Drug Treatments for Metastasis of the Lewis Lung Carcinoma ... · Lilly Research Laboratories, Lilly Corporate Center, Indianapolis, IN 46285. aggregation inhibitors have been examined

(CANCER RESEARCH 49, 4509-4516, August 15. 1989]

Drug Treatments for Metastasis of the Lewis Lung Carcinoma: Lack of Correlationbetween Inhibition of Lung Metastasis and SurvivalRonald L. Merriman,1 Katherine A. Shackelford, Lee R. Tanzer, Jack B. Campbell, Kerry G. Bemis,

and Ken Matsumoto

Lilly Research Laboratories, Lilly Corporate Center, Indianapolis, Indiana 46285

ABSTRACT

The abilities of the Eli Lilly compounds LY150310, LY189332, andLY13S305 to inhibit spontaneous metastasis and to increase animalsurvival were evaluated. These compounds represent widely varied structures and were evaluated because they have been found to inhibit throm-

boxane synthetase, cyclooxygenase, and thrombin activation, respectively. These biochemical processes have been proposed in the literatureas targets for antimetastatic drugs. The purpose of this investigation wastwofold: (a) to compare the antimetastatic activities of the Eli Lillycompounds to those of the reference antimetastatic compounds nafaza-

trom and RA233, and (b) to examine the correlation between inhibitionof spontaneous lung metastasis and survival. Spontaneous metastasis ofthe Lewis lung carcinoma was used to evaluate the antimetastatic activityof the compounds. In this model 5 x 10s tumor cells were implanted into

the gastrocnemius muscle, the primary tumor was resected on Day 14,and metastatic lung lesions were counted on Day 25. Compounds wereadministered every 12 h on Days 5 through 19. Nafazatrom, LY150310,LY189332, and IA 135305 were found to inhibit spontaneous lung metastasis in a dose-dependent manner. The Kl>â€ž,values for the respective

inhibitions with these compounds were 50, 0.5, 2, and 0.35 mg/kg/day;the respective therapeutic indexes (I.I),,,/F.D.,,) were 7, 180, 255, and511. To evaluate the effect of nafazatrom, LY150310, LY189332, andLY135305 on animal survival, the compounds were given at maximallyantimetastatic doses of 200, 60, 20, and 6 mg/kg/day, respectively. Twodosing schedules were used: (a) on Days 5 through 19 and (b) on Day 5until death. Neither the median survival times nor the numbers of long-

term survivors were significantly changed with any of the compounds atany dosing schedule. RA233, given to a maximally tolerated dose of 200mg/kg/day on Day 5 until death, did not inhibit lung metastasis and didnot increase median survival time. Postmortem examination of animalsdosed with nafazatrom, LY 1503 Ml, LY189332, and LY135305 showedcomplete inhibition in lung lesions and the appearance of lesions in theliver, kidney, spleen, and brain. The results of this investigation showthat the effect a compound has on the number of metastatic lesions in atarget organ may not be predictive of its effect on survival. To successfullytranslate laboratory data into the clinic, survival should be considered asa predictor of a compound's potential clinical utility.

INTRODUCTION

Metastasis, the hallmark of malignant cancers, is responsiblefor major problems in the diagnosis, staging, and treatment ofcancers (1). This has resulted in a considerable amount ofresearch on the biological processes involved in metastasis (forreviews on this research see Refs. 2-5). The ultimate goal ofthis research is to develop better therapies for malignant diseases.

Since Wood's original observations (6) in vivo on the involve

ment of thrombi in the metastasis of V2 rabbit carcinoma cells,there have been numerous studies implicating the involvementof fibrin and platelets in metastasis (7-9). Because of thesestudies, it is not surprising that anticoagulants and platelet

Received 12/28/87; revised 12/28/88, 5/2/89; accepted 5/18/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1To whom requests for reprints should be addressed, at Department MC7R1,Lilly Research Laboratories, Lilly Corporate Center, Indianapolis, IN 46285.

aggregation inhibitors have been examined as antimetastaticagents. The results of these numerous studies have been conflicting. Some investigators found an inhibition of metastasis,others an increase, and still others found no effect. Part of thecontroversy, especially in the older investigations, may be dueto poor experimental design because of an incomplete understanding of the pharmacology of the agents studied. Severalinvestigators, however, have pointed out that a major reasonfor confusion in the literature is due to the metastasis modelsused to evaluate an agent. In many studies the antimetastaticaction of anticoagulants or platelet aggregation inhibitors hasbeen observed in models where the tumor cells are administeredi.v. Useful insights have been gained from these "artificial"

metastasis models. However, these models can be criticized asassays of tumor cell arrest that occur under abnormally highcell burdens. In some cases, the results obtained in "artificial"models cannot extrapolate to those obtained in "spontaneous"

metastasis models where all the steps of the metastatic cascadeare in place (10). A more important flaw in the experimentaldesign of most reported antimetastasis studies is that they donot evaluate an agent's effect on survival. As discussed later,

survival data may be a more relevant measure of an experimental drug's potential clinical utility.

In this report we present the results of our investigation onthe antimetastatic action of the Eli Lilly compounds LY 150310,LY189332, and LY135305. These compounds were discoveredin screening assays designed to identify inhibitors of thrombox-ane synthetase, cyclooxygenase, and thrombin activation, respectively. The purpose of this investigation was twofold.First, the antimetastatic actions of compounds LY 150310,LY189332, and LY135305 were compared to those of RA233and nafazatrom. RA233, an inhibitor of platelet aggregation(see citations in Ref. 11), is marketed in West Germany as anantimetastatic agent under the trade name Rapenton. It is beingevaluated in the U.S. as an antimetastatic agent (12). Nafazatrom (Bay g 6575), an antithrombotic agent (see citations inRefs. 13, 14), has been reported by Honn et al. to inhibit theartificial and spontaneous metastasis of B16 melanoma andLewis lung tumors (14). It has also been evaluated clinically asan antimetastatic agent ( 15 and citations in Ref. 16). The secondpurpose of this investigation was to examine the correlationbetween the inhibition of metastasis and animal survival.

MATERIALS AND METHODS

Drugs. The structures of the drugs used in this study are shown inFig. 1. RA233 (Rapenton, mopidamole) was purchased in West Germany from Boehringer Ingleheim. Nafazatrom, LY 150310, andLY135305 were made as described in the patent literature (17-19).Compound LY189332 was prepared from the intermediate, 4'-amino-4,4"-dimethoxy-l,l':2',l"-terphenyl, as described in a patent by Mat

sumoto et al. (20). All drugs were dissolved in a sterile vehicle composedof 1 part Emulphor 620 in 39 parts 0.9% saline. Emulphor 620(polyoxyethylated vegetable oil) was purchased from General Anilineand Film, Wayne, NJ. The antimetastatic action of these Lilly compounds were discovered in screens looking for agents that inhibit tumor

4509

on March 19, 2021. © 1989 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

INHIBITION OF LUNG METASTASIS AND SURVIVAL

CH30

b

N-(CH2)2-O

H^Y\^ N(CH2-CH2OHfe

(HO.CH2-CH2)2N^Ãg^^N

RA233

Fig. 1. Structures of the drugs evaluated.

cell induced thrombotic death. The biochemical pharmacology of theseagents will be discussed in later publications.

