- Email: [email protected]
Systemic Distribution and Tumor Localization of Adoptively Transferred Lymphocytes in Mice: Comparison with Physiologically Based Pharmacokinetic Model
BRIEF ARTICLE
Neoplasia . Vol. 4, No. 1, 2002, pp. 3 – 8
3
www.nature.com/neo
Systemic Distribution and Tumor Localization of Adoptively Transferred Lymphocytes in Mice: Comparison with Physiologically Based Pharmacokinetic Model1 Robert J. Melder *,2, Lance L. Munn *, Brian R. Stoll *, Edgardo M. Marecos y, Laurence T. Baxter *,3, Ralph Weissleder y and Rakesh K. Jain * *Department of Radiation Oncology, yCenter for Molecular Imaging Research, Massachusetts General Hospital, Boston, MA, USA Abstract The mechanisms by which tumors are able to evade cellular immune responses are still largely unknown. It is likely, however, that the initial recruitment of lymphocytes to tumor vessels is limited by cell retention in normal tissue, which results in a low flux of these cells into the tumor vasculature. We grew MCaIV ( mouse mammary carcinoma ) tumors in the leg of SCID mice and injected 111In - oxine – labeled, primed T lymphocytes directed against the tumor intravenously. The systemic distribution of cells in normal organs was similar between mice injected with primed and control lymphocyte populations, except for a delayed clearance of primed lymphocytes from the lungs. Kinetics of lymphocyte localization to the tumor were identical between the primed and control lymphocyte populations. Splenectomy before the injection of primed lymphocytes increased delivery of cells to the lungs and liver after 1 hour with no significant improvement in tumor localization. Within 24 to 168 hours after injection, localization of cells in the liver of splenectomized mice was higher than in the control group. However, no significant difference in tumor localization was observed between groups. A physiologically based compartmental model of lymphocyte distribution predicted the compartmental sequestration and identified model parameters critical for experimental planning and therapeutic optimization. Neoplasia ( 2002 ) 4, 3 – 8 DOI: 10.1038/sj/neo/7900209 Keywords: lymphocyte, trafficking, pharmacokinetic, mathematical model, tumor.
Introduction Significant progress has been made in understanding the nature of the immune response to tumors and the role of tumor antigens in this process [ 1 – 3,4 ]. In some cases lineage - specific differentiation antigens may be expressed in the tumor cell population, providing potential targets for activated lymphocytes [ 5,6 ]. Recognition of a tumor associated antigen is likely to depend on recognition of the dominant epitopes as observed in both human [ 7 ] and
murine [ 8 ] systems. However, the importance of lymphocyte specificity in the initial recruitment of antitumor effector cells remains unclear. Does lymphocyte localization in a tumor require specific recognition of tumor antigen in the vascular compartment or is initial recruitment dependent on the characteristics of the tumor - associated vasculature? Tumor infiltration by lymphocytes, monocytes, and eosinophils is common [ 9 ], especially at the tumor periphery [ 10,11 ]. Although the presence of tumor - specific lymphocytes within tumor tissue suggests specific recruitment, it may also reflect a more generalized process of recruitment involving cells of multiple specificities and lineages. The ability of the leukocyte population to identify tumor associated antigens and kill the cells that express them is not sufficient to effect a complete antitumor response in most patients. One possible reason for this may be insufficient homing of cells with required specificity to the tumor tissue. Although tumor localization of cultured tumor - infiltrating lymphocytes has been demonstrated [ 12 ], it appears that in some cases the number of delivered cells was inadequate to produce a therapeutic response. Adoptive immunotherapy with lymphokine - activated and expanded effector cells, such as interleukin - 2 ( IL - 2 ) – activated tumor - infiltrating lymphocytes, has had limited success in advanced patients with cancer [ 1,13 – 15 ], yet the mechanism of delivery remains unclear. In vitro findings and limited in vivo data suggest that the tumor - specific cytolytic activity of the effector cells is preceded by the delivery of effector cells to the target tissues [ 16 – 18 ]. If the ability of lymphocytes to eradicate an established tumor is determined by the tumor burden, growth pattern, and the intensity of the immune response at the site of tumor [ 16,17,19 – 22 ], the antitumor response may be insignificant when the delivery of a specific cytotoxic cell population to a developing tumor is limited. Address all correspondence to: Dr. Rakesh K. Jain, Department of Radiation Oncology, Massachusetts General Hospital, 100 Blossom Street, Cox - 7, Boston, MA 02114, USA. E - mail: [email protected] 1 This work was supported by National Institutes of Health grants PO1 CA80124 ( R.K.J., L.L.M. ), and R01 HL64240 ( L.L.M. ) and RO1 CA86782 ( R.W. ). 2 Current address: Preclinical Development, Human Genome Science, Inc., 9410 Key West Avenue, Rockville, MD 20850, USA. 3 Current address: Advanced Process Combinatorics, West Lafayette, IN, USA. Received 22 August 2001; Accepted 2 October 2001. Copyright # 2002 Nature Publishing Group All rights reserved 1522-8002/02/$25.00
4
Localization of Lymphocytes
Melder et al.
Delivery of cells to a tumor may be influenced by at least three different factors: cellular adhesion molecule ( CAM ) expression within the tumor vasculature, the flux of circulating cells entering the tumor microcirculation, and the ability of the leukocyte population to engage the appropriate CAMs to arrest and extravasate into the tissue [ 23 ]. The present study uses experimental and mathematical models to quantify trafficking of injected effector cells to identify parameters that limit localization in the tumor.
Materials and Methods Activated Tumor - Specific Lymphocytes Mouse antitumor lymphocytes were prepared from immunization of Balb / c mice ( H - 2 d, Mls b ) with lymphocytes from C3H / sed mice ( H - 2 k, Mls c ). Briefly, C3H mice were sacrificed and their spleens were removed and placed in culture medium RPMI 1640 ( Fisher Scientific, Pittsburgh, PA ) with 10% FCS ( Gibco, Grand Island, NY ). Single - cell suspensions were prepared by disruption of the spleens followed by lysis of the erythrocytes in distilled water. The resulting cells were resuspended in saline and exposed to three freeze – thaw cycles. The resulting material was washed once in PBS, mixed with Freunds incomplete adjuvant and refrozen at 708C. Aliquots were removed and injected into the Balb / c mice at 1 - week intervals for 3 weeks. The resulting sensitized lymphocyte populations were obtained by removal of spleens from the donor populations. Single cell suspensions were prepared as before. The mononuclear population was washed once in culture medium and depleted of the monocyte population through column adhesion for 1 hour at 378C. The cells were then eluted with three volumes of warm medium, washed once and examined. Giemsa staining of the cytofuged cells revealed that 98% of the cells in the final cell suspension were small lymphocytes. Tumor Model The syngeneic carcinoma, MCaIV, was grown in female C3H mice. The MCaIV cell line was maintained by in vivo passage from frozen stock and only low passage tumors ( less than five passages ) were used. Cells were collected from harvested tumors, washed once in complete medium, and resuspended in medium without serum to a concentration of approximately 108 / ml. A volume of 10 to 20 l was injected subcutaneously in the leg of an SCID ( H - 2 d, Mls b ) mouse. Within 8 days, this procedure resulted in tumors of 0.7 - cm diameter. The leg tumor model was selected for this study because ( a ) the tumor is distant from the internal organs, simplifying interpretation of the results and ( b ) it provides convenient access for monitoring growth. In one group of mice, splenectomy was performed ( after anesthesia with ketamine, 9 mg / 100 g, and xylazine, 0.9 mg / 100 g i.m. ) under aseptic conditions. Tumor implantation was performed 3 to 5 days following surgery.
