Neuro-oNcology
`
`VEGF Trap induces antiglioma effect
`at different stages of disease
`
`Candelaria Gomez-Manzano, Jocelyn Holash, Juan Fueyo, Jing Xu,
`Charles A. Conrad, Kenneth D. Aldape, John F. de Groot, B. Nebiyou Bekele,
`and W. K. Alfred Yung
`Departments of Neuro-Oncology (C.G.-M., J.F., J.X., C.A.C., J.F.G., W.K.A.Y.), Pathology (K.D.A.), and
`Biostatistics and Applied Mathematics (B.N.B.), The University of Texas M. D. Anderson Cancer Center,
`Houston, TX; Regeneron Pharmaceuticals, Inc., Tarrytown, NY (J.H.); USA
`
`Pathological angiogenesis is a hallmark of cancer, specif-
`ically of glioblastomas, the most malignant and common
`primary brain tumor. Vascular endothelial growth factor
`(VEGF) is the key protein in the regulation of the hyper-
`vascular phenotype of primary malignant brain tumors.
`In this study, we tested VEGF Trap, a soluble decoy recep-
`tor for VEGF, in an intracranial glioma model. VEGF
`Trap was administered in short or prolonged schedules
`to animals bearing human gliomas at different stages
`of disease. Of importance, VEGF Trap treatment was
`efficacious in both initial and advanced phases of tumor
`development by significantly increasing overall survival.
`Furthermore, this effect was enhanced in animals treated
`with more prolonged regimens. In addition, we observed
`the emergence of a VEGF Trap-resistant phenotype char-
`acterized by tumor growth and increased invasiveness.
`Our results suggest that VEGF Trap will be effective
`in treating both patients with recurrent or progressive
`resectable glioblastoma and patients that have under-
`gone extensive initial surgery. Finally, our results indi-
`cate that the clinical success of VEGF Trap may depend
`on a prolonged treatment in combined therapy aiming to
`simultaneously inhibit angiogenesis and tumor invasion.
`Neuro-Oncology 10, 940–945, 2008 (Posted to Neuro-
`Oncology [serial online], Doc. D08-00085, August
`14, 2008. URL http://neuro-oncology.dukejournals.org;
`DOI: 10.1215/15228517-2008-061)
`
`Received April 7, 2008; accepted June 4, 2008.
`
`Address correspondence to C. Gomez-Manzano, Department of
`Neuro-Oncology, Unit 1002, The University of Texas M. D. Anderson
`Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA
`(cmanzano@mdanderson.org).
`
`940Neuro-Oncology■JULY 2006
`Copyright 2008 by the Society for Neuro-Oncology
`
`Keywords: glioblastoma, therapy, VEGF, VEGF Trap
`
`The striking induction of angiogenesis in glioblas-
`
`toma multiforme (GBM) has fueled the speculation
`that progression to GBM requires the activation
`of angiogenesis, a finding that has stimulated significant
`efforts to develop angiogenesis-blocking agents. Vas-
`cular endothelial growth factor (VEGF) is critical for
`promoting the earliest stages of vasculogenesis, which
`includes endothelial cell proliferation, differentiation,
`migration, and tubular formation. Clinical trials of spe-
`cific VEGF inhibitors for the treatment of patients with
`gliomas are ongoing, and preliminary analy ses showed
`beneficial effects in patients with malignant gliomas.1–4
`Recently, a new anti-VEGF agent, VEGF Trap/aflibercept
`(henceforth referred to as VEGF Trap), has been devel-
`oped by incorporating domains of both VEGF recep-
`tor 1 (VEGFR-1) and VEGFR-2 fused to the constant
`region of human immunoglobulin G1, which acts as a
`soluble decoy receptor for VEGF. VEGF Trap has very
`high affinity for all isoforms of VEGF-A (,1 pM), as
`well as placental growth factor, a closely related angio-
`genic factor.5 VEGF Trap was engineered to have mini-
`mal interactions with the extracellular matrix, and this
`property apparently accounts for its satisfying pharma-
`cokinetic profile superior to soluble forms of VEGFR-1.5
`Its efficacy has been proven in preclinical studies in
`several types of solid tumor5–9 and in a subcutaneous
`glioma model.10 Because tumor progression and angio-
`genesis are greatly dependent on the existent micro-
`environment of the tumor,11,12 we undertook this study
`to characterize the effect of VEGF Trap in an orthotopic
`glioblastoma model in several stages of the disease. We
`
`Celltrion Exhibit - 1029
`Celltrion, Inc. v. Regeneron Pharmaceuticals, Inc.
