Simplified Method for the Preparation of Human Lymphokine Activated Killer Cells

This invention pertains to adoptive immunotherapy, more particularly to the in vitro generation of human lymphokine activated killer cells for use in such therapy.

Adoptive immunotherapy has recently produced encouraging clinical results against some forms of cancer. See articles in the Wall Street Journal, Apr. 9, 1987, and Time Magazine, Apr. 20, 1987. The therapy involves removing peripheral blood from a patient, removing red blood cells (RBC's) from the blood to produce a lymphocyte-containing white blood cell (WBC) fraction, incubating the blood fraction in culture medium with interleukin-2 (IL-2) to produce activated, tumor-destroying lymphocytes called LAK cells, and injecting the LAK cells and additional IL-2 into the patient. In some cases IL-2 is injected into the patient before removal of the blood in order to stimulate production of lymphocytes.

One objection to adoptive immunotherapy is that it is very expensive. One reason it is expensive is that the current procedure for producing LAK cells is labor-intensive and time consuming. This procedure is described in Muul et al., "Large scale production of human lymphokine activated killer cells for use in adoptive immunotherapy," Journal of Immunological Methods, 88: 265-275 (1986). As described in Muul et al., in order to generate enough LAK cells for a single treatment about 2.times.10" lymphocytes were obtained by 10 successive leukaphereses of peripheral blood. In each leukapheresis, about 10-12 liters of whole blood were processed in an automated cell separator over a 4-hour period to produce a 400-500 ml leucocyte fraction. This fraction was diluted with 2 parts of a salt solution, then poured into 50 ml conical centrifuge tubes (40 ml/tube, approx. 30-40 tubes) and underlayed with 10 ml Ficoll-Hypaque solution. The contents were centrifuged, causing separation into a platelet-rich supernatant layer, a lymphocyte-rich layer, a Ficoll-Hypaque layer, a granulocyte layer and an RBC layer. The supernatant was removed from each tube and discarded. The lymphocyte-rich fraction floating on the Ficoll-Hypaque was removed from each tube; these fractions were pooled and washed three times by suspension in salt solution and centrifugation. Since these steps must be repeated for each leukapheresis, 300-400 tubes must be handled for a single treatment.

Haemonetics Corporation of Braintree, Mass., markets an automated blood cell separator known as the Haemonetics V-50, which utilizes a 2-port conically-shaped centrifuge bowl similar to the bowl described in U.S. Pat. No. 3,145,713. The V-50 can be operated according to a standard leukapheresis protocol or according to a Surge.RTM. lymphocytopheresis protocol. The latter procedure, as described in U.S. Pat. Nos. 4,464,167 and 4,416,654, involves intermittent elutriation with previously-separated plasma, and is capable of providing more precise fractions of platelets, WBC's and RBC's than can be achieved with standard leukapheresis; it is referred to hereinafter elutriation leukapheresis.

For LAK cell processing, Haemonetics recommends use of the V-50 to separate a Buffy coat composed mostly of platelets and WBC's, followed by a secondary separation using Ficoll-Hypaque to provide a density gradient in the same centrifuge bowl for isolation of mononuclear cells (lymphocytes and monocytes) from the Buffy coat. Although this procedure is much less time-consuming and labor-intensive than the standard ficoll centrifugation described in Muul et al., it would be desirable to eliminate the ficoll separation step because it adds to the cost and can cause a reduced yield of lymphocytes. However, up to now it has been considered essential by those skilled in the art to conduct a ficoll separation in order to obtain a lymphocyte fraction sufficiently free of RBC's and granulocytes to be useful for production of LAK cells. It was assumed that RBC's and granulocytes would unduly interfere with the activation of the lymphocytes.

We have discovered that the step of ficoll density gradient centrifugation can be eliminated without unduly interfering with lymphocyte activation. Thus, our invention is an improvement in the method of producing LAK cells in vitro which comprises removing RBC's from whoe blood to produce a lymphocyte-containing WBC-rich fraction and incubating the WBC-rich fraction in culture medium with IL-2 to activate the lymphocytes. The improvement comprises using the lymphocyte-containing WBC-rich fraction without intermediate separation of a lymphocyte and monocyte layer on a ficoll gradient.

