Antitumor Activity of CEP-751 (KT-6587) on Human Neuroblastoma and Medulloblastoma Xenografts1

3594 Vol. 5, 3594–3602, November 1999 Clinical Cancer Research

Furthermore, inhibition of growth was greater .(0.049 ؍ Audrey E. Evans, Kristin D. Kisselbach, (P Darrell J. Yamashiro, Naohiko Ikegaki, in the SY5Y cell line transfected with TrkB compared with Anna Marie Camoratto, Craig A. Dionne, and the untransfected parent cell line expressing no detectable 2 TrkB. Terminal deoxynucleotidyl transferase-mediated nick Garrett M. Brodeur end labeling studies showed CEP-751 induced apoptosis in Division of Oncology at The Children’s Hospital of Philadelphia and the treated CHP-134 tumors, whereas no evidence of apop- Department of Pediatrics at the University of Pennsylvania, tosis was seen in the control tumors. Finally, there was no Philadelphia, Pennsylvania 19104 [A. E. E., K. D. K., N. I., G. M. B.]; Cephalon, Inc., West Chester, Pennsylvania 19380 [A. E. E., K. D. K., apparent toxicity identified in any of the treated mice. These N. I., A. M. C., C. A. D., G. M. B.]; and College of Physicians and results suggest that CEP-751 may be a useful therapeutic Surgeons of Columbia University, Division of Pediatric Oncology, agent for NBL or MBL. New York, New York 10032 [D. J. Y.]

INTRODUCTION ABSTRACT NBL3 is one of the most common tumors in childhood. It Neuroblastoma (NBL) and medulloblastoma (MBL) are occurs at an early age, i.e., 50% of patients are Ͻ2 years of age. tumors of the neuroectoderm that occur in children. NBL The disease in infants tends to have a more benign course and and MBL express Trk family tyrosine kinase receptors, sometimes regresses spontaneously. However, children Ͼ2 which regulate growth, differentiation, and cell death. CEP- years of age usually have metastatic disease and a poor prog- 751 (KT-6587), an indolocarbazole derivative, is an inhibitor nosis. These “unfavorable” tumors have only partially benefited of Trk family tyrosine kinases at nanomolar concentrations. from the advances in multimodality therapy that have markedly This study was designed to determine the effect of CEP-751 improved survival of other malignancies in childhood, such as on the growth of NBL and MBL cell lines as xenografts. In acute lymphoblastic leukemia and Wilms’ tumor. Presently, the vivo studies were conducted on four NBL cell lines (IMR-5, treatment of these children is very aggressive, including inten- CHP-134, NBL-S, and SY5Y) and three MBL cell lines sive chemotherapy with marrow or stem cell rescue. However, (D283, D341, and DAOY) using two treatment schedules: (a) new and more specific agents are needed, ideally targeted to treatment was started after the tumors were measurable NBL cells and nontoxic to hematopoietic and other normal cells. (therapeutic study); or (b) 4–6 days after inoculation, before MBL is also a tumor of the neuroectoderm, arising in the tumors were palpable (prevention study). CEP-751 was cerebellum, and MBL has a poor prognosis in children with given at 21 mg/kg/dose administered twice a day, 7 days a more advanced disease. Patients with this tumor would also week; the carrier vehicle was used as a control. In thera- benefit from more specific and less toxic forms of treatment. peutic studies, a significant difference in tumor size was seen Most NBL tumor cells express one or more of the tyrosine between treated and control animals with IMR-5 on day 8 kinase receptors TrkA, TrkB, and/or TrkC, which bind the ؍ ؍ (P 0.01), NBL-S on day 17 (P 0.016), and CHP-134 on neurotrophin nerve growth factor, brain-derived neurotrophic ؍ day 15 (P 0.034). CEP-751 also had a significant growth- factor, and neurotrophin-3, respectively. The level of expression ؍ inhibitory effect on the MBL line D283 (on day 39, P of the individual Trk receptors may regulate growth, differen- 0.031). Inhibition of tumor growth of D341 did not reach tiation, and cell death in NBL tumors. The presence of high statistical significance, and no inhibition was apparent with TrkA expression (with or without TrkC) is associated with a DAOY. In prevention studies, CEP-751 showed a modest favorable outcome in NBL tumors (1–6). In contrast, however, ؍ growth-inhibitory effect on IMR5 (P 0.062) and CHP-134 TrkB is expressed in unfavorable NBL tumors, especially those with MYCN amplification (7). MBLs also express receptors of the TRK family, and TrkC expression is associated with a favorable outcome (8, 9). Received 11/25/98; revised 6/2/99; accepted 6/25/99. The Trk tyrosine kinase inhibitor CEP-751 (KT-6587) is an The costs of publication of this article were defrayed in part by the indolocarbazole derivative that preferentially inhibits autophos- payment of page charges. This article must therefore be hereby marked phorylation and signaling of Trk family receptors at nanomolar advertisement in accordance with 18 U.S.C. Section 1734 solely to concentration in vitro (10). The purpose of this study was to indicate this fact. 1 This work was supported in part by NIH Grant NS-34514 (to G. M. B.), a grant from the Audrey E. Evans Endowed Chair (to G. M. B.), and funds from the Children’s Cancer Research Foundation. 2 To whom requests for reprints should be addressed, at Division of Oncology, Children’s Hospital of Philadelphia, Abramson Research 3 The abbreviations used are: NBL, neuroblastoma; MBL, medulloblas- Center, Room 902-D, 3516 Civic Center Boulevard, Philadelphia, PA toma; CHOP, Children’s Hospital of Philadelphia; b.i.d., twice a day; 19104-4318; Phone: (215) 590-2817; Fax: (215) 590-3770; E-mail: TUNEL, terminal deoxynucleotidyl transferase-mediated nick end la- [email protected]. beling; DAPI, 4,6-diamidino-2-phenylindole.

