Neoadjuvant anti-PD1 immunotherapy for surgically accessible recurrent glioblastoma: clinical and molecular outcomes of a stage 2 single-arm expansion cohort

Abstract

Glioblastoma is immunologically "cold" and resistant to single-agent immune-checkpoint inhibitors (ICI). Our previous study of neoadjuvant pembrolizumab in surgically-accessible recurrent glioblastoma identified a molecular signature of response to ICI and suggested that neoadjuvant pembrolizumab may improve survival. To increase the power of this observation, we enrolled an additional 25 patients with a primary endpoint of evaluating the cell cycle gene signature associated with neoadjuvant pembrolizumab and performed bulk-RNA seq on resected tumor tissue (NCT02852655). Neoadjuvant pembrolizumab was associated with suppression of cell cycle/cancer proliferation genes and upregulation of T-cell/interferon-related gene expression. This signature was unique to patients treated with neoadjuvant pembrolizumab and was an independent positive risk factor for survival. Our results demonstrate a clear pharmacodynamic effect of anti-PD1 therapy in glioblastoma and identify pathways that may mediate resistance. However, we did not confirm a survival benefit to neoadjuvant pembrolizumab in recurrent glioblastoma and our secondary endpoint of PFS-6 was 19.5% (95% CI: 9.29-41.2%) for the pooled neoadjuvant cohorts. Our new data suggests some patients may exhibit innate resistance to pre-surgical ICI and require other concomitant therapies to sensitize effectively.

Fig. 1. Pembrolizumab treatment is associated with increased T cell/interferon-γ-gene-related signatures and decreased cell cycle/cancer proliferation gene signatures.

a Heat map of bulk tumor mRNA gene expression data showing gene signature analysis enrichment (GSAE) scores of gene sets with interquantile range ≥ 0.7. b Gene signature scores for Farmer_Breast_Cancer_Cluster_2 comparing pooled adjuvant and neoadjuvant cohorts. Each dot represents a patient, the lower and upper box bounds indicate the 25th and 75th percentile and the middle box line indicates the median value (adjuvant n = 15; neoadjuvant n = 34). P values were calculated using a two-sided Wilcoxon rank-sum test. c Gene signature scores for Ayers_et_al_IFN_genes comparing pooled adjuvant and neoadjuvant cohorts. Each dot represents a patient, the lower and upper box bounds indicate the 25th and 75th percentile and the middle box line indicates the median value (adjuvant n = 15; neoadjuvant n = 34). P values were calculated using a two-sided Wilcoxon rank-sum test. d Forest plot of Cox proportional hazard model analyzing the effect of Farmer_Breast_Cancer_Cluster_2 and clinical variables on overall survival. Diamonds mark the hazard ratio (HR) estimates while error bars denote the 95% confidence interval (CI) of the HR. The P value of each covariate was based on its Wald statistics; the P values are not adjusted. e Correlation between tumor fraction as estimated by Cibersort X and Farmer_Breast_Cancer_Cluster_2 gene signature score. Each dot represents a patient (n = 49). The Spearman’s rank correlation coefficient (ρ) and p value are calculated using a two-sided Spearman correlation test. f Correlation between myeloid fraction as estimated by Cibersort X and Ayers_et_al_IFN_genes gene signature score. Each dot represents a patient (n = 49). The Spearman’s rank correlation coefficient (ρ) and p value are calculated using a two-sided Spearman correlation test. Source data are provided as a Source Data file.

Fig. 2. Clinical trial design and survival outcomes.

Patients in the first stage of this clinical trial (S1) were randomized to no treatment prior to surgical resection of recurrent glioblastoma (Adjuvant cohort, S1, blue) or single dose of pembrolizumab 200 mg prior to surgery (Neoadjuvant cohort, S1, red). An additional 25 patients were enrolled in a second stage as a neoadjuvant expansion cohort (Neoadjuvant cohort, S2, black). a Clinical trial schema. b Kaplan–Meier plot of overall survival in the modified intention-to-treat population. Median survival of the neoadjuvant expansion cohort (black) was 209.5 days, compared to 400 days in the stage 1 neoadjuvant cohort (red) and 192 days in the stage 1 adjuvant cohort (blue) (p = 0.0858 logrank test). c Kaplan–Meier plot of progression-free survival in the modified intention-to-treat population. Median progression-free survival of the neoadjuvant expansion cohort (black) was 75 days, compared to 99.5 days in the stage 1 neoadjuvant cohort (red) and 72.5 days in the stage 1 adjuvant cohort (blue) (p = 0.0543 logrank test). Source data are provided as a Source Data file.

Fig. 3. Cell cycle/cancer proliferation and T cell and interferon signatures define survival groups in patients treated with neoadjuvant pembrolizumab.

a Participants were classified on the basis of their cell cycle/cancer proliferation and T cell/interferon-γ-related gene signatures into 4 groups. Gene signature high/low was defined as having gene signature enrichment scores above or below the median, respectively b Distribution of patients in the neoadjuvant arm versus the adjuvant arm of our clinical trial into the 4 groups defined in (a). There was statistically significant uneven distribution of patients among the 4 groups, with all patients in Group 3 coming from the neoadjuvant cohort (p = 0.027, Chi-square test). c Kaplan–Meier plot of overall survival by group as defined in a (p = 2.77 × 10⁻⁶, log-rank test). d Boxplots for gene signature scores for STAT3 and TNF-α signaling GSEA sets in groups 1-4 (n = 49). Each dot represents a patient sample. Middle box line depicts the median, while lower and upper box bounds represent the 25% and 75th percentiles, respectively. Vertical line between data points represents the data range. P values calculated using a Wilconxon rank-sum test. e Scatter dot plot of tumor-infiltrating lymphocyte (TIL) density expressed as a percentage of total nucleated cells and stratified by T cell/interferon and cell cycle gene expression signature groups as defined in a (n = 12; n = 13; n = 12; n = 11 for groups 1–4, respectively). Line is at mean TIL density with error bars denoting the standard error of the mean (n.s. not significant, two-tailed unpaired non-parametric test, Mann–Whitney). f Scatter dot plot of tumor-infiltrating lymphocyte (TIL) density expressed as a percentage of total nucleated cells and stratified T cell/interferon gene expression signature alone (n = 25 for low T cell/IFN, n = 23 for high T cell/IFN). Line is at mean TIL density with error bars denoting the standard error of the mean (p = 0.0032, two-tailed unpaired non-parametric test, Mann–Whitney). Source data are provided as a Source Data file.