Cell Culture Reagents. Minimum essential medium with Eagle's salts

and phosphate buffered saline were obtained from M.A. Bioproducts,Walkerville, MD. Fetal bovine serum (characterized grade) was obtained from Hyclone Laboratories, Logan, UT. Trypsin and collagenasewere obtained from Worthington Biochemicals, Freehold, NJ. DNasewas obtained from Sigma Chemical Co., St. Louis, MO.

Cells and Culture Conditions. The tumor used in this study was theLewis lung carcinoma. This tumor originated as a spontaneous lungcarcinoma in a C57BL/6 mouse in Dr. Margret A. Lewis' laboratory

at the Wistar Institute in 1951 (cited in Ref. 21). It is widely used asan experimental model for metastasis. A vial containing frozen 0.3-0.4-mm pieces from a solid Lewis lung tumor was obtained from Dr.Kenneth V. Honn (Department of Radiation Oncology, Wayne StateUniversity, Detroit, MI). According to Dr. Honn the tumor had originally been obtained from the Frederick Cancer Research Facility,Frederick, MD. To initiate the tumor in vitro, the pieces were thawedand placed into a 75-cm2 flask containing 18 ml of antibiotic-free

minimum essential medium supplemented with 10% fetal bovine serum.The tumor was incubated at 37Â°Cunder a humidified atmosphere of

5% CO2 in air. After a 3-day incubation, a single cell suspension wasprepared by treating with 0.16 units of trypsin/cm2 for 10 min at 37Â°C.

After this initial use of trypsin, the cells were passed using a solutionof collagenase (24 units/cm2) and DNase (7 units/cm2). The collagenase

and DNase were dissolved in phosphate buffered saline. Cells from theoriginal trypsinized culture were propagated to passage 4 and stored inliquid nitrogen. These cryopreserved cells served as the source for thecells used in this investigation. The tumor was used between in vitropassages of 5 to 7. Single cell suspensions were prepared for implantation into the mice using the collagenase/DNase solution describedabove. All culture manipulations were done under gold fluorescentlight.

Animals and Surgical Procedure. C57BL/6N female mice were obtained from the Portage, MI, breeding facilities of Charles River Laboratories, Inc. Like other investigators (22), we found an increase inthe variability and a decrease in the number of metastatic lung coloniesas the age of the host animals increased. Therefore, only mice between4 and 6 weeks old were used in this investigation. A group of fiverandomly chosen animals from each shipment was evaluated uponarrival, and at the end of each experiment, for murine pathogens. Theseevaluations are important because we have found that the variabilityincreases and the number of lung mÃ©tastasesdecreases when animalsare infected with a pathogen, such as murine hepatitis virus. Animalswere housed in plastic shoebox cages covered with a microisolator filtertop. All animal feed, water, bedding, and cages were sterilized. Theanimals were provided ad libitum water (pH 3.0) and a 4% fat mousediet manufactured by Tecklad, Madison, WI. All animal manipulations

were done aseptically in a HEPA, 100% exhaust, vertical laminar flowhood.

Surgical procedures were done with sterile instruments and aseptictechnique. A 100% oxygen atmosphere saturated with methoxyflurane(metofane) and water was used to obtain surgical anesthesia (23). Bodytemperature was maintained at 37Â°Cwith an incandescent heat source.

A skin incision was made completely around the leg just above thetumor and the skin was then pulled toward the calcanei tendon. Vesselswere ligated with 4-0 surgical thread and the leg was amputated justbelow the head of the femur. The wound was closed with stainless steelsurgical clips. In animal survival studies, the surgical clips were removedafter 14 days to prevent ulcÃ©rationand infection of the wound. Usingthis procedure recurrence of the tumor was rarely seen at the incisionsite. After amputation the animals were injected s.c. with 1 ml of sterile0.9% saline to prevent dehydration. Sterile saline was also used toirrigate the eyes and prevent blindness. During postanesthetic recoverythe animals were kept at 37Â°C.

Spontaneous Metastasis Model and Drug Treatments. The followingprocedure was used to evaluate the effect of drugs on spontaneous lungmetastasis of Lewis lung carcinoma. On Day 0, 5 x 10s tumor cells (in

a volume of 25 /il) were implanted into the gastrocnemius muscle andthe animals were then randomly divided into treatment groups (10animals per group). In agreement with others (see citations in Ref. 24),we have found that a small amount of artificial seeding occurs when atumor cell suspension is implanted intramuscularly. To determine theamount of artificial seeding, the tumor bearing leg in a group of animalswas amputated on Day 1. These animals were also treated with drugvehicle. On Day 14, the primary tumor was resected in the vehicle-treated and drug-treatment groups. Day 14 was the maximum timethat surgery could be delayed and still remove all the primary tumor.The corrected mean number of spontaneous mÃ©tastaseswas determinedby subtracting the number of tumor colonies in the Day 1 amputationcontrol group from the number of colonies in animals that had theprimary tumor resected on Day 14 and were treated with vehicle ordrug. Legs were amputated on Day 14 in one group of nontumorbearing animals. These animals were used as surgical controls for anormal tissue background in evaluation of metastatic colonies in theorgans of the vehicle-treated and drug-treated animals. These nontumorbearing animals were also treated with drug vehicle. All animals weresacrificed on Day 25 by CO2 asphyxiation and the number of macroscopic lung colonies were determined after fixation in Bouin's fixative.

The animals were sacrificed on Day 25 because our initial studies hadshown that the animals would begin to die on Day 28.

The animals were weighed at the start of an experiment and on Day25 to determine if the compounds had any effect on animal weight. Thetumor bearing legs amputated on Day 14 were used to determine if acompound had any effect on the growth of the primary tumor. Toestimate tumor weight the amputated legs were first dried in a vacuumoven. Tumor weight was then estimated by subtracting the mean legweights for the nontumor bearing animals from the mean leg weightsfor the vehicle-treated and the drug-treated tumor bearing animals. Toevaluate antimetastatic activity, the drugs (or vehicle) were given i.p.every 12 h on Days 5 through 19. This treatment period was chosenfrom our initial studies that determined the time-course for lung metastasis of the Lewis lung carcinoma after intramuscular implantation(data not shown). In these studies the Lewis lung tumor was found tostart metastasizing on Day 6. By Day 10, 100% of the animals hadmetastatic disease. Treatment was continued through Day 19 to allowfor the clearance of any tumor cells released from the primary tumoron Day 14. On Day 19 treatment was stopped to allow for the growthof tumor cells and, thus, determine whether the antimetastatic activitywas because of a cytostatic activity of the compounds.

Animal Survival Studies. In these studies 5 x IO5cells were implanted

into the gastrocnemius muscle on Day 0 and the animals were randomlydivided into treatment groups. For the studies with nafazatrom,LY189332, LY150310, and LY135305 there were 20 animals in eachgroup. For the study with RA233 there were 10 mice in each group.Amputations were done on Days 1 and 14 as discussed above. Exceptfor RA233, animals were treated with a compound at the dose foundto maximally inhibit spontaneous lung metastasis. Two dosing sched-

4510




ules were used: (a) on Days 5 through 19 and (b) on Day 5 until death.RA233 was dosed at 60 and 200 (maximally tolerated dose) mg/kg/day on Days 5 until death. The animals were checked every 4 to 12 hfor survival. The lungs, liver, kidney, spleen, and brain were removedas close to death as possible and evaluated for macroscopic metastaticlesions. The lungs and brain were fixed in Bouin's solution and theliver, kidney, and spleen were fixed in Fekete's solution.

