ABT-888 Confers Broad In vivo Activity in Combination with Temozolomide in Diverse Tumors
Joann P. Palma, Yi-Chun Wang, Luis E. Rodriguez, Debra Montgomery, Paul A. Ellis, Gail Bukofzer, Amanda Niquette, Xuesong Liu, Yan Shi, Loren Lasko,
Gui-Dong Zhu, Thomas D. Penning, Vincent L. Giranda, Saul H. Rosenberg, David J. Frost, and Cherrie K. Donawho

Abstract Purpose: ABT-888, currently in phase 2 trials, is a potent oral poly(ADP-ribose) poly- merase inhibitor that enhances the activity of multiple DNA-damaging agents, includ- ing temozolomide (TMZ). We investigated ABT-888+TMZ combination therapy in multiple xenograft models representing various human tumors having different responses to TMZ.
Experimental Design: ABT-888+TMZ efficacy in xenograft tumors implanted in subcu- taneous, orthotopic, and metastatic sites was assessed by tumor burden, expression of poly(ADP-ribose) polymer, and O6-methylguanine methyltransferase (MGMT).
Results: Varying levels of ABT-888+TMZ sensitivity were evident across a broad histo- logic spectrum of models (55-100% tumor growth inhibition) in B-cell lymphoma, small cell lung carcinoma, non–small cell lung carcinoma, pancreatic, ovarian, breast, and prostate xenografts, including numerous regressions. Combination efficacy in other- wise TMZ nonresponsive tumors suggests that TMZ resistance may be overcome by poly(ADP-ribose) polymerase inhibition. Profound ABT-888+TMZ efficacy was seen in experimental metastases models that acquired resistance to TMZ. Moreover, TMZ re- sistance was overcome in crossover treatments, indicating that combination therapy may overcome acquired TMZ resistance. Neither tumor MGMT, mismatch repair, nor poly(ADP-ribose) polymer correlated with the degree of sensitivity to ABT-888+TMZ. Conclusions: Robust ABT-888+TMZ efficacy is observed across a spectrum of tumor types, including orthotopic and metastatic implantation. As many TMZ nonresponsive tumors proved sensitive to ABT-888+TMZ, this novel combination may broaden the clinical use of TMZ beyond melanoma and glioma. Although TMZ resistance may be influenced by MGMT, neither MGMT nor other mechanisms of TMZ resistance (mis- match repair) precluded sensitivity to ABT-888+TMZ. Underlying mechanisms of TMZ resistance in these models are not completely understood but likely involve mechanisms independent of MGMT. (Clin Cancer Res 2009;15(23):7277–90)

ABT-888 is a new poly(ADP-ribose) polymerase (PARP) inhib- itor with excellent potency (KI, 5.2 and 2.9 nmol/L, PARP-1/
PARP-2) and oral bioavailability (1). As shown previously, ABT-888 enhanced the activity of multiple DNA-damaging agents, including cisplatin, carboplatin, cyclophosphamide, irinotecan, radiation, and temozolomide (TMZ), in various
xenograft and syngeneic preclinical models (2). In this study, we evaluated in detail the activity of TMZ and the ability of ABT-888 to enhance TMZ antitumor activity in multiple xenograft tumors implanted in subcutaneous, orthotopic, and metastatic sites.
The PARP family of enzymes is characterized by the ability to ADP ribosylate protein substrates, implicated in many cellular processes (differentiation, gene regulation, protein degradation,

Authors’Affiliation: AbbottLaboratories,CancerResearch,AbbottPark, Illinois Received 5/15/09; revised 8/27/09; accepted 9/1/09; published OnlineFirst 11/24/09.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).
Requests for reprints: Cherrie K. Donawho, Abbott Laboratories, In vivo Tumor Biology, Department R4N2, Building AP3, Abbott Park, IL 60064. Phone: 847-937-6556; Fax: 847-938-4777; E-mail: [email protected].
F 2009 American Association for Cancer Research. doi:10.1158/1078-0432.CCR-09-1245
replication, transcription, cell cycle progression, and methyla- tion), and overall maintenance of genomic stability (3, 4). PARP is also critical for single-strand break repair by base exci- sion repair (BER) that can lead to radioresistance and chemore- sistance (e.g., to alkylating agents; refs. 5, 6). Single-strand breaks induced by endogenous DNA damage, radiation, or al- kylation can become lethal double-strand breaks (DSB) during replication if PARP is inhibited (7, 8) and the accumulation of DSBs acts as triggers of cell death. Thus, the use of PARP in- hibitors to enhance DNA-damaging agents, such as TMZ, is be- ing evaluated in the clinic and shows promise as a therapeutic

www.aacrjournals.org 7277 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

rapidly inactivated on transfer of the alkyl group from the

Translational Relevance
ABT-888, a potent poly(ADP-ribose) polymerase (PARP) inhibitor, is currently in phase 2 clinical trials. PARP inhibitors broadly enhance the activity of DNA- damaging agents owing to the critical function of PARP-1 and PARP-2 in base excision repair. Thus, the use of PARP inhibitors to enhance DNA-damaging agents [e.g., temozolomide (TMZ)] shows promise as a therapeutic strategy for cancer. ABT-888+TMZ showed robust antitumor activity in a broad histolog- ic spectrum of subcutaneous, orthotopic, and meta- static xenograft models. These models attempt to recapitulate clinical settings that are difficult to treat and often result in severe morbidity/mortality. In addition, ABT-888+TMZ was efficacious in tumors resistant to TMZ monotherapy (inherent and ac- quired). Importantly, O6-methylguanine methyltrans- ferase expression does not preclude sensitivity to ABT-888+TMZ therapy. Our data define a rational combination regimen with direct clinical application in indications often hampered by resistance to chemotherapies (metastatic disease) and have the potential to positively affect patient outcome.

strategy for cancer (9). In addition, PARP has been implicated in DSB repair (10). The absence of PARP can lead to persistent and lethal DSBs resulting from unrepaired single-strand breaks encountered by the replication fork in cells deficient for DSB repair or homologous recombination repair mediated by BRCA1/2 (11–13). Furthermore, evidence for an alternate path- way to nonhomologous end joining in DSB repair can also be dependent on PARP and other BER proteins.
TMZ is an alkylating agent approved for anaplastic astrocyto- ma and newly diagnosed glioblastoma (14) that has also exhib- ited clinical activity in metastatic melanoma (15). It is currently under clinical investigation for multiple indications, including leukemia, lymphoma, aerodigestive tract, pancreatic, and neu- roendocrine tumors. TMZ induces DNA methylation of guanine at O6 (O6-MeG, the most cytotoxic lesion that accounts for ∼5% of the lesions) and causes other less cytotoxic lesions, gua- nine at N7 (N7-MeG, 70%) and adenine at N3 (N3-MeA, 9%; refs. 16, 17). The N3-MeA and N7-MeG adducts are rapidly re- paired by BER, but during replication, O6-MeG incorrectly pairs with thymine and triggers the mismatch repair (MMR) system leading to futile cycles of repair, generating DSBs that result in arrest and apoptosis (16, 17). A deficiency in MMR leads to tol- erance of O6-MeG, resulting in cell survival despite persistent DNA damage (18). Removal of alkylation damage mediated by the DNA repair protein O6-alkylguanine-DNA alkyltransfer- ase [O6-methylguanine methyltransferase (MGMT)] is also a mechanism of TMZ resistance (19). Disruption of BER through inhibition of PARP will block the repair of the N3-MeA and N7- MeG adducts, rendering them toxic and overcoming the above mechanisms of TMZ resistance.
Although TMZ has shown promise in melanoma and glioma, its clinical effectiveness is hampered by development of chemoresistance (14). MGMT levels and MMR deficiency are factors clinically implicated in TMZ resistance. As MGMT is
O6-MeG adducts, this has led to evaluation of protracted dosing regimens to deplete MGMT activity as a strategy to overcome resistance. Although high levels of MGMT are prevalent in tu- mors (20) and associated with TMZ resistance in vitro and in vivo (21), clinical studies have not always shown a clear association of low MGMT activity (by methylation or expres- sion) with better response or prognosis (14, 22), suggesting the existence of other resistance mechanisms. To date, pseudo- substrates of MGMT are being exploited as a therapeutic strate- gy to increase TMZ effectiveness and potentially overcome resistance by depleting MGMT (23, 24). PARP inhibition has been shown to overcome TMZ resistance in MMR-deficient cells (16, 25) and represents an attractive strategy for reversing additional mechanisms of TMZ resistance.
Results presented here show the ability of ABT-888 to en- hance TMZ activity in a broad spectrum of tumor types. Efficacy was not dependent on sensitivity to TMZ, as tumors initially refractory to TMZ showed a significant antitumor response to ABT-888+TMZ combination therapy. Of 10 models tested, TMZ was effective in 7, all of which eventually acquired resis- tance to TMZ. Three models showed very little TMZ monother- apy efficacy. In contrast, ABT-888+TMZ was significantly active in all 10 models [6 showed profound efficacy with >90% tumor growth inhibition (TGI)]. Furthermore, in a subcutaneous and intratibial model, acquired resistance to TMZ monotherapy could be overcome by the addition of ABT-888 to TMZ in cross- over treatment for a second cycle of therapy. In a second model, the efficacy of ABT-888+TMZ not only affected the primary tu- mor but also suppressed development of spontaneous lung me- tastases. Potential determinants of the in vivo response were evaluated; tumor poly(ADP-ribose) polymer (PAR) as well as MGMT expression did not correlate with in vivo sensitivity, sug- gesting MGMT-independent mechanisms of TMZ resistance.

