`
`Responsiveness of Human Gastric Tumors Implanted in Nude Mice to Clinically
`Equivalent Doses of Various Antitumor Agents
`
`Makoto INABA,“ Tazuko TASHIRO,“ Tomowo KOBAYASHI,*1' *3 Yoshio SAKURAI,*1' *4
`Koji MARUO,"‘2 Yasuyuki OHNISHI,"‘2 Yoshito UEYAMAM and Tatsuji NOMURA*2
`*’ Cancer Chemotherapy Center, Japanese Foundation for Cancer Research, Kami—Ikebukuro,
`Toshfma-ku, Tokyo 1 70 and “Central Institute for Experimental Animals, Nogawa, Takatsu-ku,
`Kawasaki, Kanagawa 2] I
`
`To reproduce clinical effects of various antitumor agents in the human tumor/nude mouse
`model, we investigated the responsiveness of 11 lines of human gastric tumor xenografts to doses
`of the agents pharmacokinetically equivalent to the respective clinical doses, which we designated
`the “rational dose” (RD). We found that the response rates to mitomycin C, 3-[(4-amino-2-
`methyl-S-pyrirnidinyl]methyl-1-[2-chloroethyl]-1-nitrosourea
`(ACNU),
`adriamycin,
`5-fluoro-
`uracil were 18%, and that to vinblastine was 30%; on the other hand, those to vincristine,
`methotrexate, and cyclophosphamide were poor. In contrast,
`in our previous study using the
`maximum tolerated doses, response rates to mitomycin C, ACNU, and vinblastine were as high as
`64—82%, and those to adriamycin and 5-fluorouracil were 18%. When these results were
`compared with the clinical response rates of gastric tumors, as a whole, the results with RD’s
`exhibited much better coincidence with the clinical data in terms of relative therapeutic potency,
`indicating the validity of the use of clinically equivalent doses instead of maximum tolerated doses
`in the human tumor model.
`
`Key words:
`Human gastric tumors — Nude mouse — Response rates — “Rational dose” —
`Clinical predictability
`
`In our study on experimental chemotherapy
`against a panel of human gastric tumors im-
`planted in nude mice,” we evaluated the anti-
`tumor effects in terms of response rates as in
`clinical chemotherapy, because it
`seemed
`dificult to reasonably evaluate the antitumor
`effect of a given agent by using a few tumor
`lines. However, even in such trials, we failed
`to reproduce the relative effectiveness of vari-
`ous antitumor agents which had been ob-
`served in clinical treatment of gastric tumors.
`These results strongly suggested to us the
`
`*3 Present address: Bioseience Research Laborato-
`ries, Sankyo Co., Ltd., Hiromachi, Shinagawa—ku,
`Tokyo 140.
`*4 Present address: Kyoritsu College of Pharmacy,
`Shiba-koen, Minato—ku, Tokyo 105.
`*5 Abbreviations used: VLB, vinblastine; VCR.
`vincristine; CPM, cyclophosphamide; S-FU, 5-
`fluorouracil; MTX, methotrexate; MMC, mito-
`mycin C; ADR, adriamycin; ACNU, 3~[(4-amin0-
`2-methyl—5-pyrimidinyl]methy-1-[2-chlor0ethyl]-
`l-nitrosourea hydrochloride; DDP, cisplatin; MTD,
`maximum tolerated dose; RD, rational dose.
`
`79(4)
`
`1988
`
`importance of using a dose of each antitumor
`agent pharmacologically equivalent to its clin-
`ical dose. In the preceding paper,” therefore,
`We described our attempt to find a dose that
`could reproduce in the nude mouse the
`clinically achievable plasma level of each
`drug, because such a dose, denoted as the
`“rational dose,” was considered to be most
`suitable for the treatment of human tumor-
`
`bearing nude mice.
`In the present study, we carried out experi-
`mental chemotherapy using the RD as a ther-
`apeutic dose against a panel of human gastric
`tumors implanted in nude mice and examined
`whether treatment with the RD could repro-
`duce the clinical effectiveness of various anti-
`
`tumor agents in terms of response rate.