`
`

`

`have previously described the development of growth
`patterns and angiogenesis in an intracranial U-87 MG
`human glioma model. Vessel cooption and remodel-
`ing were present at the early stages of disease, whereas
`the advanced stages are distinguished by high vascu-
`lar density.13 These two phases were similar to stages
`described in other previous reports.14,15 Based on this
`tumoral angiogenesis and kinetic pattern, we adminis-
`tered VEGF Trap to animals bearing U-87 MG intracra-
`nial xenografts at several phases of tumor development.
`In the present study, we demonstrated that VEGF Trap
`treatment in animals bearing human gliomas resulted
`in significant prolonged survival. Of importance, our
`results indicate that VEGF Trap was equally effective
`against initial or advanced disease, and that the response
`was enhanced when VEGF Trap was administered in a
`prolonged schedule.
`
`Material and Methods
`
`Cell Line
`
`The human glioma cell line U-87 MG was purchased
`from the American Type Culture Collection (Manassas,
`VA, USA). Cells were maintained in Dulbecco’s modi-
`fied Eagle/F12 medium (1:1, vol:vol) (The University of
`Texas M. D. Anderson Cancer Center Media Core Facil-
`ity, Houston, TX, USA) supplemented with 10% fetal
`calf serum and 1% antibiotic/antimycotic agent (Invit-
`rogen, Carlsbad, CA, USA) in a humidified atmosphere
`containing 5% CO2 at 37°C.
`
`Drugs
`
`VEGF Trap and human Fc (hFc, constant region of
`human IgG1) were kindly provided by Regeneron Phar-
`maceuticals (Tarrytown, NY, USA). Stocks of 50 mg/ml
`in aqueous solution were kept at –80°C.
`
`In Vivo Experiments
`
`The U-87 MG human glioma cells (5 3 105) were
`engrafted in the caudate nucleus of athymic mice (Har-
`lan Sprague Dawley Inc., Indianapolis, IN, USA), as
`previously described.13 At 0, 4, and 10 days after cell
`implantation, we administered VEGF Trap (25 mg/kg
`subcutaneously, twice a week, for a total of 3 or 6 weeks)
`to separate groups of 10–15 animals per treatment bear-
`ing U-87 MG intracranial xenografts. Either phosphate-
`buffered saline (PBS) or hFc was blindly administered
`as a control agent in randomly selected subgroups of
`glioma-bearing animals. Animals showing generalized
`or local symptoms of disease were euthanized. Brains
`were fixed in 4% formaldehyde for 24 h and embedded
`in paraffin. Slides were stained with hematoxylin and
`eosin. All animal studies were performed in the veteri-
`nary facilities of The M. D. Anderson Cancer Center in
`accordance with institutional guidelines.
`
`Gomez-Manzano et al.: Antiglioma effect of VEGF Trap
`
`Enzyme-Linked Immunosorbent Assays
`
`Blood was collected from the tail vein of glioma-bearing
`mice 3 days after the initial dose of VEGF Trap, hFc, or
`vehicle, and VEGF Trap was quantified in the serum by
`enzyme-linked immunosorbent assays (ELISA), as previ-
`ously reported.16
`
`Statistical Analyses
`
`The in vivo anticancer effect of different treatments was
`assessed by plotting Kaplan-Meier survival curves, and
`treatment groups were compared using the log-rank test.
`The effects of VEGF Trap when administered in differ-
`ent treatment schedules were analyzed using a permuta-
`tion test.