Cell Labeling 111 In oxine was incubated with the sensitized or nonsensitized lymphocytes for 1 hour at 378C. The cells were then washed with PBS three times and resuspended to a concentration of 107 cells / 150 l. Total activity of the cells was checked in a gamma - counter before use ( approximately 100 Ci / 107 cells ). Because of variable release rates of In from labeled cells, we used a previously optimized labeling procedure in which In oxine was synthesized and purified de novo before each labeling procedure; optimum ratios of cell / isotope and incubation times were used, cells were repeatedly washed before injection ( to remove not transchelated oxine ), and nonviable cells were excluded from the injectate [ 24,25 ]. Cell Injection and Imaging Procedure Ten minutes before injection of the cell suspension, the mice were given a mixture of ketamine ( 9 mg / 100 g ) and xylazine ( 0.9 mg / 100 g ) intramuscularly. After anesthesia, a heparinized ( 100 IU / ml ) 30 - gauge cannula was placed into the lateral tail vein. A 1 - ml syringe containing 107 labeled cells in 0.15 ml PBS was connected to the intravenous line and the cells were injected over approximately 1 minute. Imaging of cell localization was initiated 1 hour after injection and continued at 24 - hour intervals. Imaging was performed using a gamma camera ( Ohio Nuclear, Solon, OH ) with a pinhole collimator and a 37 PMT array. Images were assembled and quantified on a Macintosh computer using the NucLear Power 2.9 software. Regions of interest ( ROIs ) were determined at the time of the initial imaging procedure and remained constant throughout the subsequent time points. Counts obtained from the ROIs were adjusted for isotope decay at each time point. Counts obtained in the liver and spleen were adjusted to compensate for proximity by removing the organs and comparing relative counts in the liver and spleen using a gamma counter. The counts obtained by planer scintigraphy were adjusted by the relative distribution of total counts in these organs. In a separate subgroup of mice, imaging findings were validated by performing comparative biodistribution studies at multiple endpoints. Confirmation of Sensitization Specificity of the lymphocyte preparation was tested by proliferation assay ( Promega, Madison, WI ). Both tumor antigen and sensitizing antigen were used to elicit a proliferative response. Concanavalin A and medium were used as positive and negative controls, respectively. Tumor antigen was obtained from sterile homogenized MCaIV tumor ( F3 passage from frozen stock ), which was immediately frozen. Tumor homogenate was diluted 1:100 in fresh culture medium before use. Immunohistology Tumor tissue was removed from mice that were not used for imaging studies. Normal skin was obtained from non – tumor - bearing mice for comparison. Tissue samples were placed in plastic tissue trays and embedded in OCT medium,
Neoplasia . Vol. 4, No. 1, 2002
Localization of Lymphocytes
then quick - frozen in isopropanol and dry ice. Frozen sections were cut at 7 - m thickness and fixed in cold acetone ( 48C ) for 5 minutes. The sections were washed once with PBS, pH 7.2, with 1% BSA ( Sigma, St. Louis, MO ). The slides were then stained in a humidified chamber at 378C for 30 minutes using 5 l of anti – ICAM - 1 - biotin ( 01542D, Pharmingen, San Diego, CA ), anti – ICAM - 2 - biotin ( 01802D, Pharmingen ), anti – VCAM - 1 - biotin ( 01812D, Pharmingen ), anti – CD31 - biotin ( 01952D, Pharmingen ) or PBS alone. Following incubation, the slides were washed three times with PBS with 1% BSA, then incubated for 30 minutes with 10 l of avidin - alkaline phosphatase ( 10 4 dilution, A - 7294, Sigma ) at 378C. The slides were washed three times with PBS / 1% BSA, then incubated with 0.1 ml of Fast Red TR / naphthol AS - MX reagent in Tris buffer ( F 4523, Sigma ). The slides were washed three times with distilled water and counterstained lightly with hematoxylin ( HHS - 16, Sigma ). The slides were then washed once in distilled water and the coverslips were mounted with glycerol. Mathematical Model The basis of the theoretical model used in these studies has been described previously [ 26 ]. Briefly, the model simulates the following key processes involved in lymphocyte trafficking following adoptive transfer: ( 1 ) transport through the systemic circulation, ( 2 ) initial reversible capture ( temporary adhesion or entrapment ) at the endothelial wall, ( 3 ) arrest ( stable adhesion ) at the endothelial wall following capture, ( 4 ) transmigration of arrested lymphocytes across the endothelial wall, ( 5 ) recirculation through the lymphatic system following transmigration in organ compartments, ( 6 ) limited recirculation in tumor compartment, ( 7 ) transient accumulation of recirculating lymphocytes in the lumped lymph node before return to the systemic circulation, and ( 8 ) possible depletion ( for example, mechanical disruption or apoptosis ) in tissues. The model assumed that free labels were excreted through urine at a much faster rate than their spontaneous release from cells, rendering their influence on the observed biodistribution negligible. Lymphocyte proliferation during the distribution phase was also neglected. The major lymphoid tissues included in the model were bone marrow, spleen, and lymph node. In addition to the lymphoid tissues, other peripheral organs and tissues such as blood, heart, lung, liver, kidney, GI tract, skin, muscle, and tumor were incorporated in the model. These organs and tissues were arranged according to their anatomic relationship. The simulations were performed as described previously varying certain parameter values as appropriate to simulate the difference between normal and sensitized cell populations.
Melder et al.
