RESULTS OF CLINICAL TRIALS DEVOTED TO DENDRITIC CELL VACCINES FOR NEURO-ONCOLOGICAL DISEASES

Dendritic cell therapy of gliomas in adults

Currently, the vast majority of clinical trials of therapeutic efficacy of DC as an adjuvant therapy are devoted to the treatment of malignant gliomas. Various antigens may be used for DC-based vaccines (antigens of tumor cell lysates and extracts, individual proteins and peptides, nucleic acids). However, the feasibility of using certain antigens is largely determined by their specific releasing by tumor tissue. Most often, intraoperative tumor tissue specimens are used as antigens for DC sensitization. It is possible to use several antigens from the “collection” of these tumors. Sensitization process itself usually involves an addition of antigens to DC culture and subsequent joint incubation. Other techniques for DC sensitizing are discussed below on specific examples.

The advantages of dendritic cell vaccines sensitized with tumor antigens are no need to identify the antigens and attractiveness of this method from the standpoint of personalized medicine. Scheffer et al. reported higher immunogenicity of tumor lysates inactivated by irradiation followed by apoptosis compared with tumor lysates inactivated by freezing and thawing followed by necrosis [7]. At the same time, there is an evidence of immunosuppressive effect of apoptotic cell antigens [8]. DCs are administered subcutaneously as a rule.

To date, some trials of phase I and II have been carried out. There are preliminary data of phase III studies. Some trials of phase I included control group and survival rates were evaluated. The vaccines based on DC sensitized by tumor lysate antigens were predominantly studied for recurrent glioblastoma in early clinical trials of phase I and II. Small sample size should be considered despite the tendency to better survival in the treatment group. Yu et al. (trial phase I) reported greater median survival after redo surgery for the first recurrence of glioblastoma (n=8, 33.25 months) compared with the control group (n=26, 7.5 months). There were 3 vaccine injections with an interval of 2 weeks in the study group. Immune response against some tumor-associated antigens has been demonstrated in some patients. Patients with anaplastic astrocytoma and tumors de novo were excluded from survival analysis. There were no data on other adjuvant therapies after surgery for recurrent tumor in the study group while radiotherapy was performed in the control group [9].

De Vleeschouwer et al. similarly analyzed larger sample of patients (n=56). Adults and children were enrolled. In contrast to previous trials, dendritic cells were additionally exposed to IL-1β, tumor necrosis factor-α (TNF-α) and prostaglandin E2 in addition to standard IL-4 and GM-CSF. These cytokines increase yield, maturation, motility and immunostimulating properties of dendritic cells [10]. According to this report, median recurrence-free survival after the second surgery was 3 months, median overall survival — 9.6 months. It is important to note that reduced interval between DC injections from 2 to 1 week within one month with subsequent monthly injections of tumor lysate and total resection of the neoplasm were associated with improved recurrence-free survival [11].

New data on safety and efficacy of DC-based vaccines in the treatment of recurrent glioma ensured analysis of this method in patients with high-grade gliomas de novo. Chang et al. (trial phase I/II) used dendritic cells sensitized by autologous tumor antigens after exposure of tumor cells with interferon-γ (IFN-γ), heat shock and radiation. The authors reported median survival of patients with newly diagnosed glioblastoma (n=8) about 12,7 months, in patients with recurrent glioblastoma (n=8) — 32.2 months (all patients with glioblastoma — near 17.3 months). In authors’ opinion, better survival in the second group was associated with earlier vaccination. Indeed, radiotherapy was immediately followed by vaccine administration in patients with newly diagnosed glioblastomas while radiotherapy was not performed in patients with recurrent tumors. Moreover, early recurrences after immunotherapy in patients with newly diagnosed glioblastoma could be associated with appearance of cellular clones with acquired mutations after radiotherapy. Therefore, these cells could have another antigenic pattern compared with cells used to sensitize dendritic cells [12].