`
`MATERIALS AND METHODS
`
`Antitumor Agents VLBM and VCR (Shionogi &
`Co., Osaka) and MTX (Lederle Japan Ltd.,
`Tokyo)
`for clinical use were purchased. 5-FU,
`MMC and ADR were kindly supplied by Kyowa
`Hakko Kogyo Co., Tokyo, as pure crystals for
`517
`
`Genentech 2098
`
`Celltrion v. Genentech
`
`|PR2017-01122
`
`Genentech 2098
`Celltrion v. Genentech
`IPR2017-01122
`
`
`
`M. INABA, ET AL.
`
`experimental use. CPM and ACNU were provided
`by Shionogi & Co. and Sankyo Co., Ltd., Tokyo,
`respectively, in a form for experimental use.
`All drugs were dissolved in sterile 0.85% NaCl
`solution just before use.
`The MTD’s of all these drugs were determined
`as maximum non-lethal doses by single intravenous
`injection (daily administration for 5 days only in
`the case of 5-FU and MTX). The test doses were
`increased at a constant ratio of 1.2.
`Human Tumor Xenograft Lines Eleven human
`gastric tumors established as xenografts in nude
`mice were used. Their characteristics,
`including
`histological types, prior chemotherapy, and growth
`rates, were presented in the previous paper.” NS-3
`and NS-S lines were kindly supplied by Dr. K.
`Nakatani, Nara Medical College, Nara, and St-4
`and St-40, by Dr. T. Kubota, Keio University,
`Tokyo.
`These tumors have been maintained by serial
`subcutaneous transplantation of 2 X 2 X 2 mm cubic
`fragments ‘ in the right
`subaxiliary region of
`athymic BALB/c-nu/nu mice (Clea Japan, Inc.,
`Tokyo) approximately every month.
`Mice were housed in ethylene oxide-sterilized
`filter-capped cages with 6°Co—irradiated (3 mega-
`rad) food and autoclaved water ad Iibi'tum. All
`cages were kept in laminar-air-flow units in our
`laboratory. Six-
`to eight-week-old female mice
`weighing about 25 g were used.
`
`Measurement of Tumor Size After the transplan-
`tation,
`the mice were observed and randomized
`into several experimental groups consisting of 6
`animals each after the tumors had reached palpable
`size. The tumor volume (V) was calculated by
`means of the equation
`V=LQXaxR
`where a and b are the experimental measurements
`in mm of length and width, respectively. Each
`tumor volume was then expressed as relative tumor
`volume (RV),
`
`RV:V,,/VD
`where V, is the tumor volume at day It and V0 is the
`initial tumor volume at the time when the treat-
`ment was started (day 0).
`Chemotherapy When the tumor volume reached
`1007300 mms, chemotherapy was initiated. The
`RD‘s or MTD’s determined in advance were used
`as therapeutic doses. Drugs were given intra-
`venously by daily injection for five days in the
`case of S-FU and MTX or by single administration
`for all other drugs. The dose of each drug (deter-
`mined as RD in the preceding study”) was as
`follows: MMC, 1.7 (mg/kg); CPM, 65; ADR, 12;
`VCR, 0.4; VLB, 2.6; S-FU, 19; MTX, 15. ACNU
`was given as 3 intermittent injections of 8 mg/kg
`(0 min), 2 (25 min) and 0.8 (70 min) in one day.
`Observation was continued for 3—4 weeks.
`
`Evaluation At any given experimental day, T/C
`(%) was expressed as the average of RV of the
`
`
`ACNU
`
`VLB
`
`CI).p
`E3:
`:2
`as
`E 8
`c,
`(DD
`L"
`>
`-.—e~e
`-H
`m s,
`.——
`g
`
`
`
`100
`
`50
`
`0
`
`100
`
`50
`
`o
`
`
`
`3
`
`9
`10
`
`"
`4
`
`7
`I
`
`0
`
`10
`
`20
`
`0
`
`10
`
`20
`
`0
`
`10
`
`20
`
`Days after treatment
`
`rag.1. Responsiveness to "rational doses” of MMC, ACNU, and VLB of a panel of human
`gastric tumor xenografts. MMC, ACNU, or VLB was intravenously injected at its “rational dose“
`when the tumor size had reached 100—300 mms. Tumor sizes were measured with calipers twice
`a week, and relative growth rates were obtained according to the procedure described in “Material
`and Methods.” Tumor lines are indicated by numbers as follows: 1, $02; 2, SC-4; 3, SC—6; 4, St-
`4; 5, NS-3, 6, NS—S; 7, 4-IST; 8, SC-7; 9, SC-9; 10, St-40; ll, St—lS.
