Autologous Tumor Lysate-pulsed Dendritic Cell Immunotherapy for Pediatric Patients with Newly Diagnosed or Recurrent High-grade Gliomas

Abstract

Immunotherapy has the potential to improve clinical outcomes with little toxicity for pediatric patients with brain tumors. We conducted a pilot feasibility study of tumor lysate-pulsed dendritic cell (DC) vaccination in pediatric patients (1 to 18 years old) with newly diagnosed or recurrent high-grade glioma (HGG). A total of nine DC vaccine doses, each containing 1×10⁶ cells per dose were administered to three out of the seven originally enrolled patients. Toxicities were limited to mild side-effects, except in one case of elevated alkaline phosphatase, which resolved without clinical consequences. Two patients with primary lesions amongst the three vaccinated were alive at the time of writing, both without evidence of disease. Pre- and post-vaccination tumor samples from a patient with an anaplastic oligoastrocytoma that recurred failed to demonstrate immune cell infiltration by immunohistochemistry. Peripheral cytokine levels were evaluated in one patient following DC vaccination and demonstrated some changes in relation to vaccination. DC vaccine is tolerable and feasible with some limitations for pediatric patients with HGG. Dendritic cell based immunotherapy may provide some clinical benefit in pediatric patients with glioma, especially for patients with minimal residual disease, but further investigation of this modality is required.

Brain tumors are the second most common malignancy in children and a leading cause of morbidity and mortality from childhood cancers (1). Great strides in the treatment of childhood brain tumors have been made over the last several decades due to more advanced neurosurgical techniques, which allow for more complete and safer resection, and due to advances in the delivery of chemotherapy and radiation therapy (2–4). However, little progress was made in the treatment of certain types of brain tumors, notably malignant gliomas, brainstem tumors and brain tumors that have recurred after standard therapy (5–7). In addition, although surgery, radiation therapy, and chemotherapy can result in long term survival for some pediatric patients with brain tumors, these children suffer from a myriad of late effects, including serious cognitive and endocrine dysfunction, renal and cardiac disease, as well as second malignancies (8–11). Thus, there is an urgent need for more targeted, less toxic, and more effective therapies for this population.

Immunotherapy has gained recent attention as a more targeted and less toxic approach to cancer treatment, including of brain tumors. Recent successes using immunotherapy against prostate cancer, melanoma, and renal cell carcinoma have prompted increased interest in also establishing these methods for brain tumors (12–14). A number of different immunotherapeutic methodologies have been attempted with regards to brain tumor immunotherapy including radioisotope-conjugated passive antibodies, peptide based vaccines, and DC-based vaccines (15). DCs are regarded as the most potent antigen-presenting cells in humans. They are able to handle a variety of antigenic mediums, including peptides, whole proteins, RNA and DNA, which they process and display as peptides on their cell surface in the context of major histocompatibility complex (MHC) class I or II molecules. Stimulation by whole proteins, or even whole tumor cell protein lysates, allows for cross-stimulation of cytotoxic T-cell responses, as well as T-helper (Th)1 and Th2 pathways (16, 17), which would theoretically be more effective than trying to stimulate immunity using MHC class I restricted peptides alone.

Several clinical trials have been conducted using DCbased immunotherapy for brain tumors but only two have involved pediatric patients (16). In the first series, Caruso and colleagues used tumor RNA-pulsed DCs as a vaccine for nine pediatric patients with a variety of brain tumor types and concluded that it was safe and feasible (18). Ardon and co-authors more recently used tumor cell lysate-pulsed DCs for 45 children with recurrent brain tumors (33 high-grade gliomas, HGG) and reported that 6 of the patients with HGG had survival lasting greater than 24 months. Five out of these six also received additional chemotherapy (19). Here, we report on a single-institution pilot trial using a tumor lysate-pulsed DC vaccine for pediatric patients with primary or recurrent HGG.