One of the options for sensitizing DCs with a complete set of tumor antigens is creation of chimeras of tumor and dendritic cells (derivatives after fusion of both types of cells). Kikuchi et al. used polyethylene glycol to obtain chimeras of irradiated glioma cells and human DC [13]. The first administration of DC was carried out in 2 weeks after temozolomide chemotherapy onset in each of the study groups (phase I/II, glioblastoma de novo or recurrent tumor). Injections were repeated at least three times in each of the 28-day chemotherapy courses. Median of recurrence-free period was 10.3 and 18.3 months in patients with recurrent glioblastoma or glioblastoma de novo, respectively. Median of overall survival was 18 and 30.5 months, respectively. Immune response against WT-1 (Wilms tumor protein 1), gp100 and melanoma-associated MAGE-A3 antigens (chemoresistance antigens) was recorded in response to vaccination with chimeric cells [14].

In our country, research of immunotherapy with dendritic cell vaccines is presented by reports performed at the Polenov Russian Research Institute of Neurosurgery together with the Konstantinov St. Petersburg Institute of Nuclear Physics, Research Institute of Fundamental and Clinical Immunology of Siberian Branch of the RAS together with the Institute of Cytology and Genetics and the Research Institute of Traumatology and Orthopedics (Novosibirsk).

The authors from St. Petersburg used destruction of irradiated fragments of gliomas with subsequent pH-dependent cellular lysis to prepare sensitizing DC mixture. A total of 141 patients (2000—2009) were included in the study of dendritic cell vaccine and activated lymphocytes for the management of high-grade gliomas. Median overall survival of patients undergoing immunotherapy with standard treatment was 25 months, in the control group — 15 months. Median recurrence-free period was 12 and 8 months, respectively [15]. According to the later report, DC administration was followed by median survival of 32 (n=18) and 24 (n=11) months in patients with anaplastic astrocytoma and glioblastoma, respectively [16].

The researchers from Novosibirsk used GM-CSF and interferon (IFN)-α for DC maturation. Three was a slightly lower yield of maturated DCs in patients with high-grade gliomas. However, DCs sensitized with antigens of tumor tissue lysates and cultured in the presence of IFN-α and GM-CSF had similar functional activity compared with cells of healthy people. These DCs more significantly activated T-helpers [17, 18]. Administration of these cells was followed by higher survival of patients with anaplastic astrocytoma and glioblastoma [19].

Analysis of clinical and immunological correlations is a separate question in the study of dendritic cell therapy. Wheeler et al. analyzed immune response of patients considering interferon-γ mRNA (IFN-γ) release by mononuclear cells (study phase II). Overall survival of patients with glioblastoma significantly correlated with clear immune response. Moreover, median and 2-year survival were higher in the group of significant response (more than 1.5-fold augmentation of IFN-γ release) compared with patients with insignificant immune response [20]. Pellegatta et al. reported increased level of blood natural killers after dendritic cell therapy that was accompanied by improved median recurrence-free and overall survival [21].

Importantly, the peptides non-specifically isolated by acid elution were used for DC sensitization besides glioma lysates in early clinical trials devoted to DC therapy of glioblastoma. Yu et al. [22] and Liau et al. [23] reported significant increase of median survival in patients with glioblastoma compared with the control groups (15.2 and 8.5 months; 23.4 and 18.3 months, respectively). DC administration caused systemic (immune cells) and intracranial (tumor infiltration by lymphocytes) cytotoxic response.

Glioma is characterized by increased release of certain antigenic peptides which are absent in normal tissues. Careful selection of these peptides contributes to creation of dendritic cell vaccines with specified antitumor properties. An important step in the development of dendritic cell vaccines sensitized with peptides is isolation of the antigens. For example, some cytomegalovirus antigens are released by more than 90% of glioblastomas (up to 90% of glioblastomas are infected with this virus). However, release of these antigens is absent in normal brain tissue [24, 25]. The role of cytomegalovirus in development of glioblastoma is insufficiently confirmed [26, 27]. Nevertheless, potential role of this virus as a target for immunotherapy looks very attractive [28]. Nair et al. analyzed immunogenicity of cytomegalovirus antigens. The authors sensitized human DCs with mRNA of pp65 phosphoprotein and then studied their ability to activate T-lymphocytes. It was shown that DC-stimulated T-lymphocytes were able to lyse autologous cells of glioblastoma [29]. Subsequently, the same authors reported high survival in patients with glioblastoma de novo undergoing dendritic cell and temozolomide therapy after tumor resection and initial course of chemoradiotherapy. Median recurrence-free period was 25.3 months, median overall survival — 41.1 months. These values were higher compared with the control group despite the expansion of T-regulatory (“suppressor”) cells after therapy with temozolomide [25].

