k~mop. J. Cancer Vol. 17, pp. 129. 142 0014-2964/81,/021/1-0129 S02 1)0/0 Pcwgamon Press Ltd. 1981. Printed in Great Britain Perspectives in Cancer Research Current Results of the Screening Program at the Division of Cancer Treatment, National Cancer Institute* ABRAHAM GOLDIN,? JOHN M. VENDITTI,? JOHN S. MACDONALD,t FRANCO M. MUGGIA,+ + JANE E. HENNEY? and VINCENT T. DEVITA, Jr.? ?Division of Cancer Treatment, National Cancer Institute, Bethesda, MD 20205, U.S.A. +Division of Oncology, New York University Medical Center, 550 First Avenue, New York, NY 10006, U.S.A. Abstract--The prospective screening program at the Division of Cancer Treatment, National Cancer Institute, has now been in operation for several years and is making steady progress in the identification of new synthetic compounds and natural products of potential interest for the clinic. Data are presented on four categor#s of drugs that have been tested in the new screening panel: (a) clinically established antitumor agents; (b) new drugs and drugs for which there is renewed clinical interest based on activity in the new screen and previously inadequate clinical trial; (c) drugs in the initial phases of clinical tria# (d) compounds in development. An analysis of the data is presented, taking into account a series of important questions that are being addresked prospectively to the new screen. Although the ability to provide definitive answers must await feedback from clinical testing of compounds recommended by the screen, some generalizations appear to be emerging, and these are discussed. A comparison is made of the activity of drugs in the treatment of human tumors growing in two sites, subcutaneously and under the renal capsule. The subrenal capsule model appears to be somewhat more sensitive to drugs than the subcutaneous model and may provide certain advantages for initial panel testing. Attention is drawn to the potential usefulness in a screening program of the newly developed clonogenic techniques for growing human tumors. The screening program at the Division of Cancer Treatment is viewed as a dynamic entity, subject to modification in accordance with acquired experience. INTRODUCTION FOR CERTAIN types of cancer, chemotherapy has been capable of rendering patients free of disease, with achievement of a normal life span (Table 1) [1, 2]. However, this re- sponsive category does not include the most frequently encountered forms of malignant tumor and although with the availability of new drugs and the use of combinations of drugs and combined modalites significant re- sponses are being obtained for the common solid tumors [1, 2], there remains a great need for new and more effective antitumor agents. Accepted 26 August 1980. *This paper was presented in part at the N.C.I.- E.O.R.T.C. Conference on New Drugs in Cancer Therapy, 18-19 October, 1980, Brussels, Belgium. 129 It was this need which in 1975 prompted a reexamination of the screening systems at the Division of Cancer Treatment, National Cancer Institute, and led to the institution of Table I. Cancers in which drugs have been responsible for a fraction of patients achieving a normal life span Acute lymphocytic leukemia--pediatric :\cute myelogenous leukemia--aduh Hodgkin's disease Diltiase histiocytic lymphoma Nodular mixed lymphomas Burkitt's lymphoma Ewing's sarcoma Rhabdomyosarcoma Wilms' tumor Choriocarcinoma Testicular cancer Ovarian cancer See [2].
