Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Assessing Early Therapeutic Response to Bevacizumab in Primary Breast Cancer Using Magnetic Resonance Imaging and Gene Expression Profiles

Assessing Early Therapeutic Response to Bevacizumab in Primary Breast Cancer Using Magnetic... Abstract Antiangiogenic therapy is a promising approach for the treatment of breast cancer. In practice, however, only a subset of patients who receive antiangiogenic drugs demonstrate a significant response. A key challenge, therefore, is to discover biomarkers that are predictive of response to antiangiogenic therapy. To address this issue, we have designed a window-of-opportunity study in which bevacizumab is administered as a short-term first-line treatment to primary breast cancer patients. Central to our approach is the use of a detailed pharmacodynamic assessment, consisting of pre- and post-bevacizumab multi-parametric magnetic resonance imaging scans and core biopsies for exon array gene expression analysis. Here, we illustrate three intrinsic patterns of response to bevacizumab and discuss the molecular mechanisms that may underpin each. Our results illustrate how the combination of dynamic imaging data and gene expression profiles can guide the development of biomarkers for predicting response to antiangiogenic therapy. Antiangiogenic agents have become key elements of cancer research over the last decade. Bevacizumab, a monoclonal antibody directed against vascular endothelial growth factor (VEGF), is the most widely used antiangiogenic therapy (1). Despite the early promise of bevacizumab to improve patient outcomes, the results to date have been mixed (2,3). In the case of breast cancer, bevacizumab has demonstrated improved progression-free survival in combination with cytotoxic chemotherapy but no impact on overall survival (4–8). These results have recently led an US Food and Drug Administration (FDA) advisory committee to recommend that the FDA revoke its approval of bevacizumab as a treatment for breast cancer (9). While bevacizumab has proven to be of limited benefit when used broadly in unselected breast cancer patients, there is increasing evidence that “specific subsets” of patients may show a significant response (10). However, such effects are largely masked in large randomized trials performed without any form of patient stratification. To date, there are no proven biomarkers of efficacy of anti-angiogenic therapies (11). A number of different biomarkers of response to antiangiogenic therapy have been investigated, including polymorphisms in the VEGF and VEGF receptor-2 (VEGFR2) genes (12), circulating VEGF levels (13), in situ analysis of various vascular markers (14–16), and pharmacokinetic parameter estimates from dynamic functional imaging scans (15,17). Imaging approaches, in particular dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), have demonstrated considerable promise in their ability to monitor the effects of antiangiogenic therapy (18,19). Specifically, patients with inflammatory or locally advanced breast cancer showed a statistically significant decrease in the DCE-MRI pharmacokinetic parameter Ktrans after one cycle of bevacizumab (15,20). The vast majority of antiangiogenic agents are investigated in patients with advanced cancers where multiple mechanisms of resistance may already have developed. An alternative approach is to use these drugs as short-term first-line treatments in combination with detailed pharmacodynamic assessments. To further explore this, we have conducted a window-of-opportunity study of single-agent bevacizumab before neoadjuvant chemotherapy in primary breast cancer. Materials and Methods This is an ongoing, phase II, nonrandomized, open-label investigator-led study sponsored by the Oxford Radcliffe Hospitals NHS Trust. The target population is previously untreated breast cancer patients scheduled to start neoadjuvant chemotherapy. The primary inclusion criteria are women with either histology proven locally advanced breast cancer or tumors greater than 3 cm in diameter, and who have not yet received prior therapy. The acceptance rate for the trial was 85%. Written informed consent was obtained from all patients before commencement of any study-related procedure. A single infusion of bevacizumab (15 mg/kg) was given 2 weeks before commencement of neoadjuvant chemotherapy. Multiparametric MRI scans, core biopsies, and blood samples for pharmacodynamic assessments were performed immediately before and 2 weeks after bevacizumab therapy. Patients were deemed to have successfully completed the study if they had successfully undergone both pre- and post-bevacizumab core biopsies together with the corresponding MRI scans. Our MRI scanning protocol was designed to interrogate multiple aspects of the tumor microenvironment. Specifically, during each scan, we performed diffusion-weighted imaging to assess tumor cellularity, blood oxygen level–dependent imaging to assess the oxygenation status of red blood cells in perfused regions of the tumor, and finally DCE-MRI to assess the physiological characteristics of the tumor vasculature. Patients were imaged on either a GE Signa 