Neurosurgical application of olaparib from a thermo-responsive paste potentiates DNA damage to prolong survival in malignant glioma

Abstract

Background: There is increased pan-cancer specific interest in repurposing the poly adenosine diphosphate-ribose polymerase-1 (PARP-1) inhibitor, olaparib, for newly diagnosed or recurrent isocitrate dehydrogenase wild type glioblastoma. We explore whether intra-cavity delivery of olaparib confers a survival benefit in a pre-clinical high-grade glioma model.

Methods: Primary tumor RNA sequencing data was used to determine PARP-1 as a target in the glioblastoma infiltrative margin. We assessed radiosensitization conferred by olaparib alone and concomitant to genotoxic insults in vitro using clonal growth assays, cell cycle analysis and immunocytochemistry, and in vivo upon post-surgical delivery from a temperature-sensitive polymeric paste.

Results: RNA-sequencing confirmed PARP-1 as a viable therapy target in glioblastoma infiltrative disease. Acute exposure of glioma cells to olaparib impaired proliferation and induced late-stage apoptosis associated with DNA damage in vitro, potentiated by radiation. Using high-grade glioma orthotopic allografts, a long-term overall survival benefit was observed upon interstitial olaparib delivery concomitant with radiotherapy, compared to systemic olaparib and standard glioblastoma treatment. Combined delivery of olaparib with either temozolomide or etoposide increased long-term survival, suggestive of olaparib functioning as DNA damage sensitizer.

Conclusions: Collectively, our data support a rationale for localized olaparib delivery concomitant with the current clinical regimen for malignant glioma treatment.

Fig. 1. PARP-1 inhibition is a clinically relevant target in primary GBM and confers high-grade glioma sensitivity to DNA damage.

a Using the R2: Genomics Analysis and Visualization Platform, PARP-1 median gene expression (log2) was retrieved from two normal brain cohorts (n = 172 and n = 44; green bar plots), and three IDH-1 WT GBM cohorts (n = 20, n = 70 and n = 74; blue bar plots and where ‘u’ and ‘m’ designates unmethylated and methylated MGMT promoter respectively). One-way ANOVA (F statistic – 27.83; p-value < 0.01) indicates variance in PARP-1 expression across groups, with higher expression in primary GBM, relative to normal brain (median gene expression range of 8.31-8.73 for normal brain regions, 8.91–9.52 for IDH-1 WT GBM). b Kaplan–Meier survival analyzes on TCGA (n = 76) IDH-1 WT GBM, based upon PARP-1 gene expression, showing no significant difference between high and low expression cohorts (log rank p-value 0.636). c PARP-1 gene expression (reads per kilobase million; RPKM) respectively across intra-tumor GBM biopsies (core, peripheral rim, unsorted invasive margin) and fluorescence-activated cell sorted invasive margin tumor (FACS pos.) and non-tumor (FACS neg.) based upon 5ALA fluorescence. No significant difference in PARP-1 mean expression was observed between the clinically relevant FACS/5ALA positive invasive GBM sub-population(s), and all intra-tumor regions and FACS/5ALA negative sub-population(s) (One-way ANOVA p-value 0.58, f-ratio 2.44). d 9 L clonogenic potential is significantly reduced after 7 days of exposure to a combination of 3 µM OLA and 3 Gy XRT, relative to control or each treatment alone. e–h The proportion of single-positive (AnnV⁺-PI^-) and double-positive (AnnV⁺-PI⁺) cells significantly increased when either U251 human GBM or 9 L rat gliosarcoma cells were exposed to a combination of OLA/XRT/TMZ and OLA/XRT/ETOP, relative to vehicle-only control, each treatment alone, or a combination of OLA/XRT (****p < 0.0001; ***p = 0.0001, **p = 0.001, *p = 0.01 (with the exception of a non-significant difference between OLA/XRT/ETOP vs. OLA/XRT or ETOP alone in U251 cells for the AnnV⁺-PI+ fraction).

Fig. 2. Cell cycle alterations in glioma cells induced by olaparib exposure in vitro.

U251 and 9 L high-grade glioma cells were exposed to OLA as single agent, combined with XRT, or combined with XRT and DNA damaging agents TMZ or ETOP. Cells were treated with drug compounds 24 h post-seeding, XRT administered 48 h post-seeding, and cell cycle analyzes conducted 72 h post-seeding. OLA alone, or in combination with XRT, XRT/TMZ or XRT/ETOP, significantly decreased the proportion of G1 cells for U251 (a, c) and 9 L treatment (b, d) relative to vehicle-only control cells (CTR). The G2/M peak was significantly decreased in both U251 and 9 L cells upon OLA/XRT/TMZ exposure and increased upon either OLA/XRT or OLA/XRT/ETOP exposure relative to control cells. A significant increase in the sub-G₀/G₁ fraction was detected only after OLA/XRT/TMZ or OLA/XRT/ETOP treatment for both U251 and 9 L cells (c, d). ****p < 0.0001; ***p = 0.0001, **p = 0.001.