Fig. 1 Treatment effect of CEP-751 on NBL cell lines grown in nude mice. A, IMR-5; B, CHP-134; C, NBL-S therapeutic studies; D, IMR-5 prevention study results; E and F, CHP- 134 prevention study comparing the treatment effect of six early- and six late-appearing tumors (see Table 1). Bars, SE.

determine the effect of CEP-751 on human NBL and MBL cell 10% povidone C30 (ISP, Bound Brook, NJ), and 2% benzoyl lines growing as xenografts in nude (nu/nu) mice. alcohol (Spectrum) in distilled water. Initial in vivo studies demonstrated that CEP-751 could be given s.c. in nude mice two MATERIALS AND METHODS times a day at a dose of up to 48 mg/kg/day for at least 4 weeks, Drug. CEP-751 (KT-6587), a derivative of K252a (8), with no apparent morbidity or mortality (11). was synthesized in the laboratories of Kyowa Hakko Kogyo Cell Lines. The NBL lines were available from the (Tokyo, Japan) and was provided for these studies by Cephalon, CHOP cell bank, and the MBL lines D283-Med and D341-Med Inc. CEP-751 was solubilized for in vivo experimentation, as were obtained from Drs. Henry Friedman (Duke) and Peter described (10). CEP-751 was dissolved in a vehicle consisting Phillips (CHOP). The DAOY MBL line was obtained from the of 40% polyethylene glycol 100 (Spectrum, Los Angeles, CA), American Type Culture Collection. The cell lines were grown in

Table 1 In vivo studies of NBL Tumor size (cm3) Analysis of slopea Cell line Day of analysis Compound Control Compound Control A. Neuroblastoma therapeutic studiesb IMR-5 12 1.96 Ϯ 0.35 (n ϭ 5) 3.91 Ϯ 1.22 (n ϭ 5) 0.17 Ϯ 0.051 (n ϭ 5) 0.28 Ϯ 0.109 (n ϭ 5) P ϭ 0.025 P ϭ 0.024 CHP 134 15 1.9 Ϯ 0.75 (n ϭ 4) 4.5 Ϯ 0.98 (n ϭ 3) 0.151 Ϯ 0.055 (n ϭ 6) 0.299 Ϯ 0.068 (n ϭ 5) P ϭ 0.034 P ϭ 0.12 NBL-S 17 0.94 Ϯ 0.24 (n ϭ 5) 5.2 Ϯ 2.46 (n ϭ 4) 0.15 Ϯ 0.034 (n ϭ 5) 0.264 Ϯ 0.014 (n ϭ 4) P ϭ 0.016 P ϭ 0.12 B. Neuroblastoma prevention studiesc IMR-5 27 1.3 Ϯ 1.63 (n ϭ 7) 2.7 Ϯ 1.55 (n ϭ 6) 0.118 Ϯ 0.036 (n ϭ 7) 0.188 Ϯ 0.041 (n ϭ 6) P ϭ 0.062 P ϭ 0.073 CHP 134 (combined) See text 0.064 Ϯ (n ϭ 6) 0.122 Ϯ (n ϭ 6) P ϭ 0.058 CHP 134 (early) 29 See text 0.148 Ϯ 0.013 (n ϭ 3) 0.169 Ϯ 0.01 (n ϭ 3) P ϭ 0.28 CHP 134 (late) 58 See text 0.022 Ϯ 0.011 (n ϭ 3) 0.140 Ϯ 0.009 (n ϭ 3) P ϭ 0.049 a Figures given for the average daily growth rate. b Treatment started after tumors measured Ͼ0.24 g. c Treatment commenced prior to tumor appearance.