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 1 of 14
`
`

`
`130 Abraham Goldin et al. a new prospective screening program [3 I. A serious possible lesion in the extant screening program appeared to be the preferential selec- tion of drugs active against rapidly growing tumors. Attention was therefore focused on the utilization of slow growing tumors for drug selection and evaluation. The availab- ility of athymic (nude) mice capable of sup- porting the growth of slow growing human tumors facilitated the institution of a balanced screening program incorporating both murine and corresponding human tumors. The new screening program has been mak- ing steady progress since 1975 in the testing of synthetic compounds and natural products and in the identification of new drugs of potential interest for further development, characterization and clinical evaluation. It is the purpose of this report to summarize the results of the program, to assess the status of its acomplishment and to indicate new direc- tions under consideration, as part of an evolv- ing dynamic approach to the screening for new and more effective antitumor agents for the clinic. A number of important questions, such as those listed below, have been ad- dressed to the new screen. (1) Does the new screen increase the yield of true positive compounds (active in the screen and active in the clinic)? (2) Does extensive and/or broad spectrum activity in the screening panel result in in- creased probability of clinical antitumor effectiveness ? (3) Do human tumor xenografts and animal tumor screens select the same or different drugs as active? (4) Are the xenograli positives more active in the clinic than those selected by animal screens ? (5) Does the screen reduce the number of false positives (active in the screen but in- active in the clinic)? (6) Does it reduce the number of false negatives (inactive in tile screen, but active in the clinic)? (7) Is there a correspondence of activity against animal tumors and/or human tumor xenografts with activity against clinical tumors for specific histologic types or specific organ systems ? (8) Are compounds that bypass the P388 prescreen because of activity in other screen- ing programs or in selected biochemical or biological assays more effective in the screen- ing panel and in the clinic than compounds initially selected for further testing by the prescreen? 19) What contribution will tile utilization of tile new screening panel make to prediction of clinical effectiveness of new drugs with respect to structure-activity analysis, analogs of known antitumor agents, and mathematical approaches to activity prediction? METHODOLOGY A schema of the new prospective screen is shown in Fig. 1 [2-5]. Prior to initiation of the new prospective screen, the testing level in the Division of Cancer Treatment program had been approximately forty thousand new materials per year, but because of the more extensive ettbrt of testing involved in tile new screen the number was reduced to fifteen thousand materials per year. The compounds to be subjected to screening are no longer selected entirely at random but rather on the basis of review of the world's literature and through voluntary submissions of compounds of potential interest. These compounds are tested in a prescreen in vivo against leukemia P388. All of the compounds demonstrating activity against leukemia P388 are then tested in a panel of tumor screens including mouse colon, human colon xenograft, mouse breast, human breast xenograft, mouse lung, human lung xenograft, B16 melanoma in the mouse and leukemia L 1210 in the mouse. Compounds of interest because of reported activity in other antitumor screening pro- grams and compounds selected on the basis of" biochemical or biological assays may bypass the P388 prescreen and go directly to testing in the screening panel. Although they are incidentally also tested in the P388 system, activity in that system is not requisite for testing in the panel. Natural product isolates are tested in vivo against leukemia P388 and also in vitro in the KB tissue culture system, and those which demonstrate activity are then tested in the entire screening panel. Approximetaly 500 or more compounds per year are becoming eligible for testing against the Division of Cancer Treatment screening panel. The tumor systems currently being em- ployed are shown in Table 2. They include leukemia P388, L1210 leukemia, B16 mel- anoma, Lewis lung tumor, colon 26 tem- ployed for special comparisons), colon 38 and CD8F 1 mammary tumor in mice, and the human tumor xenografts mammary MX-1, lung LX-1 and colon CX-1. Included also are