In the improved method for this invention, the RBC's can be removed in various ways. These include standard leukapheresis, elutriation leukapheresis, and centrifugation without use of ficoll. Ficoll is a synthetic water-soluble polysaccharide that has a weight average molecular weight of about 400,000 and that is widely used for the preparation of density gradients. It is available as such and in admixture with other substances under registered trademarks such as Ficoll-Paque, Ficoll-Hypaque and Ficoll-Isopaque. Leukapheresis refers to a process in which peripheral blood is withdrawn from a patient or donor, a WBC-rich fraction is separated out, and other blood fractions (plasma, platelets and RBC's) are returned to the source. Standard centrifugation is used to separate blood from donors into plasma, WBC-rich and RBC fractions which are stored for later use. (The term "Buffy coat" as used hereinafter refers specifically to the WBC-rich fraction obtained by standard centrifugation, although the term is also used in the art to refer to a platelet-rich, WBC-rich fraction obtained by leukapheresis.)

The various methods of removing RBC's produce WBC-rich fractions with varying amounts of residual RBC's and varying differentials. (The term "differential" or "diff" refers to the number percent of lymphocytes, monocytes and granulocytes based on the total number of those three cell types in a WBC-rich fraction.) Compositions of the various fractions will also vary depending upon the source. For example, a patient who has been primed with IL-2 may have a very high lymphocyte count. Typical ranges for the WBC-rich fractions obtained by various methods are compared with typical ranges for whole blood in the following table.

From the above table, it can be seen that lymphocyte-containing WBC-rich fractions usable in this invention can have RBC/WBC ratios from about 0.2 to about 300 and granulocyte contents from about 1% to about 50%. As a practical matter Buffy coats would principally be used for screening to determine whether a patient is capable of developing LAK cells. For generating LAK cells for use in adoptive immunotherapy the leukapheresis products having RBC vol. % of about 1-20% and RBC/WBC ratio of about 0.2-250 would be preferred.

At the present time, it is preferred to use the elutriation leukapheresis product because it is more nearly like the ficoll-separated products in both RBC and granulocyte content, and therefore would probably be more readily accepted by workers in the art. In addition, it appears that elutriation leukaphersis products can be cultured at a somewhat higher cell density (e.g., 1.times.10.sup.7 /ml or higher) than can standard leukapheresis products on a routine basis. From the above table, it can be seen that elutriation leukaphersis products typically have a RBC/WBC ratio of about 0.2 to about 50, a RBC vol. percent of about 1-6 and a granulocyte content of 1-5. More typical ranges are RBC/WBC of about 0.5-25 and RBC vol. % of about 2-4.

Standard leukapheresis can be performed using instruments available from various manufacturers, including Haemonetics, Fenwall, and Cobe and following the manufacturers' instructions. The only instrument now available for performing elutriation leukapheresis is the Haemonetics V-50. Following the teaching of U.S. Pat. Nos. 4,464,167 and 4,416,654 or the instructions provided by Haemonetics, the V-50 can be used to provide a WBC-rich fraction having low RBC and granulocyte content.

Monocyte content of the WBC-rich fraction can be reduced below the figures shown in the table by treatment of the leukapheresis product with an L-amino acid lower alkyl ester or hydrogen chloride salt thereof, e.g., methyl, ethyl, propyl, isopropyl, butyl, isobutyl or t-butyl ester of phenylamaine, glutamic acid, glutamine or tyrosine. Phenyl alanine methyl ester is preferred. Further details are given in copending U.S. application Ser. No. 868,697, filed May 30, 1986, and in the examples below.

Activation of the lymphocytes by incubation with IL-2 is accomplished in this invention in the same manner as in the prior art. Containers such as conventional flasks and roller bottles can be used, but the preferred containers are 0.2-5 liter tissue culture bags made from flexible copolymeric film materials as disclosed in copending application Ser. No. 008,273, filed Jan. 29, 1987. Most preferred is a bag made of a copolymer of 97 mol % ethylene and 1-octene. Any suitable culture medium can be used, but the preferred culture medium is RPMI 1640, which is described in "Culture of Animal Cells," Freshney, 72-73, Alan R. Liss, Inc., NY. supplemented with serum. Initial cell concentrations of up to about 1.times.10.sup.7 cells/ml can be used with an elutriation leukapheresis product and up to about 1.times.10.sup.7 cells/ml with a standard leukapheresis product. A concentration of at least 1.times.10.sup.6 cells/ml should be used for reason of economy. Preferred ranges would be 5.times.10.sup.6 to 10.sup.7 cells/ml for elutriation leukapheresis products and 1-5.times.10.sup.6 cells/ml for standard leukapheresis products. The cells are incubated with IL-2 for about 2-7 days, preferably about 3-5 days, at a temperature of about 35.degree.-39.degree. C., preferably 37.degree. C.