Data Analysis. Lung metastasis was analyzed by descriptive (plotsand summary values) and inferential (/' values and significance tests)

statistics. Percentage inhibition for each treatment group was calculatedas follows:

Day 14 mean â€”treatment meanInhibition = â€”

Day 14 mean â€”Day 1 mean x 100

where Day 1 mean is the mean number of lung lesions in the Day 1amputation group, Day 14 mean is the mean number of lesions in theDay 14 amputation group and treatment mean is the mean number oflesions in a treatment group. The ED5o2for inhibition of lung lesions

was calculated by a graphical interpolation of the linear portion of thedose-response curve. Inferential statistics were used to determinewhether each dose of a compound significantly decreased the numberof metastatic lesions. For this analysis a two-sided Wilcoxon rank-sumtest was used to compare the number of lung lesions in each treatmentgroup to the number in the Day 14 amputation control group. Computations for this test were done with the aid of the NPAR1WAYprocedure in SAS (25).

Descriptive and inferential statistics were also used to analyze animalsurvival studies. Animal survival time was measured from the day oftumor implantation until death. The "product -limit" method of Kaplan

and Meier, as described by Simon (26), was used to estimate and plotanimal survival. The Kaplan-Meier plots and median survival timeswere computed with the aid of the LIFETEST procedure in SAS. Thelog rank test was used to determine the significance of the differencebetween survival of each treatment group and survival of the controlgroup. Computations for the log rank test were done with the aid ofthe LIFETEST procedure.

Inferential statistics were used to analyze the anatomical distributionof metastatic lesions. For this analysis a two-sided Wilcoxon rank-sumtest was used to compare the number of lesions in each treatment groupto the number in the Day 14 control group. The NPAR1WAY procedure was used for these computations.

RESULTS

Effects on Spontaneous Lung Metastasis. In Table 1 is shownthe effect of nafazatrom on the number of metastatic lunglesions for the individual animals in each treatment group. Thenumber of lesions for the individual animals in each controlgroup represent the number that were observed in the controlgroups for the other compounds. Nafazatrom inhibited spontaneous lung metastasis in a dose-dependent manner (Fig. 2).This inhibition was significant (P < 0.05) at 200 mg/kg/day.However, the curve that described the inhibition was steep. A54 and a 100% inhibition in the number of metastatic lunglesions was obtained at doses of 60 and 200 mg/kg/day, respectively. The ED50for the antimetastatic action of nafazatromwas approximately 50 mg/kg/day. Some toxicity that resultedin a 10% mortality was also seen at 200 mg/kg/day. At concentrations greater than this, nafazatrom was more toxic and hada LD50 value of 335 mg/kg/day. Nafazatrom was estimated tohave a therapeutic index (LD5o/ED5o) of 7.

In Fig. 2 the antimetastatic action of LY150310 is alsoshown. At 2 mg/kg/day or higher, this compound significantly(P < 0.05) inhibited spontaneous lung metastasis in a dose-

2The abbreviations used are: ED50, dose producing a 50% inhibition of

metastasis; LD50, dose that kills 50% of the mice; TXA2, thromboxane A2; PGI2.prostacyclin.

Table 1 Inhibition of spontaneous lung metastasis of the Lewis lung carcinomaby nafaiatrom

On Day 0 animals were implanted with 5 x IO5 Lewis lung carcinoma cells

into the gastrocnemius muscle and randomly divided into treatment groups (10per group). Amount of artificial seeding was determined by resecting the tumorbearing leg on Day 1. The primary was resected in the control (vehicle) and drug-treated animals on Day 14. Animals were dosed every 12 h on Days 5 through19. The number of metastatic lung lesions was determined on Day 25. Astatistically significant (P < 0.05) dose-dependent reduction in metastatic lunglesions was obtained with nafazatrom.

Treatment Number toxic deaths/ Metastatic lunggroup mg/kg/day number evaluated lesions per mouse

Day 1controlDay

14controlNafazatromVehicleVehicle20602006000/100/100/100/91/1010/100,

1,3.4,4,5,5,5,6,69.

10, 11, 11. 12.16,22,23,26.310,

1,5,6,7,9,16,27,31,330,0,

1,2, 12, 16,17.20,210.

0, 0. 0, 0, 0. 2,6.17Toxic

100-

80-

Â§

jH

I 40-

20-

00.01 0.1 1 10 100 1000

Mg/Kg/Day

Fig. 2. Dose-response relationships for the inhibition by nafazatrom (d),LY150310 (A), LY189332 (â€¢),and LY135305 (A) of spontaneous mÃ©tastasesofthe Lewis lung carcinoma. Animals were implanted with 5x10' cells on Day 0and randomly divided into treatment groups as described in "Materials andMethods." Drugs were administered i.p. at one-half the daily dose every 12 h.

Treatment was given on Days 5 through 19. The number of metastatic lunglesions were determined on Day 25.

dependent manner. However, the curve that described the inhibition was not as steep as that for nafazatrom. A plateau of100% inhibition in lung metastasis was obtained at doses of 20and 60 mg/kg/day. The ED5Ufor LY 150310 was 0.5 mg/kg/day. The doses of LY 150310 that inhibited lung metastasis by50 and 100% were 1/100th and 1/1 Oth, respectively, of thosefor nafazatrom. At doses from 0.6 to 60 mg/kg/day there wasno effect on animal weights or growth of the primary tumor.At doses greater than 60 mg/kg/day LY 150310 was toxic andhad an LD50 value of 90 mg/kg/day. However, the antimetastatic action of LY 150310 did not decrease at toxic doses. Thetherapeutic index of LY 150310 was approximately 180, whichis 26 times greater than that for nafazatrom.

LY189332 inhibited metastasis in a dose-dependent manner(Fig. 2). This inhibition was significant (P < 0.05) at a dose of2 mg/kg/day or higher. Like LY 150310, this compound wasmore potent and had a better therapeutic index than nafazatrom. However, the dose-response relationship for the antimetastatic action of LY189332 was biphasic. A 100% inhibition inlung mÃ©tastaseswas seen at doses of 6 and 20 mg/kg/day.Below 6 mg/kg/day, the antimetastatic action decreased sharplylike nafazatrom. The ED50 value for the antimetastatic actionof LY189332 was 2 mg/kg/day. The doses that inhibited lungmetastasis by 50 and 100% were l/25th and l/33rd, respectively, of those for nafazatrom. At 200 mg/kg/day only a 66%inhibition in the number of lung mÃ©tastaseswas seen. At doses

4511




from 0.2 to 200 mg/kg/day, no inhibition in the growth of theprimary tumor, no effect on animal weights, or toxic deathswere observed. LY189332 was toxic at higher doses and had aLD50 value of 510 mg/kg/day. LY189332 had a therapeuticindex of 255, which is 36 times greater than that for nafazatrom.