`
`518
`
`Jpn. J. Cancer Res. (Gann)
`
`
`
`RESPONSES OF HUMAN GASTRIC TUMORS TO RATIONAL DOSE
`
`treated mice with respect to the control. The effec-
`tiveness of each drug was evaluated in terms of the
`T/C(%) value at day 14. Evaluation as “effective”
`was based on a T/C(%) of 50% or less, with
`statistical significance as determined by the Mann-
`Whitney U-test (P< 0.01, one-sided).
`
`RESULTS
`
`Responses of human gastric tumors to
`RD’s of MMC, ACNU, and VLB are pre-
`sented in Fig.
`1;
`these drugs demonstrated
`potent therapeutic effectiveness against most
`of these tumors when the MTD’s of the drugs
`were employed.” Since their RD’s were found
`to be much lower than the MTD’s,” tumors
`
`which did not significantly respond to the
`MTD of the drug were expected to be unre-
`sponsive to its RD as well. Therefore, in Fig.
`1, only the responses to the RD’s of those
`tumors that had responded positively to the
`MTD’s are shown.
`
`With MMC, only 2 out of 9 tumor lines
`examined exhibited a statistically significant
`decrease in growth rate to 50% or less of the
`control. In the case of ACNU, 2 out of 7
`tumor lines responsive to its MTD also sig-
`nificantly responded to its RD. 011 the other
`hand, 3 out of 7 tumor lines were responsive
`to VLB at its RD, although responsiveness
`of St-lS was at a relatively low level.
`
` 0
`
`n 7
`
`100
`
`50
`
`CPM
`
`0
`
`o
`
`10
`
`20
`
`(J1 O
`
`
`
`
`
`
`
`Relativegrowthrate(%ofcontrol)
`
`0
`
`10
`
`20
`
`0
`
`10
`
`20
`
`Days after treatment
`
`Fig. 2. Responsiveness to “rational doses” of CPM, ADR, S-FU and MTX of a
`panel of human gastric tumor xenografts. Experimental conditions and numbering
`of tumor lines were the same as shown in Fig. 1, except for 5 daily injections of 5-
`FU and MTX.
`
`79(4)
`
`1933
`
`519
`
`
`
`M. INABA, ET AL.
`
`In Fig. 2, tumor responses to the other 4
`drugs are presented. The tumor lines did not
`respond well to CPM even at its MTD, and
`neither of 2 tumor lines that responded to its
`MTD was responsive to its RD. Since the RD
`of ADR was estimated to be approximately
`equivalent to its MTD,“ responses of all 11
`human gastric tumor lines to 12 mg/kg of
`ADR are shown. Only two of the lines ex-
`hibited significant
`responses. When plasma
`levels of 5-FU and MTX were compared be-
`tween man and nude mouse, the level of nei-
`ther agent in the nude mouse reached those
`levels found in human patients given the clin-
`ical dose, even when the MTD’S were injected
`
`in the case of
`into the animals.“ Therefore,
`these drugs, it was impossible to observe the
`responses to their RD’s. Alternatively, we
`were obliged to examine the responses to their
`MTD’s. Two and 1 out of 11 tumor lines
`
`responses to S-FU and
`showed significant
`MTX,
`respectively, although their growth-
`inhibitory efl'ects were not so remarkable.
`With VCR, as previously reported," none
`of the human gastric tumors responded even
`to its MTD. Since its RD is estimated to be
`
`one-fourth of the MTD,” the tumors would
`be expected to be unresponsive to its RD as
`well. Therefore, the experiment with the RD
`of VCR was not done.
`
`Table 1.