5
injected cell number within 24 hours. Redistribution of these cells at 24 hours post injection resulted in a concurrent increase in liver and spleen localization, which remained stable throughout the observation period. Lymphocyte localization within the tumor ranged from 1% to 2.5% ( 4 to 7 days post injection ). Sensitized and normal lymphocyte populations were tested for proliferation in response to stimulation by sensitizing antigen, tumor homogenate or Concanavalin A ( Figure 1 ). Lymphocyte proliferation in response to both tumor and sensitizing antigen was equal to the mitogenic response to Concanavalin A, indicating successful priming. Scintigraphic studies of cell biodistribution revealed that tumor - specific and nonspecific lymphocytes had approximately equal levels of delivery to the tumor during the first week following administration ( Figure 2; A and B ). Although this finding suggests a diminished importance of antigen specificity for tumor rejection [ 1 ], the number of lymphocytes in the injected population that were specific for the foreign antigen may have been relatively low and below the ability of our system to distinguish them from the labeled nonspecific lymphocytes. However, using similar methods, Kjaergaard and Shu [ 27 ] were also able to show that lymphocyte specificity is not a major determinant of localization. These results further suggest that the initial phase of cell recruitment to tumors is nonspecific. Our previous studies of NK cell localization in tumors also indicated that some tumor vessels support adhesive interactions with injected effector cells [ 28 ]. However, a second stage of recruitment, mediated by the infiltrating T cell population may be required to initiate tumor rejection. Unfortunately, the isotope half - life ( 2.8 days ) restricted the duration of these studies to a week, preventing observation of the longer - term kinetics. Because both the liver and spleen act as principal retention sites for injected lymphocyte populations, we hypothesized that removal of one of these compartments would alter the distribution of the injected populations.
Results and Discussion Systemic and Tumor Distributions The general pattern of lymphocyte distribution following injection was similar for all mice used in these studies. Lymphocyte localization in the lungs peaked immediately after injection and dropped to between 1% and 10% of the
Neoplasia . Vol. 4, No. 1, 2002
Figure 1. Sensitization of H - 2 d mice ( Balb / c ) with H - 2 k ( C3H ) antigen. Cell proliferation assays indicate that sensitized T lymphocytes from H - 2 d mice proliferated in response to H - 2 k antigen expressed on tumor cells or spleen cells to levels comparable with a nonspecific mitogen ( maximal stimulation ). Approximately the same level of cell proliferation ( as indicated by the staining intensity of the assay ) could be observed between cells stimulated with tumor cells, spleen cells, and Concanavalin A mitogen. Means ± SD of five wells are shown.
6
Localization of Lymphocytes
Melder et al.
Figure 2. Experimental and theoretical biodistribution profiles of sensitized ( A ) and normal ( B ) lymphocyte populations in tumor - bearing mice and in tumor - bearing mice with splenectomy ( C ). Overall patterns of systemic distribution and accumulation of lymphocytes within the tumors were similar for both populations of lymphocytes, suggesting that the initial phase ( < 7 days ) of cell recruitment to tumors may be nonspecific. Splenectomy of mice before the implantation of the tumor results in tumor accumulation kinetics comparable to the nonsensitized lymphocyte populations in the nonsplenectomized mice. Retention of transferred lymphocytes within the liver is greater than in nonsplenectomized mice, suggesting a shift in compartmental distribution following the removal of a major filtration compartment, but only a slight increase in tumor localization can be observed. Means ± SD are shown, n = 6 mice / group. The simulations of the mathematical model are shown overlying the experimental data. The predictions agree well with the experiments and capture the salient features of the experimental biodistribution profiles.
Removal of the spleen ( Figure 2C ) resulted in increased uptake within the liver as early as 1 hour after cell injection. Increased retention by the liver was approximately equal to the expected uptake by the missing spleen. These findings suggest that modification of compartmental localization of injected cells can alter distribution but does not provide a dramatic improvement in tumor delivery.
Comparison with Pharmacokinetic Model The mathematical model was used to simulate lymphocyte distribution in mice, both with and without splenectomy. All administered cell populations were assumed to have an equal probability of tumor localization and no bias was given to retention of lymphocyte subpopulations within various microvascular or lymphatic compartments. The model predicted the initial lymphocyte uptake in lungs and the subsequent redistribution to the liver and spleen ( Figure 2; A and B ). Lymphocyte retention rates were slightly overestimated for the liver and spleen, resulting in higher values than observed. However, the model predicted well the overall trends in the biodistribution profiles with time and was able to capture the increased liver uptake following splenectomy. The agreement between model predictions and experimental data ( Figure 2; A C ) supports the validity of the assumption that different lymphocyte populations have an equal likelihood of recruitment by the tumor vasculature. The expression of constitutive and inducible CAMs on normal endothelium in vitro and in vivo is tissue / organ dependent [ 29 ]. Several CAM expression studies in tumor vessels have been conducted ( for a review, see Ref. [ 30 ] ) and these studies demonstrate that CAM expression during tumor angiogenesis is highly heterogeneous. Variation in CAM expression can be observed within different tumor types, and within different regions of the same tumor. Human mammary cancer endothelium, for example, shows increased expression of various adhesion molecules ( e.g., E - selectin, VCAM - 1, and ICAM - 3 ) [ 31 ], whereas in primary melanomas and various other tumors, VCAM - 1 expression is focal or downregulated [ 32 ]. These findings help to explain how the heterogeneous infiltration of leukocytes into tumors can occur in terms of variation in adhesion of cells to tumor vessels. Although other mechanisms contribute to the migration and activity of infiltrating leukocytes within tumors, the accumulation of cells within the tumor vessels is an initial limiting factor on the system. Immunohistology performed on sections from the MCaIV tumor demonstrated moderate staining for ICAM - 1, ICAM 2, and CD31 but strong staining of VCAM - 1 throughout the tumor tissue, indicating the potential for 1 integrin – mediated adhesion in these vessels ( Table 1 ). We have shown [ 28,33 ] preferential adhesion of IL - 2 – activated natural killer ( A - NK ) cells in some vessels within human and murine ( MCaIV ) tumors. This suggests that the relative
Table 1. Expression of Cellular Adhesion Molecules in the MCaIV Tumor and Normal Skin
Antibody
Tumor tissue*
Normal tissue
Isotype control ICAM - 1 ICAM - 2 VCAM - 1 CD31
+ + ++++ ++
+ + ++
*Results indicate expression observed on frozen sections of tissue and stained by immunohistology.