1.5T scanner (Churchill Hospital, Oxford, UK) or a Siemens Symphony 1.5T scanner (Mount Vernon Hospital, London, UK). For the DCE-MRI scan, we used a high temporal resolution T1-weighted acquisition (GE FAME/Siemens VIBE sequence; temporal resolution 5.4 seconds) to image 8–12 slices taken through the central tumor region (slice thickness = 5 mm, FOV = 260 mm, acquisition/reconstruction matrix = 256 × 128/256 × 256, TR/TE = 1.5/4.2 ms, flip angle = 18°, 1 × NEX). Contrast agent (Gadopentetate dimeglumine/Gd-DTPA; dose 0.1 mmol/kg body weight) was injected using a power injector (3 ml/s) at the start of the sixth dynamic volume, followed by a chasing bolus of 40 ml of saline. Immediately before the DCE-MRI scan, we performed a variable flip angle scan (flip angles of 2° and 8°, 2 × NEX) for T1 mapping (21). To analyze each DCE-MRI scan, spline-based regions of interest were first annotated on each image slice using the 2-minute post-contrast subtraction image that contained enhancing tumor. We then applied pharmacokinetic modeling techniques to quantify the volume transfer constant Ktrans, the rate constant kep, and the fractional volume of the extravascular extracellular space ve on a voxel-wise basis within each three-dimensional tumor region of interest. Specifically, we used the Tofts model with a population-based arterial input function (modified Fritz-Hansen) to model the contrast agent concentration time course at each tumor voxel (22). For gene expression analysis, we performed ultrasound-guided core biopsies and collected fresh samples in RNAlater (Applied Biosystems/Ambion, Austin, TX). Biopsy samples were left in RNAlater for 24 hours at 4°C and then removed and kept frozen at −80°C. From these frozen samples, we extracted mRNA using an in-house trizol chloroform method. After checking the RNA quality and quantity using the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE) and the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA), we analyzed the resulting samples using Affymetrix Human Exon 1.0 ST arrays. These arrays have the advantage of using multiple probe sets for different exonic regions of each gene and provide a much greater genomic coverage than traditional arrays. Results To date, we have enrolled 43 patients with locally advanced breast cancer (age range: 28–77 years, median 49 years; size range: 25–100 mm, median 49 mm). From our preliminary analysis, the data for three patients out of 24 could not be analyzed due to problems with either the MRI scans or the core biopsy samples. There has been only one serious adverse event of grade 3 hypertension that was relieved through the use of an antihypertensive drug. No other adverse events or toxicity greater than grade 1–2 have occurred. Our initial analysis of the DCE-MRI data for the first 21 patients shows three key patterns of response to bevacizumab (Figure 1). The first pattern is characterized by a decrease in Ktrans values over the extent of the tumor. The second type of response pattern shows a large amount of cell death with a dramatic reduction in vessel permeability, blood flow, and the development of a large central necrotic core. The final pattern of response shows no change in the vascular patterns over the 2 weeks of therapy. Figure 1 View largeDownload slide DCE-MRI images of representative patients showing three patterns of response to bevacizumab. [For color image please see http://jncimono.oxfordjournals.org/.] Figure 1 View largeDownload slide DCE-MRI images of representative patients showing three patterns of response to bevacizumab. [For color image please see http://jncimono.oxfordjournals.org/.] Our initial gene expression analysis for the first 21 patients in the study showed a high variability in the response to bevacizumab. We have initially focused on “pathways” with multiple genes in each pathway as a robust approach to gene expression analysis. A number of research groups, including our own, have previously used such approaches to analyze hypoxia, proliferation, and other breast cancer pathways (23–25). In particular, we have found that the expression fold changes in hypoxia, and proliferation signatures after bevacizumab ranged from a minimum of 0.6-fold decrease to a maximum of 4.3-fold increase (Figure 2). A surprising result is that there is a highly significant correlation of the induction of hypoxia, as defined by the hypoxia metagene, with the induction of proliferation (Note: there is no overlap of the genes in the hypoxia and proliferation metagene signatures; Spearman ρ = 0.81, P < .001). Figure 2 View largeDownload slide Exon array results showing the pre-/post-bevacizumab fold change (bars) and baseline expression (lines) for the proliferation (blue) and hypoxia (red) metagene scores. The results are ranked according to the fold change in expression of the proliferation metagene score (blue bars). Patient coded samples are shown on the horizontal axis. Note that the baseline metagene scores do not correlate with the pre-/post-bevacizumab fold induction in the corresponding metagenes. Figure 2 View largeDownload slide Exon array results showing the pre-/post-bevacizumab fold change (bars) and