[CANCER RESEARCH 45, 3735-3741, August 1985]

Successful Immunotherapy of Murine Experimental Hepatic MétastaseswithLymphokine-activated Killer Cells and Recombinant Interleukin 2

Rene Lafreniere1 and Steven A. Rosenberg

Surgery Branch, Division of Cancer Treatment, National Cancer Institute, Bethesda, Maryland 20205

ABSTRACT INTRODUCTION

Lymphokine-activated killer (LAK) cells are generated in vitro

by the incubation of normal murine splenocytes in interleukin 2.We have shown previously that the systemic injection of LAKcells in conjunction with recombinant interleukin 2 can reducethe number of established pulmonary métastasesin mice. In anattempt to study this approach in the treatment of hepaticmétastases, we developed a technique for the induction ofhepatic métastasesin mice based on the intrasplenic injection oftumor cells and have tested the effects of LAK cells and recombinant interleukin 2 produced in Escherichia coli (RIL-2) therapy

on these métastases. Treatment with LAK cells alone in 14consecutive experiments rarely produced significant reduction inmétastases over control (mean percentage reduction, 12%).Therapy with RIL-2 alone produced a dose-dependent reduction

in the number of liver métastases.In 20 consecutive experimentswhen RIL-2 was administered i.p. three times a day at doses

varying from 1,000 to 5,000,10,000 to 15,000, and 25,000 units,a statistically significant (P < 0.05) reduction in liver métastaseswas seen in 2 of 12, 2 of 4, and 8 of 12 determinations,respectively (percentage reduction, 0 to 97; mean, 42%). Atdoses greater than 25,000 units, the reduction in métastaseswas highly reproducible (percentage reduction, 66 to 95; mean,83%) and was statistically significant in 14 of 14 determinations.When LAK cells were given i.v. in addition to RIL-2 administration

in 16 consecutive experiments, the percentage reduction in livermétastaseswas markedly increased over that seen with RIL-2

alone (mean percentage reduction, 77% at doses of 5,000 to25,000 units of RIL-2 and mean reduction, 97% for doses greaterthan 25,000 units of RIL-2). At doses of 5,000, 10,000, 25,000,and greater than 25,000 units of RIL-2 plus LAK cells, significant

reduction of liver métastases(P < 0.05) was achieved in 3 of 7,2 of 2, 8 of 8, and 6 of 6 determinations, respectively.

When animals were given fresh splenocytes or splenocytescultured in complete medium without RIL-2 instead of LAK cells,

no reduction in liver métastaseswas seen except for that attributable to the administration of RIL-2 alone. Sublethal total body

irradiation of the mice prior to therapy abrogated the therapeuticeffects of RIL-2, but the effects of treatment with LAK cells plusRIL-2 were maintained. Thus, treatment with RIL-2 alone or in

combination with LAK cells is effective in reducing the numberof established hepatic micrometastases in a murine model. Thesestudies are in accord with our previous observations concerningthe effective therapy of established pulmonary métastaseswithRIL-2 plus LAK cells and provide a rationale for the extension of

these observations to the treatment of metastatic cancer inhumans.

1To whom requests for reprints should be addressed, at Surgery Branch,

National Cancer Institute, Building 10, Room 2B42, Bethesda. MD 20205.Received 2/8/85; revised 4/29/85; accepted 5/9/85.

Passive immunotherapy involves the transfer to the tumor-bearing host of previously sensitized immunological reagents,such as cells or antibodies that have the ability to mediateantitumor responses. The term adoptive immunotherapy is usually used to connote passive immunotherapy with cells such aslymphocytes or macrophages. In general, the use of cells obtained from highly immunized animals has been essential to thesuccess of adoptive immunotherapy in animal tumor models (1,2).

We have recently described a simplified method for generatinglymphoid cells with selective antitumor reactivity that can beused in the adoptive immunotherapy of tumors (3-6). The incubation of normal murine or human splenocytes in IL-22, a lym-

phokine produced by lectin or antigen-activated T-cells (7, 8),

gives rise to lymphoid cells that are specifically cytotoxic to fresh,non-cultured, autologous, syngeneic, and allogeneic primary and

metastatic tumor cells regardless of their natural killer susceptibility but are not toxic to normal cells. We have called thesecytotoxic cells LAK cells. They bear the Thy-1+, Lyt-1~2+ cellsurface phenotype in the mouse (5) and are OKT-3+4~8+ in the

human (3).The recent availability of large amounts of RIL-2 with known

biological activity (9) has made it possible to evaluate the anti-

tumor effectiveness of this lymphokine in vivo either alone or inconjunction with LAK cells. In a recent report, Mule ef al. (10)documented the effectiveness of transfused LAK cells plus repetitive injections of RIL-2 in the treatment of established pul

monary micrometastases from multiple sarcomas. The effects oflow doses of RIL-2 or of LAK cells alone were not significant;

however, concomitant administration of both caused a significantreduction in the number of metastatic foci in the lungs. In laterstudies, we showed that administration of RIL-2 alone was

effective in reducing the number of métastasesbut only whengiven in very high doses (11).