Patients and Methods

Eligibility criteria and study design

The trial design was a pilot/feasibility trial using tumor lysate-pulsed DCs as an intradermal vaccine for pediatric patients with malignant glioma. Eligibility criteria for inclusion included age 1–18 years, with pathologically confirmed WHO grade III or IV glioma. Tumors with mixed and oligodendroglial components were allowed. Patients with primary or recurrent tumors were eligible. Additional eligibility requirements included Karnofsky/Lansky score of >60, adequate bone marrow function (e.g., hemoglobin ≥8 gm/dl, absolute granulocyte count ≥1,500/µl and platelet count ≥100,000), adequate liver function [alanine aminotransferase (ALT), aspartate aminotransferase(AST), and alkaline phosphatase ≤2 times institutional normals and bilirubin ≤1.5 mg/dl], and adequate renal function (serum creatinine ≤1.5 times institutional normal levels) prior to starting therapy. Patients were required to have recovered from the toxic effects of prior radiation therapy and/or chemotherapy (at least two weeks since administration of chemotherapy or six weeks for nitroso-urea compounds; two weeks after radiation therapy; two weeks after corticosteroid therapy; and one week after other biologic therapy) prior to initiating treatment. A schematic of the study is presented in Figure 1. All patients consented to the study, using consent forms approved by the University of California Los Angeles (UCLA) Institutional Review Board (IRB), prior to or immediately following tumor resection. The trial was registered with the National Cancer Institute (NCI) as NCT00107185.

Figure 1.

Schematic for phase I clinical trial: Autologous tumor lysate-pulsed dendritic cell (DC) immunotherapy for pediatric malignant glioma.

Preparation of autologous tumor lysate

Tumor specimens were collected from brain tumor surgeries performed at UCLA Medical Center and transported in a sterile fashion to the UCLA–Jonsson Cancer Center GMP facility and used to generate autologous tumor lysate as previously described (20, 21). For the preparation of tumor lysate, the brain tumor tissue was minced with a sterile scalpel, rinsed with distilled phosphate buffer saline (dPBS), and incubated with collagenase (Advanced Biofactures, Lynbrook, NY, USA) for 8 to 12 hours at room temperature. To generate lysates, tumor cell suspensions were subjected to five freeze-thaw cycles, centrifuged at 800 × g for 10 minutes, and the lysate then removed from the tube. A portion of the lysate was monitored for tumor growth and microbiological studies. The total protein concentration of each tumor lysate were determined using a Bio-Rad DC protein assay (Bio-Rad Corp., Hercules, CA, USA), and lysate aliquots with 100 µg of measured total protein were used to pulse DCs for each injection.

Leukapheresis

Ten to 30 days prior to the first injection, eligible patients underwent leukapheresis at the UCLA Clinical Hemapheresis Unit to isolate peripheral blood mononuclear cells (PBMCs). Corticosteroid therapy was stopped seven days before leukapheresis and was not used during the course of the DC vaccinations. For patients who had received post-surgical chemotherapy and/or radiation, the leukapheresis was performed when the post nadir white blood count (WBC) was greater than 1.0×10³/µl. To prevent the development of hypocalcemia from the citrate used for leukapheresis, all patients were given intravenous calcium during the leukapheresis procedure. The leukapheresis material was transported under sterile conditions from the UCLA Hemapheresis Unit to the UCLA Jonsson Cancer Center cGMP suite and further processed. Blood (120– 180 ml) was also drawn at various time points up to 14 days prior to leukapheresis as a source of autologous serum for the cell cultures.