"Interleukin-2" (IL-2) as used herein means human IL-2. It includes natural and recombinant IL-2 (rIL-2) and biologically functional equivalents thereof, such as the rIL-2 muteins disclosed in U.S. Pat. No. 4,518,584. Preferably, the IL-2 is a rIL-2 composition consisting essentially of water, rIL-2 and, optionally, a polyol as described in assignee's copending application Ser. No. 825,133, filed on Jan. 31, 1986. Preferably, the IL-2 concentration in the culture medium is in the range of about 5.times.10.sup.2 to about 5.times.10.sup.4 pM, most preferably 1000 to 2000 pM.

The LAK cells prepared by the process of the invention can be suspended in a pharmaceutically acceptable carrier, such as saline, saline containing 5% normal human serum albumin, or Hank's balanced salt solution, to provide a composition which can be infused into a patient afflicted with a tumor. The patient is concurrently treated with rIL-2 as further described by Rosenburg et al., The New England Journal of Medicine 313, 1485-1492 (1985). In that modality, the patient's blood is withdrawn, subjected to leukapheresis and harvested cells are immediately cultured for 3 days to generate LAK cells. The LAK cells are then infused into the patient. Typically, about 3.times.10.sup.10 to 14.times.10.sup.10 LAK cells are infused in 4-9 doses. Interleukin-2 is administered every eight hours at doses such as 10,000, 30,000 or 100,000 units per kilogram of weight. The treatment consists of a two-week regime of leukapheresis and reinfusion and generally repetition starting the third week. Recombinant IL-2 can be included in the LAK cell composition.

In the following examples, unless otherwise stated, a 4 hour .sup.51 Cr release assay was used to measure cytotoxicity of LAK cells for tumor cells (LAK activity). Tumor cells at a concentration of about 2.times.10.sup.6 to 10.times.10.sup.6 per ml were incubated with 100 .mu.Ci of Na.sub.2.sup.51 CrO.sub.4 in 0.4 mL of Tris-phosphate buffered saline for 1 hour at 37.degree. C. The cells were washed 3 times with RPMI 1640 containing 5% or 10% fetal calf serum (FCS) and resuspended to 10.sup.5 cells/mL in RMPI-20% FCS or RPMI-10% FCS. The effector cells (LAK cells) were suspended to various concentrations of 0.1 mL was added in to wells round bottom microliter plates. The .sup.51 Cr labelled target cells (0.1 mL) were added to all wells. After 4 hours of incubation at 37.degree. C., the plates were centrifuged and 0.1 mL of resulting supernatant was removed from each well and counted in a gamma counter. Percent cytolysis is calculated from the following formula: ##EQU1## Each variable was tested in triplicate and the resulting data are expressed as % cytolysis. This cytotoxicity test is further described in "Selected Methods in Cellular Immunology," Mishell and Shiigi, eds., 124-137, W. H. Freeman and Co., San Francisco (1980).

In other experiments, the results of the assays are presented as "Lytic Units" (LU or LU30) which are the number of target cells per 100 effector cells when 30% of the target cells are killed when LAK cells and target cells are incubated together for 4 hours at 37.degree. C. The calculation of LU is based upon the method of Pross et al., Journal of Immunological Methods 68, 235-249 (1984). The greater the number of LU, the greater the potency of the LAK cell preparation.

The disclosure of U.S. Pat. Nos. 4,464,167 and 4,416,654 relating to the production of a WBC-rich fraction by elutriation leukapheresis using previously separated plasma as elutriant are incorporated herein by reference.

(1) To study the LAK activity of cells obtained from a Haemonetics V50 using the elutriation technique to obtain white blood cells.

(2) To study the effects of phenyl alanine methyl ester (.phi.Ala) treatment and Ficoll treatment on the LAK activity of cells obtained from the Haemonetics V50 elutriation technique.

Human lymphocytes (obtained from Haemonetics Corporation using V50 elutriation technique). Raji cells.