baseline expression (lines) for the proliferation (blue) and hypoxia (red) metagene scores. The results are ranked according to the fold change in expression of the proliferation metagene score (blue bars). Patient coded samples are shown on the horizontal axis. Note that the baseline metagene scores do not correlate with the pre-/post-bevacizumab fold induction in the corresponding metagenes. Discussion The development and validation of novel biomarkers for the prediction of patient response to antiangiogenic therapy is an area of increasing importance, particularly given the recent concerns surrounding the effectiveness of such therapies to prolong patient survival and their potential effects on metastasis (8,26,27). We have investigated response to bevacizumab as a short-term first-line treatment in primary breast cancer, through the use of multiparametric MRI and exon array gene expression profiling. Our initial imaging results illustrate the considerable heterogeneity in the patient responses to bevacizumab therapy. We have identified three intrinsic patterns of early response to bevacizumab, namely: 1) significant reduction in permeability and blood flow over the extent of the tumor, 2) the development of a large central necrotic core, and 3) little or no change in the tumor vasculature. In the first response group, it is possible that these patients possess tumor cells that are resistant to death from hypoxia due to bevacizumab. Analysis of the array data for this group may help to reveal the biological mechanisms governing this effect. The second response group may ultimately correspond to the subset of patients who receive the greatest benefit from bevacizumab. However, these patients may also show an induction in genes that aid tumor regrowth as a result of increased hypoxia, which can in turn activate multiple genes involved in angiogenesis and metabolism via hypoxia inducible factor-1α (28). The third group of nonresponders may result from tumor vessels that are not dependent on VEGF for growth or survival, or alternatively that are protected by pericytes, and hence bevacizumab has no effect (29). In our initial gene array results, the interesting finding of a correlation between induction of hypoxia and proliferation in the same tumors may have important implications for combination therapy. This finding was unexpected but provides a strong argument to support the need for chemotherapy (or drugs blocking proliferation) with antiangiogenesis therapy. Of equal note is the smaller group of patients where tumor proliferation and hypoxia decrease; one may postulate that this group is obtaining direct benefit of the agent alone. Although these results are preliminary and need to be confirmed at study completion, they provide support for the idea that there exists considerable heterogeneity among patient responses to bevacizumab and that only a subset of patients with primary breast cancer show a significant response to first-line antiangiogenic therapy. This knowledge may help in patient selection for antiangiogenic agents and could lead to greater cost-effectiveness of these drugs. Funding This work was supported by funding from the Breast Cancer Research Foundation [S.M.], Roche, Cancer Research UK [A.L.H.], Engineering and Physical Sciences Research Council grant EP/E042880/1 [N.P.H.], European Union Framework 7 ACGT [F.M.B.], and the Oxford NHS Biomedical Research Centre [R.F.A., N.C.L., A.L.H.]. References 1. Ignoffo RJ.  Overview of bevacizumab: a new cancer therapeutic strategy targeting vascular endothelial growth factor,  Am J Health Syst Pharm. ,  2004, vol.  61  suppl 5(pg.  S21- S26) 2. de Gramont A,  Van Cutsem E.  Investigating the potential of bevacizumab in other indications: metastatic renal cell, non-small cell lung, pancreatic and breast cancer,  Oncology. ,  2005, vol.  69  suppl 3(pg.  46- 56) 3. Ellis LM,  Rosen L,  Gordon MS.  Overview of anti-VEGF therapy and angiogenesis. Part 1: angiogenesis inhibition in solid tumor malignancies,  Clin Adv Hematol Oncol. ,  2006, vol.  4  1(pg.  1- 10)  ; quz 11-2 4. Miller KD.  E2100: a phase III trial of paclitaxel versus paclitaxel/bevacizumab for metastatic breast cancer,  Clin Breast Cancer. ,  2003, vol.  3  6(pg.  421- 422) 5. Miller KD,  Chap LI,  Holmes FA, et al.  Randomized phase III trial of capecitabine compared with bevacizumab plus capecitabine in patients with previously treated metastatic breast cancer,  J Clin Oncol. ,  2005, vol.  23  4(pg.  792- 799) 6. Miller K,  Wang M,  Gralow J, et al.  Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer,  N Engl J Med. ,  2007, vol.  357  26(pg.  2666- 2676) 7. Smith IE,  Pierga JY,  Biganzoli L, et al.  First-line bevacizumab plus taxane-based chemotherapy for locally recurrent or metastatic breast cancer: safety and efficacy in an open-label study in 2251 patients,  Ann Oncol. ,  2010, vol.  22 (pg.  595- 602) 8. O’Shaughnessy J,  Miles D,  Gray RJ, et al.  A meta-analysis of overall survival data from three randomized trials of bevacizumab (BV) and first-line chemotherapy as treatment for patients with metastatic breast cancer (MBC),  J Clin Oncol ,  2010, vol.  28   ASCO Meeting Abstract #1005 9. Pollack A.  