Potentiation of tumor responses to DNA damaging therapy by the selective ATR inhibitor VX-970

Abstract

Platinum-based DNA-damaging chemotherapy is standard-of-care for most patients with lung cancer but outcomes remain poor. This has been attributed, in part, to the highly effective repair network known as the DNA-damage response (DDR). ATR kinase is a critical regulator of this pathway, and its inhibition has been shown to sensitize some cancer, but not normal, cells in vitro to DNA damaging agents. However, there are limited in vivo proof-of-concept data for ATR inhibition. To address this we profiled VX-970, the first clinical ATR inhibitor, in a series of in vitro and in vivo lung cancer models and compared it with an inhibitor of the downstream kinase Chk1. VX-970 markedly sensitized a large proportion of a lung cancer cell line and primary tumor panel in vitro to multiple DNA damaging drugs with clear differences to Chk1 inhibition observed. In vivo VX-970 blocked ATR activity in tumors and dramatically enhanced the efficacy of cisplatin across a panel of patient derived primary lung xenografts. The combination led to complete tumor growth inhibition in three cisplatin-insensitive models and durable tumor regression in a cisplatin-sensitive model. These data provide a strong rationale for the clinical evaluation of VX-970 in lung cancer patients.

Figures

Figure 1. VX-970 is a potent and selective inhibitor of ATR

(A) Chemical structure of VX-970. (B) Exponentially growing H2009 cells were treated overnight (17 h) with the indicated concentrations of VX-970 alone (lanes 1-6) or in combination with 20 μM cisplatin (lanes 7-12). Cells were then harvested and analyzed for the expression of P-Chk1-S345, P-H2AX-S139 and P-KAP1-S824 by immunoblotting.

Figure 2. ATR inhibition sensitizes lung cancer cells to DNA damaging agents

(A) Analysis of shifts in the concentration of DNA damaging agent required to inhibit cell viability by 50% (IC50) was used to determine the synergistic or antagonistic effects of a DDR inhibitor. Heat map representing the IC50 shift with the largest absolute value observed with VX-970 or AZD7762 in combination with cisplatin, etoposide, gemcitabine, oxaliplatin and irinotecan across a panel of 36 lung cell lines at 96 h. The colors represent a shift range from −10 (antagonism-blue) to +10 (synergy-red). (B) Histograms showing the percentage of cell lines with > 3-fold (top panel) or 10-fold (bottom panel) synergy with VX-970 (ATR) or AZD7762 (Chk1/2) in combination with cisplatin, oxaliplatin, irinotecan, gemcitabine and etoposide. (C) Impact of p53 on response to VX-970 in A549 cells. Histogram depicts maximum IC50 shift in vector control and p53 knockdown cells observed with VX-970 in combination with cisplatin, etoposide, gemcitabine, oxaliplatin and irinotecan.

Figure 3. VX-970 synergizes with cisplatin across a range of human primary lung tumor models in vitro

Dissociated tumor cells were treated in triplicate in vitro with a matrix of VX-970 and cisplatin concentrations, and synergy or antagonism was analyzed at the 95% confidence interval with MacSynergy II software. Degree of synergy, shown as positive log volume, and antagonism, shown as negative log volume, are shown.

Figure 4. VX-970 enhances the therapeutic efficacy of cisplatin in patient-derived lung tumor xenografts

(A-G) Human primary tumor tissues were passaged in SCID mice. Treatment started when the average tumor size was approximately 200 mm3. Tumor bearing mice were treated with vehicle, VX-970 (30 mg/kg in all models except 60 mg/kg in OD26749 and OD26131) PO, 4 consecutive days a week, alone and in combination with cisplatin (3 mg/kg IP, q7d), and cisplatin alone. Tumor volume and body weight were measured twice a week. Studies were terminated one or two days after the final dose of VX-970. Points show the mean tumor volume (mm3) for each treatment group (n=5-10).

Figure 5. VX-970 inhibits ATR and promotes sustained regression in a human primary lung tumor xenograft

(A) Mice bearing OD26749 tumors were treated with a single dose of VX-970 (60 mg/kg PO) and cisplatin (3 mg/kg IP) either as monotherapy or in combination. P-Chk1 and P-H2AX were analyzed by western blot 4 and 48 h after treatment, respectively, and were corrected to total H2AX levels. (B) Representative PET/CT images of mice at baseline (top row) and six days following the indicated treatment (bottom row). Concentration of 18FLT is indicated by the intensity of the orange color in the PET image. An X-ray CT image of the skeletal system (white) is superimposed for anatomical reference. The tumor is circled (green). Treatment with VX-970 alone had no discernable effect on PET signal (image not shown). (C) Average standard uptake value (SUV) of 18FLT. (D) Tumor size changes as determined by CT. (E) Examination of prognostic value of 18FLT imaging. (F) Assessment of the durability of response to VX-970. OD26749 tumors were passaged in SCID mice to P4. Treatment started when the average tumor size was approximately 200 mm3. Tumor bearing mice were treated with vehicle, VX-970 alone (60 mg/kg PO, 4 consecutive days a week), cisplatin alone (3 mg/kg, IP, q7d) and the combination, for two weeks. Vehicle and VX-970 groups were terminated after the final dose of VX-970. Following treatment discontinuation, tumor growth in the cisplatin alone and combination groups was followed twice a week until the average tumor volume reached 1000 mm3. The arrowhead on the graph marks the end of treatment. Error bars are standard errors for all graphs.