Neoplasia . Vol. 4, No. 1, 2002
Localization of Lymphocytes
degree of lymphocyte interaction with tumor and normal vessels depends on the regional variation in adhesion molecule expression within the tumor. Further insight into lymphocyte binding to tumor vessels has come from PET studies of the localization of IL - 2 A - NK cells in normal and tumor tissues in mice [ 34 ]. These findings are in agreement with the current observation that lymphocyte localization to tumor vessels is adhesion and delivery dependent. That these previous studies were performed with A - NK cell populations precludes antigen - specific recognition mechanisms for localization and supports a nonspecific adhesion mechanism for the accretion of lymphocytes within a tumor. Populations of NK cells may have high affinity to regions of tumor vessels [ 28,33 ], but may be limited to a single pass through ( or around ) the tumor vasculature due to entrapment in other organs [ 35 ]. In contrast, T lymphocytes may have lower adhesion efficiencies to the tumor vessels but are not limited to single - pass delivery. It is reasonable that T cells may localize within tumor vessels through a similar adhesive mechanism, using repeated recirculation of the tumor - specific cells to facilitate long - term accumulation of lymphocytes within tumors. The parallels between the kinetics of lymphocyte localization and the model simulations ( Figure 2; A C ), in which all injected cell populations were assumed to be recruited to the tumor vasculature with equal probability, supports the hypothesis of nonspecific recruitment of lymphocytes to a tumor. Modulation of Biodistribution We have previously shown that modulation of lymphocyte - endothelial adhesion accompanies angiogenesis [ 36 ]. The extent to which this will influence the accumulation of lymphocytes within the tumor depends on the number of cells in the tumor microcirculation available for recruitment. The data from this study and the mathematical simulations of lymphocyte biodistribution ( Figure 2 ) indicate that high retention of lymphocytes within the liver and spleen restricts the availability of a large fraction of the injected cells for recruitment to the tumor. Although lymphocytes within these organs may still be available for recirculation, the rate at which this occurs appears to be low, because the number of cells within the liver is constant after 72 hours. Removal of the spleen does not increase the number of cells available for tumor recruitment because these are acquired and retained by the liver, as observed experimentally and predicted by the theoretical model. Thus, our work suggests that improved lymphocyte delivery to tumors will require increased systemic availability of circulating lymphocytes at steady state to facilitate adhesion and infiltration of tumor - specific T - cell populations.
References [1] Rosenberg SA ( 2000 ). Principles and Practice of the Biological Therapy of Cancer. Lippincott Williams and Wilkins, Philadelphia. [2] Rosenberg SA (2001). Progress in human tumor immunology and immunotherapy. Nature 411, 380 – 384. [3] Fabre JW (2001). The allogeneic response and tumor immunity. Nat Med 7, 649 – 52.
Neoplasia . Vol. 4, No. 1, 2002
Melder et al.
7
[4] Nestle FO, Banchereau J, and Hart D (2001). Dendritic cells: on the move from bench to bedside. Nat Med 7, 761 – 765. [5] Disis M, Gralow J, Bernhard H, Hand S, Rubin W, and Cheever M (1996). Peptide based, but not whole protein, vaccines elicit immunity to HER - 2 / neu, oncogenic self - protein. J Immunol 156, 3151 – 3158. [6] Roth J, Dittmer D, Rea D, Tartaglia J, Paoletti E, and Levine A (1996). p53 as a target for cancer vaccines: recombinant canarypox virus vectors expressing p53 protect mice against lethal tumor cell challenge. Proc Natl Acad Sci USA 93, 4781 – 4786. [7] Correale P, Walmsley K, Nieroda C, Zaremba S, Zhu M, Schlom J, and Tsang K (1997). In vitro generation of human cytotoxic T lymphocytes specific for peptides derived from human prostate - specific antigen. J Natl Cancer Inst 89, 293 – 300. [8] Matsumura T, Sussman JJ, Krinock RA, Chang AE, and Shu S (1994). Characteristics and in vivo homing of long - term T - cell lines and clones derived from tumor - draining lymph nodes. Cancer Res 54, 2744 – 2750. [9] Boheim K, Denz H, Boheim C, Glassl H, and Huber H (1987). An immunohistologic study of the distribution and status of activation of head and neck tumor infiltrating leukocytes. Arch Oto - Rhino - Laryngol 244, 127 – 32. [10] Dvorak HF, and Dvorak A (1982). Immunohistological characterization of inflammatory cells that infiltrate tumors. In Tumor Immunity in Prognosis: The Role of Mononuclear Cell Infiltration S Haskill ( Ed ). Marcel Dekker, New York. pp. 279 – 307. [11] Jain RK (1996). Delivery of molecular medicine to solid tumors. Science 271, 1079 – 80. [12] Rosenberg SA (1990). Adoptive immunotherapy for cancer. Sci Am 262, 62 – 69. [13] Hwu P, and Rosenberg SA (1994). The use of gene - modified tumor - filtrating lymphocytes for cancer therapy. Ann NY Acad Sci 716, 188 – 97. [14] Rosenberg SA (1994). The gene therapy of cancer. Prev Med 23, 624 – 626. [15] Rosenberg SA (1992). The immunotherapy and gene therapy of cancer. J Clin Oncol 10, 180 – 99. [16] Basse P, and Goldfarb RH (1994). Localization of immune effector cells to tumor metastases. In Tumor Immunology and Cancer Therapy R Goldfarb and T Whiteside ( Eds ). Marcel Dekker, New York. [17] Whiteside TL, and Goldfarb RH (1994). Antitumor effector cells: extravasation and control of metastasis. In Tumor Immunology and Cancer Therapy R Goldfarb and T Whiteside ( Eds ). Marcel Dekker, New York. [18] Topalian SL, and Rosenberg SA (1991). Adoptive cellular therapy: basic principles. In Biological Therapy of Cancer V DeVita, Jr, S Hellman and S Rosenberg ( Eds ). JB Lippincott, Philadelphia. [19] Greenberg PD (1991). Adoptive T cell therapy of tumors: mechanisms operative in the recognition and elimination of tumor cells. Adv Immunol 49, 281 – 355. [20] De Boer RJ, Hogeweg P, Dullen HF, De Weger RA, and Den Otter W (1985). Macrophage T lymphocyte interaction in the anti - tumor immune response: a mathematical model. J Immunol 134, 2748 – 58. [21] De Boer RJ, and Hogeweg P (1986). Interaction between macrophages and T lymphocytes: tumor sneaking through intrinsic to helper T cell dynamics. J Theor Biol 120, 331 – 51. [22] Kuznetsov VA, Makalkin IA, Taylor MA, and Perelson AS (1994). Nonlinear dynamics of immunogenic tumors: parameter estimation and global bifurcation analysis. Bull Math Biol 56, 295 – 321. [23] Jain RK (1998). The next frontier of molecular medicine: delivery of therapeutics. Nat Med 4, 655 – 657. [24] Schoepf U, Marecos E, Melder R, Jain R, and Weissleder R (1998). Intracellular magnetic labeling of lymphocytes for in vivo trafficking studies. BioTechniques 24, 642 – 646. [25] Lewin M, Carlesso N, Tung C, Tang X, Cory D, Scadden D, and Weissleder R (2000). Tat peptide derivatized magnetic nanoparticles allow in vivo tracking and recovery of progenitor cells. Nat Biotechnol 18. [26] Zhu H, Melder RJ, Baxter LT, and Jain RK (1996). Physiologically based kinetic model of effector cell biodistribution in mammals: implications for adoptive immunotherapy. Cancer Res 56, 3771 – 81. [27] Kjaergaard J, and Shu S (1999). Tumor infiltration by adoptively transferred T cells is independent of immunologic specificity but requires down - regulation of L - selectin expression. J Immunol 163, 751 – 759. [28] Sasaki A, Melder RJ, Whiteside TL, Herberman RB, and Jain RK (1991). Preferential localization of human adherent lymphokine activated killer cells in tumor microcirculation. J Natl Cancer Inst 83, 433 – 437.