In a new experimental model designed to produce métastasessolely in the liver, we undertook studies to assess the effectiveness of adoptive therapy with RIL-2 alone and LAK cells plusRIL-2 on metastatic foci in this location. This model also tested

the ability of i.v. injected LAK cells to mediate antitumor effectsin an organ that did not represent the first capillary bed (i.e., thelung) encountered by LAK cells after i.v. injection. In this paper,we have demonstrated that treatment with high doses of RIL-2

alone can mediate the successful therapy of hepatic micrometastases. The therapeutic effect is enhanced by the concomitanti.v. injections of LAK cells. The factors responsible for successfultherapy in this model have been analyzed.

2The abbreviations used are: IL-2, interleukin 2; HBSS, Hanks' balanced salt

solution; RIL-2, recombinant interleukin 2 produced in Escherichia coli; LAK,lymphokine-activated killer cells; MCA, 3-methylcholanthrene.

CANCER RESEARCH VOL. 45 AUGUST 1985

3735

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

REDUCTION OF LIVER METASTASES IN VIVO BY RIL-2 AND LAK PLUS RIL-2

MATERIALS AND METHODS

Animals. C57BL/6 mice were obtained from the Animal ProductionColonies of the National Institutes of Health, Bethesda, MD. Mice werefed standard mice chow and water ad libitum and were used in experiments when 12 weeks or older.

Splenocytes. Spleens were harvested aseptically from C57BL/6 female retired breeders and crushed with the hub of a syringe in HBSS(Biofluids, Rockville, MD). The cell suspension and spleen fragmentswere passed through a single layer of 100-gauge nylon mesh (Nitex;

Lawshe Industrial Co., Bethesda, MD), and the erythrocytes were lysedosmotically with 10% buffered ammonium chloride solution (NIH MediaUnit) at room temperature for 2 min. The cells were then centrifuged andwashed 3 times with HBSS.

RIL-2. The gene for IL-2 (12), isolated from a high-producer Jurkat cell

line, was expressed at high levels in Escherichia coli and purified toapparent homogeneity as recently described (9). This human RIL-2 waskindly supplied by the Cetus Corp. (Emeryville, CA). Purified RIL-2 hada specific activity of 3 to 4 x 106 units/mg. The endotoxin level in thepurified preparation was less than 0.1 ng/106 units of RIL-2 as measured

in a standard limulus assay. RIL-2 was lyophilized and reconstituted with

distilled water. The preparation contained 5% mannitol and sodiumdodecyl sulfate, 131 pg/mg RIL-2 as a vehicle.

Tumor. MCA-105, a sarcoma syngeneic in C57BL/6 mice was used

in these experiments. This tumor had been induced in our laboratory bythe i.m. injection of 0.1 ml of 1% 3-methylcholanthrene in sesame oil asdescribed previously (13). A large number of vials of MCA-105 from the

first passage generation were cryopreserved. After thawing from storageat -70°C, tumor was injected s.c. in C57BL/6 mice for serial passage

and was always used within the first 7 transplant generations, at whichtime a new vial was thawed and used. Single-cell suspensions of tumor

were prepared as described previously (10). Briefly, fresh sarcomatumors were excised, minced with scissors, and stirred in a triple enzymesolution of deoxyribonuclease, hyaluronidase, and collagenase (SigmaChemical Co., St. Louis, MO) for 3 h at room temperature, filteredthrough 100-gauge nylon mesh (Nitex), washed 3 times in HBSS without

calcium or magnesium, and resuspended at the appropriate cell concentration for injection in HBSS without calcium or magnesium.

Generation of LAK Cells. LAK cells were generated by placing 5 x10" fresh splenocytes (prepared as described above) into 175-sq cm

(750 ml) flasks (Falcon Labware, Becton, Dickinson and Co., Oxnard,CA) containing 175 ml of complete medium, which consisted of RPMI1640 (Biofluids) with 10% fetal calf serum (Biofluids), 0.03% fresh glu-

tamine, streptomycin (100 ßg/m\),penicillin (100 units/ml) (all from theNIH Media Unit), 0.1 HIM nonessential amino acids, 0.1 ¿IMsodiumpyruvate (all from Microbiological Associates, Walkersville, MD), 5 x 10~5

M of 2-mercaptoethanol (Aldrich Chemical Co., Milwaukee, Wl), genta-

micin (50 Mg/ml) (Shearing, Kenilworth, NJ), and fungizone (0.5 M9/ml)(Flow Laboratories, McLean, VA). RIL-2 (175,000 units) was added toeach flask. The flasks were incubated supine at 37°Cin a moist atmos

phere with 5% CO2 for 72 h. The cells (LAK) were then harvested intosterile 250-ml centrifuge tubes (Corning No. 25350; Corning GlassWorks, Corning, NY) passed over Ficoll (Lympholyte-M; Cedarlane Lab

oratories Ltd., Hornby, Ontario, Canada) to remove dead cells, andwashed 3 times with HBSS before resuspending in HBSS for i.v. injection.Aliquots of LAK cells were tested for cytotoxicity in vitro in a standard4-h 51Cr release assay against a fresh MCA-102 sarcoma target as

described previously (5, 14). The specific cytotoxicity of the adoptivelytransferred LAK cells ranged between 35 and 70% at an effectortargetratio of 100:1, unless otherwise noted.