Materials and Methods

Compound. Enantiomerically pure ABT-888 was synthesized at Abbott. The synthesis and cell-based evaluation is published elsewhere (1, 8).
Cell lines for in vivo studies. Capan-1 and Calu-6 were obtained from the American Type Culture Collection, HeyA8 and MDA-231-LC3 from M. D. Anderson, and NCI-H526 from the National Cancer Institute Department of Tumor Repository. The Capan-1 cells and tumors used in our studies were confirmed to harbor the published 6174delT deletion mutation resulting in BRCA2 deficiency (data not shown; ref. 26). All lines were cultured in RPMI 1640 with 10% fetal bovine serum (FBS; Hyclone), except HeyA8 and MDA-231-LC3 (DMEM with 10% FBS) and Capan-1 (Iscove’s modified Dulbecco’s medium with 20% FBS). DOHH-2 was purchased from the Deutsche SammLung von Mikroorganismen und Zellkulturen and cultured in RPMI 1640 (In- vitrogen) supplemented with nonessential amino acids (Invitrogen) and 10% FBS. Cells were maintained at 37°C in a humidified atmosphere equilibrated with 5% CO2, 95% air, confirmed to be free of Mycoplasma, and used between passages 3 and 7 when in log phase for tumor cell inoculation.
The PC3M-Luc-C6 (PC3M-Luc; Caliper Corp.) is a PC3M luciferase- transfected line from an orthotopically, in vivo–selected PC3 line de- rived from a grade 4 prostate adenocarcinoma bone metastases, and the MDA-MB-231-luc-D3H2LN (MDA-231-Luc; Bioware, Caliper) is a luciferase line derived from a spontaneous lymph node metastases of MDA-MB-231-luc breast cells implanted orthotopically into the mam- mary fat pad. Both cell lines were cultured in MEM with 10% FBS.

Clin Cancer Res 2009;15(23) December 1, 2009 7278 www.aacrjournals.org

ABT-888+TMZ Therapy Efficacy in Preclinical Models

Xenograft subcutaneous flank studies. All animal studies were con- ducted in a specific pathogen-free environment in accordance to the internal Institutional Animal Care and Use Committee, accredited by the American Association of Laboratory Animal Care, which meet or exceed the standards set by the U.S. Department of Agriculture Animal Welfare Act, Public Health Service policy on humane care and use of animals, and the NIH guide on laboratory animal welfare. In all studies shown, no significant health concerns were observed (e.g., ≥20% weight loss, ruffled coat, hunched posture, dehydration, and lethargy) unless otherwise noted.
Mice were obtained from Charles River Laboratories at 5 to 6 weeks of age and used when greater than 8 weeks of age and/or ∼20 g in weight. One million cells (2 × 106 for Calu-6) were mixed 1:1 with Ma- trigel (BD Biosciences) and injected s.c. (0.1-0.2 mL) into the shaved flank of C.B.-17 scid mice. Tumors were size matched (181-520 mm3) before initiation of dosing and measured two to three times weekly. Two bisecting diameters were measured with calipers and tumor vo- lumes were estimated from the following formula: (length × width2)/2.
See Supplementary Data for details on orthotopic, brain, and intra- tibial site injections.
Bioluminescent imaging and data acquisition for PC3M-Luc and MDA- 231-LN-Luc tumors. In vivo bioluminescent imaging (BLI) was done using an IVIS Spectrum Imaging System with Living Image 3.0 software (Caliper). Briefly, a 15 mg/mL solution of luciferin was prepared fresh in PBS. Mice were injected i.p. with 150 mg/kg and imaged 10 min after luciferin administration. Uniform regions of interest were used across all groups and time points to achieve quantification of bioluminescent signal. The background signal from a naive mouse was subtracted from the total flux (photons/second) obtained in each region of interest to normalize values. Mice were staged into treatment groups based on ini- tial BLI levels to provide normal distributions with similar means.
Tumor growth calculations. Calculations were derived from either caliper measurements or BLI tumor measurements. BLI tumor measure- ments were normalized against naive mice (background): [tumor BLI measurement - naive mouse (background)]. Percentage tumor change was calculated using each individual mouse initial tumor volume as its own control and calculated by the following formula: ([BLI daily mea- surement] – [size-match BLI on d:0 of the same mouse]) / [size-match BLI (d:0) of the same mouse] × 100. Treatment effect on tumor growth rate was assessed by determining %T/Cday x calculated by the following formula: ([mean tumor volume of treated group on day x / mean tumor volume of control vehicle group on day x] × 100), where tumor volume may be represented by caliper measurements or the normalized BLI measurement for luciferase lines (time point A). Enhancement of TMZ efficacy (%T/CTMZ) was assessed using tumor growth effect in the TMZ monotherapy group as control (CTMZ) compared with the TMZ combination group (time point B). %T/CTMZ was assessed at a lat- er time point than %T/C when the vehicle control groups have reached tumor end point and can no longer be analyzed.
X-ray image analysis of tibiae and area of decreased calcification calcu- lation. The tibiae were X-rayed using a Faxitron (Faxitron X-Ray Corp.). The area of decreased calcification (ADC) of tibiae between the knee and fibula joint was analyzed using the Automatic Measure- ment Program Wizard image analysis program (AxioVision 4, Zeiss). The difference between background calcification of normal tibia and tumor-bearing areas equals the ADC. A reduction in ADC or bone density in the treated groups compared with the vehicle group repre- sents a decrease in osteolytic reaction or tumor burden expressed as
% ADC inhibition calculated by the following formula: [vehicle ADC - treated ADC] / vehicle ADC × 100 (i.e., a greater % ADC inhibition, the more effective the treatment). Both the PC3M-Luc and MDA-231- LN-Luc lines result in an osteolytic lesion when implanted in the tibiae.
ABT-888 was administered orally (twice daily for 3-6 days) in a ve- hicle containing 0.85% NaCl adjusted to pH 4.0 using HCl. TMZ (Schering-Plough) was formulated in 0.2% hydroxypropyl methylcellu- lose and administered orally (once daily for 5 days) concurrently with
ABT-888. Zoledronic acid (ZA; Novartis) was formulated in 0.85% NaCl and administered twice a week.
PAR, MGMT, hMLH1, and hMSH2 Western blot. A reduction in tu- mor PAR after ABT-888 treatment reflecting the inhibition of PARP ac- tivity in vivo (2) and MGMT expression, a potential factor in TMZ resistance mechanism (21), were evaluated by Western blot. Tumors were excised from humanely euthanized mice treated in vivo, as previ- ously described. Tumor lysates were analyzed by immunoblot using SuperSignal West Dura (Pierce). The following rabbit polyclonal anti- bodies were used: PAR from Trevigen (1:1,000) and MGMT (1:1,000) and β-tubulin (1:1,000) from Cell Signaling Technology. The following mouse monoclonal antibodies were used: hMLH1 (1 μg/mL) from BD Biosciences and hMSH2 (1 μg/mL) from EMD.
Statistical analysis. Differences between groups were analyzed us- ing the Student’s t test (two tailed) or ANOVA/Fisher’s protected least significant difference for comparing groups with P ≤ 0.05 considered statistically significant.

