of currently running CAR T-cell trials for which MB patients are eligible will use intraventricular/intracavital cell delivery, which circumvents the blood–brain barrier.
An exciting alternative to CAR T-cells is the use of NK cells that may offer certain advantages in comparison, such as lower side effects and increased resistance to immune evasion strategies of the tumor [139]. Several preclinical studies have shown that in principle, NK cells are able to recognize and eliminate MB cells; however, additional stimulation may be needed to arrive at clinically meaningful cytotoxic activity levels [140,141,142,143,144]. Interestingly, one study showed a higher and more consistent sensitivity to NK cells in vitro for non-WNT/non-SHH as compared to SHH MB cell lines [141]. In contrast, another study reported significantly higher expression levels of CD1d, an antigen recognized by NK cells, on SHH as compared to Gr. 4 MB [142]. Clearly, further studies are needed to arrive at a conclusive answer concerning which MB groups are the most promising target for NK cell therapy. The safety and feasibility of NK cell therapy for pediatric brain cancer has recently been shown by a phase I trial that also enrolled five MB patients, although without reporting molecular diagnoses (NCT02271711) [145].
2.8.3. Tumor Vaccinations
Cancer vaccine approaches harness the ability of off-the-shelf or patient-specific antigens to induce an antitumoral immune reaction [146]. Several different strategies have been developed, for instance using antigen pulsed dendritic cells, peptides, and nucleic acid vectors. Tumor vaccines have been studied for a long time; however, several early phase trials have recently shown great potential in adult glioblastoma patients [147,148,149]. One study tested dendritic cells that were pulsed with tumor lysate-derived antigens. As compared to AT/RT and high grade glioma, the five included MB patients showed only modest therapy response, albeit no molecular information is available [150]. The results of another phase I trial that used RNA-pulsed dendritic cells and also enrolled patients with recurrent MB and glioma are pending (NCT03615404). An additional trial applies another strategy that uses a peptide-based approach (NCT03299309). Lastly, MB patients are currently also eligible for a first-in-pediatrics trial that investigates a long peptide vaccine that targets the apoptosis inhibitor survivin (NCT04978727) (Table 1). Similar to most of the treatment strategies presented so far, none of these trials is restricted to non-WNT/non-SHH MB. Furthermore, so far, no definitive conclusions can be drawn from the reported data concerning differences in the efficacy of vaccination therapies for MB molecular subgroups.
2.8.4. Other Immunotherapeutic Approaches
Apart from checkpoint inhibition, increasingly more immunomodulatory approaches are entering the clinical stage. Two studies for a mixed pediatric population with CNS malignancies are currently testing immunostimulatory agents (Table 1): firstly, the antibody APX005M that targets CD40 (NCT03389802) and secondly the immune modulator WP1066, which inhibits the transcription factor STAT3 and therefore interferes with the JAK2/STAT3-pathway (NCT04334863). Additionally, oncolytic viruses, which have recently shown promising results for pediatric high grade glioma, have been proposed as a therapeutic option for MB [151,152]. Preclinical studies suggest that Gr. 3 MB might be a potential candidate for oncolytic virus therapy, and several different viral vectors have been tested throughout the last decade [153,154,155,156,157,158,159,160,161]. Oncolytic virus studies that are currently recruiting MB patients include the investigation of a modified measles vaccine (NCT02962167), Herpes Simplex Virus (NCT03911388), and one open phase IB trial that will assess possible agonistic effects between mesenchymal allogenic cells and an adenovirus-based virotherapy (NCT04758533) (Table 1). Two additional early phase studies that are open to MB patients are currently active, but not recruiting; these are testing reovirus in combination with Sagramostim and a modified poliovirus (NCT02444546, NCT03043391).
2.9. Other Molecular Therapeutic Approaches
In addition to the treatment approaches discussed thus far, several other concepts should be mentioned briefly (Table 1). Firstly, the inhibition of placental growth factor (PGF) has been proposed as a promising strategy across all MB subgroups and was tested in a phase I trial using the monoclonal antibody TB-403 (NCT02748135) [162]. The results have been presented at the AACR annual meeting 2021 and were encouraging both in terms of safety and efficacy according to the producing company; however, a peer-reviewed publication is not available to date [163]. Another interesting treatment approach represents the inhibition of DNA damage response pathways using poly (ADP-ribose) polymerase (PARP) inhibitors, a relatively new drug class that has received numerous approvals for breast and ovarian cancer patients with germline BRCA1/2 mutations in the last years [164]. This could be especially relevant to a subset Gr. 3/4 MB, since these are the molecular groups that are enriched for (germline) BRCA2- and PALB2-mutations [13,16]. Thus, it will be interesting to see whether non-WNT/non-SHH patients will be included in currently running trials using PARP inhibitors (NCT03233204 and NCT04236414). Lastly, several preclinical studies with different compounds indicate that the concept of sensitizing cancer cells to radiotherapy using small molecules could be potentially interesting for non-WNT/non-SHH MB [165,166,167]
(责任编辑：佳学基因)