F.D.A. panel urges limits for avastin,  New York Times.   July 21, 2010:EO1 10. Miles DW,  Chan A,  Dirix LY, et al.  Phase III study of bevacizumab plus docetaxel compared with placebo plus docetaxel for the first-line treatment of human epidermal growth factor receptor 2-negative metastatic breast cancer,  J Clin Oncol. ,  2010, vol.  28  20(pg.  3239- 3247) 11. Quesada AR,  Medina MA,  Alba E.  Playing only one instrument may be not enough: limitations and future of the antiangiogenic treatment of cancer,  Bioessays. ,  2007, vol.  29  11(pg.  1159- 1168) 12. Schneider BP,  Wang M,  Radovich M, et al.  Association of vascular endothelial growth factor and vascular endothelial growth factor receptor-2 genetic polymorphisms with outcome in a trial of paclitaxel compared with paclitaxel plus bevacizumab in advanced breast cancer: ECOG 2100,  J Clin Oncol. ,  2008, vol.  26  28(pg.  4672- 4678) 13. Denduluri N,  Yang SX,  Berman AW, et al.  Circulating biomarkers of bevacizumab activity in patients with breast cancer,  Cancer Biol Ther. ,  2008, vol.  7 (pg.  15- 20) 14. Baar J,  Silverman P,  Lyons J, et al.  A vasculature-targeting regimen of preoperative docetaxel with or without bevacizumab for locally advanced breast cancer: impact on angiogenic biomarkers,  Clin Cancer Res. ,  2009, vol.  15  10(pg.  3583- 3590) 15. Wedam SB,  Low JA,  Yang SX, et al.  Antiangiogenic and antitumor effects of bevacizumab in patients with inflammatory and locally advanced breast cancer,  J Clin Oncol. ,  2006, vol.  24  5(pg.  769- 777) 16. Yang SX,  Steinberg SM,  Nguyen D, et al.  Gene expression profile and angiogenic marker correlates with response to neoadjuvant bevacizumab followed by bevacizumab plus chemotherapy in breast cancer,  Clin Cancer Res. ,  2008, vol.  14  18(pg.  5893- 5899) 17. Jackson A,  O’Connor JP,  Parker GJ,  Jayson GC.  Imaging tumor vascular heterogeneity and angiogenesis using dynamic contrast-enhanced magnetic resonance imaging,  Clin Cancer Res. ,  2007, vol.  13  12(pg.  3449- 3459) 18. O’Connor JP,  Jackson A,  Parker GJ,  Jayson GC.  DCE-MRI biomarkers in the clinical evaluation of antiangiogenic and vascular disrupting agents,  Br J Cancer. ,  2007, vol.  96  2(pg.  189- 195) 19. Flaherty KT,  Rosen MA,  Heitjan DF, et al.  Pilot study of DCE-MRI to predict progression-free survival with sorafenib therapy in renal cell carcinoma,  Cancer Biol Ther. ,  2008, vol.  7  4(pg.  496- 501) 20. Thukral A,  Thomasson DM,  Chow CK, et al.  Inflammatory breast cancer: dynamic contrast-enhanced MR in patients receiving bevacizumab—initial experience,  Radiology. ,  2007, vol.  244  3(pg.  727- 735) 21. Buckley DL,  Parker GJM.  Jackson A,  Buckley DL,  Parker GJM.  Measuring contrast agent concentration in T1-weighted dynamic contrast-enhanced MRI,  Dynamic Contrast-Enhanced Magnetic Resonance Imaging in Oncology ,  2005 Berlin, Germany Springer(pg.  69- 79) 22. Tofts PS,  Brix G,  Buckley DL, et al.  Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities and symbols,  J Magn Reson Imaging. ,  1999, vol.  10  3(pg.  223- 232) 23. Winter SC,  Buffa FM,  Silva P, et al.  Relation of a hypoxia metagene derived from head and neck cancer to prognosis of multiple cancers,  Cancer Res. ,  2007, vol.  67  7(pg.  3441- 3449) 24. Buffa FM,  Harris AL,  West CM,  Miller CJ.  Large meta-analysis of multiple cancers reveals a common, compact and highly prognostic hypoxia metagene,  Br J Cancer. ,  2010, vol.  102  2(pg.  428- 435) 25. Desmedt C,  Haibe-Kains B,  Wirapati P, et al.  Biological processes associated with breast cancer clinical outcome depend on the molecular subtypes,  Clin Cancer Res. ,  2008, vol.  14  16(pg.  5158- 5165) 26. Ebos JM,  Lee CR,  Cruz-Munoz W, et al.  Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis,  Cancer Cell. ,  2009, vol.  15  3(pg.  232- 239) 27. Paez-Ribes M,  Allen E,  Hudock J, et al.  Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis,  Cancer Cell. ,  2009, vol.  15  3(pg.  220- 231) 28. Harris AL.  Hypoxia—a key regulatory factor in tumour growth,  Nat Rev Cancer. ,  2002, vol.  2  1(pg.  38- 47) 29. Azam F,  Mehta S,  Harris AL.  Mechanisms of resistance to antiangiogenesis therapy,  Eur J Cancer. ,  2010, vol.  46  8(pg.  1323- 1332) Consent and Approval: Informed consent was obtained from all patients before any study-related procedure after approval by the local institutional review board, National Research Ethics Services, UK. S. Mehta and N. P. Hughes contributed equally to this work. In addition, we thank many people who have contributed to the running of this study, including L. Hamilton (Roche), M. Greenhall, J. Clarke, H. Sheldon, A. McIntyre, J. Stirling, I. Simcock, V. Parulekar, D. Roskell, B. Kessler, C. Wright, J. Smythe, L. McRae, N. Fisher, S. Watt, R. Leek, G. Steers, H. Sheldon, K. Kaur, M. Taylor, S. Wellman, and L. War. © The Author 2011. Published by Oxford University Press. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI Monographs Oxford University Press

Assessing Early Therapeutic Response to Bevacizumab in Primary Breast Cancer Using Magnetic Resonance Imaging and Gene Expression Profiles

Loading next page...