`
`Results and Discussion
`
`Antiglioma Effect of VEGF Trap on Initial Disease
`
`The VEGF Trap-mediated antiglioma effect was assessed
`in vivo using an intracranial human glioma xenograft
`model. We selected the U-87 MG cell line for this study
`because it produces gliomas in nude mice with highly
`predictable growth kinetics and well-characterized path-
`ological features13; in addition, U-87 MG cells express
`high levels of VEGF and, when implanted intracranially
`in immunocompromised mice, develop as highly vascu-
`larized tumors.11,13 Our group has previously character-
`ized the kinetics of tumor growth and vascularization
`of human U-87 MG xenografts implanted intracranially
`in nude mice. Of interest to the present study, U-87 MG
`intracranial tumors exhibited initially minimal tumor
`growth, but changes in the host vessels were evident as
`soon as day 1 and definitely by day 4 after implantation;
`these changes included significant vessel co-option, as
`illustrated by the existence of engorged smooth-muscle
`actin (SMA)-positive vascular structures in the periph-
`ery of the xenograft.13
`To test the effect of VEGF Trap in the initial phases
`of the disease, we planned two different treatment
`schedules (Figs. 1 and 2) consisting of the subcutaneous
`administration of 25 mg/kg VEGF Trap twice weekly
`over 3 weeks, starting on either day 0 (schedule A) or
`day 4 (schedule B) after the intracranial implantation of
`human glioma cells in nude mice. Control groups were
`treated with PBS or hFc at doses and volumes similar to
`those used for the test drug. The agents were adminis-
`tered in a double-blinded manner; that is, the identity of
`the test groups was concealed from both the personnel
`preparing the drugs and the animal caretakers.
`Animals treated with VEGF Trap starting on day 0
`or day 4 after implantation had significantly prolonged
`survival compared to the hFc- or PBS-treated animals
`(p , 0.0001 and p , 0.005, respectively). In animals
`treated with schedule A, the median overall survival of
`the control-treated animals (treated with either hFc or
`PBS) was 30 days, with all animals dying by day 33.
`Treatment with VEGF Trap prolonged the mean survival
`
`Neuro-oNcology • D E C E M B E R 2 0 0 8
`
`
`
` 941
`
`

`

`Gomez-Manzano et al.: Antiglioma effect of VEGF Trap
`
`Fig. 1. Schematic representation of the treatment schedule used with the anti-vascular endothelial growth factor (VEGF) agent VEGF Trap,
`which was based on our previous studies of the kinetics of growth and vascularization in the U-87 MG intracranial model. U-87 MG cells
`were implanted in the brains of the animals on day 0, and VEGF Trap was administered starting on day 0 (schedule A), day 4 (schedule
`B), or day 10 (schedules CS and CL) after cell implantation. The two schedule C subgroups were treated in either a 3-week (schedule CS)
`or 6-week (schedule CL) schedule. Schedules A and B followed a 3-week treatment regimen. Animals were euthanized when signs of neu-
`rological or generalized disease appeared.
`
`by 8 days. In animals treated with schedule B, the mean
`survival in the PBS- and hFc-treated animals was 27.5
`and 30 days, respectively, but it was increased to 36 days
`in the group treated with VEGF Trap. No treatment-
`schedule-dependent differences in survival duration were
`observed in animals receiving VEGF Trap, suggesting
`VEGF Trap is efficacious in initial phases of disease that
`
`were characterized by active vessel co-option and remod-
`eling. Analysis performed 3 days after the first VEGF
`Trap doses were administered revealed high VEGF Trap
`levels (approximately .50 μg/ml) in the serum of all
`these animals, suggesting an efficient systemic biodistri-
`bution (data not shown).
`
`Fig. 2. Effect of the anti-vascular endothelial growth factor (VEGF) agent VEGF Trap on initial phases of disease: survival analysis of glioma-
`bearing animals treated with VEGF Trap since day 0 (A) or day 4 (B), as pictured in Fig. 1. Kaplan-Meier survival curves begin on the day
`of U-87 MG intracranial implantation following the subcutaneous injection of VEGF Trap or of vehicle or human Fc (control). The p-values
`(determined by log-rank test) show significant overall survival differences between VEGF Trap-treated and control-treated animals. Abbre-
`viations: E, events; N, number of animals.