Induction of Hepatic Métastases. C57BL/6 mice were anesthetizedwith an i.p. injection of pentobarbital (Somnifer; Richmond VeterinarySupply Co., Richmond, VA); 0.15 ml were given from a stock solutioncontaining 6 grains in 45 ml of phosphate-buffered saline (PBS; Biofluids).

Under a laminar flow hood, the mice were then positioned on a board inthe right lateral position, prepped with 70% ethanol, dried with sterile

gauze, and then using sterile scissors, a 5-mm left subcostal incision

was made. The spleen was exposed, and its short gastric vessels alongwith the gastrosplenic ligaments were cut allowing the spleen to bebrought onto the animal's abdominal wall attached to its splenic pedicle.

We had initially titered the dose of tumor cells required to give a countablenumber of liver métastases at 13 to 15 days and found 3 x 105 cells/

injection to be optimal. One ml of a single-cell suspension containing 3x 105 MCA-105 cells in HBSS prepared as described above was then

injected into the upper pole of the exposed extraabdominal spleen usinga 27-gauge needle (American Hospital Supply, McGaw Park, IL). During

the injection, the spleen swelled slightly, and then the injected suspensionflowed into the splenic vein with virtually no leakage of fluid. The injectioncompleted, the needle was removed, and a period of 1 min was allowedto elapse to permit all tumor cells to be flushed into the portal circulation.The splenic pedicle was then clipped with a medium hemoclip (EdwardWeek and Co., Inc., Research Triangle Park, NC), the spleen wasremoved, and the splenic pedicle was then dropped i.p. The abdominalwall musculature and skin were then closed in one layer using 9 mmautoclip wound clips (Clay Adams, Parsippany, NJ). The animals werethen randomly allocated to their respective therapy group and allowedto recover. By Day 3 after tumor injection, micrometastases were apparent histologically (Fig. 1).

Adoptive Immunotherapy Model. For treatment of micrometastases,LAK cells were suspended at 1 x 10°cells in 1 ml of HBSS and injected

into the tail vein on Days 3 and 6 after tumor injection. Mice were giveninjections of either HBSS or RIL-2 i.p., 0.5 ml/injection every 8 h. These

injections were given from Days 3 through 10 for HBSS and RIL-2 doses< 25,000 units/injection and from Days 3 through 7 for RIL-2 doses >25,000 units. The units of RIL-2 activity were determined using an IL-2

dependent cell line as described previously (9, 14, 15). The amount ofRIL-2 injected (in units) varied depending upon the design of each

experiment and is noted in the legends to the charts and tables.At least 6 mice were included in each treatment group. At 14 days

after tumor injection, the mice were ear tagged and randomized. Themice were then given a tail vein injection consisting of 0.5 ml of a 15%solution of India ink (Higgins Black No. 4417; A. W. Faber-Castell) inphosphate-buffered saline. The carbon particles from the ink were found

to be phagocytized by the Kupffer cells of the liver, rendering the liverblack. The mice were then killed by cervical dislocation, and their liverswere harvested and bleached by Fekete's solution (16), allowing the

métastases to become easily countable as they formed discrete whitenodules on the surface of the liver, since they did not incorporate theIndi^ ink (Figs. 2 and 3). Nodules were counted in a blind fashion withoutknowledge of the treatment of the mouse. Liver métastasestoo numerous to count upon autopsy were assigned an arbitrary value of 250,since we were able to count reliably only numbers of métastases approaching 250/liver. After all data were recorded, the code was broken.

Statistical Analysis. The significance of differences in numbers ofliver métastasesbetween groups was determined by the Wilcoxon ranksum test (17). Two-tailed Ps are presented in all experiments.

Fig.3. Representativelobes of liver (see Experiment 1, H in Table 1) from thegroup given HBSS (left) or LAK plus RIL-2 (right).

CANCER RESEARCH VOL. 45 AUGUST 1985

3736

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

REDUCTION OF LIVER METASTASES IN VIVO BY RIL-2 AND LAK PLUS RIL-2

RESULTS

Effective Reduction of the Number of Established HepaticMétastases by RIL-2 and LAK Cells Plus RIL-2. Mice with

established Day 3 liver métastases were treated with varyingdoses of RIL-2 alone or in combination with LAK cells. LAK cells(108) were given i.v. on Days 3 and 6, and RIL-2 was given i.p.every 8 h as described in "Materials and Methods." Results of 2

characteristic experiments are shown in Table 1 and Chart 1.Mice receiving RIL-2 alone had significantly fewer numbers of

established liver métastasesby Day 14 when treated with dosesgreater than 5000 units of RIL-2 (P < 0.02) compared to mice

receiving HBSS alone. Treatment with LAK cells alone (combinedwith HBSS) was not usually capable of significantly reducing thenumber of liver métastases.However, when treatment with LAKcells was combined with the administration of RIL-2, the greatest

reductions in liver métastaseswere seen. Combining LAK cellswith RIL-2 also produced significant-reduction in liver métastaseswhen compared to RIL-2 alone.The extent of reduction of livermétastases was dependent upon the dose of RIL-2. The ap

pearance of the livers at the time of harvest from a representativeexperiment is shown in Fig. 3.

