Development of In Vitro Assays for Advancing Radioimmunotherapy against Brain Tumors

Abstract

Glioblastoma (GBM) is the most common primary brain tumor. Due to high resistance to treatment, local invasion, and a high risk of recurrence, GBM patient prognoses are often dismal, with median survival around 15 months. The current standard of care is threefold: surgery, radiation therapy, and chemotherapy with temozolomide (TMZ). However, patient survival has only marginally improved. Radioimmunotherapy (RIT) is a fourth modality under clinical trials and aims at combining immunotherapeutic agents with radiotherapy. Here, we develop in vitro assays for the rapid evaluation of RIT strategies. Using a standard cell irradiator and an Electric Cell Impedance Sensor, we quantify cell migration following the combination of radiotherapy and chemotherapy with TMZ and RIT with durvalumab, a PD-L1 immune checkpoint inhibitor. We measure cell survival using a cloud-based clonogenic assay. Irradiated T98G and U87 GBM cells migrate significantly (p < 0.05) more than untreated cells in the first 20−40 h post-treatment. Addition of TMZ increases migration rates for T98G at 20 Gy (p < 0.01). Neither TMZ nor durvalumab significantly change cell survival in 21 days post-treatment. Interestingly, durvalumab abolishes the enhanced migration effect, indicating possible potency against local invasion. These results provide parameters for the rapid supplementary evaluation of RIT against brain tumors.

Keywords: brain cancers; durvalumab; glioblastoma; immune checkpoint inhibitors; immunoradiotherapy; immunotherapy; radioimmunotherapy; radiotherapy; temozolomide.

Figure A1

Choice of regions of interest for rate of migration (ROM) and late resistance (LR) analysis of T98G and U87 following treatment. Time frames used for both T98G and U87 were selected for each experiment and rested consistently within the migration region for all trials. The cell behavior was considered in the selection of the migration region. Timescales shown were shortened for ease of visualization.

Figure A2

U87 MG seeded at high density ($5.0\times {10}^5$ cells/mL), in a 24-well plate, demonstrating preferential cell–cell adherence. Images captured on the CytoSmart Omni system. Top: Cells 2 h after initial seeding. Cells have begun adhering to the bottom surface of the culture assay and have already started migrating. Bottom: Cells 24 h following initial seeding. Cells have migrated significantly, preferentially adhering to one another to form neurospheres. Notice also that there is significant empty space on the bottom surface of the well, showing that cluster formation is preferred in some aspect over formation of a cell monolayer for U87 MG cells, leading to reduced resistance with migration, in ECIS, following initially increased resistance (Figure A1). It should be noted that imaging of the ECIS array at 168 h post-treatment showed slightly smaller cluster size and increased array coverage in treated cells as compared to untreated cells.

Figure A3

Preparation of clonogenic assays. Step 1: Cells were taken from culture vessels and seeded in intermediate vessels at reduced cell density to ensure count consistency with the variable seeding strategy. Following seeding, cells were incubated for 2–3 days prior to treatment to allow for recovery. Step 2: Cells were treated and lifted for plating. Step 3: Cells were plated into clonogenic assay well plates at the variable seeding density. The plate was placed into the incubator for the duration of the experiment. Step 4: Medium was refreshed every 5 days for the duration of the experiment to ensure ample access to nutrients. Step 5: At the endpoint, cells were treated with fixing and staining solutions for final manual colony counts.

Figure A4

Images of T98G cells taken on the CytoSmart Omni. (a) T98G cells in a typical Omni image in brightfield viewing mode. (b) T98G cells under colony viewing mode. Colonies of sufficient size are displayed in the orange highlight and were included in colony analysis.

Figure 1

Typical ECIS plot of normalized resistance (64,000 Hz). Cells begin by adhering to the bottom surface of the array, forming an initial buildup region. After initial attachment to the substrate, cells migrate toward the preferred distribution. For T98G, cells migrated to cover the available space in the array, increasing resistance until reaching a peak, dubbed the “plateau region,” at which time cells have reached maximal coverage. The late region is characterized by continued proliferation and/or cell death, which increases or decreases resistance, respectively.