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 2 of 14
`
`

`
`Results of the Screening Program at N.C.L 131 SYNTHETIC COMPOUNDS t ~15,000 PER YEAR CURRENT DCT STANDARD SCREEN ACTIVE IN PM8 PRESCNEEN OR ACTIVE IN OTHER BIOLOGIC OR BIOCHEMICAL SYSTEMS -COdiP~J~NDS t~ PANEL MOUSE TUMORS LUNG COLOR BREAST LEUKEMIA MELANOMA HUMAN TUMOR XENOGHAFTS LUNG COLOR BREAST Fig. 1. Flow o3" drugs through the Division of Cancer Treatment screens. Table 2. Tumor panel systems Mammary Lung Cohm B16 Lewis CDSF] xenografi xenograft xenograft LI210 P388 Melanoma hmg Colon 26 Colon 38 Malnmary MX-1 LX-I CX-I Host CDF, or BDF1 Inoeulum IO s Ascites Site IP Parameter Mean survival time Activity T/C criteria _> 125% CDF x BDF] BDF 1 CDF t BDF x CD8F l or or BDF 1 B6C3 106 1:10 Homo- 1 x l0 s 1% Fragment 5 x l0 s Ascites genate Viable Brei cells cells IP IP IV IP SC SC Median Tumor Survival r ~ weight time inhibition T/C T/C T/C T/C T/C T/C _--_ 120°o ~ 125°.o => 140°o > 130°,o <42'!o =<420o Nu/Nu Nu/Nu Nu/Nu Swiss Swiss Swiss Fragment Fragment Fragment SC; SC; SC; SRC SRC SRC T/C T/C T/C <42% <42'~b; <42'!~,; <=20". <20% < 20"/,, the site of inoculation, the parameter of effect and criteria of activity. The protocols for screening against leukemias L1210 and P388, B16 melanoma and Lewis lung carcinoma have been de- scribed previously [6]. The origins and the experimental methods employed in the screen- ing against the carcinogen-induced transplant- able tumors colon 26 and 38 were reported by Corbett et al. [7] and the spontaneous mam- mary carcinoma in CD8F 1 mice was de- scribed by Martin et al. [8]. In the screening with the CD8F 1 mammary carcinoma, the first generation transplant is employed. The human tumor xenografts CX-1, MX- 1 and LX-1 are carried in serial transplantation in athymic mice. The CX-1 tumor model was initiated by A. Bogden at the Mason Research Institute. The MX-1 and LX-I xenografts were developed by B. Giovanella at the Stehlin Foundation for Cancer Research. The biological characteristics of the tumors that are included in the Division of Cancer Treatment tumor panel are shown in Table 3 [3]. With the human tumor xenografts, the pri- mary parameter of response is exten.t of in- hibition of tumor growth as compared with controls, with treatment initiated when the tumors are well established and palpable at the site of implantation. Because of the re- latively slow growth of the human tumor xenografts at the subcutaneous site of in- oculation each test requires approximately 60- 90 days for accomplishment. This demand in time of observation necessitated a reduction of the number of models for chemotherapy trials for established tumors. In order to minimize the time required for testing, and to permit a broadening of the base of drug evaluation and more detailed study of the matching of therapy to individual patients, further investigations are ongoing in the program, employing human tumors grow- ing in various sites in the athymic animal. Attention is focused on optimization of tumor
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 3 of 14
`
`

`
`132 Abraham Goldin et al. Table 3. Biological characterization oJ tumors included in the DCT tumor panel Tumor and code Host of origin Origin of tumor Histological description Miscellaneous Human Colon CX-I Isolated in tissue Human co}on Adenocarcinoma culture, suhse- of the colon quently maintained in nude mice Breast-MN-I isolated in nude Hmuan breast lntihrating duct cell mice carcinollqa I,ung-LX-1 Isolated in nude Human lung ()at cell carcinoma mice Mouse C(,h~n-C 26 BALB/c mouse Induced by chemical Unditti:rcntiated colon Very high rate of carcinogen a%methyl- mucosal carcinoma metastases ~-nitrosourethane Colon-C 38 Cs,BL/6 