 
/lp/oxford-university-press/assessing-early-therapeutic-response-to-bevacizumab-in-primary-breast-qDU0SS0051

References (34)

Publisher
Oxford University Press
Copyright
© The Author 2011. Published by Oxford University Press.
ISSN
1052-6773
eISSN
1745-6614
DOI
10.1093/jncimonographs/lgr027
pmid
22043045
Publisher site
See Article on Publisher Site

Abstract

Abstract Antiangiogenic therapy is a promising approach for the treatment of breast cancer. In practice, however, only a subset of patients who receive antiangiogenic drugs demonstrate a significant response. A key challenge, therefore, is to discover biomarkers that are predictive of response to antiangiogenic therapy. To address this issue, we have designed a window-of-opportunity study in which bevacizumab is administered as a short-term first-line treatment to primary breast cancer patients. Central to our approach is the use of a detailed pharmacodynamic assessment, consisting of pre- and post-bevacizumab multi-parametric magnetic resonance imaging scans and core biopsies for exon array gene expression analysis. Here, we illustrate three intrinsic patterns of response to bevacizumab and discuss the molecular mechanisms that may underpin each. Our results illustrate how the combination of dynamic imaging data and gene expression profiles can guide the development of biomarkers for predicting response to antiangiogenic therapy. Antiangiogenic agents have become key elements of cancer research over the last decade. Bevacizumab, a monoclonal antibody directed against vascular endothelial growth factor (VEGF), is the most widely used antiangiogenic therapy (1). Despite the early promise of bevacizumab to improve patient outcomes, the results to date have been mixed (2,3). In the case of breast cancer, bevacizumab has demonstrated improved progression-free survival in combination with cytotoxic chemotherapy but no impact on overall survival (4–8). These results have recently led an US Food and Drug Administration (FDA) advisory committee to recommend that the FDA revoke its approval of bevacizumab as a treatment for breast cancer (9). While bevacizumab has proven to be of limited benefit when used broadly in unselected breast cancer patients, there is increasing evidence that “specific subsets” of patients may show a significant response (10). However, such effects are largely masked in large randomized trials performed without any form of patient stratification. To date, there are no proven biomarkers of efficacy of anti-angiogenic therapies (11). A number of different biomarkers of response to antiangiogenic therapy have been investigated, including polymorphisms in the VEGF and VEGF receptor-2 (VEGFR2) genes (12), circulating VEGF levels (13), in situ analysis of various vascular markers (14–16), and pharmacokinetic parameter estimates from dynamic functional imaging scans (15,17). Imaging approaches, in particular dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), have demonstrated considerable promise in their ability to monitor the effects of antiangiogenic therapy (18,19). Specifically, patients with inflammatory or locally advanced breast cancer showed a statistically significant decrease in the DCE-MRI pharmacokinetic parameter Ktrans after one cycle of bevacizumab (15,20). The vast majority of antiangiogenic agents are investigated in patients with advanced cancers where multiple mechanisms of resistance may already have developed. An alternative approach is to use these drugs as short-term first-line treatments in combination with detailed pharmacodynamic assessments. To further explore this, we have conducted a window-of-opportunity study of single-agent bevacizumab before neoadjuvant chemotherapy in primary breast cancer. Materials and Methods This is an ongoing, phase II, nonrandomized, open-label investigator-led study sponsored by the Oxford Radcliffe Hospitals NHS Trust. The target population is previously untreated breast cancer patients scheduled to start neoadjuvant chemotherapy. The primary inclusion criteria are women with either histology proven locally advanced breast cancer or tumors greater than 3 cm in diameter, and who have not yet received prior therapy. The acceptance rate for the trial was 85%. Written informed consent was obtained from all patients before commencement of any study-related procedure. A single infusion of bevacizumab (15 mg/kg) was given 2 weeks before commencement of neoadjuvant chemotherapy. Multiparametric MRI scans, core biopsies, and blood samples for pharmacodynamic assessments were performed immediately before and 2 weeks after bevacizumab therapy. Patients were deemed to have successfully completed the study if they had successfully undergone both pre- and post-bevacizumab core biopsies together with the corresponding MRI scans. Our MRI scanning protocol was designed to interrogate multiple aspects of the tumor microenvironment. Specifically, during each scan, we performed diffusion-weighted imaging to assess tumor cellularity, blood oxygen level–dependent imaging to assess the oxygenation status of red blood cells in perfused regions of the tumor, and finally DCE-MRI to assess the