Preparation of autologous DCs

Mononuclear cells were isolated from cells obtained by leukapheresis with use of Ficoll-Hypaque centrifugation and used immediately to generate DCs. The cells were labeled and coded to ensure proper tracking of the specimens and patient confidentiality. Approximately 6×10⁹ PBMC were resuspended in a solution of RPMI-1640 (Gibco-BRL, Carlsbad, CA, USA) with 10% autologous serum. The cell suspension was added to 150 cm² tissue culture flasks at 15 ml/flask and 2×10⁸ cells/flask. The flasks were incubated at 37°C with 5% carbon dioxide (CO₂) for 2 hours to allow the cells to adhere to the flask. After the 2-hour incubation, the media in the flasks were removed. The cell layer was washed twice with 10 ml of PBS (Bio Whittaker, Walkersville, MD, USA) to remove non-adherent cells. Culture medium containing clinical-grade RPMI-1640 (Gibco-BRL), and 10% autologous serum, with 500 units/ml each of recombinant human interleukin-4 (IL-4, R&D Systems, Minneapolis, MN, USA) and recombinant human granulocyte macrophage-colony stimulating factor (GM-CSF, Leukine^®; Amgen Corporation, Thousand Oaks, CA, USA) was added to each flask after washing. The flasks were placed in the incubator for 7 days of culture. The morphology of the cells was observed under microscopy to ensure proper formation of DCs. The DCs were cryopreserved in RPMI-1640 with 20% autologous serum and 10% dimethyl sulfoxide (DMSO) until the vaccination time points. Aliquots were thawed at days −1, 13, and 27 for the first, second, and third immunizations, respectively. Thawed DC were washed twice in sterile saline.

Co-culture of DCs with tumor lysate

After thawing, a cell count was carried out, viability assessed, purity evaluated, and an aliquot was removed for sterility testing. The DCs were then centrifuged in preparation for loading with autologous tumor lysate. The DC pellet was resuspended in 1 ml of tumor lysate in RPMI medium (no serum) and incubated for 18 to 24 hours at 37°C with 5% CO₂. Following incubation, the cell suspension was washed three times to remove excess lysate and the DCs were resuspended in 1 ml of sterile dPBS for injection. Quality assurance for the tumor lysate-pulsed DCs was tested by culturing 50 µl of the final product for standard bacterial and fungal pathogens prior to injection into patients. An aliquot of the final pulsed product was also tested by flow cytometry using a FACScan machine by BD (Becton Dickinson, Franklin Lakes, NJ, USA) using antibodies to cluster of differentiation (CD)14 and human leukocyte antigen(HLA)-DR. Preparations containing >50% of large CD14-/HLA-DR+ DCs were considered as acceptable (Figure 2).

Figure 2.

Flow cytometry of autologous dendritic cell culture prior to tumor lysate pulse. We desired a ‘large cell’ population that was >50% human leukocyte antigen-DR positive and >50% cluster of differentiation 14 (CD14)-negative to meet lot release criteria for tumor lysate-pulsing and administration. FSC, Forward scatter; SSC, side scatter.

Method of DC injection

All patients received intradermal (i.d.) injections at biweekly intervals. Patients received their assigned dose of brain tumor lysate-pulsed DC i.d. in 1 ml sterile saline (using a 25-gauge needle) in alternating arms. Patients were monitored for two hours after immunization at the General Clinical Research Center (GCRC) at UCLA. Patient 4 was eligible to receive an additional vaccination, which was administered 13 months after the last of the original three vaccinations following a protocol modification allowing for booster vaccinations.