Results

Significant efficacy of ABT-888+TMZ in various tumor types regardless of TMZ sensitivity. Our previous preclinical studies showed that ABT-888 enhanced the antitumor activity of mul- tiple DNA-damaging agents, including TMZ in vitro (8) and in vivo (2). Although glioma and melanoma are primary clinical indications for TMZ (27, 28), we show that ABT-888 (25 mg/
kg/d, orally, twice daily for 5-6 days) enhances TMZ efficacy (50 mg/kg/d, orally, once daily for 5 days) when administered concurrently in tumors from divergent histologic tumor types, including B-cell lymphoma, small cell lung carcinoma, non– small cell lung carcinoma, ovarian, breast, pancreatic, and pros- tate models (Tables 1 and 2; Fig. 1; Supplementary Data S1). The TMZ dose used in our studies approximates the human clinical exposure (29). Further, the estimated minimum effica- cious dose targeted [ABT-888 doses used in this study (2)] was achieved in humans and its ability to significantly inhibit PAR in patients was shown (1, 30). No overt signs of toxicity or sig- nificant health concerns were observed other than temporary and recoverable weight loss where noted. In the BRCA2- deficient Capan-1 model, whereas ABT-888 had no single-agent activity at 50 mg/kg/d, moderate TGI was observed at higher doses (≥100 mg/kg/d; data not shown), albeit significantly less than was achieved with the TMZ combination. This result in Capan-1 is consistent with that seen using other PARP inhibi- tors and likely due to its homologous recombination deficiency (11–13, 26). ABT-888 also did not show single-agent activity at 25 mg/kg/d in other models; however, ABT-888+TMZ exhibited significant activity (68-100% TGIABT-888+TMZ) compared with vehicle in both solid and hematologic tumor types. Moreover, efficacious combination therapy was not dependent on the initial effectiveness of TMZ monotherapy, suggesting that the degree of TMZ sensitivity was not a predictor of ABT-888 combi- nation activity. In a subset of tumors (Capan-1, DOHH-2, and NCI-H526), TMZ monotherapy activity was minimal (29-41% TGI), yet ABT-888 showed marked in vivo efficacy in combina- tion with TMZ (68-89% TGIABT-888+TMZ; Table 1). Overall, our results illustrate the broad-spectrum activity of ABT-888+TMZ in a diverse panel of histologic types and show that lack of TMZ single-agent activity does not preclude sensitivity to ABT- 888+TMZ combination therapy.
In some cases, significant TMZ monotherapy activity was ob- served at early time points (HeyA8, orthotopic MDA-231-LC3,

www.aacrjournals.org 7279 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

Table 1. Efficacy of ABT-888+TMZ in subcutaneous xenograft tumors

Tumor model* TIMEPOINT A TIMEPONT B
% TGI† Number of PRs§ % TGI† Number of PRs§

P vs vehicle‡ P vs TMZ‡
DOHH-2
TMZ 29∥ 0/10

ABT-888+TMZ 68¶ NCI-H526††
8/10 55¶ 8/10

TMZ 33∥ 0/10
ABT-888+TMZ 89¶ 9/10 80¶ 9/10 Capan-1
TMZ, 1 cycle 41** 0/10
ABT-888+TMZ, 1 cycle 89¶ 9/10 81¶ 9/10 HeyA8
TMZ 92¶ 10/10
ABT-888+TMZ 92¶ 10/10 94¶ 10/10 MDA-231-LC3 Orthotopic
TMZ 90¶ 0/10
ABT-888+TMZ 97¶ 1/10 80¶ 2/10 Calu-6
TMZ, 1 cycle 79¶ 10/10
ABT-888+TMZ, 1 cycle 79¶ 10/10

TMZ then ABT-888+TMZ crossover 82¶ 7/8
ABT-888+TMZ, 2 cycles 95¶ 10/10

*n = 7-10 mice/group.
†% TGI: percentage tumor growth inhibition (100 – %T/C) at time points A and B as indicated in Fig. 1 except for DOHH-2 and Capan-1 where only time point A was evaluated.
‡Student’s t-test calculated against appropriate controls for subcutaneous xenografts (vehicle or TMZ monotherapy groups as indicated). §Number of partial responses (PR) defined as three or more measurements below initial tumor size at size-match.
∥P ≤ 0.05.
¶P ≤ 0.0001. **P ≤ 0.001.
††Alternate dose schedule of ABT-888 and TMZ (standard 5-day schedule is TMZ/daily × 5 and ABT-888/twice daily × 5 concurrently), see Figure 1.

and Calu-6; Fig. 1; time point A). However, after a period of time, resistance to TMZ or the regrowth of surviving cells after TMZ treatment became evident (HeyA8 and MDA-231-LC3, time point B) and a second cycle of TMZ proved ineffective (Calu-6, time point B). In contrast, the combination of ABT-888+TMZ produced sustained activity in the HeyA8, ortho- topic MDA-231-LC3, Capan-1, and Calu-6 models, indicating the ability to not only maintain sensitivity but also induce par- tial responses (PR; defined as three or more tumor measure- ments below the size match; Table 1). Although similar in size to tumors treated in the first cycle, tumors treated in the second cycle were insensitive to TMZ treatment, supporting the notion of acquired resistance in Calu-6 rather than simply a result of the regrowth from surviving cells. Further, a crossover treatment to combination therapy after the first cycle of TMZ showed profound activity (Calu-6), resulting in PRs similar to that achieved by two cycles of combination therapy, whereas tumors treated with a second cycle of TMZ monotherapy were resistant. These results show the ability of ABT-888+TMZ com- bination therapy to maintain efficacy in tumors inherently re- sistant to TMZ and those that acquire TMZ resistance.
Sustained combination therapy efficacy in xenografts implanted in orthotopic and metastatic sites. Clinical resistance to chemo-
therapy is frequently observed in metastatic tumors (31). Moreover, the failure to respond to TMZ is attributed to both inherent and acquired resistance mechanisms (32), similar to the development of the refractory response seen after TMZ treatment in our studies (Fig. 1). To evaluate the efficacy of ABT-888+TMZ in orthotopic and metastatic settings, several luciferase-labeled lines were implanted in orthotopic or common metastatic sites (intratibial and brain ectopic). The PC3M-Luc human prostate carcinoma was used both orthotopically and intratibially (Figs. 2 and 4; Supplementary Data S2); the MDA-231-LN-Luc, human breast carcinoma, was implanted in the brain ectopically and intratibially (common metastatic sites for breast cancer; Figs. 3 and 5; Supplementary Data S3). Both the orthotopic MDA-231-LC3 (Fig. 1; Table 1) and orthotopic PC3M-Luc (Fig. 2; Table 2) show marked and equal sensitivity to TMZ monotherapy and ABT-888+TMZ after one cycle of dosing (>90% TGI for TMZ and ABT-888+TMZ). In the orthotopic MDA-231-LC3 model, the substantial im- provement in efficacy with ABT-888+TMZ was not evident until at least 13 days after cessation of dosing when TMZ-treated tu- mors began to grow progressively. Further, ABT-888+TMZ treat- ment not only showed prolonged regression (80% TGI; P ≤ 0.0001 versus TMZ monotherapy on day 80) but also resulted