mouse Induced by chemical Colon adenocarcinoma Very low rate of carcinogen, 1,2- metastases dimethylhydrazine Melanoma BIg Cs,BL/6 mouse Spontaneous at base Melanoma of ear Irons z O~cwis lung) CsTBL/6 mouse Spontaneous in the lun~ Anaplastic carcinoma Metastases Breast CD8F 1 mouse Spontaneous Mammary adenocarcinoma Leukemia L1210 DBA/2 mouse Chemically induced wilh Lymphocytic leukemia methylchnlanthrenc Leukemia P388 DBA/2 mouse Chemically induced with Lymphocytic leukemia methylcholanthrene take, rate of growth, precision of measure- ment, extent of metastasis, uniformity of sur- vival time and other parameters that may lend themselves to precise quantitation of the inhibitory effect of antitumor agents. One of these systems, the subrenal capsule model, is under intensive investigation. The technique employed and preliminary data for the subre- nal capsule system have been reported by Bogden et al. [9]. The technique [9] involves insertion of small fragments (approximately 1.0mm 3) of human tumor xenografts under the renal capsule, where there is a rich vas- cular bed, ensuring adequate nutrient for tu- mor growth and ready drug delivery. Employing a Stereoscopic microscope in which a micrometer disc is inserted into one eye- piece, it is possible to measure, in situ, the size of the initial graft and the ultimate size achieved at the termination of the experiment. An assay time fi~ame of eleven days was selected since it was long enough to permit measurement of ex[ent of growth and of drug- induced inhibition of the human tumor xenografts. The screening data tbr the xenograt~ models in which the tumors are inoculated subcu- taneously were obtained from D. Houchens and T. Ovejera at the Battelle Columbus Laboratories. The screening data tor the xeno- grafts inoculated under the renal capsule were obtained from A. Bogden at the Mason Research Institute. In the present analysis the criteria tor drug activity against human tumor xenografts im- planted subcutaneously and under the renal capsule are those in current use by tile Division of Cancer Treatment. These are 58°{, inhibition from controls ~ T/C°o <= 42) tbr the sub- cutaneous model and 80"~, inhibition (T/C",, <20) for the subrenal capsule model. The investigators who have used these models most extensively--Ovejera et al. [10] in the case of the subcutaneous model and Bogden et al. [9] in the case of the subrenal capsule model-- have employed various cutoff points to dis- tinguish 'activity' from 'inactivity'. Also, the activities listed herein (Tables 4, 5-8, 10 and 11 and Figs. 24) as reported by the in- vestigators were, derived using diit~rent methods of computation. For the subcutaneously implanted tumor model, Ovejera et al. [10] estimated tumor weight (W) in mg from caliper measurements according to the for- lnula W= ~aZx b)/2, where a is the width and b is the length in mm. In addition, in an effort to standardize variability in tumor size among test groups at the initiation of treat- ment, these authors calculated relative weights (RW) using the formula RW= Wi/Wo, where Wo is the mean tumor weight of a group at the beginning of treatment and Wi is the mean tumor weight at any subsequent time. A significant response to treatment is indicated when a test group shows an RIV=<42°Jo of that of the control at any time during a specified range of days after the last treatment. In contrast, Bogden et al. [9] using the subrenal capsule model measured tumor length (b) and width (a) in ocular micrometer units (OMU), the micrometer disc of the
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 4 of 14
`
`

`