physiological characteristics of the tumor vasculature. Patients were imaged on either a GE Signa 1.5T scanner (Churchill Hospital, Oxford, UK) or a Siemens Symphony 1.5T scanner (Mount Vernon Hospital, London, UK). For the DCE-MRI scan, we used a high temporal resolution T1-weighted acquisition (GE FAME/Siemens VIBE sequence; temporal resolution 5.4 seconds) to image 8–12 slices taken through the central tumor region (slice thickness = 5 mm, FOV = 260 mm, acquisition/reconstruction matrix = 256 × 128/256 × 256, TR/TE = 1.5/4.2 ms, flip angle = 18°, 1 × NEX). Contrast agent (Gadopentetate dimeglumine/Gd-DTPA; dose 0.1 mmol/kg body weight) was injected using a power injector (3 ml/s) at the start of the sixth dynamic volume, followed by a chasing bolus of 40 ml of saline. Immediately before the DCE-MRI scan, we performed a variable flip angle scan (flip angles of 2° and 8°, 2 × NEX) for T1 mapping (21). To analyze each DCE-MRI scan, spline-based regions of interest were first annotated on each image slice using the 2-minute post-contrast subtraction image that contained enhancing tumor. We then applied pharmacokinetic modeling techniques to quantify the volume transfer constant Ktrans, the rate constant kep, and the fractional volume of the extravascular extracellular space ve on a voxel-wise basis within each three-dimensional tumor region of interest. Specifically, we used the Tofts model with a population-based arterial input function (modified Fritz-Hansen) to model the contrast agent concentration time course at each tumor voxel (22). For gene expression analysis, we performed ultrasound-guided core biopsies and collected fresh samples in RNAlater (Applied Biosystems/Ambion, Austin, TX). Biopsy samples were left in RNAlater for 24 hours at 4°C and then removed and kept frozen at −80°C. From these frozen samples, we extracted mRNA using an in-house trizol chloroform method. After checking the RNA quality and quantity using the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE) and the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA), we analyzed the resulting samples using Affymetrix Human Exon 1.0 ST arrays. These arrays have the advantage of using multiple probe sets for different exonic regions of each gene and provide a much greater genomic coverage than traditional arrays. Results To date, we have enrolled 43 patients with locally advanced breast cancer (age range: 28–77 years, median 49 years; size range: 25–100 mm, median 49 mm). From our preliminary analysis, the data for three patients out of 24 could not be analyzed due to problems with either the MRI scans or the core biopsy samples. There has been only one serious adverse event of grade 3 hypertension that was relieved through the use of an antihypertensive drug. No other adverse events or toxicity greater than grade 1–2 have occurred. Our initial analysis of the DCE-MRI data for the first 21 patients shows three key patterns of response to bevacizumab (Figure 1). The first pattern is characterized by a decrease in Ktrans values over the extent of the tumor. The second type of response pattern shows a large amount of cell death with a dramatic reduction in vessel permeability, blood flow, and the development of a large central necrotic core. The final pattern of response shows no change in the vascular patterns over the 2 weeks of therapy. Figure 1 View largeDownload slide DCE-MRI images of representative patients showing three patterns of response to bevacizumab. [For color image please see http://jncimono.oxfordjournals.org/.] Figure 1 View largeDownload slide DCE-MRI images of representative patients showing three patterns of response to bevacizumab. [For color image please see http://jncimono.oxfordjournals.org/.] Our initial gene expression analysis for the first 21 patients in the study showed a high variability in the response to bevacizumab. We have initially focused on “pathways” with multiple genes in each pathway as a robust approach to gene expression analysis. A number of research groups, including our own, have previously used such approaches to analyze hypoxia, proliferation, and other breast cancer pathways (23–25). In particular, we have found that the expression fold changes in hypoxia, and proliferation signatures after bevacizumab ranged from a minimum of 0.6-fold decrease to a maximum of 4.3-fold increase (Figure 2). A surprising result is that there is a highly significant correlation of the induction of hypoxia, as defined by the hypoxia metagene, with the induction of proliferation (Note: there is no overlap of the genes in the hypoxia and proliferation metagene signatures; Spearman ρ = 0.81, P < .001). Figure 2 View largeDownload slide Exon array results showing the pre-/post-bevacizumab fold change (bars) and baseline expression (lines) for the proliferation (blue) and hypoxia (red) metagene scores. The results are ranked according to the fold change in expression of the proliferation metagene score (blue bars). Patient coded samples are shown on the horizontal axis. Note that the baseline metagene scores do not correlate with the pre-/post-bevacizumab fold induction in the corresponding metagenes. Figure 