Fig. 3. Induction of γH2AX-induced DNA damage upon olaparib exposure.

U251 and 9 L cells were exposed to OLA, TMZ or ETOP as single agents, OLA combined with XRT, or OLA combined with XRT and TMZ or ETOP. Cells were treated with drug compounds 24 h post-seeding, XRT administered 48 h post-seeding, and γH2AX immunofluorescence analyzes conducted 72 h post-seeding (a, b). A significantly higher number of nuclear γH2AX-positive foci was observed after combined OLA/XRT/TMZ in both U251 and 9 L cells relative to OLA/XRT (*p = 0.01). A marked, but non-significant increase in γH2AX-positive foci was observed after combined OLA/XRT/ETOP treatment in U251 and 9 L cells, relative to OLA/XRT (c, d). CTCF, corrected total cell fluorescence; Blue – DAPI, Red – γH2AX; CTR, vehicle-only control. Primary antibody – anti-phospho-histone H2A.X (Ser139) 1:500 (Merk-Millipore), secondary antibody conjugated to Alexa fluor 594, absorbance/emission 590/617; counterstain – Hoechst 33342 absorbance/emission 345/478; scale bar, 100μm.

Fig. 4. In vivo efficacy of PLGA/PEG-mediated interstitial delivery of olaparib as mono or combination therapy in orthotopic malignant glioma allografts.

Exponential plateau fitted in vitro cumulative % (a) and amount (mg) (b) release of OLA from PLGA/PEG matrices loaded with 500 mg of drug. The release study was performed in PBS (pH = 7.4) at 37 °C and OLA quantified by HPLC over a 21-day period. Error bars indicate the standard deviation from six independent matrices. c Kaplan–Meier overall survival plots of randomized F344 rats implanted with 9 L allografts and treated 5-days post allograft implant as follows (n = 7 per arm): (Control arms) surgery alone; surgery + intraperitoneal (IP) OLA (50 mg/kg daily for 5 days) + XRT; surgery + XRT + oral TMZ by gavage (50 mg/kg daily for 5 days) (clinical standard-of-care); (Treatment arms) surgery + PLGA/PEG loaded with 10% w/w OLA; surgery + PLGA/PEG loaded with 20% w/w OLA; surgery + PLGA/PEG loaded with 10% w/w OLA + XRT; surgery + PLGA/PEG loaded with 20% w/w OLA + XRT; surgery + PLGA/PEG loaded with 10% w/w OLA / 20% w/w TMZ + XRT; surgery + PLGA/PEG loaded with 20% w/w OLA / 20% w/w TMZ + XRT; surgery + PLGA/PEG loaded with 10% w/w OLA / 50% w/w ETOP + XRT; surgery + PLGA/PEG loaded with 20% w/w OLA / 50% w/w ETOP + XRT. d Depiction of comparisons only between post-surgical combination treatment groups followed by XRT, relative to clinical standard chemotherapy and XRT, demonstrating a significant long-term survival benefit. XRT administered at 10 Gy, 5 days after polymer/drug implant. Animals still alive after 120 days post-surgery and polymer implant were designated LTS. (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).

Fig. 5. Whole-brain histological confirmation of efficacy upon PLGA/PEG-mediated interstitial delivery of olaparib to orthotopic 9 L gliosarcomas.

Animals treated with (a) Surgery/intraperitoneal OLA/XRT (Day 22), (b) surgery/oral TMZ/XRT (Day 33) and (c) surgery/ OLA 20% w/w (Day 15), show recurrent tumor extent and cellular dense regions (denoted by *) within the tumor resection cavity (delineated by arrowhead) with visible infiltration of adjacent brain parenchyma. Animals treated with (d) surgery/OLA 20% w/w/XRT (Day 120), (e) surgery/OLA 20% w/w/TMZ 20% w/w/XRT (Day 120) and (f) surgery/OLA 20% w/w/ETOP 50% w/w/XRT (Day 15), show glial scar formation but with no visible recurrent tumor cells within the surgical resection site (denoted by arrowhead) and brain parenchyma beyond. All images were taken at x40. Scale bar a–h = 2.5 mm. ‘Days’ = number of days post-polymer implant; oral TMZ administered at 50 mg/kg/day for 5 days (Days 5–9); radiotherapy (XRT) administered as an external beam single dose of 10 Gy directly post-surgery.

Fig. 6. Schematic illustration of combination chemotherapy with OLA, ETOP and TMZ, with concomitant radiotherapy, in post-resection high-grade glioma.

The resection cavity is filled with drug-impregnated biodegradable polymeric paste, with subsequent diffusion of OLA, ETOP and TMZ towards the tumor margins and infiltrating cells. DNA damage through multiple putative mechanisms is achieved in combination with radiotherapy.