8
Localization of Lymphocytes
Melder et al.
[29] Kuzu I, Bicknell R, Fletcher CD, and Gatter KC (1993). Expression of adhesion molecules on the endothelium of normal tissue vessels and vascular tumors. Lab Invest 69, 322 – 328. [30] Jain RK, Koenig GC, Dellian M, Fukumura D, Munn LL, and Melder RJ (1996). Leukocyte - endothelial adhesion and angiogenesis in tumors. Cancer Metastasis Rev 15, 195 – 204. [31] Fox SB, Turner GDH, Gatter KC, and Harris AL (1995). The increased expression of adhesion molecules ICAM - 3, E - and P - selectins on breast cancer endothelium. J Pathol 177, 369 – 376. [32] Piali L, Fichtel A, Terpe H, Imhof BA, and Gisler RH (1995). Endothelial vascular cell adhesion molecule 1 expression is suppressed by melanoma and carcinoma. J Exp Med 181, 811 – 816. [33] Melder RJ, Elmaleh D, Brownell AL, Brownell GL, and Jain RK (1994).
A method for labeling cells for positron emission tomography ( PET ) studies. J Immunol Methods 175, 79 – 87. [34] Melder RJ, Brownell AL, Shoup TM, Brownell GL, and Jain RK (1993). Imaging of activated natural killer cells in mice by positron emission tomography: preferential uptake in tumors. Cancer Res 53, 5867 – 5871. [35] Melder RJ, Kristensen CA, Munn LL, and Jain, RK. Modulation of A - NK cell rigidity: in vitro characterization and in vivo implications for cell delivery. Biorheology 38, 151 – 159. [36] Melder RJ, Koenig GC, Witwer B, Safabakhsh N, Munn LL, and Jain RK (1996). During angiogenesis, vascular endothelial growth factor and basic fibroblast growth factor regulate natural killer cell adhesion to tumor endothelium. Nat Med 2, 992 – 997.
Neoplasia . Vol. 4, No. 1, 2002
Neoplasia . Vol. 4, No. 1, 2002, pp. 3 – 8
3
www.nature.com/neo
Systemic Distribution and Tumor Localization of Adoptively Transferred Lymphocytes in Mice: Comparison with Physiologically Based Pharmacokinetic Model1 Robert J. Melder *,2, Lance L. Munn *, Brian R. Stoll *, Edgardo M. Marecos y, Laurence T. Baxter *,3, Ralph Weissleder y and Rakesh K. Jain * *Department of Radiation Oncology, yCenter for Molecular Imaging Research, Massachusetts General Hospital, Boston, MA, USA Abstract The mechanisms by which tumors are able to evade cellular immune responses are still largely unknown. It is likely, however, that the initial recruitment of lymphocytes to tumor vessels is limited by cell retention in normal tissue, which results in a low flux of these cells into the tumor vasculature. We grew MCaIV ( mouse mammary carcinoma ) tumors in the leg of SCID mice and injected 111In - oxine – labeled, primed T lymphocytes directed against the tumor intravenously. The systemic distribution of cells in normal organs was similar between mice injected with primed and control lymphocyte populations, except for a delayed clearance of primed lymphocytes from the lungs. Kinetics of lymphocyte localization to the tumor were identical between the primed and control lymphocyte populations. Splenectomy before the injection of primed lymphocytes increased delivery of cells to the lungs and liver after 1 hour with no significant improvement in tumor localization. Within 24 to 168 hours after injection, localization of cells in the liver of splenectomized mice was higher than in the control group. However, no significant difference in tumor localization was observed between groups. A physiologically based compartmental model of lymphocyte distribution predicted the compartmental sequestration and identified model parameters critical for experimental planning and therapeutic optimization. Neoplasia ( 2002 ) 4, 3 – 8 DOI: 10.1038/sj/neo/7900209 Keywords: lymphocyte, trafficking, pharmacokinetic, mathematical model, tumor.
Introduction Significant progress has been made in understanding the nature of the immune response to tumors and the role of tumor antigens in this process [ 1 – 3,4 ]. In some cases lineage - specific differentiation antigens may be expressed in the tumor cell population, providing potential targets for activated lymphocytes [ 5,6 ]. Recognition of a tumor associated antigen is likely to depend on recognition of the dominant epitopes as observed in both human [ 7 ] and
murine [ 8 ] systems. However, the importance of lymphocyte specificity in the initial recruitment of antitumor effector cells remains unclear. Does lymphocyte localization in a tumor require specific recognition of tumor antigen in the vascular compartment or is initial recruitment dependent on the characteristics of the tumor - associated vasculature? Tumor infiltration by lymphocytes, monocytes, and eosinophils is common [ 9 ], especially at the tumor periphery [ 10,11 ]. Although the presence of tumor - specific lymphocytes within tumor tissue suggests specific recruitment, it may also reflect a more generalized process of recruitment involving cells of multiple specificities and lineages. The ability of the leukocyte population to identify tumor associated antigens and kill the cells that express them is not sufficient to effect a complete antitumor response in most patients. One possible reason for this may be insufficient homing of cells with required specificity to the tumor tissue. Although tumor localization of cultured tumor - infiltrating lymphocytes has been demonstrated [ 12 ], it appears that in some cases the number of delivered cells was inadequate to produce a therapeutic response. Adoptive immunotherapy with lymphokine - activated and expanded effector cells, such as interleukin - 2 ( IL - 2 ) – activated tumor - infiltrating lymphocytes, has had limited success in advanced patients with cancer [ 1,13 – 15 ], yet the mechanism of delivery remains unclear. In vitro findings and limited in vivo data suggest that the tumor - specific cytolytic activity of the effector cells is preceded by the delivery of effector cells to the target tissues [ 16 – 18 ]. If the ability of lymphocytes to eradicate an established tumor is determined by the tumor burden, growth pattern, and the intensity of the immune response at the site of tumor [ 16,17,19 – 22 ], the antitumor response may be insignificant when the delivery of a specific cytotoxic cell population to a developing tumor is limited. Address all correspondence to: Dr. Rakesh K. Jain, Department of Radiation Oncology, Massachusetts General Hospital, 100 Blossom Street, Cox - 7, Boston, MA 02114, USA. E - mail: [email protected] 1 This work was supported by National Institutes of Health grants PO1 CA80124 ( R.K.J., L.L.M. ), and R01 HL64240 ( L.L.M. ) and RO1 CA86782 ( R.W. ). 2 Current address: Preclinical Development, Human Genome Science, Inc., 9410 Key West Avenue, Rockville, MD 20850, USA. 3 Current address: Advanced Process Combinatorics, West Lafayette, IN, USA. Received 22 August 2001; Accepted 2 October 2001. Copyright # 2002 Nature Publishing Group All rights reserved 1522-8002/02/$25.00
4
Localization of Lymphocytes
Melder et al.