Of the compounds evaluated, LY135305 had the greatestpotency and the greatest therapeutic index. At 2 mg/kg/day orhigher, this compound significantly (P < 0.05) inhibited spontaneous lung metastasis in a dose-dependent manner; but, likeLY189332, the response curve was biphasic (Fig. 2). At dosesof 2 and 6 mg/kg/day, a 100% inhibition in metastasis wasseen, and below 2 mg/kg/day the antimetastatic action ofLY135305 decreased sharply. The ED50 value for the antimetastatic action of LY135305 was 0.35 mg/kg/day. Compared tothe doses of nafazatrom needed to inhibit lung metastasis by50 and 100%, this compound was 143 and 100 times morepotent, respectively. At 20 mg/kg/day the antimetastatic effectdecreased and only a 70% inhibition in spontaneous lung metastasis was seen. No toxic deaths, no decrease in animalweights, or inhibition in the growth of the primary tumor wereseen at doses of 0.02 to 20 mg/kg/day. At higher doses thiscompound was toxic and had a LD50 value of 179 mg/kg/day.The therapeutic index for this compound was 511, which is 73times greater than that for nafazatrom.

Effects on Animal Survival. To evaluate the relationship between inhibition of lung metastasis and survival, tumor bearinganimals were treated with the maximally effective antimetastatic doses of nafazatrom, LY 150310, LY 189332, andLY135305 and survival was determined. These compoundswere given in two schedules: (a) on Days 5 through 19 or (b)on Day 5 until death. Amputations for the various control andtreatment groups were the same as the experiments discussedabove. Results of these studies are shown in Fig. 3 and sum

marized in Table 2. Animals that were dosed with nafazatromon Days 5 through 19, or on Day 5 until death, died soonerthan the control animals. The decrease in survival for theanimals dosed on Days 5 through 19 was statistically significant(P = 0.031). There were no 90-day survivors in the animals

that were dosed with nafazatrom at either dosing schedule. Themedian survival time was not significantly changed, and therewere no 90-day survivors for the animals dosed with LY 189332at either dosing schedule. The median survival times were alsonot significantly changed for the animals dosed with LY 150310and LY 135305 at either dosing schedule. Seventeen percent ofthe animals survived 90 days when they were dosed withLY 150310 on Days 5 through 19; when given on Day 5 untildeath the surviving fraction was only 10%. Similar toLY 150310, only 24% of the animals survived 90 days whenthey were dosed with LY135305 on Days 5 through 19, andthis decreased to 13% in the animals dosed on Day 5 untildeath. The number of 90-day survivors observed with

LY150310 or with LY135305 was not statistically significant(P > 0.10). In these studies the antimetastatic agent RA233was also evaluated. RA233 was given at 200 mg/kg/day (itsmaximally tolerated dose) and at 60 mg/kg/day on Day 5 untildeath. This compound had no effect on median survival timeor the number of 90-day survivors (Table 2). Postmortemexamination showed no inhibition in lung metastasis in animalstreated with either dose of RA233 (data not shown).

Effects on Anatomical Distribution of Metastatic Lesions. Asdiscussed above, treatment with antimetastatic agents at theirmaximally effective doses did not increase mean survival timesor produce a high proportion of 90-day survivors (Table 2). Tobetter understand this paradox, major organs from animalstreated with vehicle or with antimetastatic drugs were evaluatedfor the presence of metastatic colonies. In Table 3 is shown the

Fig. 3. Kaplan-Meier survival curves forthe effect of nafazatrom on survival of animalsbearing the Lewis lung carcinoma. Animalswere implanted with 5x10' cells on Day 0

and randomly divided into treatment groupsas described in "Materials and Methods." Tu

mor bearing legs were resected from controlanimals on Day 1 (A) and Day 14 (A). In ,/.animals were treated with nafazatrom at a doseof 200 mg/kg/day on Days 5 through 19 (D)or Day 5 until death (â€¢).In II. animals weretreated with LY150310 at a dose of 60 mg/kg/day on Days 5 through 19 (D) or Day 5until death (â€¢).In C, animals were treatedwith LY 189332 at a dose of 20 mg/kg/day onDays 5 through 19 (D) or Day 5 until death(â€¢).In D, animals were treated with LY135305at a dose of 2 mg/kg/day on Days 5 through19 P) or Day 5 until death (O).

100

80

60

40

20

100

80

60

40

20

30 60 0 30 60 90

D.

30 60 0 30

Days After Tumor Implantation60 90

4512




Table 2 Summary of effect of treatment with nafazatrom, L Y150310, L Y189332, L Y135305, and RA233 on survival of mice bearingLewis lung carcinoma

On Day 0 animals were implanted with 5 x 10* Lewis lung carcinoma cells into the gastrocnemius muscle and randomly divided into treatment groups. The

primary tumor was resected in the control and drug treated animals on Day 14. Except for RA233, animals were treated every 12 h with a dose of drug that wasfound to maximally inhibit spontaneous lung metastasis. Two dosing schedules were used: (a) on Days 5 through 19 and (Â¿>)on Day 5 until death. RA233 was dosedat 60 and 200 (maximally tolerated dose) on Day 5 until death. Animals were checked every 4 to 12 h for survival.

TreatmentgroupExperiment

I (n =20)Day14controlNafazatromLY

189332Experiment

2 (n =20)Day14controlLY150310LY

135305Experiment

3 (n =10)Day14controlRA233mg/kg/dayVehicle20020Vehicle606Vehicle60200Treatment

days5-death5-195-death5-195-death5-death5-195-death5-195-death5-death5-death5-deathMediandays

survival(range)30.0

(25.0-37.0)27.5(24.0-31.0)26.0

(24.0-36.0)29.5(25.0-39.0)33.0(24.0-38.0)33.0

(28.0-53.0)34.5(29.0-113.0*)35.5(31.1-112.0*)38.0(28.0-113.0*)32.0(26.0-113.0*)30.5

(25.0-32.0)28.5(24.0-33.0)32.0

(25.0-34.0)%

90-daysurvivors00000017102413000pvalue"0.0310.0930.9710.5750.3100.2940.1870.9520.5940.266

Â°The log rank test was used to determine P values for the difference between the survival curves for drug treated animals relative to the curves for vehicle treatedanimals. The statistical analysis was done as described in "Materials and Methods."

* Experiments were ended on these days.

Table 3 Effect of treatment with nafazatrom, LY Â¡89332,LY150310, and L Y135305 on organ distribution of Lewis lung carcinoma cellsOn Day 0 animals were implanted with 5 x 10s Lewis lung carcinoma cells into the gastrocnemius muscle and randomly divided into treatment groups. The

primary tumor was resected in the control (vehicle) and drug treated animals on Day 14. Animals were treated every 12 h from Day 5 until death with a dose of drugthat was found to maximally inhibit spontaneous lung metastasis. Animals were checked every 4 to 12 h, organs were removed as close to death as possible, andevaluated for macroscopic metastatic lesions.