`Responses of a Panel of Human Gastric Tumor Xenografts to Rational Doses of
`
`Various Antitumor Drugs
`
`Relative tumor growth rate (T/C%)"J
`
`VCR
`VLB
`S-FU
`MTX
`ACNU
`ADR
`
`MMC
`
`CPM
`
`Human
`tumor line
`
`67
`:19
`E")
`(—)
`76
`65
`(— )°’
`84
`SC-2
`101
`83
`(i)
`(i)
`51
`(—)
`(—)
`66
`SC-4
`78
`58
`65
`(i)
`87
`g
`(i)
`E
`SC-6
`72
`55
`56
`(—)
`72
`81
`(—)
`(n)
`St-4
`71
`63
`(A)
`(~)
`Q
`(—)
`(—)
`110
`NS-3
`93
`66
`(’)
`(r)
`91
`(‘1
`(‘)
`(—)
`NS'3
`@
`70
`_2
`( i )
`51
`E
`73
`57
`4- 1 ST
`98
`74
`(7)
`(re)
`58
`(——)
`(—)
`81
`SC-7
`70
`61
`72
`(7)
`fl
`77
`(i)
`97
`SC—9
`68
`g
`74
`(—)
`63
`74
`(m)
`52
`St-40
`
`E
`72
`109
`(—)
`88
`83
`79
`12
`St-15
`a) Determined as ratio (%) of mean relative tumor volume of the treated group to that of the untreated one
`at day 14 (see “Materials and Methods" for details).
`1)) Judged “ineffective” from data on response to MTD.
`c) Underlined values are “effective" according to our evaluation criteria: T/C value of 50% or less with a
`statistical significance by the Mann—Whitney U—test (P< 0.01, one-sided).
`
`Table II. CompariSOn of Experimental and Clinical Response Rates of Gastric
`
`Tumors to Various Antitumor Agents
`
`Experimental response rate (%)
`Clinical”
`I‘CSPOI‘ISC
`
`RD
`rate (%)
`
`Antitumor
`
`agent
`
`MTD
`
`MMC
`CPM
`ACNU
`ADR
`VCR
`VLB
`S-FU
`MTX
`
`82 (9/11)
`18 (2/11)
`64 (7/11)
`18 (2/11)
`0 (0/11)
`64 (7/11)
`18 (2/11)
`9(1/11)
`
`31
`7
`11
`18
`0
`
`(70/227)
`18 (2/11)
`(5/72)
`0 (0/11)
`(4/37)
`18 (2/11)
`(38/208)
`18 (2/11)
`(0/11)
`0 (0/11)
`._
`30 (3/11)
`23 (122/527)
`18 (2/11)
`
`9(1/11)
`10
`(3/30)
`
`520
`
`Jpn. J. Cancer Res. (Gann)
`
`
`
`RESPONSES OF HUMAN GASTRIC TUMORS T0 RATIONAL DOSE
`
`its RD against
`Efficacy of each drug at
`each tumor line was expressed as T/C(%)
`of growth rate on day 14, and evaluation of
`effectiveness according to our criteria was also
`made (Table I). From these data, effective-
`ness of various antitumor agents examined at
`their RD was obtained as a response rate.
`Those values are shown together with re—
`sponse rate to the MTD,” and both are com-
`pared with clinical response rate cited from
`other reports (Table 11).“) It is clear that
`response rates to RD’s are in much better
`accordance with the clinical response rates
`than are the experimental response rates to
`MTD’s.
`
`DISCUSSION
`
`trials on gastric tumors, rela-
`In clinical
`tively high response rates were observed with
`MMC, S-FU, and ADR (Table II). As a
`matter of fact, these agents are the ones most
`frequently used for clinical treatment of gas-
`tric tumors. 0n the other hand, when a panel
`of human tumors implanted in nude mice
`were treated with various antitumor agents at
`their MTD’s, MMC, ACNU, and VLB ex-
`hibited extremely high response rates of 82, 64
`and 64%, respectively. Other drugs such as
`ADR and S-FU showed relatively low efi‘ec-
`tiveness. Such results are obviously not
`in
`good accordance with the clinical data.
`However,
`in the present study using the
`dose pharmacokinetically equivalent
`to the
`clinical dose,
`i.e., the RD, MMC, ADR,
`5—
`FU, ACNU, and VLB demonstrated almost
`equal
`therapeutic effectiveness.
`In terms of
`relative therapeutic efficacy,
`these results,
`except for VLB, seem to show much better
`coincidence with the clinical ones. For VLB,
`we could not find any report presenting clini-
`cal data for gastric tumors. Accordingly, it is
`difficult to evaluate this result; it might predict
`some clinical activity of this agent against
`gastric tumors, or the result may simply be
`due to an overestimated RD.
`In contrast,
`none of the tumors exhibited any significant
`response to VCR even at its MTD.