Delivery of cells to a tumor may be influenced by at least three different factors: cellular adhesion molecule ( CAM ) expression within the tumor vasculature, the flux of circulating cells entering the tumor microcirculation, and the ability of the leukocyte population to engage the appropriate CAMs to arrest and extravasate into the tissue [ 23 ]. The present study uses experimental and mathematical models to quantify trafficking of injected effector cells to identify parameters that limit localization in the tumor.
Materials and Methods Activated Tumor - Specific Lymphocytes Mouse antitumor lymphocytes were prepared from immunization of Balb / c mice ( H - 2 d, Mls b ) with lymphocytes from C3H / sed mice ( H - 2 k, Mls c ). Briefly, C3H mice were sacrificed and their spleens were removed and placed in culture medium RPMI 1640 ( Fisher Scientific, Pittsburgh, PA ) with 10% FCS ( Gibco, Grand Island, NY ). Single - cell suspensions were prepared by disruption of the spleens followed by lysis of the erythrocytes in distilled water. The resulting cells were resuspended in saline and exposed to three freeze – thaw cycles. The resulting material was washed once in PBS, mixed with Freunds incomplete adjuvant and refrozen at 708C. Aliquots were removed and injected into the Balb / c mice at 1 - week intervals for 3 weeks. The resulting sensitized lymphocyte populations were obtained by removal of spleens from the donor populations. Single cell suspensions were prepared as before. The mononuclear population was washed once in culture medium and depleted of the monocyte population through column adhesion for 1 hour at 378C. The cells were then eluted with three volumes of warm medium, washed once and examined. Giemsa staining of the cytofuged cells revealed that 98% of the cells in the final cell suspension were small lymphocytes. Tumor Model The syngeneic carcinoma, MCaIV, was grown in female C3H mice. The MCaIV cell line was maintained by in vivo passage from frozen stock and only low passage tumors ( less than five passages ) were used. Cells were collected from harvested tumors, washed once in complete medium, and resuspended in medium without serum to a concentration of approximately 108 / ml. A volume of 10 to 20 l was injected subcutaneously in the leg of an SCID ( H - 2 d, Mls b ) mouse. Within 8 days, this procedure resulted in tumors of 0.7 - cm diameter. The leg tumor model was selected for this study because ( a ) the tumor is distant from the internal organs, simplifying interpretation of the results and ( b ) it provides convenient access for monitoring growth. In one group of mice, splenectomy was performed ( after anesthesia with ketamine, 9 mg / 100 g, and xylazine, 0.9 mg / 100 g i.m. ) under aseptic conditions. Tumor implantation was performed 3 to 5 days following surgery.
Cell Labeling 111 In oxine was incubated with the sensitized or nonsensitized lymphocytes for 1 hour at 378C. The cells were then washed with PBS three times and resuspended to a concentration of 107 cells / 150 l. Total activity of the cells was checked in a gamma - counter before use ( approximately 100 Ci / 107 cells ). Because of variable release rates of In from labeled cells, we used a previously optimized labeling procedure in which In oxine was synthesized and purified de novo before each labeling procedure; optimum ratios of cell / isotope and incubation times were used, cells were repeatedly washed before injection ( to remove not transchelated oxine ), and nonviable cells were excluded from the injectate [ 24,25 ]. Cell Injection and Imaging Procedure Ten minutes before injection of the cell suspension, the mice were given a mixture of ketamine ( 9 mg / 100 g ) and xylazine ( 0.9 mg / 100 g ) intramuscularly. After anesthesia, a heparinized ( 100 IU / ml ) 30 - gauge cannula was placed into the lateral tail vein. A 1 - ml syringe containing 107 labeled cells in 0.15 ml PBS was connected to the intravenous line and the cells were injected over approximately 1 minute. Imaging of cell localization was initiated 1 hour after injection and continued at 24 - hour intervals. Imaging was performed using a gamma camera ( Ohio Nuclear, Solon, OH ) with a pinhole collimator and a 37 PMT array. Images were assembled and quantified on a Macintosh computer using the NucLear Power 2.9 software. Regions of interest ( ROIs ) were determined at the time of the initial imaging procedure and remained constant throughout the subsequent time points. Counts obtained from the ROIs were adjusted for isotope decay at each time point. Counts obtained in the liver and spleen were adjusted to compensate for proximity by removing the organs and comparing relative counts in the liver and spleen using a gamma counter. The counts obtained by planer scintigraphy were adjusted by the relative distribution of total counts in these organs. In a separate subgroup of mice, imaging findings were validated by performing comparative biodistribution studies at multiple endpoints. Confirmation of Sensitization Specificity of the lymphocyte preparation was tested by proliferation assay ( Promega, Madison, WI ). Both tumor antigen and sensitizing antigen were used to elicit a proliferative response. Concanavalin A and medium were used as positive and negative controls, respectively. Tumor antigen was obtained from sterile homogenized MCaIV tumor ( F3 passage from frozen stock ), which was immediately frozen. Tumor homogenate was diluted 1:100 in fresh culture medium before use. Immunohistology Tumor tissue was removed from mice that were not used for imaging studies. Normal skin was obtained from non – tumor - bearing mice for comparison. Tissue samples were placed in plastic tissue trays and embedded in OCT medium,
Neoplasia . Vol. 4, No. 1, 2002
Localization of Lymphocytes
then quick - frozen in isopropanol and dry ice. Frozen sections were cut at 7 - m thickness and fixed in cold acetone ( 48C ) for 5 minutes. The sections were washed once with PBS, pH 7.2, with 1% BSA ( Sigma, St. Louis, MO ). The slides were then stained in a humidified chamber at 378C for 30 minutes using 5 l of anti – ICAM - 1 - biotin ( 01542D, Pharmingen, San Diego, CA ), anti – ICAM - 2 - biotin ( 01802D, Pharmingen ), anti – VCAM - 1 - biotin ( 01812D, Pharmingen ), anti – CD31 - biotin ( 01952D, Pharmingen ) or PBS alone. Following incubation, the slides were washed three times with PBS with 1% BSA, then incubated for 30 minutes with 10 l of avidin - alkaline phosphatase ( 10 4 dilution, A - 7294, Sigma ) at 378C. The slides were washed three times with PBS / 1% BSA, then incubated with 0.1 ml of Fast Red TR / naphthol AS - MX reagent in Tris buffer ( F 4523, Sigma ). The slides were washed three times with distilled water and counterstained lightly with hematoxylin ( HHS - 16, Sigma ). The slides were then washed once in distilled water and the coverslips were mounted with glycerol. Mathematical Model The basis of the theoretical model used in these studies has been described previously [ 26 ]. Briefly, the model simulates the following key processes involved in lymphocyte trafficking following adoptive transfer: ( 1 ) transport through the systemic circulation, ( 2 ) initial reversible capture ( temporary adhesion or entrapment ) at the endothelial wall, ( 3 ) arrest ( stable adhesion ) at the endothelial wall following capture, ( 4 ) transmigration of arrested lymphocytes across the endothelial wall, ( 5 ) recirculation through the lymphatic system following transmigration in organ compartments, ( 6 ) limited recirculation in tumor compartment, ( 7 ) transient accumulation of recirculating lymphocytes in the lumped lymph node before return to the systemic circulation, and ( 8 ) possible depletion ( for example, mechanical disruption or apoptosis ) in tissues. The model assumed that free labels were excreted through urine at a much faster rate than their spontaneous release from cells, rendering their influence on the observed biodistribution negligible. Lymphocyte proliferation during the distribution phase was also neglected. The major lymphoid tissues included in the model were bone marrow, spleen, and lymph node. In addition to the lymphoid tissues, other peripheral organs and tissues such as blood, heart, lung, liver, kidney, GI tract, skin, muscle, and tumor were incorporated in the model. These organs and tissues were arranged according to their anatomic relationship. The simulations were performed as described previously varying certain parameter values as appropriate to simulate the difference between normal and sensitized cell populations.