2 View largeDownload slide Exon array results showing the pre-/post-bevacizumab fold change (bars) and baseline expression (lines) for the proliferation (blue) and hypoxia (red) metagene scores. The results are ranked according to the fold change in expression of the proliferation metagene score (blue bars). Patient coded samples are shown on the horizontal axis. Note that the baseline metagene scores do not correlate with the pre-/post-bevacizumab fold induction in the corresponding metagenes. Discussion The development and validation of novel biomarkers for the prediction of patient response to antiangiogenic therapy is an area of increasing importance, particularly given the recent concerns surrounding the effectiveness of such therapies to prolong patient survival and their potential effects on metastasis (8,26,27). We have investigated response to bevacizumab as a short-term first-line treatment in primary breast cancer, through the use of multiparametric MRI and exon array gene expression profiling. Our initial imaging results illustrate the considerable heterogeneity in the patient responses to bevacizumab therapy. We have identified three intrinsic patterns of early response to bevacizumab, namely: 1) significant reduction in permeability and blood flow over the extent of the tumor, 2) the development of a large central necrotic core, and 3) little or no change in the tumor vasculature. In the first response group, it is possible that these patients possess tumor cells that are resistant to death from hypoxia due to bevacizumab. Analysis of the array data for this group may help to reveal the biological mechanisms governing this effect. The second response group may ultimately correspond to the subset of patients who receive the greatest benefit from bevacizumab. However, these patients may also show an induction in genes that aid tumor regrowth as a result of increased hypoxia, which can in turn activate multiple genes involved in angiogenesis and metabolism via hypoxia inducible factor-1α (28). The third group of nonresponders may result from tumor vessels that are not dependent on VEGF for growth or survival, or alternatively that are protected by pericytes, and hence bevacizumab has no effect (29). In our initial gene array results, the interesting finding of a correlation between induction of hypoxia and proliferation in the same tumors may have important implications for combination therapy. This finding was unexpected but provides a strong argument to support the need for chemotherapy (or drugs blocking proliferation) with antiangiogenesis therapy. Of equal note is the smaller group of patients where tumor proliferation and hypoxia decrease; one may postulate that this group is obtaining direct benefit of the agent alone. Although these results are preliminary and need to be confirmed at study completion, they provide support for the idea that there exists considerable heterogeneity among patient responses to bevacizumab and that only a subset of patients with primary breast cancer show a significant response to first-line antiangiogenic therapy. This knowledge may help in patient selection for antiangiogenic agents and could lead to greater cost-effectiveness of these drugs. Funding This work was supported by funding from the Breast Cancer Research Foundation [S.M.], Roche, Cancer Research UK [A.L.H.], Engineering and Physical Sciences Research Council grant EP/E042880/1 [N.P.H.], European Union Framework 7 ACGT [F.M.B.], and the Oxford NHS Biomedical Research Centre [R.F.A., N.C.L., A.L.H.]. References 1. Ignoffo RJ.  Overview of bevacizumab: a new cancer therapeutic strategy targeting vascular endothelial growth factor,  Am J Health Syst Pharm. ,  2004, vol.  61  suppl 5(pg.  S21- S26) 2. de Gramont A,  Van Cutsem E.  Investigating the potential of bevacizumab in other indications: metastatic renal cell, non-small cell lung, pancreatic and breast cancer,  Oncology. ,  2005, vol.  69  suppl 3(pg.  46- 56) 3. Ellis LM,  Rosen L,  Gordon MS.  Overview of anti-VEGF therapy and angiogenesis. Part 1: angiogenesis inhibition in solid tumor malignancies,  Clin Adv Hematol Oncol. ,  2006, vol.  4  1(pg.  1- 10)  ; quz 11-2 4. Miller KD.  E2100: a phase III trial of paclitaxel versus paclitaxel/bevacizumab for metastatic breast cancer,  Clin Breast Cancer. ,  2003, vol.  3  6(pg.  421- 422) 5. Miller KD,  Chap LI,  Holmes FA, et al.  Randomized phase III trial of capecitabine compared with bevacizumab plus capecitabine in patients with previously treated metastatic breast cancer,  J Clin Oncol. ,  2005, vol.  23  4(pg.  792- 799) 6. Miller K,  Wang M,  Gralow J, et al.  Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer,  N Engl J Med. ,  2007, vol.  357  26(pg.  2666- 2676) 7. Smith IE,  Pierga JY,  Biganzoli L, et al.  First-line bevacizumab plus taxane-based chemotherapy for locally recurrent or metastatic breast cancer: safety and efficacy in an open-label study in 2251 patients,  Ann Oncol. ,  2010, vol.  22 (pg.  595- 602) 8. O’Shaughnessy J,  Miles D,  Gray RJ, et al.  