RPMI 1640 containing 10% fetal bovine serum maintained in tic study). Animals were assigned to treatment or control groups 2 75-cm Costar flasks in a humidified atmosphere of 5% CO2 by matching tumor sizes in pairs, usually when at least six and air. For in vivo studies, cells were grown in 150-cm2 culture animals had measurable tumors; and (b) the treatment was flasks to ϳ75% confluence, split to insure they were growing in started 4–6 days after the implantation, before tumors were log phase, and harvested in 24 h. Cells were harvested using palpable (prevention study). The day of initiating treatment 0.2% tetrasodium EDTA in PBS. varied with the growth rate of the cell line; D341 had the Animals. Four-week-old female athymic NCR (nu/nu) shortest time to tumor appearance (7 days), and treatment was mice were obtained from the National Cancer Institute (Fred- started day 4 from inoculation, whereas CHP-134 tumors ap- erick, Maryland). The Institutional Animal Care and Use Com- peared in 15–17 days, therefore treatment was started day 6. mittee of the Joseph Stokes, Jr. Research Institute at CHOP Treatment with CEP-751 or vehicle control was administered approved all animal studies. Mice were maintained at four per b.i.d., 7 days/week. cage under humidity- and temperature-controlled conditions and Trk Studies. The level of Trk expression in the cell lines a light/dark cycle that was set at 12-h intervals. All animals were used was measured by a semiquantitative reverse transcription- fed autoclaved Purina mouse chow and water ad libitum. PCR method.4 All NBL and MBL cell lines (except DAOY) In Vivo Antitumor Experiments. Tumors were meas- expressed low levels of TrkA and/or TrkB. To assess the im- ured in unanesthetized animals two times a week using a portance of Trk expression for responsiveness to CEP-751, the Vernier caliper. Tumors were measured in three dimensions SY5Y cell line was transfected with TrkB (14) in a constitutive ϫ ϫ d1 d2 d3, and calculations of tumor volume were made by expression vector (15). ␲ multiplying the product of three dimensions by /6. Body Apoptosis Detection by the TUNEL Method. The ap- weights were obtained, and the dose of compound was modified optosis detection system measures the fragmented DNA of ϫ as the weight of the animal changed. For NBL cell lines, 2.5 apoptotic cells by catalytically incorporating fluorescein-12 7 10 cells/inoculate in RPM1 1640 were injected s.c. in the flank, dUTP at the 3Ј-OH DNA ends using the enzyme terminal together with an equal volume of Matrigel (Becton Dickinson; deoxynucleotidyl transferase, which forms a polymeric tail ϫ 7 Refs. 12 and 13). In the MBL lines D341 and D283, 1 10 (TUNEL assay; Ref. 16). Using tumor touch preparations, the cells were injected in a similar fashion. The success rate of cells harvested from tumors on days 2, 4, and 6 after treatment tumor development after injection with tumor cells was gener- were fixed with formaldehyde (4% in PBS), permeabilized with ally from 75 to 100%. However, DAOY cells required passage Triton X-100, and incubated with equilibration buffer, nucleo- as a xenograft through mice before their growth was sufficiently tide mix, and terminal deoxynucleotidyl transferase enzyme. rapid to be used in the in vivo study. Tumor tissue was minced, After incubation, DAPI was added to stain the cells. Slides were homogenized in a tissue homogenizer, and then transferred to analyzed at 460 nM to see the blue DAPI-stained cells and at 350 10-ml syringes and forced through hypodermic needles of de- nm to visualize the green apoptotic cells. The proportion of the creasing size down to 22 gauge, prior to injection. Subsequently, the time required for tumors to appear decreased from 80 to 24 days. Two study approaches were used: (a) the treatment was started after the s.c. tumor was palpable at ϳ0.2 cm3 (therapeu- 4 N. Ikegaki et al., unpublished observations.

Table 2 In vivo studies of MBL Tumor size (cm3) Analysis of slopea Cell line Day of analysis Compound Control Compound Control A. Medulloblastoma therapeutic studiesb D341 15 1.98 Ϯ 1.86 (n ϭ 5) 2.24 Ϯ 0.790 (n ϭ 5) 0.199 Ϯ 0.091 (n ϭ 6) 0.238 Ϯ 0.068 (n ϭ 6) P ϭ 0.60 P ϭ 0.75 DAOY 16 1.51 Ϯ 0.689 (n ϭ 5) 1.97 Ϯ 0.603 (n ϭ 3) 0.152 Ϯ 0.045 (n ϭ 5) 0.151 Ϯ 0.069 (n ϭ 3) P ϭ 0.39 P ϭ 0.96 D283 39 1.044 Ϯ 0.297 (n ϭ 8) 1.54 Ϯ 0.51 (n ϭ 8) 0.017 Ϯ 0.008 (n ϭ 8) 0.031 Ϯ 0.012 (n ϭ 8) P ϭ 0.031 P ϭ 0.021 B. Medulloblastoma prevention studyc D341 16 0.91 Ϯ 0.92 (n ϭ 5) 2.17 Ϯ 1.42 (n ϭ 6) 0.092 Ϯ 0.072 (n ϭ 8) 0.186 Ϯ 0.039 (n ϭ 9) P ϭ 0.10 P ϭ 0.126 a Figures given for the average daily growth rate. b Treatment started after tumors measured Ͼ0.24 g. c Treatment commenced prior to tumor appearance.