Clin Cancer Res 2009;15(23) December 1, 2009 7280 www.aacrjournals.org

ABT-888+TMZ Therapy Efficacy in Preclinical Models

in 2 of 10 cures (confirmed by histology). In the orthotopic PC3M-Luc model, a second round of treatment at day 42 provided no additional efficacy for TMZ monotherapy, whereas the ABT-888+TMZ maintained continued regression until day 61 (100% TGI; P ≤ 0.05 versus TMZ monotherapy on day 55, time point B) with a significant survival advantage com- pared with TMZ alone (P ≤ 0.05 versus TMZ monotherapy; Fig. 2C).
Because ABT-888 crosses the blood-brain barrier, we investi- gated the efficacy of ABT-888+TMZ in the brain. MDA-231-Luc cells were injected ectopically in the brain and monitored for tumor growth (1.3 × 107 photons/s, day 11) followed by
TMZ or ABT-888+TMZ (Fig. 3A and B). Significant efficacy was observed with both treatment regimens following the first cycle of therapy. Although resistance to TMZ monotherapy de- veloped during the second treatment cycle, profound efficacy continued to be observed with the ABT-888+TMZ through cy- cle 3. The effect on tumor growth (tumor sizes remaining below baseline throughout) correlated clearly with the marked im- provement in survival to end point seen with combination treatment (100% TGI; P ≤ 0.0001 versus TMZ monotherapy on day 43; Fig. 3C).
Metastatic bone lesions can be osteolytic (PC3M-Luc or MDA-231-Luc) or osteoblastic in nature. Bisphosphonates such

Table 2. Efficacy table of ABT-888+TMZ in orthotopic and metastatic xenograft tumors

Tumor model* Time point A Time point B
% TGI† No. PRs‡ % TGITMZ† No. PRs‡

P vs vehicle§ P vs TMZ§
PC3M-Luc, prostate
Orthotopic vs TMZ × 2
TMZ, 1 cycle 95∥ 5/9
ABT-888+TMZ, 1 cycle 100∥ 9/9
TMZ, 2 cycles 0/8
ABT-888+TMZ, 2 cycles >100∥ 9/9
Intratibial vs (TMZ+ZA) × 2
TMZ, 1 cycle 78∥ 0/10
ABT-888+TMZ, 1 cycle 98∥ 8/10

TMZ then
ABT-888+TMZ+ZA
81∥
0/10

ABT-888+TMZ then
ABT-888+TMZ+ZA
97∥
6/10

TMZ+ZA, 2 cycles 77∥ 0/10 0/10

ABT-888+TMZ+ZA, 2 cycles >100¶ 8/10 99∥ MDA-231-LN-Luc, breast
Brain ectopic vs TMZ × 2
TMZ, 1 cycle 97**
ABT-888+TMZ, 1 cycle >100** 11/11 TMZ, 2 cycles
7/10

ABT-888+TMZ, 2 cycles >100∥
Intratibial vs (TMZ+ZA) × 2
ZA, 1 cycle 0 0/6
TMZ, 1 cycle 0 0/9
TMZ+ZA, 1 cycle 2 0/10
ABT-888+TMZ, 1 cycle 63∥ 2/10
ABT-888+TMZ+ZA, 1 cycle 61∥ 2/9
11/11

TMZ, 2 cycles 0 0/9
TMZ+ZA, 2 cycles 0/10
ABT-888+TMZ, 2 cycles >100¶ 10/10
ABT-888+TMZ+ZA, 2 cycles 92∥ 6/9
Spontaneous lung metastases vs (TMZ+ZA) × 2
TMZ, 1 cycle 12 8/9
TMZ+ZA, 1 cycle >100∥ 10/10
ABT-888+TMZ, 1 cycle >100∥ 9/9
ABT-888+TMZ+ZA, 1 cycle >100∥ 9/9
TMZ, 2 cycles 0 7/9
TMZ+ZA, 2 cycles 5/10
ABT-888+TMZ, 2 cycles >100∥ 9/9

ABT-888+TMZ+ZA, 2 cycles >100∥

*n = 6-11 mice per group.
†Percentage TGI (100 – %T/C) at time points A and B as indicated in Figs. 2 to 5.
‡Number of PRs defined as three or more measurements below initial tumor size at size match.
9/9

§ANOVA or Fisher’s protected least significant difference for orthotopic, intratibial, or brain ectopic xenograft models calculated against appro- priate controls (vehicle or TMZ monotherapy groups as indicated).
∥P ≤ 0.05. ¶P ≤ 0.001.
**P ≤ 0.0001.

www.aacrjournals.org 7281 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

Fig. 1. Efficacy of ABT-888 in combination with TMZ in various xenograft models. ABT-888 was administered at 25 mg/kg/d, orally, twice daily (BID) for 5 d, concurrently with TMZ at 50 mg/kg/d, orally, once (QD) daily for 5 d, for one or more cycles as indicated. A, Capan-1, pancreatic carcinoma.
B, HeyA8, ovarian carcinoma. C, MDA-231-LC3, orthotopic breast carcinoma. D, Calu-6, non–small cell lung carcinoma. All are subcutaneous models except for MDA-231-LC3. Points, mean of 8 to 10 mice per treatment group; bars, SEM.

as ZA, which inhibit osteoclast action and bone resorption, are current standard of care for patients with bone metastases (33); therefore, ZA was used in conjunction with TMZ and ABT-888 to maintain bone integrity. The addition of ZA to treatments in the intratibial models was well tolerated, and as expected, the X-ray analysis revealed a better maintenance of bone density or reduced ADC in groups receiving ZA (Fig. 4B; Supplementary Data S2 and S3). However, ZA monotherapy did not decrease tumor burden or increase survival without the addition of ABT-888 to TMZ treatment.
Metastasis to the bone is a common occurrence with ad- vanced prostate and breast cancers (34, 35); thus, we assessed the efficacy of ABT-888+TMZ in the bone. The intratibial in- jection of PC3M-Luc and MDA-231-Luc cells resulted in osteo- lytic lesions; therefore, tumor burden was evaluated by both BLI and X-ray analysis. As was observed for this tumor in the orthotopic setting (Fig. 2B), while TMZ monotherapy was initially active, tumors quickly regrew and were resistant to a second cycle of treatment. In contrast, ABT-888+TMZ pro- duced sustained activity through both cycles (Fig. 4; Table 2). Furthermore, following one cycle of TMZ monotherapy, ABT-888+TMZ in the second treatment cycle overcame TMZ resistance. The potent activity of ABT-888+TMZ was also shown by an increase in time to end point by Kaplan-Meier analysis (Fig. 4C and D; P ≤ 0.05 versus vehicle or TMZ+ZA).
Likewise, the intratibial MDA-231-Luc model was largely resis- tant to TMZ monotherapy (0-2% TGI versus vehicle) but was sensitive to ABT-888+TMZ (>60% TGI; P ≤ 0.05 versus vehicle on day 37, time point A; Fig. 5; Table 2), with regression of tu- mors continuing through a second cycle of combination therapy (>90% TGI; P ≤ 0.05 versus TMZ monotherapy on day 55, time point B; Fig. 5).
Interestingly, mice injected intratibially with MDA-231-Luc cells also developed spontaneous lung metastases resulting in mortality after day 42 in the vehicle group (Fig. 5A and B). Both the TMZ and ABT-888+TMZ groups showed a marked benefit and potent antitumor activity, resulting in regressions of lung metastases. In contrast, acquired TMZ resistance or a significant regrowth of intratibial tumors was observed without the addi- tion of ABT-888 after the second cycle of therapy on day 55 (Fig. 5B; Table 2, time point B). Therefore, TMZ monotherapy and combination treatment seem to be uniformly more effec- tive (with PRs) in the spontaneous lung metastases than in the intratibial tumors within the same mouse and regrowth of lung metastases was not shown until after the second cycle of TMZ. Albeit a small sample size, this finding implies that sensitivity to chemotherapy may be influenced by the site as a reflection of the microenvironment or drug delivery as well as acquisition of molecular changes (MGMT, MMR induction, etc.). Although the nature of this difference in degree of sensitivity to treatment

Clin Cancer Res 2009;15(23) December 1, 2009 7282 www.aacrjournals.org

ABT-888+TMZ Therapy Efficacy in Preclinical Models

within different organ sites is unclear, our results show that ABT-888+TMZ provided strong efficacy in multiple tissue compartments, including the lung. Overall, these results from two different tissue types (prostate and breast) implanted in different organ sites (orthotopic or metastatic) exhibit the effec- tiveness of ABT-888+TMZ combination therapy, including tu- mors that rapidly acquire TMZ resistance.