`Results of the Screening Program at N.C.I. 133 microscope eye-piece having been calibrated so that l00MU = 1.0 mm. Treatment activity was based on the change in average tumor diameter over the prescribed course of treat- ment compared with the change in average control diameter. Thus, T/C% = D T/DC x 100, where DT is the mean tumor diameter (a+b)/2 of the treated group at the end of treatment less the mean tumor diameter at the beginning of treatment, and where DC is the change in mean tumor diameter of con- trols over the same period. RESULTS AND DISCUSSION The data in the screening panel for a series of the more established antitumor agents are summarized in Table 4. In a previous ret- rospective analysis it had been suggested that compounds that are active in a number of screening systems in rodents could have more likelihood of demonstrating activity against hematologic malignancies and solid tumors in the clinic [11, 12]. Also, the more extensive the response in the tumor systems, the greater the possibility appeared to be that the com- pounds would exert clinical antitumor ac- tivity. That such high and broad spectrum activity in the screening panel could be in- dicative of greater probability of antitumor effectiveness in the clinic is reflected also in the data of the new screening panel where for the more established clinically active anti- tumor agents high and broad spectrum ac- tivities were obtained ~Table 4). Taking into account the total number of animal tumors plus human tumor xenografts, including the subcutaneous and subrenal capsule sites, all but one of the drugs (L-asparaginase) were active in greater than 45% of the tumor systems, ranging from 46% of the tumor systems for methotrexate, 6-mercaptopurine, adriamycin and bleomycin to over 80% for nitrogen mustard, melphalan, cyclo- phosphamide, mitomycin C, CCNU and cis- platinum II (Table 4). Overall, the animal tumor systems rated a higher percentage of drugs as active than did the human tumor xenografts in either the subcutaneous or sub- renal capsule sites (Table 4). The re- duced sensitivity qf the human tumor xeno- grafts could provide an important advantage in drug selection if it is also accompanied by the identification of new types of antitumor drugs. Table 4. Activity in the new screening panel for clinically active antitumor agents* NSC number Drug BI 6 Active systems Mela- Lewis Colon Colon Subcutaneous Subrenal capsule Per 1,1210 P388 noma hmg 26 38 CD8F 1 MX-I LX-I CX-1 MX-1 t,X-I CX-1 Number cent 740 Methotrexate 272 296 120 752 6-Thioguanine 228 14__5 12__.88 755 6-Mercaptopurine 263 15.....0 13..._4 762 Nitrogen mustard 30_.._4 25_._!1 23_..)5 3053 Actinomycin D 17..33 61_.88 19_..[1 3088 Chlorambucil 14.99 171 140 8806 Melphalan 23._._7 28_..!1 25_..27 13875 Hexamethyl- melamine 132 117 126 19893 5-Fluorouracil 180 220 14.O 26271 Cyclophosphamide 236 > 300 17_...66 26980 Mitomycin C 17....88 242 18....11 45388 I)TIC 160 130 14_.~5 49842 Vinblastine 154 252 280 63878 Cytosine arabinoside 285 255 15_...99 67574 Vincristine 14..__7 30_..00 18...._9 77213 Procarbazine 188 180 168 79037 GCNU 24._..33 27__.88 28_._7 82151 Daunomycin 161 >266 >350 148 106 76 192 > 200 l0 121 246 10 125 262 30 124 154 15 125 19__0 59 154 >309 20 3.4 41 66 93 37 54 6/13 46 6_ 71 76 81 70 NT 121 7/13 54 21 75 77 99 90 61 103 6/13 46 11 NTt NT NT NT NT NT 6/7 86 0 70 46 73 5 NT IO 9/12 7:) 16 24 46 60 - 5....._~5 16 NT 8/12 67 _.1 ~ 35 101 - 2_...~5 47 -13 11/13 85 202 150 12 15___oo 20o o 222 16___~5 2 14___22 18___7 2 267 126 13 111 18_~8 o_ 16 10 81 85 - 1_..._7 72 12 9/13 69 0 73 70 88 56 36 60 7/13 54 0 1 37 113 -3......27 41 0 11/13 85 16 1 41 62 NT NT NT 9/10 90 20 55 40 92 37 NT NT 7/11 64 3 NT 117 119 NT NT NT 6/9 67 14....33 164 34 15 71 102 73 NT NT NT 7/10 116 130 23 _7 8 69 89 NT NT NT 7/10 15...._4 115 38 102 46 8 60 NT - 1.,..._.7 NT 7/11 253 >363 0 15 43 15 77 19 --11 NT 10/12 122 ~ 88 55 NT NT NT NT NT NT 4/7 95441 Methyl CCNU 109229 L-Asparaginase 119875 C/s-Platinum 122819 VM 26 123127 Adriamycin 125066 Bleomycin 178248 Chlorozotocin 409962 BCNU Number of drugs active Percentage >310 >275 >279 >242 345 4 7 30 48 83 NT NT NT 8/10 117 15_.._4 104 NT 113 109 55 NT NT NT NT NT NT 1/7 207 264 28.88 26....]1 245 2_7 0 20 86 66 - 1_.....7 6 NT 10/12 239 > 350 > 285 113 2~ 48 16 NT NT NT NT NT NT 5/7 > _3..~ >300 > 30_...Q >252 310 68 1 68 73 72 59 43 37 6/13 120 193 144 14_..22 116 IO 9 2_7 83 51 66 26 51 6/13 > 43__.9 > 25__!1 > 35._.~6 164 > 322 15 _9 51 68 75 NT NT NT 7/10 >._5,fi.3. >298 267 > 30....~5 >3~ 36 6 43 85 68 17 -- 28 73 9/13 24/26 25/26 24/26 17/25 21/26 20/26 23/26 10/21 7/22 0/22 8/15 5/13 4/11 92 96 92 68 81 77 88 44 32 0 53 38 36 70 70 64 83 57 80 14 83 71 46 46 70 69 *Underlining means drugs are active. tNT = Not tested.