`
`942
`
`
`
` Neuro-oNcology • D E C E M B E R 2 0 0 8
`
`

`

`Antiglioma Effect of VEGF Trap on Disease Burden
`
`To test the effect of VEGF Trap on tumor burden, and
`based on our previous study of U-87 MG intracranial
`growth and angiogenesis, we decided to start treat-
`ment on day 10 after cell implantation in one subgroup
`of mice (Fig. 1, schedule CS). According to our previ-
`ous studies, by day 10, increased microvascular density
`(MVD) was associated with exponential tumor growth
`and a decrease in the rate of induced angiogenesis within
`the host and the tumor periphery.13 Twelve days after
`implantation, the tumors consisted of spherical masses
`of cells with a high MVD and large, distorted, SMA-
`positive vessels. The tumor limits were clearly defined,
`and the cancer cells did not exhibit the invasive pattern
`into host tissue seen in preceding days.13
`In the present study, glioma cells were implanted
`intracranially, and 10 days later, VEGF Trap was admin-
`istered subcutaneously at a dose of 25 mg/kg twice
`weekly for 3 weeks. Control groups were treated with
`PBS or hFc at doses and volumes similar to those of the
`test drug. Treatment of the glioma-bearing animals with
`VEGF Trap resulted in a significant increase in the sur-
`vival of these animals (p , 0.005) (Fig. 3A). In particu-
`lar, the median overall survival of control-treated (PBS
`or hFc) animals was 31 days, with all the animals dead
`by day 33, whereas the mean survival of VEGF Trap-
`treated animals was 45 days. We observed no significant
`difference in the effect of VEGF Trap on prolonging sur-
`vival at different stages of the disease (comparing effects
`of schedules A and B with schedule CS) (p . 0.1, permu-
`tation test), suggesting that VEGF Trap can be similarly
`effective in both the initial and burden disease stage.
`These data further suggest that targeting circulating
`levels of VEGF is equally effective in challenging tumor
`
`Gomez-Manzano et al.: Antiglioma effect of VEGF Trap
`
`growth under both initial and established tumoral vas-
`culature phases.
`
`Antiglioma Effect of Prolonged VEGF Trap Treatment
`
`We next explored the effect in vivo of more prolonged
`VEGF Trap treatment. In this experiment, animals bear-
`ing intracranial human gliomas were treated with VEGF
`Trap (25 mg/kg) twice weekly for 6 weeks starting on
`day 10 after cell implantation (Fig. 1, schedule CL). Con-
`trol animals were treated with vehicle or hFc (25 mg/
`kg) twice weekly until they showed signs of disease, at
`which time they were euthanized according to institu-
`tional regulations. Animals treated with VEGF Trap
`for 6 weeks survived longer than did animals treated
`with hFc (median overall survival, 55 days and 21 days,
`respectively; Fig. 3B) (p , 0.0001). We also analyzed
`the difference in median survival times between the
`animals treated with VEGF Trap for 6 weeks and those
`treated for 3 weeks. Using the permutation test and after
`adjusting for overall survival on PBS-treated groups, we
`found the increase in survival obtained with the 6-week
`VEGF Trap treatment to be significantly greater than the
`increase in survival obtained with the 3-week treatment
`(p , 0.05). These data suggest that VEGF Trap is more
`effective in prolonging overall survival when adminis-
`tered in a prolonged treatment schedule.
`
`Histological Examination of VEGF Trap-Treated
`Tumors
`
`Microscopic analysis of histological sections from
`formalin-fixed, paraffin-embedded brains revealed that
`control- and VEGF Trap-treated animals eventually suf-
`fered from the lethal growth of their tumors. Because of
`
`Fig. 3. Effect of the anti-vascular endothelial growth factor (VEGF) agent VEGF Trap on advanced glioma disease: survival analyses of
`glioma-bearing animals that were treated with VEGF Trap starting on day 10 after cell implantation in either a 3-week (schedule CS) or
`6-week (schedule CL) regimen, as pictured in Fig. 1. Kaplan-Meier survival curves begin on the day of U-87 MG intracranial implantation
`following the subcutaneous injection of VEGF Trap or control agent (vehicle or human Fc). The p-values (determined by log-rank test)
`show significant overall survival differences between VEGF Trap-treated and control-treated animals. Abbreviations: E, events; N, number
`of animals.
`
`Neuro-oNcology • D E C E M B E R 2 0 0 8
`
`
`
` 943
`
`

`

`Gomez-Manzano et al.: Antiglioma effect of VEGF Trap
`
`A.
`
`B.
`
`Fig. 4. Histological examination of brain sections from animals treated with the anti-vascular endothelial growth factor (VEGF) agent VEGF
`Trap. (A) Hematoxylin-eosin staining of mouse brains bearing U-87 MG xenografts treated with human Fc (hFc) or VEGF Trap according
`to schedule B. No signs of hemorrhagic areas or an enhanced invasive phenotype were observed after VEGF Trap treatment. N, normal
`tissue; T, tumor tissue. Original magnification, 3100. (B) Histological examination of brain sections from animals treated with VEGF Trap
`as described for schedule CL. Sections stained with hematoxylin and eosin show the presence of an invasive phenotype with satellitosis
`characterized by glioma clustering around vascular vessels and accumulation of invasive glioma cells far from the main tumor mass (arrows).