Melder et al.
5
injected cell number within 24 hours. Redistribution of these cells at 24 hours post injection resulted in a concurrent increase in liver and spleen localization, which remained stable throughout the observation period. Lymphocyte localization within the tumor ranged from 1% to 2.5% ( 4 to 7 days post injection ). Sensitized and normal lymphocyte populations were tested for proliferation in response to stimulation by sensitizing antigen, tumor homogenate or Concanavalin A ( Figure 1 ). Lymphocyte proliferation in response to both tumor and sensitizing antigen was equal to the mitogenic response to Concanavalin A, indicating successful priming. Scintigraphic studies of cell biodistribution revealed that tumor - specific and nonspecific lymphocytes had approximately equal levels of delivery to the tumor during the first week following administration ( Figure 2; A and B ). Although this finding suggests a diminished importance of antigen specificity for tumor rejection [ 1 ], the number of lymphocytes in the injected population that were specific for the foreign antigen may have been relatively low and below the ability of our system to distinguish them from the labeled nonspecific lymphocytes. However, using similar methods, Kjaergaard and Shu [ 27 ] were also able to show that lymphocyte specificity is not a major determinant of localization. These results further suggest that the initial phase of cell recruitment to tumors is nonspecific. Our previous studies of NK cell localization in tumors also indicated that some tumor vessels support adhesive interactions with injected effector cells [ 28 ]. However, a second stage of recruitment, mediated by the infiltrating T cell population may be required to initiate tumor rejection. Unfortunately, the isotope half - life ( 2.8 days ) restricted the duration of these studies to a week, preventing observation of the longer - term kinetics. Because both the liver and spleen act as principal retention sites for injected lymphocyte populations, we hypothesized that removal of one of these compartments would alter the distribution of the injected populations.
Results and Discussion Systemic and Tumor Distributions The general pattern of lymphocyte distribution following injection was similar for all mice used in these studies. Lymphocyte localization in the lungs peaked immediately after injection and dropped to between 1% and 10% of the
Neoplasia . Vol. 4, No. 1, 2002
Figure 1. Sensitization of H - 2 d mice ( Balb / c ) with H - 2 k ( C3H ) antigen. Cell proliferation assays indicate that sensitized T lymphocytes from H - 2 d mice proliferated in response to H - 2 k antigen expressed on tumor cells or spleen cells to levels comparable with a nonspecific mitogen ( maximal stimulation ). Approximately the same level of cell proliferation ( as indicated by the staining intensity of the assay ) could be observed between cells stimulated with tumor cells, spleen cells, and Concanavalin A mitogen. Means ± SD of five wells are shown.
6
Localization of Lymphocytes
Melder et al.
Figure 2. Experimental and theoretical biodistribution profiles of sensitized ( A ) and normal ( B ) lymphocyte populations in tumor - bearing mice and in tumor - bearing mice with splenectomy ( C ). Overall patterns of systemic distribution and accumulation of lymphocytes within the tumors were similar for both populations of lymphocytes, suggesting that the initial phase ( < 7 days ) of cell recruitment to tumors may be nonspecific. Splenectomy of mice before the implantation of the tumor results in tumor accumulation kinetics comparable to the nonsensitized lymphocyte populations in the nonsplenectomized mice. Retention of transferred lymphocytes within the liver is greater than in nonsplenectomized mice, suggesting a shift in compartmental distribution following the removal of a major filtration compartment, but only a slight increase in tumor localization can be observed. Means ± SD are shown, n = 6 mice / group. The simulations of the mathematical model are shown overlying the experimental data. The predictions agree well with the experiments and capture the salient features of the experimental biodistribution profiles.
Removal of the spleen ( Figure 2C ) resulted in increased uptake within the liver as early as 1 hour after cell injection. Increased retention by the liver was approximately equal to the expected uptake by the missing spleen. These findings suggest that modification of compartmental localization of injected cells can alter distribution but does not provide a dramatic improvement in tumor delivery.