Dose Titration of RIL-2 in Vivo. The reduction in the numberof liver métastasesresulting from RIL-2 administration alone was

highly dose dependent. Results of 42 determinations from 20consecutive experiments are shown in Chart 2. The reduction inmétastaseswas variable at RIL-2 doses less than or equal to

25,000 units with a percentage reduction ranging from 0 to 97%(mean, 42% in 28 determinations). At doses from 1,000 to 5,000,10,000 to 15,000, and 25,000 units, the percentage reductionover HBSS was statistically significant (P < 0.05) in 2 of 12,2 of4, and 8 of 12 determinations, respectively. At doses greaterthan 25,000 units, the reduction in liver métastaseswas highlyreproducible with a range of percentage reduction of 66 to 95%(mean, 83% in 14 determinations); 14 of 14 determinations hada statistically significant decrease in the number of métastases(P < 0.0005). At doses equal to or greater than 100,000 unitsgiven i.p. approximately every 8 h, significant toxicity was seenby 5 days that limited the amount of RIL-2 that could be given

in any one experiment. This toxicity was manifested as weightgain of the animals and ascites with pleural effusions seen atautopsy.

When 1 x 10" LAK cells/ml were given i.v. on Days 3 and 6

after tumor injection in addition to RIL-2, the percentage reduc

tion of liver métastaseswas markedly increased over the corresponding reduction when RIL-2 alone was used. Results of 23

determinations from 16 consecutive experiments are shown inChart 3. At doses of 5,000 to 25,000 units, the correspondingrange of percentage reduction was 6 to 99% (mean, 75% in 17determinations). At doses of 5,000, 10,000, and 25,000 units,the percentage reduction over HBSS control was statisticallysignificant (P < 0.05) in 3 of 7, 2 of 2, and 8 of 8 determinations,respectively. At doses greater than 25,000 units, the range inpercentage reduction was 95 to 99% (mean, 97% in 6 determinations); 6 of 6 determinations resulted in a statistically significantreduction in the number of métastasescompared to the HBSScontrol. Statistically significant differences were seen betweenmice treated with LAK cells plus RIL-2 compared to RIL-2 alone,both in the <25,000 unit group (P < 0.0005) and >25,000 unitgroup (P < 0.0005).

Necessity of in Vitro Activation of Fresh Splenocytes byRIL-2 to Achieve Immunotherapeutic Effect On Liver Métastases. The increased reduction in liver métastasesobtained byLAK cells plus RIL-2 compared to RIL-2 alone was dependent

upon the activation of LAK cells. The i.v. adoptive transfer of 1x 10" fresh splenocytes or splenocytes cultured for 3 days

without RIL-2 failed to achieve any significant antitumor effectwhen given in combination with RIL-2 compared to RIL-2 given

alone (Table 2). However, when LAK cells were combined withRIL-2, a significant reduction in the number of métastasesoverRIL-2 alone was achieved (P < 0.01 in both experiments).

Elimination of Therapeutic Effect of RIL-2 Alone but not ofLAK Cells Plus RIL-2 by Irradiation of the Host. We examined

the effect of a sublethal dose of irradiation upon the reduction ofliver métastasesby RIL-2 and LAK cells plus RIL-2. Mice weregiven 500 rads total body irradiation from a 137Cssource. One h

later, they were given injections of tumor cells intrasplenically,and therapy was begun 3 days later (Chart 4).

In the normal nonirradiated host, treatment with 50,000 unitsRIL-2 i.p. 3 times a day significantly reduced the number of liver

métastaseswhen used alone (mean number of métastases inHBSS group was 247 compared to 17 in the RIL-2 group; P <

0.005) and when given with LAK cells (mean, 0.2 métastases;P< 0.004). However, in irradiated mice, treatment with RIL-2 alone

was ineffective in reducing the number of liver métastases(250in the HBSS group compared to 227 in the RIL-2 group), whereas

Table 1Dose titration of RIL-2 versus reduction of established Day 3 MCA-105 liver micrometastases with or without

LAK cell treatmentThe number of MCA-105 cells injected i.v. «as3 x 105; the number of mice used was 6. In both experiments,

RIL-2 or HBSS was injected i.p. approximately every 8 h from Days 3 to 10 after tumor injection.

Number of métastases(mean) with8

RIL-2 alone (units) LAK + RIL-2 (units)6

Experi-ment Dayc 0 5,000 10,000 25,000 100,000 0 5,000 10,000 25,000 100,000

14 242 140 61 95

14 241 213 Not done 100

43 198 30 8

82 240 182 Notdone 130 Statistical significance of differences. Experiment 1, A versus: B, P < 0.003; C, P < 0.005; D, P < 0.02;

E, P < 0.005; F, P < 0.03; G, P < 0.005; H, P < 0.005; I, P < 0.01; J, P < 0.02. Experiment 2, A versus: B,nonsignificant; D, P < 0.004; E, P < 0.006; F, nonsignificant; G, nonsignificant; I, P < 0.002; J, P < 0.004.

0 LAK cells (1 x 10") were injected i.v. on Days 3 and 6 after tumor injection.c Time at which livers were removed after tumor injection.

