Dendritic cell immunotherapy for glioblastoma

Abstract

Dendritic cell immunotherapy is emerging as a promising addition to the multimodal treatment of patients with glioblastoma multiform. Initial Phase I and II trials have demonstrated favorable outcomes with minimal toxicity. In this editorial, the current status and the future challenges of this therapy are discussed.

Keywords::

• dendritic cell
• glioblastoma multiforme
• immunotherapy
• overall survival
• side-effect

Despite the integration of optimized multimodality therapy, glioblastoma multiforme (GBM) remains a treatment challenge with a mean survival of 14.6 months with the current standard of care [1]. Although numerous novel therapies have been tried and tested, a major improvement in prognosis is yet to be achieved.

An emerging treatment modality that has gained increasing attention in the management of GBM is active dendritic cell (DC) immunotherapy. The concept underpinning the therapy is of a personalized, targeted, anticancer vaccine based on DCs, the antigen-presenting cells of the immune system, with the aim of stimulating the immune system against the tumor. This has already been tested in the treatment of other malignancies, such as melanoma and hematopoietic cancer [2], while since 2010, it has been formally approved for the management of metastatic prostate cancer [3].

Brain tumors represent an immunotherapeutic challenge, because of the brain’s unique and effective immune surveillance [4] and because of the immunosuppressive effect [5,6] that tumors generate in their environment. Since 2001, various research groups have attempted to address this challenge through Phase I/II clinical trials evaluating DC vaccination on patients with newly diagnosed or recurrent GBM or other high-grade gliomas.

While the immunotherapeutic protocols vary, the underlying approach to overcome the obstacles and immunologically access the tumors remains the same. The vaccines thus consist of two components: the patient’s DCs, deriving from peripheral blood monocytes obtained via leukapheresis [2,7], which are then pulsed with the patient’s tumor lysate or peptides (representing the tumor-associated antigens [TAA]) obtained and procured during surgical resection. This autologous DC–TAA fusion is subsequently introduced back to the patient, usually intradermally, since the skin is one of the common sites where DCs reside. DCs then display TAAs, interacting with activating CD4 and CD8 T cells, which then cross the blood–brain barrier to enter the brain parenchyma, target and kill the tumor cells.

Within the limitation of small numbers and potential biases, DC vaccine trials have revealed clear evidence that this customized treatment has a low toxicity while it can improve clinical outcome by provoking highly specific antitumor immune response, as reflected by infiltration of the tumor with activated T cells on histological examination.

Mean overall survival (OS) in trials recruiting patients with newly diagnosed GBM varied between 16.0 and 38.4 months, while for recurrent GBM, it ranged between 9.6 and 35.9 months [7]. Furthermore, this advantage in OS seems to be more pronounced when the recursive partitioning analysis classification of the European Organisation for Research and Treatment of Cancer is applied. Thus, at least a subgroup of GBM populations (recursive partitioning analysis III: age <50; KPS ≥90) appears particularly responsive to DC vaccines with survival reaching 39.7 months for newly diagnosed GBM [8]. In addition, advances in genetic profiling and molecular subclassification of the GBMs are likely to identify other subgroups of patients with predisposition to added benefit from DC vaccination [9,10]. Indeed, Prins et al. recently showed particularly better survival advantage in GBM patients with mesenchymal gene expression [11].

The majority of DC protocols showed a favorable safety profile, which compared well with the other modalities commonly used in these patients. Vaccination-related side effects were minor, mostly consisting of redness and itching at the site of injection [7]. Of equal significance, no cases of autoimmune encephalitis or exacerbation of a concomitant autoimmune disease elsewhere in the body were reported. Serious adverse events such as neurotoxicity due to increased peritumoral edema were very rare [12,13]. Mode of administration of the vaccine, namely intradermally or subcutaneously, makes the treatment delivery easy and minimally invasive.