Median number of metastatic lesions(range)GroupExperiment

1Day 14 controlNafazatromLY189332Experiment

2Day 14 control

LY150310Experiment

3Day 14 controlLY135305TreatmentVehicle

200 mg/kg/day20mg/kg/dayVehicle

60mg/kg/dayVehicle

6 mg/kg/dayLung36(10-80)

0 (0-0)Â°1(0-3)Â°24(16-51)

0(0-1)Â°40

(20-89)2(0-7)Â°The

following /"-values comparing the number of lesions in the drug treated an:Wilcoxon rank-sum test as described in "Materials and Methods": Â°P < 0.001, " PLiver0

(0-0)6 (2-25)Â°

10(2-20)Â°0(0-1)

6(5-20)Â°0(0-1)

8(4-19)Â°Kidney0

(0-0)2 (0-6)Â°1(0-3)c0

(0-0)1(0-5)Â°0(0-1)

3(0-6)fÂ¡mais

relative to the number in the Day= 0.005, c P = 0.006, d P = 0.01 5, ' P =Spleen0

(0-0)1 (0-2)f0(0-Of0

(0-0)0(0-5/0

(0-0)0 (0-0)*Brain0

(0-0)0 (0-2/1(0-2)'0

(0-0)1(0-4)"0

(0-0)0(0-4)'14

control groups were computed by the0.035, fP = 0.078, and * P > 0. 100.

effect of treatment with nafazatrom, LY 189332, LY 150310, mechanisms of action, were evaluated. The purposes of thisand LY135305 on the organ distribution of Lewis lung carci- investigation were twofold. First, the antimetastatic actions ofnoma cells. Only macroscopic, pulmonary mÃ©tastaseswere compounds LY135305, LY150310, and LY189332, were corn-observed in animals that were treated with vehicle and had their pared to those of RA233 and nafazatrom. Second, the correla-primary tumor removed on Day 14. (Note, organs were not tion between the inhibition of spontaneous lung metastasis andhistorically evaluated for the presence of microscopic metas- animal survival was examined. This research was done as parttases). Consistent with previous studies, there was a significant of a program to discover noncytotoxic drugs to treat cancer(P < 0.001) decrease in the number of pulmonary mÃ©tastases metastasisin animals whose primary tumor was resected on Day 14 and RA233 was used as a reference antimetastatic drug. RA233were treated with either nafazatrom, LY189332, LY150310, or dÂ¡dnot Â¡nhjbitspontaneous lung metastasis nor did it have anyLY135305 on Day 5 until death. In animals treated with these effec, on medÂ¡ansurvÂ¡va,dme These resuhs have a,so beencompounds, however, there was also a marked increase in observed by other investigators using the Lewis lung carcinomaextrapulmonary mÃ©tastases.The most significant number of dtadons jn Rrf_ ,n ^ Ua ^ fl/ hayemetastatic lesions were observed in the liver (P < 0.001) and , , . ,.,.,.. . , ., .. . ., , , ,.., ,_ . --,. â€ž . . , , . . , observed an inhibition in both the incidence and the frequencykidney (P < 0.006). Fewer lesions were observed in the spleen . , , _,,- , ,__.

... , , -n . â€¢ , -r- of spontaneous lung mÃ©tastasesof the B16 melanoma (27).and brain, and only occasionally was their number significant. â€ž.-%,,. . ,.,. r ,

RA233 is an inhibitor of platelet aggregation (see citations inRef. 11). Its mechanism of action appears to be the inhibition

DISCUSSION Â°fcyclic AMP phosphodiesterase. Maniglia et al. concluded,however, that the inhibition of metastasis of the B16 melanoma

In this investigation the antimetastatic action of several dif- was not related to inhibition of platelet aggregation. In a smallferent drugs, representing widely varied structures and different clinical study by Gastpar et al. (28), RA233 was found to

4513




increase survival in patients with sarcoma or lymphoma of thehead and neck when used as an adjunct to surgery or radiotherapy. Currently, RA233 is being studied in a larger clinical trialby Zacharski et al. (12). It will be interesting to see if RA233 isfound to have clinical activity in this larger study.

Nafazatrom was also used as a reference antimetastatic drug.In agreement with other studies (14, 29), nafazatrom was foundto inhibit the spontaneous formation of lung mÃ©tastases.However, other investigators, using the same and different tumortypes, have not observed an antimetastatic action with nafazatrom (29-31). The dose of nafazatrom for 100% inhibition was25 times greater than that reported by Honn et al. (14). Manyfactors could explain this difference in the therapeutic efficacyfor nafazatrom. For example, in the study by Honn et al. theLewis lung carcinoma was grown s.c. and the drug was givenS.C.;whereas, in our investigation the tumor was grown intramuscularly and the drug was given i.p. Intraperitoneal administration was used because it approximates oral administration(32). For nafazatrom this could be important because after oraladministration it is extensively metabolized during first passagein the liver (33). Other factors, such as drug vehicle and treatment duration, could also be important.

Nafazatrom's ability to increase levels of prostacyclin in the

vascular endothelium has been proposed as a mechanism forits antimetastatic action (34). This increase is thought to resultfrom both a stimulation in synthesis and from an inhibition indegradation of prostacyclin. Nafazatrom's prostacyclin enhanc

ing action is thought to inhibit tumor cell-platelet interactionsby changing the balance between the proaggregating and vaso-constrictive actions of TXA2 and the antiaggregating andvasodilative actions of PGI2.

LY150310 was evaluated as an antimetastatic agent becauseit had been found to inhibit TXA2 synthetase (18). Agents thatinhibit TXA2 synthetase have been proposed as antimetastaticagents because of their ability to change the ratio of TXA2 toPGI2 and thus inhibit tumor cell-platelet interactions (34).Honn et al. have reported that TXA2 synthetase inhibitors haveantimetastatic activity in vivo (34). However, other investigatorshave not found a correlation between inhibition of TXA2 synthetase and inhibition of spontaneous metastasis (35).

LY189332 was evaluated as an antimetastatic agent becauseit had been found to inhibit cyclooxygenase (20). In sometumors a significant elevation in the levels of prostaglandinshas been observed (see citations in Refs. 35 and 36). This hasled to the concept that prostaglandins may be important intumor dissemination. Some prostanoids are thought to be important in tumor cell-platelet interactions, in modulating thefunction of the immune system, and possibly in the growth oftumor cells (36, 37). It should be noted, however, that therelative amounts and ratios of different prostanoids can varygreatly among tumors. Thus, in some tumors prostaglandinmetabolism may not play an important role in their growth anddissemination. The Lewis lung carcinoma used in this studyhas been shown to make high amounts of prostanoids, especially prostaglandin E; and prostaglandin F2a(38). The importance of this to the metastatic activity of this tumor is still inquestion (39). Several experimental studies have examined theeffect of cyclooxygenase inhibitors on metastasis (see citationsin Ref. 7). The results of these studies can be summarized asinconclusive. The reasons for this confusion may be due todifferences in procedures used to evaluate antimetastatic action,or because tumors in which prostaglandins have no role inmetastasis may have been used.