Fig. 2. Efficacy of ABT-888 in combination with TMZ in orthotopic prostate PC3M-Luc xenografts. ABT-888 was administered at 25 mg/kg/d, orally, twice daily for 5 d, concurrently with TMZ at 50 mg/kg/d, orally, once daily for 5 d, for two cycles as indicated.
A, representative BLI. B, corresponding graph. C, Kaplan-Meier analysis based on 1,000% tumor change end point by BLI. A striking increase in survival of the ABT-888+TMZ group when compared with TMZ monotherapy group was evident with none of the mice in the ABT-888+TMZ group reaching end point when the study ended. Points, mean of 9 to 10 mice per treatment group; bars, SEM.
Significant PAR inhibition in tumors after ABT-888 treatment regardless of sensitivity to ABT-888+TMZ combination therapy. To delineate potential underlying mechanisms of sensitivity to ABT-888+TMZ combination therapy, we initially assessed the inhibition of tumor PAR after ABT-888 treatment in vivo by Western blot (Fig. 6). ABT-888 was administered concurrently with TMZ for 3 to 5 days at doses similar to the efficacy studies

www.aacrjournals.org 7283 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

Fig. 3. Efficacy of ABT-888 in combination with TMZ in MDA-231-Luc breast xenografts implanted in the brain metastatic site. ABT-888 was administered at 25 mg/kg/d, orally, twice daily for 5 d, concurrently with TMZ at 50 mg/kg/d, orally, once daily for 5 d, for three cycles as indicated.
A, representative BLI. B, corresponding graph of the brain ectopic breast at various time points showing antitumor efficacy of ABT-888+TMZ. Whereas loss of sensitivity to TMZ is observed after a second cycle, sustained efficacy is observed in the ABT-888+TMZ group after three cycles. C, Kaplan-Meier analysis based on 300% tumor change end point by BLI. Although treatment with TMZ monotherapy significantly increased survival compared with vehicle, ABT-888+TMZ markedly improved survival compared with TMZ treatment alone. A greater weight loss (17% from baseline) was observed in the ABT-888+TMZ group at the end of the third cycle but recovered weight
12 d later (6% weight loss from baseline). Points, mean of 11 mice per treatment group; bars, SEM.

described above and tumors harvested at 2 h after last dose. This treatment schedule and harvest time point was chosen because it correlates with initiation of efficacy and the presence of signifi- cant tumor drug levels (36). We initially examined PAR in repre- sentative models where differential enhancement of efficacy was observed with ABT-888 combination therapy compared with TMZ therapy alone (DOHH-2, NCI-H526, Capan-1, and Calu-6). Significant tumor PAR inhibition was observed in all tumor lines after ABT-888 treatment independent of the degree
of sensitivity to combination therapy, therefore indicating that ABT-888 effectively inhibits PAR in a diverse array of tumor types. To date, of 12 tumors tested, all express high levels of PAR and treatment with ABT-888 effectively reduces PAR levels in each case (data not shown).
MGMT expression before and after TMZ or ABT-888+TMZ combination therapy indicates MGMT-independent mecha- nism. Because MGMT is reported to have an important role in TMZ resistance due to its ability to repair DNA methylation,

Clin Cancer Res 2009;15(23) December 1, 2009 7284 www.aacrjournals.org

ABT-888+TMZ Therapy Efficacy in Preclinical Models

we next sought to determine if the tumor xenograft MGMT status influenced the sensitivity to TMZ and ABT-888+TMZ. Interestingly, the MGMT status of tumors showed no clear cor- relation to the TMZ sensitivity profile by Western blot (Supple- mentary Data S4; also confirmed by RNA microarray and methylation status; data not shown). Note that some TMZ- refractory xenografts (29-41% TGI) with varying sensitivities to ABT-888+TMZ (68-89% TGI; DOHH-2, NCI-H526, and Capan-1) were found to be MGMT expressing (Fig. 6B). In con- trast, some xenografts were negative or low for MGMT protein but divergent in TMZ sensitivity profiles: initially TMZ sensitive but became insensitive to TMZ and/or acquired TMZ resistance (78-97% TGI; Calu-6, orthotopic and intratibial PC3M, ortho- topic MDA-231-LC3, and brain ectopic MDA-231-LN-Luc) or TMZ resistant (2% TGI; intratibial MDA-231-LN-Luc). These re- sults suggest that although MGMT may be a determinant for TMZ resistance in gliomas (14), MGMT-independent factors
may be important in other tumor types. We have also investi- gated the molecular changes to some extent in the MDA-231- Luc line using accessible tumors and have shown no changes in MGMT expression in tumors implanted s.c. or in the brain (Supplementary Data S5), indicating that although molecular changes such as MGMT may influence sensitivity profiles of cells implanted in different organ sites, the exact mechanisms are not yet clear. More importantly, our results show that ABT-888 can significantly enhance the activity of TMZ in tumors independent of MGMT expression and that MGMT expression does not preclude the ability to achieve marked efficacy with ABT-888+TMZ treatment.
MGMT expression in untreated xenografts may not be truly reflective of the dynamic alkylation repair process; thus, we explored the functional consequence of treatment in vivo to MGMT expression. The subsequent inactivation or degradation of MGMT on transfer of the O6-methyl group during DNA

Fig. 4. Efficacy of ABT-888 in combination with TMZ in intratibial prostate PC3M-Luc xenografts. ABT-888 was administered at 25 mg/kg/d, orally, twice daily for 5 d, concurrently with TMZ at 50 mg/kg/d, orally, once daily for 5 d, for two cycles as indicated. A, X-ray images of tibia showing a normal tibia (A1), PC3M-Luc–bearing tibia with osteolytic lesion (A2), image analysis area outlined in red (A3), and quantitation area outlined in blue (A4) as a measure
of tumor burden. B, representative BLI and X-ray images of mice. C, corresponding graphs for the BLI. D, Kaplan-Meier analysis based on 1,000% tumor change end point by BLI. Whereas loss of sensitivity to TMZ+ZA is observed after a second cycle, sustained efficacy is observed in the ABT-888+TMZ+ZA groups. Resistance was overcome by treatment with TMZ in the first cycle that crossed over to ABT-888+TMZ+ZA in the second cycle compared with
the TMZ+ZA group treated for two cycles. Addition of ABT-888 to TMZ+ZA for two cycles resulted in a striking increase in survival compared with
either the vehicle or TMZ+ZA two-cycle groups. Eighty percent of the mice in the triple-combination group did not reach end point at the end of study. Points, mean of 10 mice per treatment group; bars, SEM.

www.aacrjournals.org 7285 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

Fig. 5. Efficacy of ABT-888 in combination with TMZ in the intratibial breast MDA-231-Luc with development of spontaneous lung metastases. ABT-888 was administered at 25 mg/kg/d, orally, twice daily for 5 d, concurrently with TMZ at 50 mg/kg/d, orally, once daily for 5 d, for two cycles as indicated.
A, representative BLI images of the intratibial tumors and resulting spontaneous lung metastases showing the antitumor efficacy of ABT-888 combination groups (± ZA). B, graphical representation of % tumor change based on BLI of intratibial tumor and lung metastases in corresponding mice at various time points. Whereas TMZ groups (± ZA) showed regrowth of intratibial tumors, ABT-888 combination groups (± ZA) showed sustained antitumor efficacy resulting in PRs. Columns, mean of 6 to 10 mice per treatment group; bars, SEM.

alkylation repair (37) leading to a decrease or depletion of MGMT after treatment may be a better indicator for the role of MGMT in the repair process, hence resistance to treatment. Our results show that although there were changes in MGMT levels, they were minimal; that is, a reduction in MGMT was observed in the MGMT+ xenografts after ABT-888+TMZ but did not correlate with the enhanced sensitivity of these lines to ABT-888+TMZ combination treatment (DOHH-2, NCI- H526, Capan-1, and Calu-6). Interestingly, although Calu-6 tumors did not express MGMT initially, significant MGMT was observed after the second cycle (Fig. 6C). However, tu- mors of different sizes reflecting treatment effect after the sec- ond cycle (tumor size: TMZ, two cycles > TMZ then ABT-888
+TMZ > ABT-888+TMZ, two cycles) show consistent increased levels of MGMT regardless of treatment, further indicating the lack of correlation of MGMT with sensitivity. Furthermore, the tumors effectively treated with two cycles of ABT-888+TMZ showed differential MGMT levels. Nevertheless, MGMT levels observed after treatment did not correlate with the in vivo phe- notype, lending support to MGMT-independent mechanisms of resistance.