CANCER RESEARCH VOL. 45 AUGUST 1985

3737

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

REDUCTION OF LIVER METASTASES IN VIVO BY RIL-2 AND LAK PLUS RIL-2

>250

liftK| 100

10.000 25.000 100.000

(Unit»of RIL-2)

K LAK LAK LAK LAK

¡S 5.000 tO.OOO 2SDOO 100000

(LAK - Units of RIL-2)

Chart 1. Effect of RIL-2 and LAK cells on therapy of MCA-105 liver métastases.The decrease in experimentally induced MCA-105 liver métastasescaused by theinjection of RIL-2 given i.p. every 8 h (/eft), was compared to the same doses ofRIL-2 administered concomitantly with 1 x 10" syngeneic LAK cells generated in

vitro. RIL-2 was given i.p. beginning on Day 3, and LAK cells were given i.v. atDays 3 and 6 posttumor injection (right). The number of métastaseswere countedon a coded fashion at Day 14 after MCA-105 injection. •,measure of the numberof liver métastasesin an individual mouse.

100801

60Q

40

i* 20-«0-20Uot-

•*•*-•

*•s.^**¥

•,, ,1,11,1 1 , 1,aß

5.000 10.000 15.000 25.000^

"*"*£*•1

1 1 il, l 1 1150.000

75.0OO 100.000 150,000

UNITS OF RIL-2

Chart 2. Dose titration of the ability of RIL-2 to reduce liver métastasesin 20consecutive experiments. Each animal was injected with 3x10* MCA-105 cellsand treated with RIL-2 i.p. every 8 h starting on Day 3. Increasing doses of RIL-2

led to increasing reduction in the number of hepatic métastases. •,separateexperimental determination.

1 ill 1 1 1 ill 1110.000 15.000

UNITS OF RIL-2

50.000 75.000 '00000 150.000

Chart3. Effect of LAK cells plus increasing doses of RIL-2 on MCA-105 livermétastasesin 16 consecutive experiments. Each animal was injected with 3 x 10sMCA-105 cells and treated with RIL-2 i.p. every 8 h starting on Day 3. Increasingdoses of RIL-2 led to increasing reduction in the number of hepatic métastaseswhen 10* LAK cells were administered i.v. concurrently on Days 3 and 6. •,

separate experimental determination.

Table 2Effects of LAK cells versus fresh cells and cultured cells with and without RIL-2

on the reduction of established Day 3 liver métastasesThe number of MCA-105 sarcoma cells injected intrasplenically was 3x10*;

groups contained 6 mice. Livers were removed 14 days after tumor injection inExperiments 1 and 2, and métastaseswere counted blindly.

GroupA

BCDE

FGHTreatmentCells"None

NoneLAKLAKCultured splenocyteCultured splenocyteFresh splenocyteFresh splenocyteNumber

of métastases(mean)3RIL-2C

Experiment1250

+ 214250

+ 95250

+ 215250

+ 240Experiment

2246

53193

1250

50245

60a Statistical significance of differences: Experiment 1, A versus: B, nonsignificant;

C, nonsignificant; D, P < 0.003; E, nonsignificant; F, nonsignificant; G, nonsignificant; H, nonsignificant; B versus: D, P < 0.01 ; F, nonsignificant; H, nonsignificant.Experiment 2, A versus: B, P < 0.005; C, P < 0.05; D, P < 0.005; E, nonsignificant;F, P < 0.005; G, nonsignificant; H, P < 0.005; B versus: D, P < 0.005; F,nonsignificant; H, nonsignificant.

61 x 10* LAK cells, fresh normal splenocytes, or splenocytes cultured for 3

days in CM without RIL-2 were given i.v. on Days 3 and 6 after tumor injection.c Mice received 5000 units/injection or 7500 units/injection of RIL-2 in Experi

ments 1 and 2, respectively, from Days 3 through 10 after tumor injection.

>250

LAK

HBSS

LAK

ML-2

IRRADIATED HOST(500 Rad» NON-IRRADIATEDHOST

Chart 4. Effect of host irradiation on RIL-2 and LAK plus RIL-2 therapy of livermétastases.At day 0, one-half of mice (6/group) were given a sublethal total bodyirradiation dose of 500 rads with '37Cs, and 1 h later, they were anesthetized andgiven injections intrasplenically of 3 x 10s MCA-105 cells (left). RIL-2 (50,000 units)

was given i.p. every 8 h for 5 days beginning on Day 3. Syngeneic LAK cells (1 x10") were given i.v. on Days 3 and 6. Mice were ear tagged and sacrificed at Day

14, and liver métastaseswere counted blindly. High doses of RIL-2 alone couldreduce the number of liver métastasesin the normal but not in the irradiated mice.LAK cells plus RIL-2 effectively reduced métastasesin both normal and irradiated

the combination of RIL-2 with LAK cells was highly effective in

irradiated mice (mean, 0.3 métastases;P < 0.009) in reducingliver métastases.

DISCUSSION

In work reported previously from our laboratory (3-6), we have

estalished that normal murine splenocytes or human peripheralblood lymphocytes cultured in vitro with the lymphokine IL-2

developed lytic activity for fresh autologous, noncultured, syngeneic, or allogeneic primary and metastatic tumor cells, regard-

CANCER RESEARCH VOL. 45 AUGUST 1985

3738

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

REDUCTION OF LIVER METASTASES IN VIVO BY RIL-2 AND LAK PLUS RIL-2

less of their natural killer susceptibility, but not to normal cells,when tested in an in vitro cytotoxicity 51Crrelease assay. Detailed

studies in the mouse and human have characterized the natureof the phenotype of both the precursor and effector cells involvedin this phenomenon. These cells have been called LAK cells.