(1) Cell culture medium (CCM)=RPMI 1640 with 10% FBS, L-glutamine and Gentamicin

(2) Phosphate buffered saline (PBS) 1.times.without Ca.sup.++ and Mg.sup.++

(3) Ficoll Hypaque (Ficoll)

(4) .phi.Ala

(5) Unopette.RTM. for WBC count

(6) Ethylene butene copolymer Bag for cell culture

(7) T25 tissue culture flasks

(8) 1% NP 40

(9) 2.times.TD buffer

(10) .sup.51 Cr (as sodium chromate)

(11) recombinant Interleukin-2, 10 units/ml in 0.5M glucose (IL-2)

(12) 96 well .mu. bottom tissue culture plate

(13) SCS Harvesting System (Skatron)

(14) Beckman Gamma 4000 Counter

(15) Trypan Blue

(A) Preparation of Cells

(1) A total of 250 ml of a white blood cell fraction was collected froma Haemonetics V50 machine using the elutriation procedure as described in U.S. Pat. Nos. 4,416,654 and 4,464,167.

(2) A WBC count was performed using a Unopette.RTM.. The fraction contained 1.36.times.10.sup.7 WBC/ml and was estimated to contain approximately 3 vol. % RBC.

(3) The cells were then brought to a concentration of 1.times.10.sup.7 WBC/ml (total volume=340 ml).

(4) 40 Ml of cells were put directly into culture (as described below).

(5) The remaining 300 ml were treated with .phi.Ala (as described below).

(B) .phi.Ala Treatment

(1) Place 300 ml of cells into T150 flask.

(2) Add 30 ml of .phi.Ala to cells

(3) Mix well (gently).

(4) Incubate at room temperature for 40 minutes.

(5) After incubation, separate blood into 2 aliquots each containing 165 ml

(a) aliquot 1 was placed into culture

(b) aliquot 2 was underlayed with Ficoll (as described below) and then placed into culture.

(C) Set Up Culture

(1) Cells Straight from V50 (No Ficoll; No .phi.Ala)

(a) place 40 ml cells into a 50 ml centrifuge tube

(b) centrifuge cells 10 minutes at 1200 rpm

(c) discard supernatant

(d) resuspend cells in CCM to a total volume of 40 ml

(e) place desired amount of cells into T25 flasks

(f) add CCM to flasks to bring white cells to desired concentration

(g) add 5 .mu.l IL-2 to each flask (final concentration 10 units/ml)

(h) place flasks in 37.degree. C. incubator with 5% CO.sub.2.

(2) Aliquot 1.fwdarw.Cells from V50(.phi.Ala and No Ficoll)

(a) place 165 ml of .phi.Ala treated cells into a 250 ml centrifuge tube

(b) centrifuge for 10 minutes at 1200 rpm

(c) discard supernatant

(d) resuspend cells in 50 ml CCM

(e) perform cell count using Unopette.RTM.; the WBC count was 1.9.times.10.sup.7 per ml

(f) set up cultures in bags and flasks according to cell concentrations desired

(g) place cultures in 37.degree. C. incubation with 5% CO.sub.2.

T25 Flask: 5.times.10.sup.6 cells/ml

Bag: 9.times.10.sup.6 cells/ml ##EQU2## (3) Aliquot 2.fwdarw.Cells from V50(.phi.Ala and Ficoll) (a) place 40 ml of .phi.Ala treated cells into 4-50 ml centrifuge tubes

(b) underlay blood with 10 ml of Ficoll

(c) centrifuge for 30 minutes at 1900 rpm

(d) collect interface layer with a sterile pasteur pipette and place cells into a sterile 50 ml centrifuge tube

(e) bring volume in the tube up to 50 ml using PBS

(f) centrifuge for 10 minutes at 1200 rpm

(g) discard supernatant

(h) resuspend pellet in 50 ml of CCM

(i) centrifuge for 10 minutes at 1200 rpm

(j) resuspend in 5 ml of CCM

(k) perform cell count using trypan blue; then

(l) set up cultures in bags and flasks according to cell concentrations desired

(m) place cultures in 37.degree. C. incubator with 5% CO.sub.2.

NOTE: No interface layer resulted after step 3; therefore, the cells were resuspended, re-underlayed with Ficoll and recentrifuged. After this, the cells in the interface were collected.