`Original magnification: left, 3100; right, 3200.
`
`previous studies describing that treatment with antian-
`giogenic agents may result in intracranial hemorrhages
`or enhance tumor invasion,2,17 we specifically examined
`the tumors for the presence of these adverse effects.
`Histological examination of the brains of the cohorts
`treated for 3 weeks did not reveal either phenomenon.
`Treated U-87 MG-derived tumors displayed a very well-
`defined border with the normal host parenchyma (Fig.
`4A). However, examination of the brains of animals that
`received prolonged treatment (6 weeks) of VEGF Trap,
`which survived longer than those treated on a 3-week
`schedule, revealed the signs of mass effect and the pres-
`ence of the so-called “secondary structures” or “satel-
`litosis” consisting of aggregations of glioma cells in the
`perivascular regions, as well as the presence of glioma
`cells along the Virchow-Robin spaces (Fig. 4B). These
`data suggest that U-87 MG-derived xenografts acquired
`an invasive phenotype in response to anti-VEGF therapy.
`These results are in agreement with a similar pattern of
`
`growth of intracranial G55 xenografts in animals treated
`with an antibody against mouse VEGFR-2, DC101,17 or
`a neutralizing VEGF antibody.18 These results may be
`likewise in agreement with those from clinical trials in
`patients with cancer treated with VEGF inhibitors, in
`that they survived longer but eventually exhibited resis-
`tance to the treatment.19,20 Of importance, the model
`described here offers us the possibility of testing com-
`bined therapies designed to counteract the emergence of
`a resistant phenotype to anti-VEGF therapies.
`Taken together, our data show that treatment with
`VEGF Trap significantly prolonged the survival of
`glioma xenograft-bearing mice. Of great interest, initial/
`residual disease and disease burden were both similarly
`affected by the antiangiogenesis treatment. In addition,
`the prolonged use of VEGF Trap (over 6 weeks) improved
`outcomes significantly more than did treatment admin-
`istered in a short schedule (over 3 weeks).
`The traits for personalized medicine are emerging
`
`944
`
`
`
` Neuro-oNcology • D E C E M B E R 2 0 0 8
`
`

`

`Gomez-Manzano et al.: Antiglioma effect of VEGF Trap
`
`for the treatment of brain tumors, and they will need to
`take into consideration the highly heterogeneous nature
`of these tumors.1,21 However, the fact that all brain
`tumor subtypes rely on blood vessels for survival and
`growth indicates the broad applicability of this strategy.
`Thus, our report provides data that encourage the test-
`ing of VEGF Trap in patients with recurrent malignant
`gliomas, and in this regard, results from a multicenter
`study consisting of a phase II clinical trial of VEGF Trap
`in patients with recurrent gliomas will soon be avail-
`able. Finally, we suggest that VEGF Trap should also be
`considered for the treatment of patients after extensive
`surgery, which we would regard as carrying minimal
`residual disease, in combination with therapies targeting
`the migratory and invasive properties of gliomas.
`
`Acknowledgments
`
`We greatly thank Drs. John S. Rudge and Risa Shapiro
`(Regeneron Pharmaceuticals, Inc.) for their useful com-
`ments and the ELISA studies, Betty Notzon (Department
`of Scientific Publications, M. D. Anderson) for editorial
`assistance, and Verlene Henry and Jennifer Edge for
`technical assistance (Brain Tumor Center, M. D. Ander-
`son). This work was supported by National Cancer Insti-
`tute grant CA-16672 (supporting the Research Histopa-
`thology and Research Animal Support core facilities at
`M. D. Anderson) and was partially sponsored by Regen-
`eron Pharmaceuticals, Inc.
`J.H. is currently at Novartis Institutes for BioMedical
`Research, Emeryville, CA, USA.
`
`References
`
`4.
`
`1. Batchelor TT, Sorensen AG, di Tomaso E, et al. AZD2171, a pan-VEGF
`receptor tyrosine kinase inhibitor, normalizes tumor vasculature and
`alleviates edema in glioblastoma patients. Cancer Cell. 2007;11:83–
`95.
`2. Pope WB, Lai A, Nghiemphu P, et al. MRI in patients with high-grade
`gliomas treated with bevacizumab and chemotherapy. Neurology.