Recently, Chang ef al. (14) reported that systemically administered RIL-2 resulted in the generation of spleen cells in vivothat were cytotoxic in vitro in a 4-h 51Cr release assay against

fresh tumor targets. Further studies by Ettinghausen et al. (18)have also shown that RIL-2 given systemically will result in the

appearance of proliferating cells in vivo in multiple organs (lung,liver, spleen, kidney, lymph nodes) that resemble activated lymphocytes on histological sections and bear the Thy-1.2 antigen.

When the organs harboring such cells were harvested and thesecells were separated on Ficoll gradients, these activated lymphocytes were cytotoxic in vitro in a 4- and 18-h 51Cr release assay

against fresh tumor targets. Thus, it appears that RIL-2 admin

istration leads to LAK cells generation in vivo.Recently Mule ef al. (10) reported on the therapy of established

pulmonary micrometastases with LAK cells plus RIL-2. In that

model, LAK cells given alone had no effect. In a separate study,Rosenberg ef al. (11) have also reported on the successfultherapy of established lung métastases and subdermal solidtumor with high doses of RIL-2 alone, whereas sodium dodecyl

sulfate given alone had no effect. LAK cells are larger thanresting lymphocytes and tend to lodge in the first capillary bedthey traverse (the lung) shortly after i.v. injection. It was thus ofinterest to study the therapeutic effects of LAK cells in organsother than the lung to determine whether LAK cells could mediatetherapeutic effects systemically. Because no reproducible modelexisted for the study of isolated liver métastases,we first developed a model for the induction of artificial métastases in thisorgan that could be easily quantitated.

By injecting tumor cell suspensions into the spleen, we avoidedthe difficulties of injecting directly into the portal vein and wereable to generate métastaseswith a single operative procedure.The spleen was removed to prevent the development of a largetumor mass in this organ. It was found that 3 x 105 MCA-105

cells injected in 1 ml of HBSS without calcium and magnesiumwould usually produce around 250 countable métastasesby 2weeks. The métastasescould be identified histologically by Day3 (Fig. 1) and were visible grossly by Day 8 after tumor injection.The model produced métastases that were easily countablewhen the liver was bleached in Fekete's solution to produce

white nodules against the black liver parenchyma generated bythe uptake of India ink injected prior to sacrifice. The number ofmétastasesproduced was highly reproducible. Gross and histological examination of other organs at autopsy failed to revealmetastatic deposits outside of the liver. Splenectomy did notimpact on the effectiveness of RIL-2 and LAK plus RIL-2 therapy

when tested in our lung métastasesmodel (results not shown).Hougen ef a/. (19) have shown that splenectomized nude micehave a comparable number of lymphocytes in lymph nodes andblood possibly due to release of lymphocytes by the lympho-

myeloid organs in compensation for the loss of the spleen.In this study, we have analyzed some of the requirements

necessary for the successful treatment of established Day 3 livermétastaseswith RIL-2 and LAK cells plus RIL-2. The productionof RIL-2 in large quantities by isolating the gene coding for IL-2

and expressing it in E. coli (9, 12) has made it possible to test

the in vivo effects of this lymphokine. As shown in Charts 1 and2 and Table 1, the repetitive administration of large doses ofRIL-2 was able to achieve reduction of established liver métas

tases. This requirement probably reflects the necessity toachieve sustained bioavailability of the lymphokine in vivo due toits short serum half-life after i.p. injection (14, 15). The fact thatthe systemic administration of RIL-2 was capable of significantly

reducing liver métastasesin vivo is in accord with the investigations of Ettinghausen ef a/.3 who showed that IL-2 can generate

cytotoxic cells in the liver in vivo.The percentage reduction in the number of liver métastases

by RIL-2 was seen in a dose-dependent fashion; doses greater

than 25,000 units i.p. 3 times a day were necessary to achievea reproducibly significant effect. These doses, however, carriedwith them significant toxicity, thus limiting the amount of RIL-2

that could be given in any one experiment. Our results are inaccord with studies of the therapy of lung métastasesby Muleef al. (10) who showed that the i.v. transfer of LAK cells givenconcomitantly with RIL-2 could significantly reduce the numberof established metastastic foci at doses of RIL-2 that had no

effect when administered alone.In the course of multiple experiments in the treatment of liver

métastases,it was found that the systemic transfer of LAK cellsgiven alone at Days 3 and 6 after tumor induction rarely wasable to mediate a significant reduction of métastases.However,as shown in Chart 3, when concomitant administration of RIL-2was given, we obtained a significant reduction in the number ofmétastases,especially at RIL-2 doses greater than 5,000 units

i.p. 3 times a day. The percentage reduction was significant inall 16 experiments when greater than 5,000 units of RIL-2 was

given in conjunction with LAK cells. Consistent reduction ofhepatic métastaseswith treatment by RIL-2 alone required doses

greater than 25,000 units and was always less effective thanwhen combined with LAK cell administration. Thus, the administration of LAK cells significantly reduced the number of metastatic foci in the liver when combined with RIL-2 at RIL-2 doses

that were much lower than those required without LAK cells. Inaddition, RIL-2 and LAK cells plus RIL-2 have been shown

recently to prolong survival and effect cures in animal bearingmurine sarcoma hepatic métastases."