green cells in the blue background was calculated, and the were four treated animals and three controls remaining. The results were compared between treated and control tumor. treated tumors were significantly smaller than tumors in control Statistical Analysis. The Exact Wilcoxon-Mann-Whit- animals (1.9 cm3 versus 4.5 cm3, P ϭ 0.034; Table 1; Fig. 1B). ney test of significance (17) was used to compare the tumor size The analysis of slope was applied to the animals that died early, at a particular test day between CEP-751 treated and control because there was more than one tumor measurement to docu- groups.5 Also, each subject’s individual slope was calculated ment early growth rate. For the CEP-751-treated animals, assuming that tumor size is a linear function of time, and the the slope was 0.151 Ϯ 0.055 cm3/day versus 0.299 Ϯ 0.068 slopes in the two groups (CEP-751 versus vehicle) were com- in controls, but this difference did not reach significance pared using Student’s t test (18). (P ϭ 0.12). Twenty animals were injected with NBL-S, and by day 13, RESULTS 15 of 20 had developed small tumors that grew slowly. Treat- NBL Therapeutic Studies. After injection with 2.5 ϫ ment was started on day 38 after injection, when most of the 107 IMR-5 cells/animal, five mice were treated with CEP-751 at animals had measurable tumors. Two animals with internal 9.3 mg/kg b.i.d., and five control animals were treated with tumors were not included. Seven were assigned to the treatment vehicle alone. Tumors appeared between 9 and 18 days from group (average tumor size, 0.23 cm3; range, 0.19–0.34 cm3), inoculation and ranged from 0.43 to 0.8 cm3 (median, 0.47 cm3) and six were assigned to the control group (average, 0.31 cm3; in control animals and 0.36 to 0.55 cm3 (median, 0.47 cm3)in range, 0.23–0.37). One animal in each arm died of accidental animals treated with CEP-751. By day 8 from the start of death. The remainder were analyzed at day 17 of treatment. The treatment, tumors in the treated animals averaged 1.21 cm3 Ϯ treated tumors were significantly smaller than controls (0.94 Ϯ 0.43, which was significantly smaller than the tumors in control 0.24 versus 5.17 Ϯ 2.46, P ϭ 0.016; Table 1 and Fig. 1C). 3 animals (2.60 cm Ϯ 0.53; P ϭ 0.008; Fig. 1A). The analysis of Similarly, there was a significant difference in the slope of the slope (average daily tumor growth rate) for the treated animals two curves (0.051 Ϯ 0.034 treated versus 0.213 Ϯ 0.143, P ϭ 3 was 0.17 cm /day Ϯ 0.051 compared with the control rate of 0.016). 3 Ϯ ϭ 0.29 cm /day 0.11 (P 0.032). Similar differences were also NBL Prevention Studies. Fourteen animals were in- apparent at day 12 (Table 1; Fig. 1A). jected with 2.5 ϫ 107 IMR-5 cells/animal. One control animal Twelve animals injected with CHP-134 developed tumors died of accidental death early in the study. Treatment with 21 within 24–34 days. The control tumors ranged in size from 0.29 mg/kg b.i.d. was started on day 5 from injection of IMR-5 cells, 3 3 to 1.3 cm (median, 0.64 cm ), and the treated tumors ranged in before the tumors appeared. In six control animals, the tumors 3 3 size from 0.35 to 0.78 cm (median, 0.41 cm ) when treatment appeared by day 13, and in the seven treated animals, the tumors was started with 21 mg/kg b.i.d. (this dose was used for all appeared by day 17. However, the overall delay in the median subsequent studies). Six animals were treated with CEP-751 and growth of tumors was not significant (P ϭ 0.27). On day 27, the six with vehicle. However, two animals in each group died of average tumor size in seven treated animals was smaller at 1.3 accidental death early in the study, and one animal in the control cm3 versus 2.7 cm3 in the six-control group (Table 1; Fig. 1D), group had to be sacrificed on day 9 because of rapid tumor but this difference did not reach statistical significance (P ϭ growth. Thus, for the analysis of tumor size on day 15, there 0.062). Also, the analysis of slope showed a clear trend (0.118 Ϯ 0.036 treated versus 0.188 Ϯ 0.041 control), but this also did not reach significance (P ϭ 0.073). Animals were injected with 2.5 ϫ 107 CHP-134 cells, and 5 The software used to perform this test was authored by C. R. Mehta and N. R. Patel, entitled “StatXact3 for Windows: Statistical Software treatment (21 mg/kg b.i.d.) was started on day 6 prior to the for Exact Nonparametric Inference. User Manual,” pp. 216–226. Cam- appearance of tumor. A total of 12 tumors developed (6 in each bridge, MA: CYTEL Software Corporation, 1995. group). In the CEP-751-treated animals, they appeared between

Fig. 2 Treatment effect of CEP-751 on 2 MBL cell lines grown in nude mice. Thera- peutic studies: A, D341; B, D283. Prevention study: C, D341. Bars, SE. (See Table 2.)