Discussion

Ourresultsshowthe potent antitumorefficacy ofABT-888+TMZ across a spectrum of histologic types, where combination

Clin Cancer Res 2009;15(23) December 1, 2009 7286 www.aacrjournals.org

ABT-888+TMZ Therapy Efficacy in Preclinical Models

efficacy was independent of TMZ activity and can overcome both inherent and acquired TMZ resistance. Although resis- tance to TMZ may be mediated by MGMT, our results suggest mechanisms independent of MGMT. These data support parallel clinical investigation of ABT-888+TMZ combination therapy in other solid tumors, in addition to glioma and melanoma.
Although the alkylating agent TMZ has a narrow approval for recurrent anaplastic astrocytoma and newly diagnosed glioblas- toma (14), it has shown activity in tumors highly resistant to chemotherapy with poor clinical outcome, thus providing po-
tential significant clinical benefit. However, TMZ efficacy can be hampered by acquired resistance after repeated treatment (15). Therefore, expanding the clinical indications for TMZ as well as overcoming resistance factors is an attractive strategy for ex- ploiting PARP inhibition. We have previously shown the con- version of single-strand breaks induced by TMZ to more lethal DSBs after PARP inhibition mediated by ABT-888 (8). Thus, the enhancement of TMZ activity by ABT-888+TMZ was not surpris- ing in TMZ-sensitive models. Interestingly, ABT-888+TMZ was significantly efficacious in most of the models examined, in- cluding those that were initially resistant or became resistant

Fig. 6. PAR and MGMT expression in xenografts before and after TMZ or ABT-888+TMZ treatment in vivo. A, Western blot analysis showing significant inhibition of PAR in representative xenograft tumors (NCI-H526, DOHH-2, and Capan-1) after treatment with ABT-888 alone (25 mg/kg/d, orally, twice
daily for 3-5 d) or in combination with TMZ (50 mg/kg/d, orally, once daily for 5 d). Tumors were harvested at 2 h after last dose. B, Western blot analysis of MGMT in xenograft tumors after treatment in vivo. A minimal reduction in MGMT expression was observed after TMZ or ABT-888+TMZ treatments, indicating the activation of MGMT due to alkylation-induced repair. ABT-888 monotherapy did not induce a significant change in MGMT expression.
C, Western blot analysis for MGMT and PAR in Calu-6 xenograft tumors after one or two cycles of treatment in vivo. Whereas no expression of MGMT was observed after the first cycle of treatment, significant expression of MGMT was detected after the second cycle of treatment regardless of treatment effect. Significant inhibition of PAR was observed after ABT-888 treatment.

www.aacrjournals.org 7287 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

to TMZ, indicating that PARP inhibition can confer sensitivity to TMZ-refractory tumors perhaps through BER-mediated repair mechanisms (16, 17, 38). Therefore, TMZ response was not an absolute predictor of enhanced sensitivity to ABT-888+TMZ combination therapy.
Our studies also show the good tolerability of multiple cy- cles of ABT-888+TMZ, supporting its feasibility as a combina- tion therapy for the clinic. In a subset of models, evaluation at later time points was necessary because profound TMZ monotherapy activity was initially observed. In these models, the acquired resistance to TMZ monotherapy was evident even after a second treatment cycle (ovarian HeyA8; non–small cell lung carcinoma Calu-6; orthotopic or intratibial prostate; and orthotopic, brain ectopic, or intratibial breast). Strikingly, ABT-888+TMZ showed a marked yet sustained activity and often resulted in PRs after repeated dosing cycles. In two cases where the second cycle of TMZ therapy was ineffective, the crossover treatment from TMZ therapy in cycle 1 to ABT-888+TMZ in cycle 2 afforded substantially increased efficacy (Calu-6 and intratibial prostate). These results indicate potential adaptive mechanisms or selective pressures that result in acquired TMZ resistance and the ability of ABT-888 to overcome this resistance. Moreover, differential efficacy of ABT-888+TMZ treatment in different organ sites was observed and underscores the impor- tance of the microenvironment or tumor-stromal interactions resulting in epigenetic changes, drug penetration, potential ability of a subpopulation to grow/expand in vivo, and/or selection of tumor cells that can influence response to treatment (39, 40). Interestingly, discordant in vitro and in vivo sensitivity to ABT-888+TMZ was observed (Supplementary Data S6), in support of the important role of many factors mentioned above that can determine in vivo sensitivity. This is the first report of the broad activity of a PARP inhibitor + TMZ therapy in multiple histologic types in subcutaneous and challenging orthotopic or metastases models (including a spontaneous lung metastases model). The ability to overcome resistance and significant efficacy in orthotopic or metastatic sites may have clinical implications for metastatic disease that is often found to be chemoresistant (31).
The ABT-888 combination response in both the intratibial prostate and breast tumors showed profound combination effi- cacy that was compatible with bisphosphonate bone-sparing therapy. Although ZA improved bone density, it did not affect tu- mor growth at the doses used. On the other hand, ABT-888+TMZ exhibited an effect on bone integrity as determined by X-ray analysis and the alleviation of bone decalcification is likely a secondary effect due to the decrease in tumor burden after treatment.
Although identification of factors contributing to combina- tion sensitivity is beyond the scope of this article, we deter- mined if known factors implicated in TMZ resistance correlated with ABT-888 combination sensitivity. The potent in- hibition of tumor PARP activity of ABT-888 shown by signifi- cant inhibition of PAR in all xenografts coincided with combination sensitivity but did not reflect the degree of sensi- tivity, pointing to factors other than tumor drug levels. Several mechanisms for TMZ resistance have been proposed that are likely to influence ABT-888 combination sensitivity. Resistance to alkylating agents such as TMZ may be mediated by DNA re- pair enzymes such as MGMT (repair of O6-MeG), BER (repair of N3-MeA and N7-MeG), and MMR mutations (tolerance to O6-
MeG; refs. 16–19, 41). MGMT, MMR, and, recently, homolo- gous-directed recombination/homologous recombination– related (BRCA2 and XRCC1) mechanisms have the ability to repair or induce tolerance to O6-MeG (42). In fact, MGMT ex- pression or activity and MMR proficiency have been shown by several studies to correlate with clinical outcome in primary central nervous system tumors, including glioblastoma (28, 41, 43), and the loss of MGMT activity by expression or meth- ylation also correlated with response to alkylation-based thera- py (44, 45). However, our preliminary evaluation yielded no clear correlation of MGMT expression with response to treat- ment. In addition, O6-benzylguanine treatment, an inhibitor of MGMT, was unable to sensitize cells to TMZ in vitro, further supporting mechanisms independent of MGMT (data not shown). These findings are consistent with a recent study show- ing that MGMT activity and MMR proficiency do not preclude enhancement of TMZ activity by ABT-888 in leukemia primary and cell lines (46). MGMT, being a suicide enzyme depleted on activation, was minimally decreased in all xenografts after treat- ment, indicating that MGMT is active in MGMT+ lines, although not sufficient to render resistance to therapy. These results are in contrast to recent studies on primary human glioma xenografts where MGMT+ lines exhibited resistance to ABT-888+TMZ com- bination therapy (38). The underlying reasons for differential sensitivity to ABT-888+TMZ in MGMT+ lines are not clear but may reflect tumor type–dependent responses. Nevertheless, our results clearly show potent combination activity and that TMZ resistance (inherent or acquired) as well as MGMT+ xenografts do not preclude sensitivity to ABT-888+TMZ combination therapy.
A role for PARP in the regulation of the DNA methyltransfer- ase, DNMT1, and its influence in overall methylation levels have been shown (4), but no significant changes in MGMT after ABT-888 treatment alone were observed, coincident with no known ability of ABT-888 to affect methylation. Regardless of MGMT levels, MMR (MSH2 and MSH6) can play a role in O6- MeG tolerance (47). Although the role of MMR in our system is not yet fully understood, none of the xenografts assessed in our studies are MMR deficient1 (48), a profile commonly found in colon lines (Supplementary Data S7; ref. 49). Another factor, p53, known to suppress MGMT activity predicting TMZ sensi- tivity (14, 50), also showed no correlation (DOHH-2 and var- ious p53 wild-type lines exhibited varying levels of sensitivity to TMZ). Our data showed significant reduction in PAR after treat- ment, indicating potent inhibition of tumor PAR activity. There- fore, the in vivo response phenotype may be mediated by factors influenced by PARP but are independent of BER or alternatively involve potential PARP-independent BER mechanisms. Al- though the exact roles of potential factors are still unclear, mul- tiple complex mechanisms that determine the in vivo response phenotype may be involved.
We have presented in vivo data that exhibit the robust and potent ability of ABT-888 to increase and maintain TMZ efficacy in several different tumor types. Further, a TMZ-refractory response and MGMT expression do not preclude sensitivity to ABT-888+TMZ combination therapy. Although the determinants