`2006;66:1258–1260.
`3. Vredenburgh JJ, Desjardins A, Herndon JE 2nd, et al. Phase II trial of
`bevacizumab and irinotecan in recurrent malignant glioma. Clin Can-
`cer Res. 2007;13:1253–1259.
`Jain RK, di Tomaso E, Duda DG, et al. Angiogenesis in brain tumours.
`Nat Rev Neurosci. 2007;8:610–622.
`5. Holash J, Davis S, Papadopoulos N, et al. VEGF-Trap: a VEGF blocker
`with potent antitumor effects. Proc Natl Acad Sci U S A. 2002;99:
`11393–11398.
`6. Byrne AT, Ross L, Holash J, et al. Vascular endothelial growth factor-
`trap decreases tumor burden, inhibits ascites, and causes dramatic
`vascular remodeling in an ovarian cancer model. Clin Cancer Res.
`2003;9:5721–5728.
`7. Kim ES, Serur A, Huang J, et al. Potent VEGF blockade causes regres-
`sion of coopted vessels in a model of neuroblastoma. Proc Natl Acad
`Sci U S A. 2002;99:11399–11404.
`8. Riely GJ, Miller VA. Vascular endothelial growth factor trap in non
`small cell lung cancer. Clin Cancer Res. 2007;13:s4623–s4627.
`9. Verheul HM, Hammers H, van Erp K, et al. Vascular endothelial
`growth factor trap blocks tumor growth, metastasis formation, and
`vascular leakage in an orthotopic murine renal cell cancer model. Clin
`Cancer Res. 2007;13:4201–4208.
`10. Wachsberger PR, Burd R, Cardi C, et al. VEGF trap in combination
`with radiotherapy improves tumor control in u87 glioblastoma. Int J
`Radiat Oncol Biol Phys. 2007;67:1526–1537.
`11. Blouw B, Song H, Tihan T, et al. The hypoxic response of tumors is
`dependent on their microenvironment. Cancer Cell. 2003;4:133–
`146.
`
`12. Roberts WG, Delaat J, Nagane M, et al. Host microvasculature influ-
`ence on tumor vascular morphology and endothelial gene expression.
`Am J Pathol. 1998;153:1239–1248.
`13. Lee OH, Fueyo J, Xu J, et al. Sustained angiopoietin-2 expression
`disrupts vessel formation and inhibits glioma growth. Neoplasia.
`2006;8:419–428.
`14. Bergers G, Benjamin LE. Tumorigenesis and the angiogenic switch.
`Nat Rev Cancer. 2003;3:401–410.
`15. Holash J, Maisonpierre PC, Compton D, et al. Vessel cooption, regres-
`sion, and growth in tumors mediated by angiopoietins and VEGF. Sci-
`ence. 1999;284:1994–1998.
`16. Rudge JS, Holash J, Hylton D, Russell M, et al. Inaugural article: VEGF
`Trap complex formation measures production rates of VEGF, provid-
`ing a biomarker for predicting efficacious angiogenic blockade. Proc
`Natl Acad Sci U S A. 2007;104:18363–18370.
`17. Kunkel P, Ulbricht U, Bohlen P, et al. Inhibition of glioma angiogenesis
`and growth in vivo by systemic treatment with a monoclonal antibody
`against vascular endothelial growth factor receptor-2. Cancer Res.
`2001;61:6624–6628.
`18. Rubenstein JL, Kim J, Ozawa T, et al. Anti-VEGF antibody treatment of
`glioblastoma prolongs survival but results in increased vascular coop-
`tion. Neoplasia. 2000;2:306–314.
`19. Kerbel RS, Yu J, Tran J, et al. Possible mechanisms of acquired resis-
`tance to anti-angiogenic drugs: implications for the use of combina-
`tion therapy approaches. Cancer Metastasis Rev. 2001;20:79–86.
` Miller KD, Sweeney CJ, Sledge GW Jr. The snark is a boojum: the
`continuing problem of drug resistance in the antiangiogenic era. Ann
`Oncol. 2003;14:20–28.
`21. Kleihues P, Burger PC, Scheithauer BW. Histological typing of tumours
`of the central nervous system. 2nd ed. Berlin: Springer; 1993:1–105.
`
`20.
`
`Neuro-oNcology • D E C E M B E R 2 0 0 8
`
`
`
` 945
`
`