Phuphanich et al. used a mixture of antigen epitopes associated with gliomas and overexposed on the surface of tumor stem cells for DC sensitization (epidermal growth factor receptor HER2, glycoprotein gp100, MAGE-1, tyrosinase-related protein (TRP-2), subunit α2 of IL-13 receptor ( IL13Rα2), interferon-inducible protein — a component of inflammosomes (AIM-2)). Injection of this vaccine (ICT-107) reduced expression of glioma stem cell marker genes in tumor samples [30]. Application of sensitized DCs was accompanied by a tendency to increase of overall survival and significant prolongation of recurrence-free period [31]. In 2016, Phuphanich et al. reported 6 patients with more than 6-year survival and 3 patients with more than 8-year survival in the treatment group [32].

Unification of vaccination conditions and greater sample size in clinical trial phase III were necessary to determine the effect of vaccines on survival rates. Currently, these tests are conducted regarding the technology of manufacturing an autologous dendritic cell vaccine DCVax. Antigens of autologous glial tumor lysates are used to sensitize DCs in this vaccine [33]. Liau et al. reported intermediate results of DCVax vaccine administration. Overall group enrolled 331 patients. Side effects grade 3—4 (National Cancer Institute Common Toxicity Criteria) at least presumably associated with therapy were observed in 7 patients (cerebral edema, seizures, severe nausea, lymphadenitis). The authors reported satisfactory tolerability of the vaccine and increased survival compared to standard therapeutic approaches for glioblastoma de novo. Median overall survival was 23.1 months, in patients with methylated MGMT gene promoter (O6-methylguanine DNA methyltransferase gene) — 34.7 months after surgery, 3-year survival — 46.4%. However, there were small subgroups of patients with median survival of 46.5 and 88.2 months, respectively [34]. Moreover, ICT-107 vaccine (NCT02546102) is currently under investigation (study phase III).

Dendritic cell therapy of intracerebral tumors in children

Clinical trials of immunotherapy in children are less developed despite successful management of brain tumors in adults. In part, this may be due to sampling of insufficient volume of tumor tissue, for example, in patients with pontine gliomas. However, Boczkowski et al. reported that about 2000 tumor cells are sufficient for amplification of the required amount of RNA in order to create a potentially immunogenic vaccine [35]. Another common problem is impaired function of mononuclear cells after cytotoxic therapy (radio- or chemotherapy) and actually on the background of tumor process that reduces yield of dendritic cells [36]. There is a technique of administering granulocyte growth factor (G-CSF) prior to obtaining mononuclear fraction in order to increase yield of dendritic cells [37].

There is evidence of obtaining blood DCs from the patients with medulloblastoma accounting up to 20% of brain tumors in children [38].

Up to 1/3 of tumors are characterized by recurrent course despite various modern methods of therapy [39]. Moreover, 3-year survival of patients with recurrent medulloblastoma is less than 20% [40]. Nair et al. reported fundamental possibility to increase yield and quality of mononuclear precursors of dendritic cells in children after chemotherapy and subsequent induction with G-CSF. Monocytic CD14+ fraction was used for differentiation of dendritic cells. DC electroporation was performed using plasmids with RNA of some proteins (matrix protein of Flu M1 virus, cytomegalovirus pp65 protein or survivin protein). Plasmids were injected through the pores in the cellular membrane created under electric field. DC maturation was achieved in the presence of TNF-α, IL-1β, IL-6, prostaglandin E2. Adequate mobilization of monocytes for differentiation into DK was observed in 4 out of 5 patients while proper amount of mature phenotype of dendritic cells was obtained in 3 out of these 4 patients. However, functional activity of DC was absent in one patient due to insufficient mobilization of T-lymphocytes. Thus, the authors reported the possibility to obtain functionally active DCs from patients with refractory medulloblastoma after chemotherapy [41].