The mechanism(s) of RIL-2-mediated antitumor effects in vivo

is currently unknown. However, as indicated by the data in Chart4, the effectiveness of high doses of RIL-2 against liver métas

tases was abrogated when the host was given a sublethal doseof total body irradiation prior to RIL-2 therapy. This suggeststhat the RIL-2 acts through a host lymphoid component rather

than by direct tumor cytotoxicity. The effects of LAK cells whengiven with RIL-2 were sustained despite host preirradiation. It is

unknown at this time whether LAK cells mediate their effects invivo by direct tumor toxicity or whether the host contributes arelatively radioresistant component to the response, as has beenshown in other tumor systems (20, 21). This finding, however,does suggest that LAK cells do not require full host immunocom-

petence to mediate antitumor effects in vivo and suggests thatthis treatment may lend itself to combination with other antican-

3S. Ettinghausen,E. H. Lipford III,J. J. Mulé,and S. A. Rosenberg.Recombinantinterleukin 2 stimulates in vivo proliferation of adoptively transferred lymphokine-activated killer (LAK)cells, submitted for publication.

* R. Lafreniereand S. A. Rosenberg.Successful therapy of hepatic métastasesfrom several murine tumors using lymphokine-activated killer (LAK) cells andrecombinant interleukin2, submitted for publication.

CANCER RESEARCH VOL. 45 AUGUST 1985

3739

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

REDUCTION OF LIVER METASTASES IN VIVO BY RIL-2 AND LAK PLUS RIL-2

cer therapies. The requirement for IL-2 in conjunction with LAK

cells suggests that the LAK cells divide in vivo and that thisproliferation is required for manifestation of the therapeutic effect.Using the transfer of LAK cells from congenie mice, Ettinghausenef a/.3 have demonstrated that this is indeed the case.

Work in our laboratory (22) had shown previously that phyto-hemagglutinin-activated lymphocytes and LAK cells were clearedby the lungs early after i.v. injection and were then distributedmainly to the liver and spleen. The finding that LAK cells plusRIL-2 therapy could successfully reduce liver métastasesmaybe related to this hepatic distribution of LAK cells.

Some of the mice treated with LAK cells plus RIL-2 when

autopsied at Day 14 failed to show any evidence of liver métastases, but most mice did have small numbers of metastaticdeposits. It is possible that the migration of LAK cells to the liveris not optimal and that more RIL-2 and/or the transfer of addi

tional LAK cells would cure additional mice. The presence ofsuppressor cells within the infused LAK cell population is anotherpossible problem (23, 24) that could be overcome by the removalof tumor-enhancing cells (25).

The findings reported here provide a rationale for the performance of clinical trials of RIL-2 alone or in conjunction with LAK

cells in the therapy of liver métastases in humans. We haverecently begun these clinical trials.

REFERENCES

1 Borberg, H., Oettgen, H. F.. Choudry, K., and Beattie, E. J. Inhibition ofestablished transplants of chemically induced sarcomas in syngeneic mice bylymphocytes from immunized donors. Int. J. Cancer, 70: 539-547, 1972.

2. Smith, H. G., Harmel, R. P., Hanna, M. G., Zwilling, B. S., Zbar, B., and Rapp,H. J. Regression of established intradermal tumors and lymph node métastasesin guinea pigs after systemic transfer of immune lymphoid cells. J. Nati. Cancerlnst.,58: 1315-1322, 1977.

3. Grimm, E. A., Mazumder, A., Zhang, H. Z., and Rosenberg, S. A. Lymphokine-activated killer cell phenomenon. Lysis of natural killer-resistant fresh solidtumor cells by interleukin-2 activated autologous human peripheral bloodlymphocytes. J. Exp. Med.. 755. 1823-1841, 1982.

4. Lotze. M. T., Grimm, E. A., Mazumder, A., Strausser, J. L., and Rosenberg,S. A. Lysis of fresh and cultured autologous tumor by human lymphocytescultured in T-cell growth factor. Cancer Res., 41: 4420-4425.1981.

5. Rosenstein, M., Yron, l., Kaufmann, Y., and Rosenberg, S. A, Lymphokine-activated killer cells: lysis of fresh syngeneic natural killer-resistant murinetumor cells by lymphocytes cultured in interleukin 2. Cancer Res., 44: 1946-

1953, 1984.6. Yron, I., Wood, T. A., Spiess, P. J.. and Rosenberg, S. A. In vitro growth of

murine T cells. V. The isolation and growth of lymphoid cells infiltratingsyngeneic solid tumors. J. Immunol., 725: 238-245,1980.

7. Gillis, S., Perm, M. M., Ou, W., and Smith, K. A. T cell growth factor: parametersof production and a quantitative microassay for activity. J. Immunol., 720:2027-2032,1978.

8. Morgan, D. A., Ruscelli, F. W., and Gallo, R. Selective in vitro growth of Tlymphocytes from normal human bone marrows. Science (Wash. DC), 792:1007-1008,1976.