1 Wellcome Trust Sanger Institute Cancer Genome Project Web site, http://
www.sanger.ac.uk/genetics/CGP, 2009.

Clin Cancer Res 2009;15(23) December 1, 2009 7288 www.aacrjournals.org

ABT-888+TMZ Therapy Efficacy in Preclinical Models

of resistance are not fully understood, it may be MGMT (or MMR) independent. Although the clinical indications for TMZ are focused, our results support the evaluation of ABT-888+TMZ effectiveness in multiple human cancer types outside of glioma and melanoma. These models suggest adap- tive resistance or selective pressures that result in a refractory response, often a challenge encountered in the clinic that fur- ther complicates analysis of clinical resistance mechanisms. Multiple mechanisms of resistance acquired by tumor cells in response to genotoxic stress have limited the effectiveness of chemotherapy (39, 40). Therefore, chemotherapeutic resis- tance, de novo or acquired, is a constant challenge in the clinic. Delineating mechanisms and/or determinants for resistance or
sensitivity are pivotal in developing therapeutic strategies that positively affect clinical outcome.

Disclosure of Potential Conflicts of Interest
C.K. Donawho, patent author/employee, Abbott Laboratories.

Acknowledgments

We thank Dr. Jonathan Hickson for his assistance as well as expert advice in optical imaging; Dr. Alexander Shoemaker for very helpful discus- sions in preparation of the manuscript; Dr. Kenneth Idler for BRCA sequenc- ing; and Velitchka Bontcheva-Diaz, Lance Kaleta, Phuong Le, Lenette Paige, and Lauren Smithee for their excellent technical assistance.

References

1.Penning TD, Zhu GD, Gandhi VB, et al. Discov- ery of the poly(ADP-ribose) polymerase (PARP) inhibitor 2-[(R)-2-methylpyrrolidin-2-yl]-1H-benz- imidazole-4-carboxamide (ABT-888) for the treatment of cancer. J Med Chem 2009;52: 514–23.
2.Donawho CK, Luo Y, Penning TD, et al. ABT- 888, an orally active poly(ADP-ribose) polymer- ase inhibitor that potentiates DNA-damaging agents in preclinical tumor models. Clin Cancer Res 2007;13:2728–37.
3.Jagtap P, Szabo C. Poly(ADP-ribose) polymer- ase and the therapeutic effects of its inhibitors. Nat Rev Drug Disc 2005;4:421–40.
4.Caiafa P, Guastafierro T, Zampieri M. Epige- netics: poly(ADP-ribosyl)ation of PARP-1 regu- lates genomic methylation patterns. FASEB J 2008;23:672–8.
5.de Murcia JM, Niedergang C, Trucco C, et al. Requirement of poly(ADP-ribose) polymerase in recovery from DNA damage in mice and in cells. Proc Natl Acad Sci U S A 1997;94:7303–7.
6.Bernstein C, Bernstein H, Payne CM, Garewal H. DNA repair/pro-apoptotic dual-role proteins in five major DNA repair pathways: fail-safe protection against carcinogenesis. Mutat Res 2002;511:145–78.
7.Plummer E. Inhibition of polyADP-ribose poly- merase in cancer. Curr Opin Pharmacol 2006;6: 364–8.
8.Liu X, Shi Y, Guan R, et al. Potentiation of temozolomide cytotoxicity by poly(ADP)ribose polymerase inhibitor ABT-888 requires a conversion of single-stranded DNA damages to double-stranded DNA breaks. Mol Cancer Res 2008;6:1621–9.
9.Plummer ER, Calvert AH. Targeting poly(ADP- ribose) polymerase: a two-armed strategy for cancer therapy. Clin Cancer Res 2007;13: 6252–6.
10.Ratnam K, Low J. Current development of clin- ical inhibitors of poly(ADP-ribose) polymerase in oncology. Clin Cancer Res 2007;13:1383–8.
11.Bryant H, Schultz N, Thomas H, et al. Specific killing of BRCA2-deficient tumours with inhibi- tors of poly (ADP-ribose) polymerase. Na- ture 2005;434:913–7.
12.Farmer H, McCabe N, Lord Christopher J, et al. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature 2005;434: 917–21.
13.McCabe N, Turner NC, Lord CJ, et al. Deficien- cy in the repair of DNA damage by homologous recombination and sensitivity to polyADP-ribose polymerase inhibition. Cancer Res 2006;66: 8109–15.
14.Mrugala MM, Chamberlain MC. Mechanisms
of disease: temozolomide and glioblastoma— look to the future. Nat Clin Pract Oncol 2008;5: 476–86.
15.Tentori L, Graziani G. Recent approaches to improve the antitumor efficacy of temozolomide. Curr Med Chem 2009;16:245–57.
16.Cheng CL, Johnson SP, Keir ST, et al. Poly (ADP-ribose) polymerase-1 inhibition reverses temozolomide resistance in a DNA mismatch re- pair-deficient malignant glioma xenograft. Mol Cancer Ther 2005;4:1364–8.
17.Trivedi RN, Almeida KH, Fornsaglio JL, Schamus S, Sobol RW. The role of base exci- sion repair in the sensitivity and resistance to temozolomide-mediated cell death. Cancer Res 2005;65:6394–400.
18.Liu L, Markowitz S, Gerson SL. Mismatch re- pair mutations override alkyltransferase in con- ferring resistance to temozolomide but not to 1,3-bis(2-chloroethyl)nitrosourea. Cancer Res 1996;56:5375–9.
19.Kokkinakis DM, Bocangel DB, Schold SC, Moschel RC, Pegg AE. Thresholds of O6-alkylgua- nine-DNA alkyltransferase which confer signifi- cant resistance of human glial tumor xenografts to treatment with 1,3-bis(2-chloroethyl)-1-nitro- sourea or temozolomide. Clin Cancer Res 2001; 7:421–8.
20.Silber JR, Mueller BA, Ewers TG, Berger MS. Comparison of O6-methylguanine-DNA methyl- transferase activity in brain tumors and adjacent normal brain. Cancer Res 1993;53:3416–20.
21.Stupp R, Hegi M. Methylguanine methyltrans- ferase testing in glioblastoma: when and how? J Clin Oncol 2007;25:1459–60.
22.Preusser M, Charles Janzer R, Felsberg J, et al. Anti-O6-methylguanine-methyltransferase (MGMT) immunohistochemistry in glioblasto- ma multiforme: observer variability and lack of association with patient survival impede its use as clinical biomarker. Brain Pathol 2008;18: 520–32.
23.Ranson M, Middleton MR, Bridgewater J, et al. Lomeguatrib, a potent inhibitor of O6-alkylgua- nine-DNA-alkyltransferase: phase I safety, phar- macodynamic, and pharmacokinetic trial and evaluation in combination with temozolomide in patients with advanced solid tumors. Clin Can- cer Res 2006;12:1577–84.
24.Khan O, Middleton MR. The therapeutic po- tential of O6-alkylguanine DNA alkyltransferase inhibitors. Expert Opin Investig Drugs 2007;16: 1573–84.
25.Liu L, Taverna P, Whiteacre CM, Chatterjee S, Gerson SL. Pharmacologic disruption of base excision repair sensitizes mismatch repair-defi- cient and proficient colon cancer cells to methyl- ating agents. Clin Cancer Res 1999;5:2908–17.
26.Xia F, Taghian DG, DeFrank JS, et al. Defi- ciency of human BRCA2 leads to impaired ho- mologous recombination but maintains normal nonhomologous end joining. Proc Natl Acad Sci U S A 2001;98:8644–9.
27.Bleehen NM, Newlands ES, Lee SM, et al. Can- cer Research Campaign phase II trial of temozo- lomide in metastatic melanoma. J Clin Oncol 1995;13:910–3.
28.Friedman HS, McLendon RE, Kerby T, et al. DNA mismatch repair and O6-alkylguanine- DNA alkyltransferase analysis and response to Temodal in newly diagnosed malignant glioma. J Clin Oncol 1998;16:3851–7.
29.Vera K, Djafari L, Faivre S, et al. Dose-dense regimen of temozolomide given every other week in patients with primary central nervous system tumors. Ann Oncol 2004;15:161–71.
30.Kummar S, Kinders R, Gutierrez ME, et al. Phase 0 clinical trial of the poly (ADP-ribose) po- lymerase inhibitor ABT-888 in patients with ad- vanced malignancies. J Clin Oncol 2009;27: 2705–11.
31.Chou AJ, Gorlick R. Chemotherapy resistance in osteosarcoma: current challenges and future directions. Expert Rev Anticancer Ther 2006;6: 1075–85.
32.Augustine CK, Yoo JS, Potti A, et al. Genomic and molecular profiling predicts response to te- mozolomide in melanoma. Clin Cancer Res 2009;15:502–10.
33.Coleman RE. Bisphosphonates: clinical experi- ence. Oncologist 2004;9:14–27.
34.Alix-Panabieres C, Muller V, Pantel K. Current status in human breast cancer micrometastasis. Curr Opin Oncol 2007;19:558–63.
35.Lein M, Miller K, Wirth M, et al. Bone turnover markers as predictive tools for skeletal complica- tions in men with metastatic prostate cancer treated with zoledronic acid. Prostate 2009;69: 624–32.
36.Palma JP, Rodriguez LE, Bontcheva-Diaz VD, et al. The PARP inhibitor, ABT-888 potentiates te- mozolomide: correlation with drug levels and reduction in PARP activity in vivo. Anticancer Res 2008;28:2625–35.
37.Srivenugopal KS, Mullapudi SR, Shou J, Hazra TK, Ali-Osman F. Protein phosphorylation is a regulatory mechanism for O6-alkylguanine-DNA alkyltransferase in human brain tumor cells. Cancer Res 2000;60:282–7.
38.Clarke MJ, Mulligan EA, Grogan PT, et al. Effective sensitization of temozolomide by ABT-888 is lost with development of temozolomide resistance in glioblastoma xenograft lines. Mol Cancer Ther 2009;8:407–14.
39.Dalton WS. The tumor microenvironment: fo- cus on myeloma. Cancer Treat Rev 2003;29:11–9.