Clinical, imaging and laboratory evaluation

In addition to baseline, interval clinical evaluations were performed, which included complete medical histories and targeted physical examinations, Karnofsky or Lansky performance status, neurological examination. Central nervous system (CNS) imaging consisted of magnetic resonance imaging (MRI) of brain (and spine if that was a known site of disease), with and without contrast, and was carried out at screening, baseline (day -2 to -14), and at months 2, 4, 6, 8, 10 and 12 of study visits. Laboratory monitoring was performed with serum chemistries, including complete metabolic panel, liver function tests and lactate dehydrogenase (LDH) at all study visits. Hematological parameters were monitored with complete blood count (CBC), differential and platelets at all study visits, and coagulation tests at screening and assignment only. Urinalysis was performed initially, at day 28 and at months 2 and 12. Serum pregnancy test was performed for all females of child-bearing age. Blood testing was carried out to assess immune competence and follow-up monitoring; baseline tests for HIV, hepatitis B and C, and syphilis serology were conducted. Serum samples were collected prior to each vaccination and at each 2-month visit time-point to assess inflammatory cytokine production. We performed cytokine analysis for patient 4. This was performed using the BD Cytometric Bead Array (CBA) Human Th1/Th2 Cytokine Kit (Cat#551809, BD Biosciences) as per the manufacturer’s instructions. An in vitro control sample using DCs from a random HLA-A*0201⁺ donor to activate autologous T-cells after pulsing with the influenza M1 peptide was also analyzed (Figure 3). Specimens for three surgical time-points (biopsy, resection, and post-recurrence) were available for immunohistochemical staining for patient 5. Routine immunohistochemical staining was performed for CD3 (1:150, clone SP7; GenWay Biotech Inc, San Diego, CA, USA), CD20 (1:100, clone SP32; Santa Cruz biotechnologies, Santa Cruz, CA, USA), and CD68 (1:100, clone KP1; Santa Cruz biotechnologies). Staining was graded as follows: 0, no lymphocytes; 1, very low density, typically, rare blood vessel may have few (e.g. 2–8) perivascular lymphocytes, neuropil may show rare single lymphocytes; 2, low density, typically few scattered blood vessels with small perivascular cuffs of lymphocytes, scattered single lymphocytes are readily seen in neuropil; 3, moderate density, typically moderate numbers of blood vessels with small to prominent perivascular cuffs of lymphocytes, neuropil exhibits single and loose clusters of lymphocytes; 4, high density, typically many blood vessels with moderate to thick, dense perivascular lymphocytic cuffs, neuropil lymphoid infiltrates are prominent and dense, approaching a sheet-like pattern in areas.

Figure 3.

Control cytokine bead analysis (CBA) using random donor HLA-A*0201 positive peripheral blood mononuclear cells stimulated with the influenza M1 peptide showing robust stimulation of interferon-γ and tumor necrosis factor-alpha (TNF-α). IL: interleukin; IFN-γ: interferon-gamma.

Results

Patient characteristics and vaccine administration

A total of seven patients with HGG were initially enrolled between 2005 and 2008, prior to, or immediately following tumor resection. Table I shows a summary of these patients’ haracteristics at enrollment. Patients were aged between 1–18 years, five of them were females and two were males. Five of the patients had recurrent tumors and two had newly diagnosed lesions. Pathology was reviewed independently by two neuropathologists, with the majority of patients (n=5) having glioblastoma (GBM) and two anaplastic astrocytoma (AA). Out of seven patients enrolled, three eventually received vaccinations (patients 1, 4 and 5), three had rapid disease progression before the vaccine was available for use (patients 2, 3 and 7), and one patient’s (6) family elected not to continue with study therapy after initially consenting for study enrollment (Table I). Patients 1, 4 and 5 received a total of three, four, and two injections of 1×10⁶ DC/injection, respectively.

Table I.

Summary of enrolled patients.

Patient	Age (years)	Gender	Pathology for trial	Assigned DC Dose 106	Survival from surgery to obtain tissue for lysate to last contact	Current status
1	14	Male	Recurrent GBM	1	9 months	DOD
2	7	Female	Recurrent GBM	1	8 months	DOD
3	16	Female	Recurrent GBM	1	1 month	DOD
4	13	Female	GBM	1	51 months	Alive, NED
5	2	Female	AOA	1	40 months	Alive, NED (receiving adjuvant therapy for recurrence)
6	5	Female	Recurrent AA	n/a	n/a (lost to FU in 1 month)	Tissue collected, further therapy refused by parents
7	7	Male	Recurrent GBM	1	6 months	DOD

GBM, glioblastoma multiforme; AOA, anaplastic oligoastrocytoma; DOD, dead of disease; NED, no evidence of disease; FU, followup; n/a, not applicable.

Treatment prior to recurrence or concomitant with DC vaccine administration

In addition to surgery, all patients received chemotherapy and/or radiation prior to, during or after the vaccination period (Table II). Patient 1 was originally treated with surgery only in another country. He then experienced recurrence within three months and was transferred to UCLA for a second resection (sub-total), and subsequently received radiation to the tumor bed (50 Gy) and whole brain (45 Gy), along with concomitant adjuvant temozolomide. The patient then received three doses of DC vaccination and continued monthly temozolomide following vaccinations until about two months prior to his death (Figure 4A–D).