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 5 of 14
`
`

`
`1,74 Abraham Goldin et al. The P388 system was the most sensitive, in- dicating activity for 96(I0 of the drugs. The L1210 and B16 melanoma systems were next in responsiveness, each indicating activity for 920. o of the drugs, whereas the Lewis lung system rated 68°o of the drugs as active. Of the three types of xenografts employed the mammary tumor xenograft MX-1 was the most responsive, yielding 44°o active with the subcutaneous tumor and 53°t, with the subre- nal capsule tumor. The LX-1 lung xenografi was second in responsiveness, giving 32°i, active with the subcutaneous site and 38'!0 active, with the subrenal capsule site. The CX-1 tumor was the least responsive, resulting in zerot!'o selection as active with the sub- cutaneous site, but 36t{() as active with the subrenal capsule site. The latter result is suggestive of overprediction for this human colon tumor with the criterion of effectiveness employed. In order to examine the question of wheth- er there is an~ correspondence of activity in animal tumors and human tumors for specific types of tumor, the screening panel data are listed for the breast tumor, lung tmnor and colon tumor models for drugs that have been reported to have activity against breast, lung and colon carcinoma in the clinic (Tables 3, 6,7). For the drugs reported as active against breast tumor in the clinic the CD8F 1 mam- mary tumor model in the mouse identified activity for 3/3 (Table 5). The MX-1 sub- cutaneous model identified 2/3 drugs, tailing to identi~" adriamycin as active. The subrenal capsule model classified only 1/3 drugs as active, missing methotrexate and adriamycin. The active and marginally active drugs were identified in essentially the same proportion. For drugs reported to be active or mar- ginally active against small cell lung cancer, the Lewis lung (IV) tumor system identified all as active whereas the lung tumor xenograft LX-1 both subcutaneously and in the subre- hal capsule indicated activity for a reduced nulnber [Table 6). In the active drug cat- cgor~, l,X-1 SC and LX-1 SR(', systems failed to identify adriamycin as active. The LX- 1 SRC system also rated cyclo- phosphamide and methotrexate as negative. For the drugs listed as marginally active, CCNU was identified as active by the LX-1 SC and LX-1 SRC systems, but hexamethyl- melamine was rated as inactive. No drugs are listed as definitely active against epidermoid lung cancer. Of the 5 drugs listed as marginally active, only nit- "l able 3. Activity in breast tumor model~ oj the screen- mg panel. Druga with reported activity in the treatment o] t:linieal brea.st cancer Breast tumor models CD8Fx MX-I MX-I SC SRC Active in clinic* Cyclophosphamide 0 + + l + -- 37 + Methotrexate 20 + 34 + 93 - Adriamycin 1 + 68 - 59 - Actives in tumor models 3/3 2/3 1/3 Marginal activity in clinic + 5-Fluorouracil 0 + 73 - 56 - Melphalan l + 2 + - 35 + Mitomycin C l + 4.1 + NT Actives in tumor models 3/3 2/3 I/2 *Active in clinic: Drug consistently produces partial regressions and at least occasionally produces complete regressions. It may improve survival in some patients. +Marginal activity in clinic: Drug has been observed to produce only partial regressions of tumor and has no definite efti~ct on patient survival. +Activity followed by rating of + (active) or - (inactive). Table 6. Activity in lung tumor models of the screening panel. Drugs with reported activity in the treatment of clinical lung cancer Lung tun/or inodels l,cvds hmg I,X-I LXd SC SRC SMALl, CELL Acti'~e in clinic* Cyclophosphamide 222 + 37 + I 1 Methotrexate 148 + 41 + 37 - Adriamycin > 232 + 73 /'I - Procarbazine 154 + 8 + 17 + \P 16 No data Attires in tumor models 4/4 3/4 1~4 Marginal activity in the clime* CCNU 253+ 15+ II + Hexamethylmelamine 202 + 81 72 - Actives in tumor models 2/2 1/2 1 ~2 EPIDERMOID Marginal activity in clinic* Cyclophosphamide 222 + 37 + 41 - Methotrexate 148 + 41 + 37 Ariamycin > 252 + 73 - t3 Nitrogen mustard 125 - NT NT C;~-Platinum 11 261 + 86-- 6+ Attires in tumor models 4/5 2/4 1/4 *See footnotes to Table 5. Table 7. Activity in colon tumor models of the screening panel. Drugs with reported activity in the treatment of clinical colon cancer (~olon tumor models CX-1 C,X-I Marginal activit?, m clinic* Colon 26 Colon 38 SC SRC 5-Fluorouracil 200 + 0 + 88 - 60 - Mitomycin C 187+ 9+ 62- NT Methyl CCNU 345 + 4 + 83 - NT Actives in tumor models 3/3 3/3 0/3 0/1 *Set' lbotnotes to Table 5.
