Phase 2 trial of personal dendritic cell vaccines in newly diagnosed glioblastoma: 3-year follow-up and correlations with survival

Abstract

Survival of glioblastoma (GBM) patients is unsatisfactory. Adding patient-specific vaccines to standard therapy may improve outcomes. Autologous dendritic cells (DC) loaded with autologous tumor antigens (ATA) from short-term cell lines established from autologous tumor are a promising therapeutic vaccine strategy. DC-ATA vaccines were tested in a single-arm phase 2 trial in newly diagnosed GBM patients. Tumor tissue was collected during surgical resection. DC were differentiated from peripheral blood monocytes. Intent-to-treat enrollment took place before standard concurrent radiation therapy and temozolomide (RT/TMZ). DC-ATA was manufactured during RT/TMZ, then injected subcutaneously over 6 months concurrently with adjuvant TMZ. DC-ATA was reliably manufactured; treatment was well tolerated. Fifty-seven patients started vaccine therapy; 69% received all eight vaccine doses. The 10.7-month median progression-free survival (PFS) is 57% greater than the average of medians from standard treatment arms in six randomized trials that also enrolled patients before RT/TMZ. Disease control and overall survival (OS) dropped precipitously once DC-ATA was completed. Survival curves stratified by prognostic variables did not separate until after DC-ATA was discontinued. Multivariate analysis identified receipt of all eight planned vaccine injections, MGMT promoter methylation, concurrent dexamethasone dose ≤2 mg, and receipt of six or more TMZ cycles as independent variables. Common features among the seven patients who were progression-free after 3 years were eight vaccine injections, ≤2 mg dexamethasone, age <60 years, and adjuvant TMZ given without concurrent bevacizumab. PFS was encouraging, and the data suggest that OS may be increased by extending vaccine treatment.

Keywords: Glioblastoma; autologous tumor antigens; dendritic cell vaccine.

Figure 1.

Schema for treatment and follow-up. After recovery from RT/TMZ, patients received three weekly AV-GBM-1 injections, then one week later, the first cycle of TMZ -based therapy. Thereafter, five more doses/cycles of each were given at 4-week intervals. If TMZ had never been started or was discontinued, AV-GBM-1 injections could continue. When AV-GBM-1 dosing was completed, patients could continue TMZ-based therapy per their managing physician. AV-GBM-1, autologous dendritic cells pulsed with lysate from irradiated autologous tumor cells that were proliferating in cell culture; PBMC, peripheral blood mononuclear cells; GBM, glioblastoma; RT, radiation therapy; TMZ, temozolomide; RX, treatment.

Figure 2.

Survival from intent-to-treat enrollment. (A) Progression-free survival Kaplan-meier curve with 95% confidence limits and number of patients at risk at each interval. (B) Overall survival Kaplan-meier curve with 95% confidence limits and number of patients at risk at each interval. PFS, progression-free survival; OS, overall survival.

Figure 3.

Median months of progression-free survival (PFS) and 95% confidence intervals for newly diagnosed glioblastoma patients from an enrollment date before starting concurrent radiation therapy and temozolomide chemotherapy RT/TMZ). The numbers of patients in each trial are shown in parentheses.

Figure 4.

Overall survival Kaplan-Meier curves stratified by maintenance temozolomide regimens. Three patients treated with both bevacizumab (Bev) and TTF are included in both the TMZ+Bev and TMZ+TTF curves in (A) and (B), but are shown separately in (C). Median OS in months: TMZ 20.9, TMZ+Bev 16.5, TMZ+TTF 15.0, TMZ+Bev+TTF 13.0, no TMZ 5.1. OS was superior for TMZ alone compared to no TMZ (p = .0003), TMZ+Bev alone (p = .056), TMZ+Bev+TTF (p = .055), and the other three TMZ-containing regimens combined (p = .069), but not TMZ+TTF alone (p = .744). Compared to no TMZ, OS was superior for TMZ alone (p = .0003), TMZ+Bev (p = .045) and TMZ+TTF (p = .050), but not TMZ+Bev+TTF (p = .316). OS was inferior (p = .022) for the two bevacizumab regimens combined (n = 14: TMZ+Bev n = 11 and TMZ+Bev+TTF n = 3) compared to the other two TMZ-containing regimens combined (n = 35: TMZ n = 28 and TMZ+TTF n = 7). TMZ+TTF alone (n = 7) was not better than TMZ alone (n = 28) (p = .744) nor all other TMZ-containing regimens (n = 42) (p = .850). PFS, progression-free survival; OS, overall survival; TMZ, temozolomide; Bev, bevacizumab; TTF, tumor-treating fields.

Figure 5.

Overall survival Kaplan-Meier curves from intent-to-treat enrollment stratified by baseline prognostic variables. (A) Age less than 60 years vs age 60 years or greater. (B) Gender, male vs female. (C) Race/ethnicity, White vs non-White (D) Karnofsky performance status of 90 or 100 vs 70 or 80. (E) Methylation status of the MGMT promoter, methylated vs unmethylated. (F) IDH gene mutation, mutated vs wild-type (non-mutated). OS, overall survival; KPS, Karnofsky performance status; MGMT, O⁶-Methylguanine-DNA methyltransferase; IDH, isocitrate dehydrogenase.

Figure 6.

Overall survival Kaplan-Meier curves stratified by clinical treatment variables potentially predictive for survival. (A) Dexamethasone dose when vaccine was initiated, 2 mg or less versus greater than 2 mg. (B) Occurrence or not of a treatment-emergent central nervous system serious adverse event or seizures. (C) Treatment clinical trial site geographic located in Southern California versus elsewhere. (D) Number of temozolomide cycles, 6 or more versus less than 6. (E) Number of AV-GBM-1 doses injected, 8 versus 2 to 7. OS, overall survival; mos, months; Dex, dexamethasone TEAE, treatment-emergent adverse event; CNS, central nervous system; SAE, serious adverse event; so Cal, Southern California; TMZ, temozolomide; AV-GBM-1, autologous dendritic cells pulsed with lysate from irradiated autologous tumor cells that were proliferating in cell culture.

Figure 7.

Elispot numbers and association with survival. (A) Interferon-gamma Elispot numbers at baseline and after two vaccinations. At baseline (day 0), right before first injection, mean number of Elispots was 2.1 (median 1.9, range 0–34). Highest number after one or two injections (day 7 & 14) was 5.5 (median 2.3, range 0.33–37.8). (B) Changes in Elispot numbers: highest mean was 11.6 (median 6.0, range 1.7–37.8), middle mean was 0.9 (median 1.0, range 0.3–1.3), lowest mean was − 1.1 (median −1.0, range–2.8 to +0.1). Median OS were 21.9, 16.0 and 13.3 months. The trend in difference among all three curves was not significant (p = .259), nor was the difference between the tertiles with the highest and lowest Elispot numbers (p = .196). OS = overall survival.