www.aacrjournals.org 7289 Clin Cancer Res 2009;15(23) December 1, 2009

Cancer Therapy: Preclinical

40.Lopez-Serra L, Esteller M. Proteins that bind methylated DNA and human cancer: reading the wrong words. Br J Cancer 2008;98:1881–5.
41.Pepponi R, Marra G, Fuggetta MP, et al. The effect of O6-alkylguanine-DNA alkyltransfer- ase and mismatch repair activities on the sensitivity of human melanoma cells to temozo- lomide, 1,3-bis(2-chloroethyl)1-nitrosourea, and cisplatin. J Pharmacol Exp Ther 2003;304:661–8.
42.RoosWP, Nikolova T,Quiros S,etal. Brca2/Xrcc2 dependent HR, but not NHEJ, is required for protection against O(6)-methylguanine triggered apoptosis, DSBs and chromosomal aberrations by a process leading to SCEs. DNA Repair (Amst) 2009;8:72–86.
43.Spiro TP, Liu L, Majka S, Haaga J, Willson JK, Gerson SL. Temozolomide: the effect of once- and twice-a-day dosing on tumor tissue levels of the DNA repair protein O6-alkylgua-
nine-DNA-alkyltransferase. Clin Cancer Res 2001;7:2309–17.
44.Esteller M. Cancer epigenomics: DNA methy- lomes and histone-modification maps. Nat Rev Genet 2007;8:286–98.
45.Brandes AA, Franceschi E, Tosoni A, et al. MGMT promoter methylation status can predict the incidence and outcome of pseudoprogre‐ ssion after concomitant radiochemotherapy in newly diagnosed glioblastoma patients. J Clin Oncol 2008;26:2192–7.
46.Horton TM, Jenkins G, Pati D, et al. Poly(ADP- ribose) polymerase inhibitor ABT-888 potenti- ates the cytotoxic activity of temozolomide in leukemia cells: influence of mismatch repair sta- tus and O6-methylguanine-DNA methyltransfer- ase activity. Mol Cancer Ther 2009;8:2232–42.
47.D’Atri S, Tentori L, Lacal PM, et al. Involve- ment of the mismatch repair system in temozo-
lomide-induced apoptosis. Mol Pharmacol 1998;54:334–41.
48.Forbes SA, Bhamra G, Bamford S, et al. The Catalogue of Somatic Mutations in Cancer (COS- MIC). In: Dracopoli NCKB, Haines JL, Moir DT, Morton CC, Seidman CE, Seidman JG, Smith DR, editors. Current protocols in human genet- ics/editorial board. Somerset, NJ: John Wiley &
Sons, Inc.; 2008, p. 10.1.1–26.
49.Chen PC, Kuraguchi M, Velasquez J, et al. Novel roles for MLH3 deficiency and TLE6-like amplification in DNA mismatch repair- deficient gastrointestinal tumorigenesis and progression. PLoS Genet 2008;4:1000092.
50.Hermisson M, Klumpp A, Wick W, et al. O6- methylguanine DNA methyltransferase and p53 status predict temozolomide sensitivity in hu- man malignant glioma cells. J Neurochem 2006;96:766–76.

Clin Cancer Res 2009;15(23) December 1, 2009 7290 www.aacrjournals.org

ABT-888 Confers Broad In vivo Activity in Combination with Temozolomide in Diverse Tumors
Joann P. Palma, Yi-Chun Wang, Luis E. Rodriguez, et al.
Clin Cancer Res 2009;15:7277-7290. Published OnlineFirst November 24, 2009.

Updated version
Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-09-1245

Supplementary
Material
Access the most recent supplemental material at: http://clincancerres.aacrjournals.org/content/suppl/2009/12/02/1078-0432.CCR-09-1245.DC1
.html

Cited Articles

This article cites by 49 articles, 29 of which you can access for free at: http://clincancerres.aacrjournals.org/content/15/23/7277.full.html#ref-list-1

Citing articles

This article has been cited by 13 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/15/23/7277.full.html#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and Subscriptions
To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected].

Permissions

To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].