Table II.

Therapies and vaccine doses for patients who received vaccinations.

Patient	Pathology	Extent of resection	Radiation pre-vaccination	Chemotherapy pre-vaccination	Vaccinations received at assigned dose	Post-vaccination therapy
1	Recurrent GBM	STR	Tumor bed to 50 Gy and whole brain to 45 Gy	Adjuvant temozolomide Concomitant with RT	3	Temozolomide
4	GBM	GTR	Tumor bed to 50.4 Gy	Concurrent and adjuvant temozolomide	4	Temozolomide, cis-retinoic acid, bevacizumab, irinotecan
5	AOA	GTR	None	High-dose chemotherapy, AHSCR	2	None (until recurrence)

AHSCR, autologous hematopoietic stem cell rescue; STR, subtotal resection; GTR, gross total resection; RT, radiation therapy.

Figure 4.

Representative images of T1 post-gadolinium magnetic resonance (MR) images. A and B: Pre-DC vaccination images of recurrent GBM for patient 1. C and D: Post-DC vaccination images of recurrent GBM for patient 1. E and F: Images for patient 4 at initial scan (E) and 13 months later after surgery, radiotherapy, chemotherapy, and DC vaccine (F).

Patient 4, with newly diagnosed GBM received radiation to the tumor bed (59.4 Gy) along with concomitant temozolomide following gross total resection (GTR). She then received her initial three DC vaccinations. Following this, she was placed on alternating courses of irinotecan/bevacizumab and temozolomide/cis-retinoic acid for one year, starting 30 days after her last vaccination. She received one booster vaccination about 13 months following the initial vaccinations. The patient then continued on temozolomide monthly for another six months and is currently off therapy and being followed up with regular MRI scans (Figure 4E and F).

Patient 5 was originally diagnosed with a primary anaplastic oligoastrocytoma and, following GTR of her tumor, received induction chemotherapy followed by high dose chemotherapy and three tandem autologous stem cell rescues. Following recovery from this therapy, she received only two doses of DC vaccination. She was taken off the study due to a grade 4 elevation in serum alkaline phosphatase, which was not clinically significant. She was subsequently monitored with MRI scans and unfortunately developed a recurrence one year later. The patient has since received further chemotherapy and radiation therapy and is doing well with no evidence of disease.

Toxicity and feasibility

Administration of all nine doses of DC vaccine was tolerated reasonably well with relatively mild adverse events without significant clinical consequences (seven instances of headaches and mild injection site erythema). The only laboratory Grade four abnormality per National Cancer Institute (NCI) criteria was observed in a 2-year-old patient (#5) with newly diagnosed anaplastic oligoastrocytoma, who experienced significant elevation of alkaline phosphatase (5447 U/l, isoenzymes: intestine 4%, bone 68%, and liver 28%), two weeks after administration of the first dose of DC vaccine, which persisted after her second dose, at which point she was taken off the study. A work up for malignancy or other processes was completely negative.

With regards to feasibility, the major issue was the delay between surgical resection of the recurrent GBMs, tapering off perioperative dexamethasone, leukapheresis and production/administration of the final vaccine product. The majority of patients with recurrence had been extensively pre-treated with radiation and a variety of chemotherapeutics [including anti-vascular endothelial growth factor (VEGF) therapy] and their disease was progressing too quickly to allow adequate time for vaccine production and administration, suggesting that this method of immunotherapy may be best suited for newly diagnosed tumors.

Clinical outcome and quality of life

As mentioned above, of the three patients who received DC vaccinations, two are alive and well 51 and 40 months following their surgery to obtain tissue for lysate (study enrollment) (Table I). Patient 1 may have had a partial response (on MRI) after DC vaccination, although he was also receiving adjuvant temozolomide at the time (Figure 4). Patient 4 suffers from some residual mild hemiparesis from tumor surgery, but is otherwise attending school at an age-appropriate level. Patient 5 also suffers from mild hemiparesis, but is otherwise thriving and doing well cognitively.