There are attempts to use dendritic cell vaccines in patients with diffuse pontine glioma. This malignancy (grade IV) is referred to diffuse midline gliomas in accordance with 2016 WHO classification [42]. Immunotherapy of this neoplasm is justified by specific mutation of H3 histone associated with replacement of lysine with methionine at the position 27 (K27M). This mutation determines potential immunogenicity of H3 histone [43]. A peptide vaccine triggering immune response against above-mentioned epitope (NCT02960230) is currently being investigated. However, there are no data on the use of this antigen for DC sensitization.

Benitez-Ribas et al. obtained autologous DCs from mononuclear cells of children with diffuse stem glioma. Dendritic cells were sensitized with antigens of cellular culture of this glioma. Specific immunogenicity and safety of dendritic cell vaccine were reported in 8 out of 9 patients (with different H3 K27M status) [44].

Research of the effectiveness of DC-based vaccines sensitized with tumor RNA and adaptive lymphocyte transfer for recurrent medulloblastoma and primitive neuroectodermal tumor is currently ongoing (NCT01326104). Ardon et al. analyzed children with high-grade glioma, medulloblastoma, primitive neuroectodermal tumor, ependymoma and atypical teratoid-rhabdoid tumor. The authors found more significant immune response in patients with glioma and atypical teratoid-rhabdoid tumor compared with medulloblastoma, primitive neuroepithelial tumor and ependymoma [45].

It should be noted that gliomas in children differ by another pattern of tumor antigen expression compared with adults. At the same time, cytomegalovirus gene expression is also increased in childhood gliomas that makes these antigens a perspective target for therapy in pediatric neuro-oncology [46]. Nevertheless, the question active immunotherapy for gliomas in children remains unclear [47].

Dendritic cell therapy of intracerebral metastases

Effective treatment of metastatic tumors is a very difficult task. Dendritic cell therapy may be applied in these patients. Brain metastases was an exclusion criterion in a large number of trials devoted to immunotherapy [48].

Clinical data on the efficacy of dendritic cell vaccines in the treatment of intracerebral metastases are presented by small studies and case reports. There is evidence of a positive interaction between radiotherapy and immunotherapy in patients with brain metastases of melanoma. Karbach et al. reported a patient with similar metastases and 10-year remission after radiotherapy and DC vaccination. Autologous dendritic cells were sensitized with antigens of metastatic tumor tissue lysate obtained from lymph nodes. There was also immune response in the form of expansion of CD8+ T cells specific for MAGE-A1 antigen of melanoma [49].

Dillman et al. demonstrated safety of autologous combination of GM-CSF and dendritic cell vaccine. Dendritic cells were sensitized with antigens of previously cultured in the presence of IFN-γ, irradiated and frozen cells of melanoma. Brain metastases were diagnosed in 3 patients of the experimental group. Individual calculations of survival were not performed in this group. Median overall survival was 13.8 months [50].

Mayordomo et al. reported clinical improvement for 8 months in a patient with skin, liver, lung and bone metastases of melanoma after administration of dendritic cells. However, brain metastasis was detected later [51]. Laurell et al. injected allogeneic DC in the primary tumor and reported the effect of dendritic cell therapy on regression of intracerebral metastases. Patients with metastatic renal cell carcinoma were predominantly analyzed. Allogeneic DCs were used due to possible development of bystander effect. It is enhanced immune response due to functioning of injected DCs and immune induction associated with DC non-sensitized with certain antigen. Patients with intracerebral metastases were excluded. However, there was a patient with intracerebral metastases occurred after therapy onset. Their regression was further noted. It should be additionally noted that this patient received tyrosine kinase inhibitor (sunitinib) [52].

Okwan-Duodu et al. reported high efficacy of radiotherapy combined with IL-2 immunotherapy for intracerebral metastases of melanoma. The possible mechanism of this effect is explained by potentiation of dendritic-cell unit after post-radiation release of melanoma-associated antigens.

Thus, combination of the above-mentioned treatment with additional administration of dendritic cell vaccines seems to be very interesting and requiring further analysis [53].

