PARP1 Inhibition Radiosensitizes Models of Inflammatory Breast Cancer to Ionizing Radiation

Abstract

Sustained locoregional control of disease is a significant issue in patients with inflammatory breast cancer (IBC), with local control rates of 80% or less at 5 years. Given the unsatisfactory outcomes for these patients, there is a clear need for intensification of local therapy, including radiation. Inhibition of the DNA repair protein PARP1 has had little efficacy as a single agent in breast cancer outside of studies restricted to patients with BRCA mutations; however, PARP1 inhibition (PARPi) may lead to the radiosensitization of aggressive tumor types. Thus, this study investigates inhibition of PARP1 as a novel and promising radiosensitization strategy in IBC. In multiple existing IBC models (SUM-149, SUM-190, MDA-IBC-3), PARPi (AZD2281-olaparib and ABT-888-veliparib) had limited single-agent efficacy (IC50 > 10 μmol/L) in proliferation assays. Despite limited single-agent efficacy, submicromolar concentrations of AZD2281 in combination with RT led to significant radiosensitization (rER 1.12-1.76). This effect was partially dependent on BRCA1 mutational status. Radiosensitization was due, at least in part, to delayed resolution of double strand DNA breaks as measured by multiple assays. Using a SUM-190 xenograft model in vivo, the combination of PARPi and RT significantly delays tumor doubling and tripling times compared with PARPi or RT alone with limited toxicity. This study demonstrates that PARPi improves the effectiveness of radiotherapy in IBC models and provides the preclinical rationale for the opening phase II randomized trial of RT ± PARPi in women with IBC (SWOG 1706, NCT03598257).

Conflict of interest statement

Disclosure of Potential Conflicts of Interest: The authors have no relevant conflicts of interest.

©2019 American Association for Cancer Research.

Figures

Figure 1:. PARP1 inhibition does not affect proliferation of IBC cell lines.

IBC cell lines were treated with either olaparib or veliparib and cell viability was measured 72 hours after treatment. In SUM-190 (A,B) and MDA-IBC-3 (C,D) cells, neither veliparib or olaparib showed significant effects on proliferation at doses up to 10μM. In SUM-149 cells (E,F), olaparib, but not veliparib, can inhibit proliferation at high doses (2.2μM). Graphs are shown as the average of three independent experiments ± SEM.

Figure 2:. Clonogenic survival of IBC cell lines decreases with olaparib treatment.

Olaparib treatment results in a dose-dependent reduction in survival fraction of SUM-190 (A), MDA-IBC-3 (C), and SUM-149 (E) cell lines. Representative data from single experiments are shown for each cell line. The surviving fraction of cells after 6 Gy (B, D, F) was calculated as the mean of three independent experiments and depicted ± SEM for each cell line. (p < 0.05 = *, p < 0.01 = **)

Figure 3:. Radiation in combination with the PARP1 inhibition leads to persistence of DNA damage in IBC cell lines.

Immunofluorescence microscopy was used to measure γH2AX foci in SUM-190 (A) and SUM-149 (B) cells. Cells were pretreated for one hour with olaparib and fixed at 0.5, 4, 12, 16, and 24 hours after radiation, then stained for DAPI and γH2AX. Cells containing ≥ 15 foci were scored as positive. In SUM-190 cells at 4, 12, and 16 hours, there were significantly higher levels of cells positive for γH2AX for those treated with the combination of 2 Gy radiation and 1μM olaparib compared to cells treated with 2 Gy radiation alone. In SUM-149 cells, 20nM olaparib and 2 Gy radiation results in a higher percentage of γH2AX positive cells compared to cells treated with radiation alone at both 12 and 16 hours. Representative images of γH2AX foci in SUM-190 (C) and SUM-149 (D) cells at 16 hours are shown for all treatment groups. Graphs represent the average of three independent experiments ± SD. (p < 0.05 = *)

Figure 4:. PARP1 inhibition increases dsDNA breaks and significantly decreases PAR formation in IBC cell lines.

Neutral comet assay in SUM-190 cells (A) shows higher levels of dsDNA damage at 4 hours in cells treated with radiation and olaparib compared to untreated cells, or cells treated with RT or olaparib alone (p < 0.05 = *). Graphs represent the average of three independent experiments ± SD and representative images for each treatment are shown. In SUM-190 (B) and MDA-IBC-3 (C) cells, radiation induced DNA damage causes an increase in PAR formation at both 6 and 24 hours after 4 Gy radiation. In the combination group that receives a one-hour pretreatment of 1μM olaparib before radiation, PAR formation is significantly lower at 6 and 24 hours after RT. In SUM-149 (D) cells, this same trend can be observed at a much lower dose of olaparib (20nM). Though the enzymatic activity of PARP1 is efficiently inhibited at these doses, total levels of PARP are not significantly different across the treatment conditions.

Figure 5:. PARP1 inhibition with radiation is more effective than radiation alone in a SUM-190 xenograft model.

SUM-190 cells were subcutaneously injected into CB17-SCID mice, and treatment was started when tumors reached approximately 80 mm3 (A). Olaparib treatment began one day before the initiation of radiation treatment and ended on the same day as the last fraction of radiation. With this paradigm, the combination treatment leads to delayed growth of tumors (B) and an increased time to tumor doubling (C) and tumor tripling (D) (p < 0.0001 = ****). The treatment did not display significant toxicities, and animal weights were not significantly different between the treatment groups (E). Using the FTV method, there was a synergistic effect with olaparib and RT treatment to antagonize tumor growth (ratios >1 indicate synergism) (F). A two-way ANOVA was performed to compare tumor volume between experimental groups.