days 15 and 46 (median, day 29) and in the control group Fourteen mice were injected with 0.3 cm3 of DAOY between days 9 and 25 (median, day 15). However, because the minced tumor tissue. The first tumor started to grow by day 3, treated tumors grew more slowly, it was not possible to compare and nine tumors were measurable by day 13. Treatment was tumor size in treated versus control on a specific day of treat- started on day 16 from injection, by which time control tumors ment. Furthermore, tumors developed at two substantially dif- ranged from 0.51 to 1 cm3 (median, 0.62 cm3) and treated ferent rates; six developed early (days 9–21), and six developed tumors ranged from 0.61 to 1 cm3 (median, 0.72 cm3). By day later (days 22–46). By analyzing the slopes of the total group, 16 of treatment, tumors in the treated animals measured 1.51 the daily growth of treated tumors was 0.06 Ϯ 0.05 versus cm3 versus 1.97 cm3 in the controls, which was not significant 0.122 Ϯ 0.02 in the controls (P ϭ 0.058). Little effect of (P ϭ 0.039; Fig. 2B), and analysis of slope (0.15 versus 0.17 treatment was seen on the tumor growth slopes in the early cm3/day) was not significant (P ϭ 0.67; Table 2). group (0.148 versus 0.169 cm3/day; P ϭ 0.28), but in the later Sixteen animals were inoculated with D283, and all grew group, treated tumors grew at a significantly slower rate (0.022 tumors; 8 control tumors size ranged from 0.3 to 0.5 cm3 versus 0.140 cm3/day; P ϭ 0.049; Table 1; Fig. 1, E and F). (median, 0.35 cm3), and 8 treated tumors measured 0.3–0.45 MBL Therapeutic Studies. Sixteen animals were in- cm3 (median, 0.34 cm3). Tumors appeared by day 11, and jected with D341 MBL cells, and 12 developed tumors by day treatment was started 6 days later. The growth rate of the control 10. These animals started treatment on day 14; the six tumors in tumors was slow, and by day 39, the largest tumor measured 2.4 the control group ranged from 0.03 to 0.29 cm3 (median, 0.14 cm3 (range, 1.0–2.4 cm3). The difference between the treated cm3), and the treated group ranged from 0.07 to 0.31 cm3 versus control tumors (1.04 and 1.54 cm3) was significant (P ϭ (median, 0.15 cm3). Two tumors developed internally (one in 0.031; Table 2 and Fig. 2B), and the analysis of slope (0.017 each group) and grew very rapidly at parallel rates, reaching 6.8 versus 0.031 cm3/day) was also significant (P ϭ 0.021). cm3 in 10 days of treatment. Excluding these two, on day 15 the MBL Prevention Studies. Eighteen animals were in- average tumor size in the five treated animals was 1.98 cm3, jected with 1 ϫ 107 D341 cells/animal, and tumors appeared whereas the average in the five control animals was 2.24 cm3, between day 3 and 13. Treatment was started on day 4 from (Table 2 and Fig. 2A), but this difference was not statistically injection. On day 16, there was a difference between the average significant (P ϭ 0.60). The analysis of slope (0.199 versus 0.238 tumor size in the treated versus control animals (0.9 cm3 and cm3/day) also did not show a significant difference (P ϭ 0.75). 2.17 cm3, respectively) with marginal significance (P ϭ 0.10).

Fig. 3 Treatment effect of CEP-751 on SY5Y NBL cells (A) and two TrkB transfected clones, SY5Y-G8 (B) and SY5Y- G12 (C).

The analysis of slope (0.092 versus 0.186 cm3/day) did not weight loss, rashes, or unusual behavior. Unexplained acute reach statistical significance (P ϭ 0.126; Table 2; Fig. 2C). deaths occurred usually during injection (presumably from cer- Mechanism of Antitumor Activity. SY5Y, an NBL line vical cord injury due to resistance to restraint), or from skin, that expresses low levels of TrkB, was transfected with TrkB. lymph node, or other infection, but there were no differences Two of the resulting clones, G8 and G12 expressing low and noted between the treated and control animals. high levels of TrkB, respectively, were inoculated into nude mice (15). Interestingly, the SY5Y-G12 clone expressing the highest level of TrkB had the fastest growth rate in untreated or DISCUSSION vehicle-treated tumors, and the effect of CEP-751 was the NBLs can be classified in at least two or three different greatest. The daily growth rates of SY5Y, SY5Y-G8, and categories (19). A favorable subset has a peak incidence SY5Y-G12 were 0.16, 0.20, and 0.21 cm3/day, and the signif- around 6 months of age and is uncommon after age two. icant difference between treated and control tumor size was 0.4, These tumors are generally responsive to treatment. More- 0.1, and 0.01 cm3, respectively (Fig. 3). over, they have a propensity to regress spontaneously in The effect of treatment on the tumors was studied by the infants or to mature to a benign ganglioneuroma in older TUNEL method to determine whether there was evidence of patients. The commitment to regress, as well as the propen- apoptosis (16). The CEP-751-treated tumors showed marked sity to differentiate, may both be mediated by the nerve evidence of apoptosis during the course of treatment (Fig. 4). A growth factor/TrkA pathway (19). In contrast, the unfavor- rare cell was seen with florescent staining on day 2 (0–3% of able MBLs have a peak age of 2–4 years. These tumors are cells per high power field) and increasing numbers on days 4 generally metastatic at the time of diagnosis and have a very (42%) and 6 (92%). No evidence of apoptosis was seen in the poor outcome. Interestingly, these tumors frequently express control group. The tumors prepared for TUNEL staining were both TrkB and its ligand, brain-derived neurotrophic factor, also studied by light microscopy. In the treated tumors, there which may represent an autocrine growth and survival path- was no evidence of necrosis or maturation of the tumor. way (21). Thus, agents that target the Trk family receptors Toxicity. Mice were observed twice daily and weighed may be useful therapeutically for patients with both favorable every 2 weeks; no toxicity was observed in terms of progressive and unfavorable diagnoses.