9. Rosenberg, S. A., Grimm, E. A., McGrogan, M., Doyle, M., Kawasaki, E.,Koths, K., and Mark, D. F. Biological activity of recombinant human interteukin-2 produced in Escherichia coli. Science (Wash. DC), 223: 1412-1415,1984.

10. Mulé,J. J., Shu, S., Schwarz, S. L., and Rosenberg, S. A. Adoptive immune-therapy of established pulmonary métastaseswith LAK cells and recombinantinterleukin-2. Science (Wash. DC), 225: 1487-1489, 1984.

11. Rosenberg, S. A., Mulé,J. J., Shu, S., Spiess, P., Reichert, C. M., and Scharw,S. Regression of established pulmonary métastasesand subcutaneous tumormediated by the systemic administration of high dose recombinant IL-2. J.Exp. Med., 767: 1169-1188, 1985.

12. Taniguchi, T., Matsui, H., Fujita, T., Takaoka, C., Kashima, N., Yoshimoto, R.,and Hamuro, J. Structure and expression of a cloned c DMA for humaninterleukin-2. Nature (Lond.), 302: 305-309,1983.

13. Parker, G. A., and Rosenberg. S. A. Serologie identification of multiple tumor-associated antigens on murine sarcomas. J. Nati. Cancer Inst., 58: 1303-1309, 1977.

14. Chang, A. E., Hyatt, C. L., and Rosenberg, S. A. Systemic administration ofrecombinant human interleukin-2 in mice. J. Biol. Resp. Modif., 3: 561-572,1984.

15. Donohue, J. H., and Rosenberg, S. A. The fate of interleukin-2 after in vivoadministration. J. Immunol.. 730: 2203-2208,1983.

16. Fekete, E. A comparative morphological study of the mammary gland in a highand a low tumor strain of mice. Am. J. Pathol., 14: 557-578, 1938.

17. Gehan, E. A generalized Wilcoxon test for comparing arbitrarily single censoredsamples. Biometrika, 52: 203-223, 1965.

18. Ettinghausen, S. E., Lipford, E. H., Ill, Mulé,J. J., and Rosenberg, S. A.Systemic administration of recombinant interleukin-2 stimulates in vivo lymphoid cell proliferation in tissues. J. Immunol.. in press, 1985.

19. Hougen, H. P., Hansen, F., Jensen, E. K., and Ropke, C. Kinetics of smalllymphocytes in normal and nude mice after splenectomy. Scand. J. Haematol.,78: 257-266,1977.

20. Mulé,J. J., Rosenstein, M., Shu, S., and Rosenberg, S. A. Eradication of adisseminated syngeneic lymphoma by systemic adoptive transfer of immunelymphocytes and its dependence upon a host component(s). Cancer Res., 45:526-531,1985.

21. Shu, S., Fonseca, L. S., Kato, H.. and Zbar, B. Mechanisms of immunologicaleradication of syngeneic guinea pig tumor. Participation of a component(s) ofrecipient origin in the expression of systemic adoptive immunity. Cancer Res.,43:2637-2643,1983.

22. Mazumder, A., Eberlein, T. J., Grimm, E. A., Wilson, D. J., Keenan, A. M.,Aamodt. R., and Rosenberg, S. A. Phase I study of the adoptive immunother-apy of human cancer with lectin activated autologous mononuclearcells.Cancer (Phila.), 53: 896-905,1984.

23. Thoman, M. L., and Weigle, W. 0. Interleukin-2 induction of antigen-nonspecificsuppressor cells. Cell. Immunol., 85: 215-224, 1984.

24. Ting, C. C., Yang, S. S., and Hargrove, M. E. Induction of suppressor T cellsby interleukin-2. J. Immunol., 733: 261-266, 1984.

25. Small, M., and Trainin, N. Separation of populations of lymphoid cells intofractions inhibiting and fractions enhancing syngeneic tumor growth in vivo. J.Immunol., 777: 292-297, 1976.

Fig. 1. Liver section at Day 3 posttumor injection showing clumps of tumor cells growing within the liver sinusoids. H&E, x 860.

Fig. 2. Section of Day 14 liver metastasis showing its relationship to the liver surface and parenchyma. Note the black particles throughout this section which representcarbon particles within Kupffer cells from the i.v. injection of India ink given prior to sacrifice. H&E, x 54.

CANCER RESEARCH VOL. 45 AUGUST 1985

3740

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

REDUCTION OF LIVER METASTASES IN VIVO BY RIL-2 AND LAK PLUS RIL-2

1

Mpl:m .¿»*•»**%'' •''V. i'r «' /

t*. ' t " ».^ ft -'aí- •»'•»«i* \'f

.s

'. \

CANCER RESEARCH VOL. 45 AUGUST 1985

3741

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from

1985;45:3735-3741. Cancer Res   Rene Lafreniere and Steven A. Rosenberg  Recombinant Interleukin 2Metastases with Lymphokine-activated Killer Cells and Successful Immunotherapy of Murine Experimental Hepatic

  Updated version

  http://cancerres.aacrjournals.org/content/45/8/3735

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

  .pubs@aacr.orgDepartment 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/45/8/3735To request permission to re-use all or part of this article, use this link

on March 22, 2020. © 1985 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from