Pathology before and after vaccination

Patient 5 with anaplastic oligoastrocytoma had an initial biopsy followed by a complete resection but then experienced disease recurrence following initial therapy and DC vaccination. Tumor samples from this patient obtained at initial diagnosis, definitive resection, and at recurrence were analyzed using immunohistochemical staining for macrophage, T-cell, and B-cell infiltration post-vaccination and demonstrated a slight increase in infiltrating leukocytes after the first surgery but prior to vaccination. At recurrence, there were scattered macrophage-type cells but no infiltrating lymphocytes (Table III).

Table III.

Immunohistochemistry results of two pre- and one post-DC vaccine tumor tissues for patient 5.

	Pre-DC vaccine		Post-DC vaccine
Marker	Sample 1 (at first diagnostic surgery)	Sample 2 (3 weeks after first surgery)	Sample 3 from surgery at relapse (23 months after initial diagnosis)
CD3 (T-cells)	Very rare cells (0)	1 in mass tumor, 2 focal edge, 3 in encephalomalacia	0 (Small fragmented tumor sample)
CD20 (B-cells)	0	0–1 patchy	0 (Small tumor sample)
CD68 (Macrophages)	1 scattered cells in tumor (often perivascular)	0–2 in tumor, 2–4 in tumor edge/encephalomalacia	1 (Small tumor sample)

0, no lymphocytes; 1, very low density immune cellular infiltrate; 2, low density immune cellular infiltrate; 3, moderate density immune cellular infiltrate; 4, high density immune cellular infiltrate.

Cytokine analysis

CBA analysis of Th1/Th2 cytokines [interferon-γ, tumor necrosis factor (TNF)-α, interleukin (IL)-2, IL-4, IL-6, and IL-10] was performed on serial samples from patient 4 as indicated in Figure 5. At baseline, the patient had high levels of TNFα and IL-6 which both declined after the first vaccination, but increased back to baseline levels after the second vaccination (day 28). However, they then dropped below detectable limits at subsequent time points. IL-10 approximately doubled after two vaccinations (day 28) and then seemed to return to baseline level. IL-2, IL-4, and interferon-γ were present at low levels during the vaccinations and then largely became undetectable.

Figure 5.

Cytokine bead array analysis of cytokine levels from patient 4. *Days of vaccination; #actual value was >5000 pg/ml (above upper limits of the standards for this cytokine). TNF-α=tumor necrosis factor-alpha; INF-γ=interferon-gamma; IL=interleukin.

Discussion

With recently approved immune-based therapies for prostate cancer, melanoma, and renal cell carcinoma among others, immunotherapy is gaining a reputation as a potentially effective cancer therapy. In spite of this increasing acceptance, true objective response rates remain low and the patients for whom immunotherapy offers the most benefit remains those with minimal residual disease. This fact rings true even in our small series, where disease in the one patient with sub-totally resected disease progressed, while those with undetectable disease by imaging remain alive. These results, combined with results from prior studies (18, 19), suggest that DC-based vaccination is feasible and may have a role in the multimodality treatment of childhood brain tumors. A form of ‘maintenance’ immune therapy has already been found effective for advanced neuroblastoma (22).

Our report is one of very few describing an active immunotherapeutic approach for children with brain tumors. The only open recruiting studies at the time of the writing of this article were a putative brain-tumor initiating cell line-pulsed DC vaccine study at the University of Minnesota and a combined DC vaccine plus adoptive T-cell approach at Duke (NCT01171469 and NCT01326104, www.clinicaltrials.gov). Caruso and colleagues reported in 2004 on nine patients with a variety of recurrent brain tumors. They used tumor-derived RNA to pulse autologous DCs which were then injected both intravenously (i.v.) and i.d. They were able to vaccinate seven out of the nine patients and demonstrated no toxicity (18). Ardon and colleagues more recently reported in 2010 on 45 children with recurrent HGG vaccinated with tumor lysate-pulsed DCs similar to our methodology. They also reported little toxicity and no decrease in quality of life measures associated with the vaccine treatment (19). The main limitation of our study is the small sample size, which limits our ability to draw any conclusions with regards to toxicity or efficacy. Our small series is novel in that two out of the three patients that received the vaccination were newly diagnosed at the time of trial enrollment, providing some support for the use of this methodology in newly diagnosed patients. In addition, we were able to look at a panel of Th1/Th2 cytokines using CBA technology on banked serum specimens, as well as test for the presence of immune cell infiltration from pre-and post-vaccination (at recurrence) tumor specimens, suggesting the feasibility of using these approaches in larger immunotherapy trials.