Fig. 4 Effect of compound on apoptosis in CHP134 tumors obtained after 1, 3, and 5 days of treatment. Left (from top to bottom), DAPI stained tumor cells on days 2, 4, and 6. Right (from top to bottom), TUNEL stained tumor cells on days 2, 4, and 6. The second and third rows show progressive increase in apoptotic cells with time, as detected by the TUNEL assay, in CEP-751 treated tumors.

CEP-751 is a relatively specific inhibitor of the Trk family We tested the effect of this compound on four NBL cell of tyrosine kinases (TrkA, TrkB, and TrkC) at low concentra- lines growing as xenografts in nude mice, either by treating tions. Indeed, this agent was shown to have antitumor activity palpable tumors or by trying to prevent tumor development after both in vitro and in vivo in prostate cancer, ovarian cancer, and inoculation. In the therapeutic studies, treatment with CEP-751 melanoma (10, 11, 22). Given the important roles that members resulted in a statistically significant decrease in tumor growth in of the Trk family play in NBLs, this tumor appeared to be a very three lines (IMR-5, CHP-134, and NBL-S) in animals. In both attractive candidate for treatment with CEP-751. models, the drug had no apparent effect for the first 7 days; then

a substantial difference in the slope of tumor growth was ob- the United States for solid tumors in adults (24, 25).6 CEP-701 served. In the prevention studies, there was a noticeable delay in is effective when given p.o., making it an attractive candidate the development of tumors in the treated animals versus the for use in a pediatric population. controls. In the CHP-134 animals, the delay in appearance of In summary, CEP-751 reduced in vivo growth of NBL cells tumors was so pronounced that many of the control animals had growing as xenografts using two different treatment paradigms. died from their tumor burden before the treated animals devel- It also resulted in a decrease in growth of MBL xenografts, with oped palpable tumors. borderline significance. Evidence of in vivo antitumor efficacy To determine to what extent the level of Trk receptors using established NBL cell lines suggests that this compound played a role in the degree of response, SY5Y cells expressing may be even more effective in primary tumors from untreated very low levels of TrkB were transfected with TrkB. Cells from patients, which express higher levels of Trk family receptors two of the resulting clones were grown in nude mice, which than cell lines. Furthermore, the apparent greater efficacy in were then treated with CEP-751. The SY5Y-G12 clone with the slower growing tumors and in retarding tumor appearance sug- highest expression of TrkB was the most sensitive to treatment; gests that its therapeutic efficacy may be greater in states of therefore, the effect of CEP-751 may be mediated at least in part minimal residual disease. Thus, CEP-751, or potential deriva- through its inhibition of TrkB. tives of this compound, appear promising as therapeutic agents It has been suggested that Trk expression supports not for tumors such as NBLs and possibly MBLs, either alone or in only the growth and differentiation of neuronal cells but is combination with other agents. also important for cell survival. If Trk receptors and/or their ligands are involved in a survival pathway for NBL cells, ACKNOWLEDGMENTS treatment with CEP-751 would be expected to result in in- We are grateful to Huaqing Zhao (Division of Biostatistics, CHOP) creased cell death. To examine this hypothesis, tumors were for assistance with the statistical analysis of the data. harvested at several times posttreatment and were examined for evidence of apoptotic cells using the TUNEL assay. There REFERENCES was no apparent difference after 1 day of treatment; however, 1. Nakagawara, A., Arima-Nakagawara, M., Scavarda, N. J., Azar, there were more apoptotic cells in the treated tumors after 4 C. G., Cantor, A. B., and Brodeur, G. M. Association between high and 6 days compared with the corresponding controls. These levels of expression of the TRK gene and favorable outcome in human results suggest that CEP-751 may induce apoptosis in these neuroblastoma. N. Engl. J. Med., 328: 847–854, 1993. tumors, presumably as a consequence of blocking the auto- 2. Suzuki, T., Bogenmann, E., Shimada, H., Stram, D., and Seeger, phosphorylation and signaling of the Trk receptor(s) being R. C. Lack of high-affinity nerve growth factor receptors in aggressive neuroblastomas. J. Natl. Cancer Inst., 85: 377–384, 1993. expressed. 