A dose-dependent inhibition in spontaneous lung mÃ©tastases

was seen with LY189332. Its dose-response relationship, however, was different from that for LY 150310 in two aspects.At doses producing less than 100% inhibition, the antimetastatic action of LY189332 decreased more sharply thanthat for LY 150310. More important, the metastatic action ofLY 189332 decreased at high doses. The reason for the biphasicdose-response curve of LY 189332 is not known. A cumulativetoxicity at high dose levels could be one explanation for thedecrease in antimetastatic action. This is not supported, however, by the observation that no difference was seen betweenthe animal weights in the control and high dose groups. Anotherpossible explanation is that this drug has different pharmacological actions at different dose levels. For example, aspirin hasa dual effect on the formation of thrombosis; low doses areantithrombotic whereas higher doses are thrombogenic (40).This may be explained by aspirin's multiple effects on arachi-

donic acid metabolism. Low concentrations inhibit the formation of TXA2 from arachidonic acid while high concentrationsinhibit PGI2 formation.

LY 135305 was evaluated as an antimetastatic agent becauseit was found to inhibit the activation of factor VII by tissuefactor (factor HI) in the extrinsic coagulation cascade (19).There is considerable experimental and clinical evidence thatsupports the concept of an association between coagulation andmalignant disease (2-4, 7, 8). The generation of thrombin bytumor cells, and the subsequent formation of fibrin, is considered important in the formation of emboli and in the protectionof tumor cells from the immune system. Since thrombin hasbeen shown to be a mitogen, it may also stimulate tumor cellgrowth. The Lewis lung carcinoma has been shown to havehigh procoagulant activity (41 and citations in Ref. 42). Studieshave shown that the spontaneous metastasis of this tumor isassociated with a low grade disseminated intravascular coagulation (see citations in Ref. 42). The hemostatic abnormalitiesseen with the Lewis lung carcinoma are similar to the laboratoryabnormalities seen in patients with solid, metastasizing tumors.

The dose-response relationship for the antimetastatic actionof LY135305 was similar to that for LY189332. At dosesproducing less than 100% inhibition, the antimetastatic actiondecreased sharply. A decrease in antimetastatic action was alsoseen at high doses. The reason for the biphasic dose-responsecurve for LY 135305 is not known, but it may be the same asthat proposed for LY 189332.

Numerous studies have been conducted to evaluate the antimetastatic efficacy of anticoagulants. An inhibition in tumorgrowth and metastasis have been observed with agents, such asheparin, warfarin, and proteinase inhibitors, that preventthrombus formation (43, 44). Other studies, however, have notfound an antimetastatic action with these agents (43, 44 andcitations in Ref. 2). As stated in the introduction, this controversy may be due to the different metastasis models used. Ofthe anticoagulant agents tested, warfarin and related compounds are the most potent. Anticoagulation may not totallyexplain its mechanism of action, however, since some couma-rins that do not have anticoagulant activity have been found toinhibit metastasis (see citations in Ref. 2).

Although the Lilly compounds, LY150310, LY189332, andLY135305 were found to inhibit spontaneous metastasis to thelung, the mechanisms for their antimetastatic actions have notbeen definitively shown. Studies on the biochemical pharmacology of these agents are currently in progress and will bereported in a later paper.

As stated earlier, numerous studies have examined the effectof a variety of agents on the metastasis of experimental tumors.

4514




These studies have measured the effect of an agent by countingthe number of metastatic lesions in a target organ; with fewexceptions, have studies measured an antimetastatic effect bydetermining survival. Although it could be argued that evaluation of tumor burden is appropriate for cytoreductive agents,this is not adequate for antimetastatic agents. For several reasons, survival data is a more important parameter for theevaluation of these agents. First, survival data in animal modelsare needed to show that an antimetastatic agent has the potential to increase survival in patients that already have metastaticdisease. In clinical oncology it is generally held that as many as50% of patients will have disseminated disease when a canceris diagnosed (45). Several studies have also established that formost human cancers generalized disease does not occur directlyfrom the primary cancer (see citations in Ref. 2). For a givenprimary cancer there may be several initial sites of metastasis.Generalized disease is produced by the sequential metastasisfrom these second sites to third and fourth sites, et cetera. Aneffective antimetastatic agent would prevent this sequence,sometimes called "metastasis from metastasis," and in effect

freeze the metastatic process. In combination with other therapeutic modalities, such an agent could increase survival inpatients that have metastatic disease when therapy is started,or aid in the cure of patients treated before metastasis begins(46).

A second reason for survival studies in animals is to showthat an anticoagulant or a platelet aggregation inhibitor has notexacerbated a cancer by increasing the dissemination of mÃ©tastases. Most blood-borne human tumor cells produce metastaticcolonies in the first capillary bed encountered (47 and citationsin Refs. 2 and 48). The same is true for animal tumors, regardless of whether the mÃ©tastasesdevelop from i.v. injected cellsor spontaneously from solid tumors (48). If an agent inhibitsmetastasis to the organ of first pass clearance, it is possible thatit could promote capillary passage and produce a greater dissemination of metastasis. It is assumed that most tumor cellsegressing from a capillary bed are not viable (see citations inRef. 2). However, these studies have been done on a limitednumber of tumor types. For each tumor and agent being evaluated, survival studies are needed to definitively show this. Thisinformation, preferably obtained in several tumors, is neededto decide on whether an agent should be tested in humans.

For the reasons given above the effect of the Lilly compoundsLY150310, LY189332, and LY135305, and nafazatrom wereevaluated for their effect on survival. These compounds weregiven at their maximally effective antimetastatic doses. RA233(Rapenton) was also evaluated to its maximally tolerated dose.None of the compounds produced a significant number of long-term survivors or significantly increased the median survivaltime. These results were surprising, especially for the Lillycompounds, because they were potent inhibitors of spontaneouslung metastasis. The lack of a correlation between inhibition ofmetastasis and survival was also disappointing because therapywas started on Day 5. This is approximately 1 day before wecould detect any metastatic lung colonies. To be relevant toclinical oncology, it is also important that an antimetastaticagent increase the life span of animals when therapy is delayeduntil metastasis has already occurred. Considering the effect onsurvival when therapy was begun just before metastasis started,it is quite possible that this effect would be even less if therapywere delayed.

To understand why there was not a correlation between theinhibition of spontaneous lung metastasis and animal survival,the anatomical distribution of tumor lesions was evaluated after

treatment with the compounds. Inhibition of metastatic lunglesions was observed with nafazatrom, LY150310, LY189332,and LY135305. However, the decrease in pulmonary lesionswas associated with an increase in metastatic lesions in extra-pulmonary sites, such as the liver, kidney, spleen, and brain.Thus, the lack of a correlation between inhibition in lungmetastasis and an increase in survival appears to result from anincreased dissemination to extrapulmonary sites. To the best ofour knowledge, this is the first report of this phenomenon usinga spontaneous metastasis model. Several studies have shownthat treatment with anticoagulants can lead to a wider distribution of i.v. injected radiolabeled tumor cells (see citations inRef. 49). Other investigators have also observed a change in theanatomical distribution of metastatic lesions with other agents.With an artificial metastasis model using B16 melanoma. Hag-mar and Boeryd found that heparin, at a dose that inhibitedlung lesions, increased the colonization of several extrapulmonary sites (48). In a recent study Willmott et al. examined theeffect of RA233 on the organ colonization of sarcoma 180 cells(47). These investigators found that RA233 treatment did notchange the survival time or the incidence of pulmonary lesionsbut it did cause a higher incidence of extrapulmonary lesions.Although not determined in this investigation, this may be whyRA233 did not increase survival in animals bearing the Lewislung carcinoma.