Figure 2

ECIS migration of T98G and U87 glioblastoma cells treated with radiation and temozolomide at 7300 ng/mL. Representative plots of normalized resistance in the first 60 h (a) and 12 h (c) post-treatment for T98G and U87, respectively, are shown. Calculated percent differences in ROM for treated conditions, relative to the untreated control, for T98G (b) and U87 (d). * p < 0.05, ** p < 0.01, NS non-significant.

Figure 3

Migration-focused results of ECIS tests on T98G and U87 glioblastoma cells treated with radiation and durvalumab. (a) Representative plot of normalized resistance in the first 40 h post-treatment for T98G. (b) Violin plot of percent differences in ROM for (a), relative to the untreated T98G control. (c) Representative plot of normalized resistance in the first 40 h post-treatment for U87. (d) Violin plot of percent differences in ROM for (c), relative to the untreated U87 control. NS non-significant.

Figure 4

Barrier function for T98G treated with radiation and TMZ. (a) A representative plot of the barrier function over time. (b) Violin plots for (a) showing the direct comparison of the late barrier function relative to the untreated condition. Overall, treatment with radiation led to decreased barrier function following the plateau region. Additionally, cells treated with 5 Gy of radiation and TMZ had further decreased barrier function relative to 5 Gy only (p < 0.05). * p < 0.05, NS non-significant.

Figure 5

Barrier function-focused results of ECIS tests on T98G glioblastoma cells treated with radiation and durvalumab. (a) Representative plot of the barrier function R_b post-treatment for T98G shown. (b) Violin plots of % differences in late R_b for treated conditions, relative to the untreated control. The addition of durvalumab did not significantly affect the barrier function compared to irradiated controls. NS non-significant.

Figure 6

Late resistance (LR) results. (a) Plots of the normalized resistance over 1 week post-treatment for T98G. (b) Violin plots of percent difference comparisons of late resistance for treated conditions in (a) relative to the untreated condition. (c) Plots of the normalized resistance over 1 week post-treatment for U87. (d) Violin plots of percent difference comparisons of late resistance for treated conditions in (a) relative to the untreated condition. Though marginal changes can be seen in (b,d), there were no statistically significant changes to LR for either cell line treated with radiation alone or with concurrent temozolomide. NS non-significant.

Figure 7

Late resistance-focused plots for T98G and U87 treated with radiation and durvalumab. (a) Representative normalized resistance plot for T98G treated with radiation and durvalumab over one week post-treatment. (b) Violin plots of comparison of late resistance (LR) between treated conditions in (a). (c) Representative normalized resistance plot for U87 treated with radiation and durvalumab over one week post-treatment. (d). Violin plots of comparison of late resistance (LR) between treated conditions in (c). The addition of durvalumab did not significantly change late resistance compared to irradiated conditions; however, cells treated with increased radiation doses showed decreased late resistance following the plateau region for T98G (p < 0.01). ** p < 0.01, NS non-significant.

Figure 8

Survival curve and calculated alpha/beta ratio. (a) Survival curve for T98G cells treated with radiation at 0, 5, and 20 Gy alone, with concurrent temozolomide, durvalumab, and both. (b) Calculated alpha/beta ratio for (a). (c) Survival curve for U87G cells treated with radiation at 0, 5, and 20 Gy alone, with concurrent temozolomide, durvalumab, and both. (d) Calculated alpha/beta ratio for (c). Overall, increased radiation dose led to greater cell death (p < 0.001) for both cell lines. The addition of each agent to radiation showed no significant change to cell survival relative to the radiation-only conditions for T98G, but for U87, the addition of TMZ decreased cell survival (p < 0.0001) at 0 Gy, and marginally, but not significantly, decreased cell survival at higher radiation doses. NS non-significant.