3. Kogner, P., Barbany, G., Dominici, C., Castello, M. A., Raschella, MBLs are neuroectodermal tumors of the central nervous G., and Persson, H. Coexpression of messenger RNA for TRK protoon- system. Recent studies have suggested that these tumors also cogene and low affinity nerve growth factor receptor in neuroblastoma express members of the Trk family, and high expression of TrkC with favorable prognosis. Cancer Res., 53: 2044–2050, 1993. was correlated with improved survival (8, 9). Therefore, we 4. Borrello, M. G., Bongarzone, I., Pierotti, M. A., Luksch, R., Gaspa- tested the effect of CEP-751 on three MBL cell lines growing as rini, M., Collini, P., Pilotti, S., Rizzetti, M. G., Mondellini, P., and DeBernardi, B. TRK and RET protooncogene expression in human xenografts in nude mice in vivo. The tumors treated with CEP- neuroblastoma specimens: high-frequency of Trk expression in non- 751 were generally smaller than those in the control group. advanced stages. Int. J. Cancer, 54: 540–545, 1993. Studies of D283, which grew more slowly, did show a thera- 5. Yamashiro, D. J., Nakagawara, A., Ikegaki, N., Liu, X-G., and peutic effect of the compound. The difference in D341 was not Brodeur, G. M. Expression of TrkC in human neuroblastoma. Onco- statistically significant, and a difference in DAOY was not seen. gene, 12: 37–41, 1996. Two of the three lines showed minimal TrkA expression, and 6. Ryden, M., Sehgal, R., Dominici, C., Schilling, F. H., Ibanez, C. F., DAOY expressed none. Thus, its effect in D283 could be and Kogner, P. Expression of mRNA for the neurotrophin receptor trkC in neuroblastomas with favourable tumour stage and good prognosis. attributable to the effect of CEP-751 on TrkA or on other Br. J. Cancer, 74: 773, 1996. tyrosine kinases. Additional studies are required to fully char- 7. Nakagawara, A., Azar, C. G., Scavarda, N. J., and Brodeur, G. M. acterize these possibilities. Expression and function of TRK-B and BDNF in human neuroblasto- Although CEP-751 was developed because of its ability to mas. Mol. Cell. Biol., 14: 759–767, 1994. inhibit autophosphorylation and signaling of Trk receptors, it 8. Segal, R. A., Goumnerova, L. C., Kwon, Y. K., Stiles, C. D., and also inhibits platelet-derived growth factor, epidermal growth Pomeroy, S. L. Expression of the neurotrophin receptor TrkC is linked to a favorable outcome in medulloblastoma. Proc. Natl. Acad. Sci. USA, factor receptor, and protein kinase C activity (10) and may 91: 12867–12871, 1994. inhibit other, as yet unidentified tyrosine, serine, or threonine 9. Grotzer, M. A., Janss, A. J., Fung, K-M., Rorke, L. B., Sutton, L. N., kinases as well. Furthermore, because this is a small, lipophilic Cnaan, A., Phillips, P. C., Lee, V. M-Y., and Trojanowski, J. Q. molecule, it should enter cells readily and could have wider Neurotrophin receptor trkC predicts good clinical outcome in medullo- effects than its ability to inhibit selected tyrosine kinases. De- spite the potential for nonselective activity, the toxicity of this compound appears to be minimal at these concentrations. The compound has a short half-life in animals, requiring the twice- 6 S. J. Miknyoczki, A. Klein-Szanto, C. A. Dionne, and B. A. Ruggeri. daily treatment, but the half-life in humans is much longer, The Trk tyrosine kinase inhibitor CEP-701 (KT-5555) exhibits signifi- which may enhance its therapeutic effect (23). Also CEP-701, cant antitumor efficacy in preclinical xenograft models of human pan- an analogue of CEP-751, is presently in Phase I clinical trials in creatic ductal adenocarcinoma, submitted for publication.