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 6 of 14
`
`

`
`Results of the Screening Program at N.C.I. 135 rogen mustard was inactive in the Lewis lung model, but this drug was not tested in the LX-1 models. The LX-1 SC model failed to identify adriamycin and cisplatinum II as active and the LX-1 SRC system failed to identify cyclophosphamide, methotrexate and adriamycin as active. For the drugs 5-fluorouracil, mitomycin C and Methyl CCNU, reported as having mar- ginal activity against colon tumor in the clinic, the colon 26 and colon 38 tumor models identified 3/3 as active and the CX-1 SC model failed to identify any as active (Table 7). Only 5-fluorouracil was tested in the CX-1 SRC system and it was inactive. The lack of activity of these antitumor agents against the CX-1 human colon tumor model suggests that this system may have important discriminatory value in drug selection for treatment of this highly refractory clinical tumor category. Thus, the data (Tables 5, 6, 7) indicate only a partial correspondence between the activity in animal tumor models and that which might be expected in the clinic for specific histologic types. Since the xenograft models , particularly the colon tumor, are resistant to ther- apy, they may prove to be important in drug selection. The human tumor xenograft models do not appear to yield as many false positives as the animal tumor models (Table 4). Overprediction by the animal models was especially indicated for the CD8F t breast tumor model for clinical breast cancer, and by the colon 26 and colon 38 animal models for colon cancer. The Lewis lung carcinoma did not appear to yield as many false positives. Data in the screening panel for new drugs and drugs for which there is renewed clinical interest based in large measure on activity in the new screen are summarized in Table 8. Clinical data for these compounds are sum- marized in Table 9. 4'- (9-acridinylamino) methanesutfon-m- anisidide tAMSA), an acridine derivative [13], showed broad spectrum activity in the new screening panel, including activity against leukemias P388 and L1210, B16 melanoma, colon tumors 26 and 38 and the CD8F 1 mammary tumor. It was inactive, however, Table 8. New drugs and drugs of renewed clinical interest*. Activity in tumor models BI6 Mela- Lewis Colon Colon SC SRC Active systems NSC L1210 P388 noma lung 26 38 CD8F t MX-I l,X-I CX-I MX-I LX-I CX-I Number Percent 249992 AMSA 185 216 224131 PALA 120 135 7365 D-O-Norleucine 174 166 32946 Methyl G 176 176 157365 Neocarzinostatin 175 190 Number of drugs Active 4/5 5/5 Per cent 80 100 243 125 24_..! 25 ~ 74 75 67 63 96 NTt 6/12 50 192 215 19_._7 11 ! 35 41 78 32 80 76 8/13 62 112 121 162 14 3 11 2 56 -25 51 52 8/13 62 96 103 106 75 88 116 83 92 57 85 39 2/13 15 17....~5 109 20.__0 37 43 75 89 NT 51 41 69 5/12 42 3/'5 l/5 4/5 4/5 3/5 2/5 2/5 0/4 1/5 0,:'5 0/4 60 20 80 80 60 40 40 0 20 0 0 *Underlining means drugs are active. tNT = Not tested. Table 9. New drugs and drugs of renewed clinical interest. Clinical activity No. Patients CR PR MR 249992 AMSA Acute leukemia in adults 22 3 Acute leukemia in adults 13 2 Breast cancer 22 5 224131 PALA Non small cell lung cancer 21 3 Bladder cancer 10 2 7365 n-O-Norleucine Breast 14 2 Lung carcinoma 9 2 Hodgkin's disease 11" 2 32946 Methyl G Transitional cell carcinoma 2 l Colon 9 2 Esophageal 2 1 Renal cell 4 1 Adenocarcinoma unknown primary 3 1 157365 Neocarzinostatin Pancreatic 88 3 8 Gastric 217 2 4 Acute leukemia 76 11 4
`
`NOVARTIS EXHIBIT 2049
`Par v Novartis, IPR 2016-00084
`Page 7 of 14
`
`

`