T25 Flask: 5.times.10.sup.6 cells/ml

T25 Flask: 10.times.10.sup.6 cells/ml

Bag: 1.9.times.10.sup.6 cells/ml ##EQU3##

(D) LAK Assay

The LAK assay was performed after cells were in culture for 4 days, according to the procedure given above.

NOTE: Due to the overabundance of red blood cells contained in the specimens, 3 specimens were treated with lysis buffer prior to the LAK assay. Lysing solution: 0.83 g Ammonium Chloride, 200 ml distilled H.sub.2 O

(1) resuspended cell pellet in 10 ml lysing solution

(2) incubate for 20 minutes at room temperature

(3) centrifuge for 10 minutes at 1200 rpm

(4) discard supernatant

(5) resuspende in 1 ml of CCM

(6) perform cell count

(7) set up E:T ratios as described in LAK assay procedure.

The following diagram summarizes the protocol for this example. ##STR11##

(A) Separation of Cells

(1) Cells were collected via elutriation technique on Haemonetics V-50.

(2) A cell count was performed=1.3.times.10.sup.7 cells/ml in 442 mls. 5.8.times.10.sup.9 total cells.

(3) Cell Volume was split in two for processing.

(B) LAB Ficoll

(1) 221 mls of cells were mixed with PBS and layered on Ficoll.

(2) Centrifuged 30 min. at 200 rpm

(3) Cells were then washed and counted

256 cells viable

99% Viability

1 nonviable cell

5.1.times.10.sup.8 cells ml

2.times.10.sup.9 /40 ml

(4) Set up cells in Culture for [email protected]

(5) The remaining of these Ficoll layered cells were set up for .phi.Ala

(a)

(b) Incubate 40 min and then wash.

(c) Perform cell count and put cells in culture for LAK

(d) Set up cells for culture

(c) LAK directly from V-50

(1) The second half of cells were used at this time. The amount of cells necessary to have cultures at a concentration of 1.times.10.sup.6, 5.times.10.sup.6 and 1.times.10.sup.7 were used and the remaining cells were diluted and treated with phenyl alanine methyl ester.

(D) .phi.Ala without Ficoll

.phi.Ala cells without separating with Ficoll first

C.sub.1 V.sub.2 =C.sub.2 V.sub.2

(200 ml) (1.2.times.10.sup.7)=(k) (1.times.10.sup.7)=x=260

#mls media=60 mls

#mls .phi.Ala=29 mls.

Incubate 40 min.

Wash.

Perform cell count and put up in culture

Viable 240

%Viable 93%

Cells/ml=4.8.times.10.sup.7

Total=1.9.times.10.sup.9 /40 ml

All cultures were incubated for 3 days at 37.degree. C. and 5% CO.sub.2. LAK-.sup.51 Cr release performed. In the tables which follow, % REL means % Release, which is the same as % Cytolysis, calculated as explained above. E:T means the ratio of effector (LAK) cells to target (tumor) cells.

Cell counts were performed on all cultures.

A standard leukapheresis product containing 225 mls human leukocytes prepared from 3600 mls whole blood collected in 550 mls anticoagulant ACD-B was obtained from Biological Specialty Corporation, Lansdale, PA. The following procedures were performed using this product.

(1) Set up a Unopette.RTM. (WBC) and a differential.

Differential:

90% Lymphocytes

6% Monocytes

4% Granulocytes

Direct=165 cells

3.3.times.10.sup.7 cells/ml

7.4.times.10.sup.9 cells/225 mls

(2) Set up cells in culture for LAK

@1.times.10.sup.6 =0.30 ml cells+9.70 ml media+5 .mu.l IL-2

@5.times.10.sup.6 =1.52 ml cells+8.48 ml media+5 .mu.l IL-2

@1.times.10.sup.7 =3.04 ml cells+6.96 ml media+5 .mu.l IL-2

Incubated @37.degree. C., 5% CO.sub.2 for 4 days.

(3) Next 20 mls of cells were removed from the remaining cells and mixed with 20 mls of PBS without Ca.sup.++ and Mg.sup.++. 40 mls of this mixture was layered onto 40 mls of Ficoll and centrifuged for 1/2 hours. Removed mononuclear cell layer and washed these cells 3 times. Performed 90 min. monocyte adherence. Washed 2 more times and performed cell count and put in culture for LAK. This is the Standard Sample.