blastoma/primitive neuroectodermal brain tumors. Proc. Am. Assoc. (KT6587) on human neuroblastoma (NBL) xenografts expressing TrkB. Cancer Res., 40: 606, 1999. Proc. Am. Soc. Clin. Oncol., 17: 352, 1998. 10. Camoratto, A. M., Jani, J. P., Angeles, T. S., Maroney, A. C., 19. Brodeur, G. M., Maris, J. M., Yamashiro, D. J., Hogarty, M. D., and Sanders, C. Y., Murakata, C., Neff, N. T., Vaught, J. L., Isaacs, J. T., and White, P. S. Biology and genetics of human neuroblastomas. J. Pediatr. Dionne, C. A. CEP-751 inhibits Trk receptor tyrosine kinase activity in Hematol. Oncol., 19: 93–101, 1997. vitro and exhibits anti-tumor activity. Int. J. Cancer, 72: 673–679, 1997. 20. Brodeur, G. M., Nakagawara, A., Yamashiro, D. J., Ikegaki, N., 11. Dionne, C. A., Camoratto, A. M., Jani, J. P., Emerson, E., Neff, N., Liu, X-G., Azar, C. G., Lee, C. P., and Evans, A. E. Expression of TrkA, Vaught, J. L., Murakata, C., Djakiew, D., Lamb, J., Bova, S., George, TrkB and TrkC in human neuroblastomas. J. Neuro-Oncol., 31: 49–55, D., and Isaacs, J. T. Cell cycle-independent death of prostate adenocar- 1997. cinoma is induced by the trk tyrosine kinase inhibitor CEP-751 (KT6587). Clin. Cancer Res., 4: 1887–1898, 1998. 21. Acheson, A., Conover, J. C., Fandl, J. P., DeChiara, T. M., Russell, 12. Sterling-Levis, K., White, L., Trickett, A. E., Gramacho, C., M., Thadani, A., Squinto, S. P., Yancopoulos, G. D., and Lindsay, R. M. Pittman, S. M., and Tobias, V. Heterotransplantation of early B-lineage A BDNF autocrine loop in adult sensory neurons prevents cell death. acute lymphoblastic leukemia using a solubilized attachment matrix Nature (Lond.), 374: 450–453, 1995. (Matrigel). Cancer Res., 53: 1222–1225, 1993. 22. Camoratto, A. M., Jani, J. P., Murakata, C., Vaught, J. L., and 13. Fridman, R., Kibbey, M. C., Royce, L. S., Zain, M., Sweeney, M., Dionne, C. A. CEP-751, a selective Trk receptor tyrosine kinase inhib- Jicha, D. L., Yannelli, J. R., Martin, G. R., and Kleinman, H. K. itor, possesses selective anti-tumor action. Proc. Am. Assoc. Cancer Enhanced tumor growth of both primary and established human and Res., 38: 470, 1997. marine tumor cells in athymic mice after co-injection with Matrigel. 23. Bhargava, P., Marshall, J., Dahut, W., Rizvi, N., Dordal, M., J. Natl. Cancer Inst., 83: 769–774, 1991. Samara, E., El-Shourbagy, T., Ness, E., Bischoff, J., and Hawkins, M. 14. Nakagawara, A., Liu, X-G., Ikegaki, N., White, P. S., Yamashiro, NCI-EORTC Symposium on new drugs in cancer therapy, Amsterdam, D. J., Nycum, L. M., Biegel, J. A., and Brodeur, G. M. Cloning and June 16–19, 1998. chromosomal localization of the human TRK-B tyrosine kinase receptor 24. George, D. J., Dionne, C. A., Jani, J., Angeles-Chundura, T., Lamb, gene (NTRK2). Genomics, 25: 538–546, 1995. J., and Isaacs, J. T. Sustained in vivo regression of Dunning H rat 15. Gavrieli, Y., Sherman, Y., and Ben-Sasson, S. Identification of prostatic cancers treated with combinations of androgen ablation and programmed cell death in situ via specific labeling of nuclear DNA Trk kinase inhibitors, CEP-751 (KT-6587) or CEP-701 (KT-5555). fragmentation. J. Cell. Biol., 119: 493–501, 1992. Cancer Res., 59: 2395–2401, 1999. 16. Wilcoxon, F. Individual comparisons by ranking methods. Biomet- 25. Mauer, A. M., Vogelzang, N. J., Janish, L., Mani, S., Samara, E., rics, 1: 80–83, 1945. and Ratain, M. J. Phase I study of CEP-2563 dihydrochloride 17. Matthews, D. E., and Farewell, V. T. Using and Understanding (CEP2563), a tyrosine kinase inhibitor, in patients with solid tumors. Medical Statistics, Ed. 3. Basel, Switzerland: S. Karger AG, 1996. NCI-EORTC Symposium on New Drugs in Cancer Therapy, Amster- 18. Evans, A. E., Kisselbach, K. D., Camoratto, A. M., Liu, X., Eggert, dam, June, 16–19, 1998. A., Ikegaki, N., Dionne, C., and Brodeur, G. M. The effect of CEP-751

Downloaded from clincancerres.aacrjournals.org on September 25, 2021. © 1999 American Association for Cancer Research. Antitumor Activity of CEP-751 (KT-6587) on Human Neuroblastoma and Medulloblastoma Xenografts

Audrey E. Evans, Kristin D. Kisselbach, Darrell J. Yamashiro, et al.

Clin Cancer Res 1999;5:3594-3602.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/5/11/3594

Cited articles This article cites 22 articles, 7 of which you can access for free at: http://clincancerres.aacrjournals.org/content/5/11/3594.full#ref-list-1

Citing articles This article has been cited by 12 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/5/11/3594.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/5/11/3594. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.