`As previously noted for 5—FU and MTX,
`their RD’s were estimated to be somewhat
`
`greater than their MTD’s, suggesting the pos-
`sibility of underestimation of their clinical
`effects in this model. Practically, the effect of
`both agents had to be evaluated at
`their
`
`79(4)
`
`1988
`
`MTD’s. Two and 1 out of 11 tumor lines
`
`significantly responded to 5-FU and MTX,
`respectively, although the inhibitory effects
`were relatively low. These results indicate that
`the degree of underestimation might be not
`so great.
`With respect to the accuracy of RD and
`number of tumor lines used for evaluation of
`response rate, the present experimental condi-
`tions are not necessarily sufficient. Neverthe—
`less, we have clearly proved that use of the
`RD as a clinically equivalent dose, as com-
`pared with the MTD, reproduces well in the
`human tumor/nude mouse model the clini-
`cally observed relative efi‘ectiveness of several
`antitumor agents.
`In considering the practical application of
`the RD, it seems most desirable to use it in
`the sensitivity testing of individual clinical
`tumors using the nude mouse model, where
`we can compare the sensitivities to various
`antitumor agents on the standard of achiev-
`able plasma levels of each drug in man, but not
`in the nude mouse. We think it probable that
`use of the RD’s of various drugs will provide
`a more accurate prediction of the relative
`sensitivities of a given tumor in situ to various
`antitumor agents tested.
`Most important is how we can apply the
`concept underlying RD to the evaluation of
`new antitumor compounds. In the preclinical
`study of new candidate compounds, therapeu-
`tic effectiveness is examined using various
`murine and human tumor models. Such drugs
`often exhibit potent activities against some of
`these tumors. However, if the clinical MTD
`of such a compound is significantly smaller
`than the effective dose in the mouse from a
`
`the drug
`pharmacokinetic point of view,
`would not be expected to be clinically effective
`despite a positive response in the mouse. In
`other words,
`to reasonably predict clinical
`effects of new compounds, we need to have
`not only therapeutic results but also knowl—
`edge of their pharmacokinetics.
`In predicting RD’s of new compounds in
`the development stage, further progress in
`kinetic analysis of cell-killing action of the
`drugs and prediction of their human pharma-
`cokinetic parameters by the method of animal
`scale-up“) is absolutely necessary. Our cur—
`rent efforts are directed toward such kinetic
`studies.
`
`521
`
`
`
`M. INABA, ET AL.
`
`ACKNOWLEDGMENTS
`
`This study was supported in part by a Grant-in-
`Aid for New Drug Development Research from
`the Ministry of Health and Welfare, Japan. The
`authors thank Ms. Reiko Emura and Michiyo
`Kuwabara for their excellent technical assistance.
`
`(Received Oct. 22, I987/Accepred Feb. 20, 1988)
`
`REFERENCES
`
`1)
`
`Inaba, M., Tashiro, T., Kobayashi, T.,
`Fujirnoto, S., Sakurai, Y., Maruo, K.,
`Ohnishi, Y., Ueyama, Y. and Nomura, T.
`Evaluation of response rates to various anti-
`tumor agents of human gastric tumors im-
`planted in nude mouse. Jpn. J. Cancer Res.
`(Garza). 77, 190—196 (1986).
`
`2)
`
`3)
`
`4)
`
`5)
`
`6)
`
`Inaba, M., Kobayashi, T., Tashiro, T. and
`Sakurai, Y. Pharmacokinetic approach to
`rational therapeutic doses for human tumor-
`bearing nude mice.
`Jpn.
`J. Cancer Res.
`(Germ), 79, 509~516 (1988).
`Ramming, K. P. and Haskell, C. M. Stomach
`cancer. In “Cancer Treatment,” ed. C. M.
`Haskell, pp. 257—271 (1985). W. B. Saunders
`Co., Philadelphia.
`Ogawa, M. Current status of chemotherapy
`in advanced esophageal and gastric cancer.
`Jpn. J. Cancer Chemother., 12, 209772103
`(1982) (in Japanese).
`Dedrick, R. L. Animal scale-up. J. Pharma-
`cokinet. Biopharm., 1, 435—46] (1973).
`Boxenbaum. H.
`Interspecies
`scaling,
`allometry, physiological time, and the ground
`plan pharmacokinetics.
`J. Pharmacokinet.
`Biopharm., 10, 201—227 (1982).
`
`522
`
`Jpn. J. Cancer Res. (Gann)
`
`