The goal of cancer drug development is to decrease themortality or morbidity of cancers. To obtain this goal, theendpoint for an animal tumor model, no matter what the model,should be clinically relevant so that the information obtainedin the laboratory can be used to justify further clinical evaluation. In this investigation we found that a variety of antithrom-botic agents could inhibit spontaneous lung metastasis of theLewis lung carcinoma. However, when animal survival wasexamined, no significant increase was seen in either the numberof survivors or in median survival times. An initial conclusionthat could be reached from this investigation is that antithrom-botic and anticoagulant drugs will not be clinically useful in theprevention of metastasis. This conclusion should be reachedwith caution, however, because only one tumor type was evaluated. Considering the variability that has been shown to occuramong tumors, it is unwise to make generalizations about alltumors based on the results obtained with one tumor. Theeffects of the agents used in this investigation could be quitedifferent if they were evaluated with a different tumor model.With this caveat aside, however, the results of this investigationshow that the effect a compound has on the number of metastatic lesions in a target organ may not be predictive of itseffect on survival. To successfully translate laboratory data intothe clinic, survival should be considered as a predictor of acompound's potential clinical utility. It is probably also best to

evaluate a compound with more than one tumor model. Theuse of survival as an endpoint in preclinical studies should beconsidered important given the high cost of doing clinical trialsand the current regulatory requirements for the approval of anew cancer drug (50).

REFERENCES

1. Fidler, I. J., and Hart, I. R. Principles of cancer biology: Biology of cancermetastasis. In: V. T. DeVita, Jr., S. Hellman. and S. A. Rosenberg (eds.),Cancer. Principles and Practice of Oncology, pp. 80-92. Philadelphia: J. B.Lippincott. 1982.

2. Leonard Weiss, Principles of Metastasis. Academic Press. Inc., 1984.3. Welch, D. R., Bhuyan, B. K., and Liotta, L. A. (eds.) Cancer Metastasis.

Experimental and Clinical Strategies. New York: Alan R. Liss Inc., 1986.4. Torisu. M., and Yoshida, T. Basic Mechanisms and Clinical Treatment of

4515




Tumor Metastasis. Academic Press, Inc., New York, 1985. 27.5. Barrett, A. J.. Baici, A., and Strauli, P, (eds.) Proteinases and Tumor Invasion.

Vol. 6 of the monograph series of the Eur. Org. Ree. Treat. Cancer. NewYork: Raven Press, 1980.

6. Wood, S., Jr. Pathogenesis of metastasis formation observed in vivo in the 28.rabbit ear chamber. Arch. Pathol., 66: 550-568, 1958.

7. Gasic, G. J. Role of plasma, platelets, and endothelial cells in tumor metastasis. Cancer Metastasis Rev., 3: 99-116, 1984. 29.

8. Dvorak, H. F., Senger, D. R., and Dvorak, A. M. Fibrin as a component ofthe tumor stroma: origins and biological significance. Cancer MetastasisRev., 2:41-73, 1983. 30.

9. Al-Mondhiry, H. Tumor interaction with hemostasis: The rational for theuse of platelet inhibitors and anticoagulants in the treatment of cancer. Am. 31.J. Hematol., 16: 193-202, 1984.

10. Maat, B., and Hilgard, P. Critical analysis of experimental models for theantimetastatic effects of anticoagulants. In: K. Hellmann, P. Hilgard, and S.Eccles (eds.), Metastasis. Clinical and Experimental Aspects, pp. 109-115. 32.The Hague: M. Nijhoff Inc., 1980.

11. Ambrus, J. I.... Ambrus, C. M., Gastpar, H., Huberman, E., Montagna, R.,Biddle, W., Leong, S., and Horoszewicz, J. Antimetastatic and antitumor 33.effect of platelet aggregation inhibitors: In: G. A. Jamieson (ed.), Interactionof Platelets and Tumor Cells, pp. 97-111. New York: Alan R. Liss Inc.,1982. 34.

12. Zacharski, L. R., Henderson, W. G., Rickles, F. R., Forman, W. B., VanEeckhout, J. P., Cornell, C. J., Jr., Forcier, R. J., and Marlon, J. F. Plateletsand malignancy: rational and experimental design for the VA cooperativestudy of RA-233 in the treatment of cancer. Am. J. Clin. Oncol., 5: 593- 35.609, 1982.

13. Wong, P. Y-K., Chao, P. H-W., and McGiff, J. C. Nafazatrom (Bay g-6575),an antithrombotic and antimetastatic agent, inhibits 15-hydroxyprostaglan-din dehydrogenase. J. Pharmacol. Exp. Ther., 233: 757-760, 1982.

14. Honn, K. V., Meyer, J., Neagos, G., Henderson, T., Westley, C., and 36.Ratanatharathorn, V. Control of tumor growth and metastasis with prosta-cyclin and thromboxane synthetase inhibitors: evidence for a new antitumor 37.and antimetastatic agent (Bay g 6575). In: G. A. Jamieson (ed.), Interactionof Platelets and Tumor Cells, pp. 295-331. New York: Alan R. Liss Inc., 38.1982.

15. Warrell, R. P., Jr., Bockman, R. S., Staszewsi, H., and Maiese, K. Clinicalstudy of a new antimetastatic compound, nafazatrom (Bay g 6575). Effectson platelet consumption and monocyte prostaglandin production in patients 39.with advanced cancer. Cancer (Phila.), 57: 1455-1460, 1986.

16. Temple, W. J., and Ketcham, A. S. Current clinical trials with anticoagulanttherapy in the management of cancer patients. In: K. V. Honn and B, F. 40.Sloane (eds.), Hemostatic Mechanisms and Metastasis, pp. 409-425. Boston:M. Nijhoff, Inc., 1984. 41.

17. Busse, W. D., Honn, K. V., Moeller, E., and Scuter, F. Treatment of themetastasis and growth of tumor cells. Eur. Patent No. 62319, October 13,1982. 42.

18. Hirsch, K. S., Jones, C. D., and Taylor, H. M. Aromatase imidazole derivatives. U.S. Patent 4,649,730. April 21, 1987.

19. Smith, G. F., and Campbell, J. B. Antithrombotic uses of substituted alkyl-amines. U.S. Patent Application filed. 43.

20. Matsumoto, K., and Ward, J. S. Terphenyl derivatives and pharmaceuticaluses thereof. U. S. Patent 4,495,202. January 22, 1985.

21. Sugiura, K., and Stock, C. C. Studies in a tumor spectrum. I. Comparison of 44.the action of methyl-bis(2-chlorethyl)amine and 3-bis(2-chlorethyl)-amino-methyl-4-methoxymethyl-5-hydroxy-6-methylpyridine on the growth of a 45.variety of mouse and rat tumors. Cancer (Phila.), 5: 382-402, 1952.

22. Fidler, 1. J. General considerations for studies of experimental cancer metas- 46.tasis. In: H. Busch (ed.), Methods in Cancer Research, pp. 399-439. NewYork: Academic Press, 1978.