`1,3'6 Abraham Goldin et al. against Lewis lung carcinoma and the human tumor xenografts. AMSA is being subjected to broad spectrum Phase II testing in the clinic and has been reported to be active in acute leukemia in adults [14, 15] and in breast cancer [15]. In initial studies it has also shown activity against lymphoma, ovarian and renal cell tumors. AMSA is a likely candidate for clinical investigation employing drug combinations. Phosphonacetyl-L-aspartic acid (PALA) is an inhibitor of aspartate transcarbamylase [16]. It showed only borderline activity in the treatment of leukemia P388 and little to no activity against leukemia L1210. It did, how- ever, show broad spectrum activity in the new screen, including activity against Lewis lung carcinoma [17, 18], B16 melanoma, co- lon tumors 26 and 38 and the mammary tumor CDSF 1. It also showed moderate ac- tivity against the mammary xenografl MX-1 and the lung xenograft LX-1. In addition, PALA has been reported to show broad spect- rum activity against a variety of experimental solid minors 119]. PALA has entered clinical trial, and to date minimal responses have been seen with non small cell lung tumor and bladder cancer [20]. Clinical interest was also renewed in the glutamine antagonist 6-diazo-5-oxo-~,- norleucine (DON) [21] as a result of its activity in the new screen. In addition to activity in leukemias P388 and LI210 it was active in colon 26, colon 38 and CD8F, mammary tumor, and the human tumor xcn- ografts MX-1 and LX-1. DON had been reported initially to produce partial remissions against breast tumor and lung carcinoma [22] and Hodgkin's disease [23] in the clinic. The drug has also evidenced activity againsl choriocarcinoma. A new Phase I study has been activated with DON, employing inter- mittent high doses. Renewal of interest in the drug methyl glyoxal bis guanyl hydrazonc (Methyl G) is attributable to the influence of scheduling characteristics. In the new screen Methyl G was observed to be active against leukemias 1,1210 and P388, with possible activity against the CX-1 xenograft in the subrenal capsule site. A complete remission has been reported recently against transitional cell carcinoma and also against adenocarcinoma, and partial remissions against colon, esophageal and renal (:ell tumors [241 . Neocarzinostatin was actiw" against tcukemias P388 and L1210, BI6 melanoina and colon tumors 26 and 38 of the new screen. It was inactive against Lewis lung carcinoma and the human tumor xenografts. In the clinic it has evidenced activity against pancreatic [25] and gastric tumors [26], acute leukemia [26], and hepatoma. As for the well established drugs, so too with the new drugs and drugs of renewed clinical interest, high and broad spectrum activity were observed in the experimental tumor systems (Table 8). Methyl G provided the only marked exception, being active in <rely 2/13 of tile test systems. Again, in gem> ral, a higher incidence of actives was observed fbr the animal test systems as. compared with the human tumor xenografts. A listing of the screening data tbr drugs about to enter or primarily in initial clinical trials is presented in Table 10. Many of these drugs elicited high and broad spectrum ac- tivity in the new screen. 'Bypass' compounds include Baker's antilol (NSC 139105) and 2'deoxycoformycin (NSC 218321), which were not active in any of the systems of the new screening panel, and dich- loroallyl lawsone (NSC 126771), which had only marginal activity in the P388 system. The remaining compounds listed in Table 10 were active in three or more test systems of the new screen. Data in the new screening panel tbr com- pounds in development are listed in Table 11. High and broad spectrum activity were obser- ved for many of these materials. With some exceptions such as the 'bypass' compounds. ADI tEHNA) {NSC 263164), an adenosine deaminase inhibitor, and trimethyl melamine (NSC 57552), an older drug of interest, the compounds were active in at least one and usually more systems of the new screen. Since in general there is greater activity against the murine tumors than against the human tumor xenografts growing in athymic mice it may be of interest to focus attention on drugs that have demonstrated activity against the human tumor xenografts. Among the drugs about