Patient 4 had apparent cytokine changes (temporally related to DC vaccination) for IL-6, TNFα, and IL-10. IL-6 is generally considered to be a pro-inflammatory cytokine, although it has been shown to play different roles depending on the immunological context. High serum levels of IL-6 have been shown to correlate negatively with survival for a number of malignancies, and in a p53-derived peptide vaccine study for breast cancer, a decrease in IL-6 levels post-vaccination predicted a more favorable outcome (23). However, another study has shown a relationship between IL-6 production and maturation of Th17 pro-inflammatory T-cells which seems to be beneficial for antitumor immunity (24). Interestingly, IL-6 is also a potent inducer of osteoclast activity and although levels were not measured for patient 5, increased IL-6 levels could have provided one explanation for elevated alkaline phosphatase activity in her case (25). TNFα is another pro-inflammatory cytokine that has been used as an immune stimulant in pre-clinical models to effectively eradicate tumors(26). Therefore, it would seem an increase in this cytokine would be advantageous for an antitumor immunological effect. IL-10 is an immune-inhibitory cytokine that takes part in the development of T-regulatory cells which are generally thought to work against antitumor immunotherapy. Induction of IL-10 responses may be preferentially stimulated by immature as opposed to mature DCs (27). This protocol called for the use of immature DCs as we believe during the pulsing of tumor lysate, sufficient maturation of the DCs does occur. However, we cannot rule out the possibility that further pre-vaccination differentiation of the DCs may be advantageous (such as by using lipopolysaccharide or CD40L). It is difficult to ascertain the meaning of these cytokine changes from a single patient. It may be that the patient’s initially elevated IL-6 levels were a baseline level of ‘ineffective’ inflammation consistent with the poor prognosis of other patients with cancer with elevated IL-6 levels as mentioned above(23). Following the first dose of vaccine, these levels declined, only to rebound prior to the third dose of vaccine. However, after all three doses, IL-6 remained at near undetectable levels, suggesting a potentially favorable antitumor immunological response. With regards to the levels of interferon-γ and TNF-α at baseline, the patient underwent a leukopheresis procedure performed only 10 days prior to these measurements, and it is possible these cytokines were stimulated as a result of recovery from that procedure. The low cytokine levels that seemed to persist following the vaccination may have been due to the adjuvant therapies she received (temozolomide, cis-retinoic acid, bevacizumab, and irinotecan) although these did not start until a month following her final vaccination. Finally, it may be true that circulating cytokines bear little relation to the intracranial immune response we were trying to elicit. Further studies looking at the relationship of peripheral cytokine levels to intracranial immune responses will be needed to examine this issue. The absence of T-cell infiltration seen in the tumor samples from patient 5 was discouraging but not unexpected given that the tumor had recurred. Whether this was due to intrinsic immune resistance provoked by the tumor(28), or simply that our methodology failed to elicit an effective intracranial immune response cannot be determined. The scattered macrophages seen may represent recently described myeloid suppressor cells (29). Of note, Prins and colleagues demonstrated T-cell infiltration following DC vaccination in adults with GBM at recurrence and that this seemed to correlate with the mesenchymal sub-type of GBM (21).

In conclusion, this report supports the tolerability and feasibility of tumor lysate-pulsed DC immunotherapy for pediatric glial tumors, at least for patients with limited prior therapies and minimal residual disease. Future investigations with larger patient numbers are warranted to investigate and further advance the use of this important modality.