Although the initial results appear promising, questions and challenges remain. The choice of best tissue to derive the TAA is a matter of debate. Arguably, the use of autologous whole tumor lysate is superior to the use of partial tumor lysate or preselected artificial peptides. Notionally, the former addresses the genetic heterogeneity of GBM [14,15] by exploiting the full array of glioma antigens to trigger the most multipotent immune response. It also spares the potential issues raised by the use of artificial peptides such as reduced specificity or immune-stimulant potency. This may explain the overall better OS seen in some studies using tumor lysate [8,11,16] compared with those using peptides [17–21]. Furthermore, the absence of any reports of autoimmune encephalitis reassures against the risk of immune damage to normal neural tissues with the use of whole tumor antigens in DC vaccines. Multipeptide vaccination or mRNA has also been used by some authors as TAA [21,22].

Enhancing immune responses to overcome tumor-induced immunosuppression, in an attempt to improve the outcome of DC vaccination, is a strategy used in a number of studies. Immunostimulants such as imiquimod [11] and polyinosinic–polycytidylic acid stabilized by lysine and carboxymethylcellulose [19] are two such examples. Further work is needed to clarify this area.

Another need, not addressed so far by clinical trials, is the identification of a physiological biomarker of the triggered immune response against the tumor. Surveillance of GBM DC immunotherapy via changes in the peripheral blood immune markers, such as CD3⁺, CD4⁺, CD8⁺, IFN-γ and TNF-α among others or delayed-type hypersensitivity skin test reactions, has not been strongly or uniformly correlated to vaccine efficacies and therefore currently lack prognostic value [7]. This could either indicate the need to identify more sensitive peripheral blood immune markers or more possibly be a consequence of the immunological ‘shield’ of the CNS, whereby the changes in the peripheral markers do not reflect the intensity of the intracranial immune response. Peripheral blood Treg [11] and neuroimaging techniques, such as [¹⁸F]-labeled 3′-fluoro-3′-deoxythymidine ([¹⁸F]FLT) PET [23], are currently two areas of active research to identify promising candidate biomarkers of response to treatment.

Another issue still requiring clarification is the optimal timing of DC vaccinations with respect to the other treatment modalities. In the current literature, the timing of commencement of DC vaccination is by no means uniform, ranging from immediately following the surgery to several weeks after completion of radiotherapy and chemotherapy. Since the wide adoption of the ‘Stupp protocol’ as standard of care, in most trials, DC vaccines have been used as an adjunct to the standard treatment. Thus, administration of the vaccines typically commenced 2–4 weeks after the completion of concomitant radiotherapy and chemotherapy, with a significant number of vaccinations coinciding with the adjuvant chemotherapy phase [7]. A number of authors [24–27] have suggested increased chemosensitivity of tumor when adjuvant temozolomide is combined with immunotherapy based on better clinical outcomes. The underlying mechanism remains unknown, though a strong link between the predominant T cell effectors and chemosensitivity has been proposed [25].

Interaction between DC vaccinations and radiotherapy remains controversial. Concern has been raised with radiotherapy before vaccination because of the risk of developing radiotherapy-induced mutant tumor cells, immunologically diverse than those obtained during surgery, which could then escape immune recognition, rendering vaccines less potent against tumors [28]. Others are, however, in favor of radiotherapy and concomitant DC vaccination, suggesting that antigen-presenting cells are attracted by the release of tumor antigens and signals caused by the irradiation-induced tumor cell necrosis and apoptosis [29]. Additionally, in a preclinical brain tumor model, it was found that radiation can provoke upregulation of MHC molecules in tumor cells, rendering them better immune targets and thus resulting in improved outcomes [30]. There is, therefore, a clear need for optimization of DC vaccine protocols so that synergy between the treatment modalities can be achieved.

In conclusion, though, the biggest challenge is the verification of the promising results of DC immunotherapy, suggested by the Phase I/II studies, in a large cohort of patients. The ongoing (recruiting initially in the USA and now in Europe) Phase III trial is a first attempt to meet this challenge [31] and is likely to be followed up by others to finally define the place of this therapy in the management of patients with GBM.

Financial & competing interests disclosure

S Polyzoidis is subinvestigator in the Phase III clinical trial evaluating DCVax^®-L, autologous dendritic cells pulsed with tumor lysate antigen for the treatment of glioblastoma multiforme (GBM). K Ashkan is chief investigator for Europe in the Phase III clinical trial evaluating DCVax^®-L, autologous dendritic cells pulsed with tumor lysate antigen for the treatment of glioblastoma multiforme (GBM). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.