Actual issues of dendritic cell therapy of gliomas

Currently, there are several unresolved issues regarding the future of dendritic cell vaccines in clinical practice. Other methods of active immunotherapy are being studied in parallel with creation of these vaccines (for example, vaccination of patients with immunogenic peptides without DCs).

In 2017, Weller et al. published a review devoted to fundamental and applied aspects of immunotherapy for glioblastoma [54]. The authors supposed that successful antitumor vaccination may be achieved by combinations of these strategies with targeted therapy, for example, monoclonal antibodies (modulators of immune response control points) and immunosuppressive metabolic pathway inhibitors. Weller et al. emphasize the concept of relationship of response to therapy with genotype. Indeed, this approach reflects personification as a trend in translational medicine. For example, potential association of effective therapy with HLA-A2 allele was shown in the above-mentioned study of ICT-107 vaccine [30]. Only HLA-A2-positive patients were enrolled in the NCT02546102 study. Some trials of dendritic cell vaccines and MGMT status found higher survival in the group with methylated MGMT promoter [34, 55]. In our opinion, correlation of MGMT promoter methylation with successful dendritic cell therapy requires confirmation, since this prognostically favorable factor and immunotherapy can act independently of each other. However, Liau et al. supposed correlation between the effect of dendritic cell vaccine and temozolomide therapy [34].

A very important aspect of clinical research is integration of dendritic cell therapy into the overall treatment plan. Features of interaction of dendritic cell vaccines with systemic administration of corticosteroids are unclear while corticosteroids are one of the main components of symptomatic therapy for intracerebral tumor. As a rule, high-dose corticosteroid therapy was an exclusion criterion considering significant immunosuppressive effect. Any prescription of corticosteroids was an exclusion criterion in some studies [11, 56], while limitation of doses was possible in other ones [12]. Some earlier researches in vitro did not show a significant effect of dexamethasone on DC maturation per se and their phagocytic activity [57]. Further studies demonstrated immunosuppressive effect of this drug due to a special pool of monocytes [58].

Adaptation of therapeutic response criteria in glioma patients with particular emphasis on immunotherapy is also significant. New RANO criteria were published in 2015 (Response Assessment in Neuro-Oncology, Okada et al.) which were called iRANO (immunotherapy RANO) [59]. Some features of tumor response against this type of therapy were taken into account during development of new criteria. For example, pseudoprogression following immunotherapy can develop later compared with standard therapy (6 vs. 3 months). Moreover, immunotherapy may be followed by response after appearance of visual signs of tumor progression. Thus, it is not established whether continuation of immunotherapy in patients with confirmed tumor progression is possible. In some cases, immunotherapeutic agents can initiate inflammatory response. These inflammatory infiltrates can mimic visual signs of tumor progression including typical contrast enhancement. Obviously, various immunotherapeutic methods probably imply different response patterns that can require additional refinement of the criteria.

An important current problem is organization of clinical trials of dendritic cell vaccines in Russia. The authors of the aforementioned national trials talk not so much a clinical study but intermediate nature of the work between research and application of dendritic cell vaccines. For example, it is unclear what patients were analyzed in these reports (tumor de novo or recurrence). There is also no clear analysis of toxicity and safety of the drugs (except for general formulations about the absence of “serious complications”). Federal Law No. 180-FZ “On Biomedical Cellular Products” dated on June 23, 2016 has entered into force since January 1, 2017. Chapter 5 of the Article 35 of this law (entered into force on January 1, 2018) regulates manufacturing of biomedical cellular preparations exclusively in accordance with Good Clinical Practice guidelines (approved by the order of the Ministry of Health of the Russian Federation No. 669n on September 22, 2017). In our opinion, correspondence of dendritic cell vaccines to the concept of “biomedical cellular preparations” is relevant issue. According to definition, this product should be a “standardized population of cells with reproducible cellular composition” obtained by cultivation. However, maturation of dendritic cells includes their differentiation from the precursors. Therefore, lost ability of cellular reproduction is implied. At the same time, modern legal practice considers dendritic-cell vaccines exactly as biomedical cellular preparations that requires additional clarifications in order to avoid legal conflicts.