A meta-analysis of overall survival data from three randomized trials of bevacizumab (BV) and first-line chemotherapy as treatment for patients with metastatic breast cancer (MBC),  J Clin Oncol ,  2010, vol.  28   ASCO Meeting Abstract #1005 9. Pollack A.  F.D.A. panel urges limits for avastin,  New York Times.   July 21, 2010:EO1 10. Miles DW,  Chan A,  Dirix LY, et al.  Phase III study of bevacizumab plus docetaxel compared with placebo plus docetaxel for the first-line treatment of human epidermal growth factor receptor 2-negative metastatic breast cancer,  J Clin Oncol. ,  2010, vol.  28  20(pg.  3239- 3247) 11. Quesada AR,  Medina MA,  Alba E.  Playing only one instrument may be not enough: limitations and future of the antiangiogenic treatment of cancer,  Bioessays. ,  2007, vol.  29  11(pg.  1159- 1168) 12. Schneider BP,  Wang M,  Radovich M, et al.  Association of vascular endothelial growth factor and vascular endothelial growth factor receptor-2 genetic polymorphisms with outcome in a trial of paclitaxel compared with paclitaxel plus bevacizumab in advanced breast cancer: ECOG 2100,  J Clin Oncol. ,  2008, vol.  26  28(pg.  4672- 4678) 13. Denduluri N,  Yang SX,  Berman AW, et al.  Circulating biomarkers of bevacizumab activity in patients with breast cancer,  Cancer Biol Ther. ,  2008, vol.  7 (pg.  15- 20) 14. Baar J,  Silverman P,  Lyons J, et al.  A vasculature-targeting regimen of preoperative docetaxel with or without bevacizumab for locally advanced breast cancer: impact on angiogenic biomarkers,  Clin Cancer Res. ,  2009, vol.  15  10(pg.  3583- 3590) 15. Wedam SB,  Low JA,  Yang SX, et al.  Antiangiogenic and antitumor effects of bevacizumab in patients with inflammatory and locally advanced breast cancer,  J Clin Oncol. ,  2006, vol.  24  5(pg.  769- 777) 16. Yang SX,  Steinberg SM,  Nguyen D, et al.  Gene expression profile and angiogenic marker correlates with response to neoadjuvant bevacizumab followed by bevacizumab plus chemotherapy in breast cancer,  Clin Cancer Res. ,  2008, vol.  14  18(pg.  5893- 5899) 17. Jackson A,  O’Connor JP,  Parker GJ,  Jayson GC.  Imaging tumor vascular heterogeneity and angiogenesis using dynamic contrast-enhanced magnetic resonance imaging,  Clin Cancer Res. ,  2007, vol.  13  12(pg.  3449- 3459) 18. O’Connor JP,  Jackson A,  Parker GJ,  Jayson GC.  DCE-MRI biomarkers in the clinical evaluation of antiangiogenic and vascular disrupting agents,  Br J Cancer. ,  2007, vol.  96  2(pg.  189- 195) 19. Flaherty KT,  Rosen MA,  Heitjan DF, et al.  Pilot study of DCE-MRI to predict progression-free survival with sorafenib therapy in renal cell carcinoma,  Cancer Biol Ther. ,  2008, vol.  7  4(pg.  496- 501) 20. Thukral A,  Thomasson DM,  Chow CK, et al.  Inflammatory breast cancer: dynamic contrast-enhanced MR in patients receiving bevacizumab—initial experience,  Radiology. ,  2007, vol.  244  3(pg.  727- 735) 21. Buckley DL,  Parker GJM.  Jackson A,  Buckley DL,  Parker GJM.  Measuring contrast agent concentration in T1-weighted dynamic contrast-enhanced MRI,  Dynamic Contrast-Enhanced Magnetic Resonance Imaging in Oncology ,  2005 Berlin, Germany Springer(pg.  69- 79) 22. Tofts PS,  Brix G,  Buckley DL, et al.  Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities and symbols,  J Magn Reson Imaging. ,  1999, vol.  10  3(pg.  223- 232) 23. Winter SC,  Buffa FM,  Silva P, et al.  Relation of a hypoxia metagene derived from head and neck cancer to prognosis of multiple cancers,  Cancer Res. ,  2007, vol.  67  7(pg.  3441- 3449) 24. Buffa FM,  Harris AL,  West CM,  Miller CJ.  Large meta-analysis of multiple cancers reveals a common, compact and highly prognostic hypoxia metagene,  Br J Cancer. ,  2010, vol.  102  2(pg.  428- 435) 25. Desmedt C,  Haibe-Kains B,  Wirapati P, et al.  Biological processes associated with breast cancer clinical outcome depend on the molecular subtypes,  Clin Cancer Res. ,  2008, vol.  14  16(pg.  5158- 5165) 26. Ebos JM,  Lee CR,  Cruz-Munoz W, et al.  Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis,  Cancer Cell. ,  2009, vol.  15  3(pg.  232- 239) 27. Paez-Ribes M,  Allen E,  Hudock J, et al.  Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis,  Cancer Cell. ,  2009, vol.  15  3(pg.  220- 231) 28. Harris AL.  Hypoxia—a key regulatory factor in tumour growth,  Nat Rev Cancer. ,  2002, vol.  2  1(pg.  38- 47) 29. Azam F,  Mehta S,  Harris AL.  Mechanisms of resistance to antiangiogenesis therapy,  Eur J Cancer. ,  2010, vol.  46  8(pg.  1323- 1332) Consent and Approval: Informed consent was obtained from all patients before any study-related procedure after approval by the local institutional review board, National Research Ethics Services, UK. S. Mehta and N. P. Hughes contributed equally to this work. In addition, we thank many people who have contributed to the running of this study, including L. Hamilton (Roche), M. Greenhall, J. Clarke, H. Sheldon, A. McIntyre, J. Stirling, I. Simcock, V. Parulekar, D. Roskell, B. Kessler, C. Wright, J. Smythe, L. McRae, N. Fisher, S. Watt, R. Leek, G. Steers, H. Sheldon, K. Kaur, M. Taylor, S. Wellman, and L. War. © The Author 2011. Published by Oxford University Press.

Journal

JNCI MonographsOxford University Press

Published: Oct 1, 2011

There are no references for this article.