Standard Cell Count:

Viable=155

Non-viable=1

% Viable=99%

Cells/ml=3.1.times.10.sup.7 /ml

Total=7.7.times.10.sup.8 /25 ml.

Dilution for cell conc. of 1.times.10.sup.6 =0.32 ml cells+9.68 ml media+5 .mu.l (IL-2).

(4) The remaining cells (200 mls) were then diluted with 460 mls CCM to bring the cell count to 1.times.10.sup.7 /ml, and treated with 72 mls .phi.Ala.

Incubated at R.T. for 40 min.

Cells clotted during incubation.

Removed as much unclotted suspension as possible--washed and counted cells with WBC Unopette.RTM..

26 cells viable

1 cell nonviable

96% viability

5.2.times.10.sup.6 cells/ml

1.04.times.10.sup.9 cells/200 mls.

Put cells up in a bag at 5.times.10.sup.6 =48 mls cells+2 ml media+25 .mu.l IL-2.

Incubated at 37.degree. C. 5% CO.sub.2 for 4 days for LAK.

(5) Cell Counts After Incubation

E:T ratio 20:1, 10:1, 5:1, 2.5:1

A standard leukapheresis product containing 230 mls human leukocytes prepared from 3600 ml whole blood collected in 520 mls anticoagulant ACD-B was obtained from Biological Specialty Corporation, Lansdale, PA. The following procedures were performed using this product.

(1) Removed 10 mls of cells

(A)

(1) Took 5 mls of this blood and mixed with 5 mls of PBS

(2) Underlayered 10 mls of Ficoll

(3) Centrifuged for 30 min @2000 rpm

(4) Washed and counted

(5) This was the Standard @1.5.times.10.sup.6 cells/ml in 10 ml flask

Standard

Viable=49

Non-viable=0

% Viable=100%

Cells/ml=9.8.times.10.sup.6 /ml

Total=1.96.times.10.sup.8 /20 ml.

Dilution: 1.5 ml cells+8.5 mls media+1 .mu.l IL-2

(b)

(1) The second 5 mls was used for direct testing

(2) A WBC (via Unopette.RTM.) and differential were performed:

(3) WBC 4.5.times.10.sup.7 /ml, 2.25.times.10.sup.8 /5 ml

Diff.: 72% Lymphocytes, 22% Granulocytes, 6% Monocytes

(4) Cells were then put up in culture at

(2) LAK Assay

E:T ratio 40:1, 20:1, 10:1, 5:1, 2.5:1 1.25:1

Cells were diluted to 4.times.10.sup.6

Raji's were diluted to 1.times.10.sup.5

Buffy coat--52 mls of blood

Cell count

4.3.times.10.sup.7 cells/ml (total WBC)

1.8.times.10.sup.7 neutrophils/ml (42%)

(est.) 1-2.times.10.sup.7 lymphocytes/ml (20-50%)

(est.) 5.times.10.sup.9 RBC/ml

(est.) 0.5-1.times.10.sup.7 monocytes/ml

Ficoll-Hypaque (Ficoll)

CCM--5% FCS--RPMI

(1) No Ficoll

(a) To 10.5 ml of Buffy coat add 215 ml CCM. Cell Count 2.times.10.sup.6 cells/ml (total WBC)

(b) Add 10 .mu.l/ml of IL-2

(c) Place 112 ml of culture mix in flask

(d) Place 112 ml of culture mix in bag

(e) Incubate at 37.degree. C. for 20 days.

(f) Sample at 3, 6, 12, 17 and 20 days for cell count and .sup.51 Cr Release (LAK) assay.

(2) Ficoll

(a) Put 42 mls Buffy Coat in 50 ml centrifuge tube

(b) Centrifuge at 800 g for 10 minutes.

(c) Discard supernatant, recover mononuclear WBC layer (Lymphocytes and monocytes) floating on Ficoll layer, wash 3.times.. 300.times.10.sup.6 total mononuclear cells isolated.

(d) Add CCM to provide mononuclear cell concentration of 2.times.10.sup.6 /ml

(e) Place 75 ml in flask

(f) Place 75 ml in bag

(g) Incubate at 37.degree. C. for 20 days

(h) Sample at 3, 6, 12, 17 and 20 days for cell count and .sup.51 Cr release (LAK) assay.