23. Cunliffe-Beamer, T. L. Biomethodology and surgical techniques. In: H. L. 47.Foster, J. D. Small, and J. G. Fox (eds.), The Mouse in BiomÃ©dicalResearch,Vol. 3. Normative Biology, Immunology, and Husbandry, pp. 402-438. NewYork: Academic Press, 1983. 48.

24. Hagmar, B., Ryd, W., and Erkell, L. J. Methods in metastasis research. Acritical review. Invasion Melasi as.. 3: 1-21, 1983. 49.

25. SAS Users Guide: Statistics. Version 5 Edition. SAS Institute Inc. Cary, NC:1985.

26. Simon, R. M. Design and conduct of clinical trials. In: V. T. DeVita, Jr., S. 50.Hellman, and S. A. Rosenberg (eds.), Cancer. Principles and Practice ofOncology, pp. 198-225. Philadelphia: J. B. Lippincott, 1982.

Maniglia, C. A., Tudor, G., Gomez, J., and Sartorelli, A. C. The effect of2,6-bis(diethanolamino)-4-piperidinopyrimido[5,4-</]pyrimidine (RA233) ongrowth, metastasis and lung colony formation of B16 melanoma. CancerLett., 16: 253-260, 1982.Gastpar, V. H. Der einfluss von antikoagulantien und aggregationschemmernauf die metastasierung. In: E. Kronowski (ed.), Neue Aspekte der KrebsbekÃ¤mpfung,pp. 111-130. Stuttgart: Geforg Thieme, 1979.Ambrus, J. L., Ambrus, C. M., Gastpar, H., and Williams, P. Study ofplatelet aggregation in vivo. IX. Effect of nafazatrom on in vivo plateletaggregation and spontaneous tumor metastasis. J. Med., 13: 35-47, 1982.Haas, J. H., CorbeÂ«,T. H., and Haas, C. D. Nafazatrom: Lack of antimetastatic effect. Proc. Am. Assoc. Cancer Res., 26: 51, 1985.Stamford, I. F., Melhuish, P. B., Carroll, M. A., Corrigan, C. J., Patel, S.,and Bennett, A. Survival of mice with NC carcinoma is unchanged by drugsthat are thought to inhibit thromboxane synthesis or increase prostacyclinformation. Br. J. Cancer, 54: 257-263, 1986.Lukas, G., Brinale. S. D., and Greengard, P. The route of absorption ofintraperitoneally administered compounds. J. Pharmacol. Exp. Ther., 178:562-566,1971.Philipp, E., Ritter, W., and Patzschke, K. The relationship between dose,pharmacokinetics, plasma-concentrations and antithrombotic effects of nafazatrom. Thrombosis Res. Suppl., 4: 129-133, 1983.Honn, K. V., Onoda, J. M., Menter, D. G., Taylor, J. D., and Sloane, B. F.Prostacyclin/thromboxanes and tumor cell metastasis. In: K. V. Honn andB. F. Sloane (eds.), Hemostatic Mechanisms and Metastasis, pp. 207-231.Boston: M. Nijhoff Inc., 1984.Vicenzi, E., Valori, V. M., Dalessandro, A. P. B., Niewiarowski, A., Rotilio,D., and Donati, M. B. The prostacyclin-thromboxane balance and experimental metastasis: Questions from pharmacological studies. In: K. Hillmannand S. A. Eccles (eds.), Treatment of Metastasis: Problems and Prospects.Proceedings, pp. 267-270. Taylor and Francis Pub., London, 1985.Powles, T. J., Bockman, R. S., Honn, K. V., and Ramwell, P. (eds.),Prostaglandins and Cancer. New York: Alan R. Liss Inc., 1982.Thaler-Dao, H., Crastes De Paulet, A., and Paoletti, R. (eds.) Icosanoids andCancer. New York: Raven Press, 1984.Chiabrando, C., Broggini, M., Castagnoli, M. N., Donelli, M. G., Noseda,A., Visintainer, M., Garattini, S., and Fanelli, R. Prostanglandin and thromboxane synthesis by Lewis lung carcinoma during growth. Cancer Res., 45:3605-3608,1985.Young, M. R., Newby, M., and Meunire, J. Relationships between morphology, dissemination, migration, and prostaglandin E2 secretion by clonedvariants of Lewis lung carcinoma. Cancer Res., 45: 3918-3923, 1985.Hirsch, J. Progress review: The relationship between dose of aspirin, side-effects and antithrombotic effectiveness. Stroke, 16:1-4, 1985.Tohgo, A., Tanaka, N. G., and Hidemasa, O. Platelet-aggregating activitiesof metastasizing tumor cells. III. Platelet aggregation as resulting fromthrombin generation by tumor cells. Invasion Metastas., 5:96-105, 1985.Donati, M. B., Rotillio, D., Delaini, F., Giavazzi, R., Mantovani, A., andPoggi, A. Animal models for the study of platelet-tumor cell interactions. In:G. A. Jamieson (ed.), Interaction of Platelets and Tumor Cells, pp. 159-176.New York: Alan R. Liss Inc., 1982.Maat, B., and Hilgard, P. Anticoagulants and experimental Metastasis-evaluation of antimetastatic effects in different systems. J. Cancer Res. Clin.Oncol., 101: 275-283, 1981.Markus, G. The role of hemostasis and fibrinolysis in the metastatic spreadof cancer. Sem. Thromb. Hemost., 10:61-70, 1984.Sugerbaker, E. V. Some characteristics of metastasis in man. A. J. Pathol.,97:623-632, 1979.Crissman, J. D. Is there clinical relevance for therapies which disrupt themetastatic cascade? In: K. V. Honn and B. F. Sloane (eds.), HemostaticMechanisms and Metastasis, pp. 1-14. Boston: M. Nijhoff Inc., 1984.Willmott, N., Malcolm, A., McLeod, T., Gracie, A., and Caiman, K. C.Changes in anatomical distribution of tumor lesions induced by platelet-active drugs. Invasion Metastas., 3: 32-51, 1983.Hagmar, B., and Boeryd, B. Disseminating effect of heparin on experimentaltumour metastasis. Pathol. Eur., 4: 274-282, 1969.Hilgard, P. Blood platelets and tumor dissemination. In: G. A. Jamieson(ed.), Interaction of Platelets and Tumor Cells, pp. 143-158. New York:Alan R. Liss Inc., 1982.Johnson, J. R., and Temple, R. Food and drug administration requirementsfor approval of new anticancer drugs. Cancer Treat. Rep., 69: 1155-1159,1985.

4516



1989;49:4509-4516. Cancer Res Ronald L. Merriman, Katherine A. Shackelford, Lee R. Tanzer, et al. SurvivalLack of Correlation between Inhibition of Lung Metastasis and Drug Treatments for Metastasis of the Lewis Lung Carcinoma:

Updated version

http://cancerres.aacrjournals.org/content/49/16/4509

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/49/16/4509To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

Drug Treatments for Metastasis of the Lewis Lung Carcinoma ... · Lilly Research Laboratories, Lilly Corporate Center, Indianapolis, IN 46285. aggregation inhibitors have been examined