Comparison with Pharmacokinetic Model The mathematical model was used to simulate lymphocyte distribution in mice, both with and without splenectomy. All administered cell populations were assumed to have an equal probability of tumor localization and no bias was given to retention of lymphocyte subpopulations within various microvascular or lymphatic compartments. The model predicted the initial lymphocyte uptake in lungs and the subsequent redistribution to the liver and spleen ( Figure 2; A and B ). Lymphocyte retention rates were slightly overestimated for the liver and spleen, resulting in higher values than observed. However, the model predicted well the overall trends in the biodistribution profiles with time and was able to capture the increased liver uptake following splenectomy. The agreement between model predictions and experimental data ( Figure 2; A C ) supports the validity of the assumption that different lymphocyte populations have an equal likelihood of recruitment by the tumor vasculature. The expression of constitutive and inducible CAMs on normal endothelium in vitro and in vivo is tissue / organ dependent [ 29 ]. Several CAM expression studies in tumor vessels have been conducted ( for a review, see Ref. [ 30 ] ) and these studies demonstrate that CAM expression during tumor angiogenesis is highly heterogeneous. Variation in CAM expression can be observed within different tumor types, and within different regions of the same tumor. Human mammary cancer endothelium, for example, shows increased expression of various adhesion molecules ( e.g., E - selectin, VCAM - 1, and ICAM - 3 ) [ 31 ], whereas in primary melanomas and various other tumors, VCAM - 1 expression is focal or downregulated [ 32 ]. These findings help to explain how the heterogeneous infiltration of leukocytes into tumors can occur in terms of variation in adhesion of cells to tumor vessels. Although other mechanisms contribute to the migration and activity of infiltrating leukocytes within tumors, the accumulation of cells within the tumor vessels is an initial limiting factor on the system. Immunohistology performed on sections from the MCaIV tumor demonstrated moderate staining for ICAM - 1, ICAM 2, and CD31 but strong staining of VCAM - 1 throughout the tumor tissue, indicating the potential for 1 integrin – mediated adhesion in these vessels ( Table 1 ). We have shown [ 28,33 ] preferential adhesion of IL - 2 – activated natural killer ( A - NK ) cells in some vessels within human and murine ( MCaIV ) tumors. This suggests that the relative
Table 1. Expression of Cellular Adhesion Molecules in the MCaIV Tumor and Normal Skin
Antibody
Tumor tissue*
Normal tissue
Isotype control ICAM - 1 ICAM - 2 VCAM - 1 CD31
+ + ++++ ++
+ + ++
*Results indicate expression observed on frozen sections of tissue and stained by immunohistology.
Neoplasia . Vol. 4, No. 1, 2002
Localization of Lymphocytes
degree of lymphocyte interaction with tumor and normal vessels depends on the regional variation in adhesion molecule expression within the tumor. Further insight into lymphocyte binding to tumor vessels has come from PET studies of the localization of IL - 2 A - NK cells in normal and tumor tissues in mice [ 34 ]. These findings are in agreement with the current observation that lymphocyte localization to tumor vessels is adhesion and delivery dependent. That these previous studies were performed with A - NK cell populations precludes antigen - specific recognition mechanisms for localization and supports a nonspecific adhesion mechanism for the accretion of lymphocytes within a tumor. Populations of NK cells may have high affinity to regions of tumor vessels [ 28,33 ], but may be limited to a single pass through ( or around ) the tumor vasculature due to entrapment in other organs [ 35 ]. In contrast, T lymphocytes may have lower adhesion efficiencies to the tumor vessels but are not limited to single - pass delivery. It is reasonable that T cells may localize within tumor vessels through a similar adhesive mechanism, using repeated recirculation of the tumor - specific cells to facilitate long - term accumulation of lymphocytes within tumors. The parallels between the kinetics of lymphocyte localization and the model simulations ( Figure 2; A C ), in which all injected cell populations were assumed to be recruited to the tumor vasculature with equal probability, supports the hypothesis of nonspecific recruitment of lymphocytes to a tumor. Modulation of Biodistribution We have previously shown that modulation of lymphocyte - endothelial adhesion accompanies angiogenesis [ 36 ]. The extent to which this will influence the accumulation of lymphocytes within the tumor depends on the number of cells in the tumor microcirculation available for recruitment. The data from this study and the mathematical simulations of lymphocyte biodistribution ( Figure 2 ) indicate that high retention of lymphocytes within the liver and spleen restricts the availability of a large fraction of the injected cells for recruitment to the tumor. Although lymphocytes within these organs may still be available for recirculation, the rate at which this occurs appears to be low, because the number of cells within the liver is constant after 72 hours. Removal of the spleen does not increase the number of cells available for tumor recruitment because these are acquired and retained by the liver, as observed experimentally and predicted by the theoretical model. Thus, our work suggests that improved lymphocyte delivery to tumors will require increased systemic availability of circulating lymphocytes at steady state to facilitate adhesion and infiltration of tumor - specific T - cell populations.
References [1] Rosenberg SA ( 2000 ). Principles and Practice of the Biological Therapy of Cancer. Lippincott Williams and Wilkins, Philadelphia. [2] Rosenberg SA (2001). Progress in human tumor immunology and immunotherapy. Nature 411, 380 – 384. [3] Fabre JW (2001). The allogeneic response and tumor immunity. Nat Med 7, 649 – 52.
Neoplasia . Vol. 4, No. 1, 2002
Melder et al.
7
[4] Nestle FO, Banchereau J, and Hart D (2001). Dendritic cells: on the move from bench to bedside. Nat Med 7, 761 – 765. [5] Disis M, Gralow J, Bernhard H, Hand S, Rubin W, and Cheever M (1996). Peptide based, but not whole protein, vaccines elicit immunity to HER - 2 / neu, oncogenic self - protein. J Immunol 156, 3151 – 3158. [6] Roth J, Dittmer D, Rea D, Tartaglia J, Paoletti E, and Levine A (1996). p53 as a target for cancer vaccines: recombinant canarypox virus vectors expressing p53 protect mice against lethal tumor cell challenge. Proc Natl Acad Sci USA 93, 4781 – 4786. [7] Correale P, Walmsley K, Nieroda C, Zaremba S, Zhu M, Schlom J, and Tsang K (1997). In vitro generation of human cytotoxic T lymphocytes specific for peptides derived from human prostate - specific antigen. J Natl Cancer Inst 89, 293 – 300. [8] Matsumura T, Sussman JJ, Krinock RA, Chang AE, and Shu S (1994). Characteristics and in vivo homing of long - term T - cell lines and clones derived from tumor - draining lymph nodes. Cancer Res 54, 2744 – 2750. [9] Boheim K, Denz H, Boheim C, Glassl H, and Huber H (1987). An immunohistologic study of the distribution and status of activation of head and neck tumor infiltrating leukocytes. Arch Oto - Rhino - Laryngol 244, 127 – 32. [10] Dvorak HF, and Dvorak A (1982). Immunohistological characterization of inflammatory cells that infiltrate tumors. In Tumor Immunity in Prognosis: The Role of Mononuclear Cell Infiltration S Haskill ( Ed ). Marcel Dekker, New York. pp. 279 – 307. [11] Jain RK (1996). Delivery of molecular medicine to solid tumors. Science 271, 1079 – 80. [12] Rosenberg SA (1990). Adoptive immunotherapy for cancer. Sci Am 262, 62 – 69. [13] Hwu P, and Rosenberg SA (1994). The use of gene - modified tumor - filtrating lymphocytes for cancer therapy. Ann NY Acad Sci 716, 188 – 97. [14] Rosenberg SA (1994). The gene therapy of cancer. Prev Med 23, 624 – 626. [15] Rosenberg SA (1992). The immunotherapy and gene therapy of cancer. J Clin Oncol 10, 180 – 99. [16] Basse P, and Goldfarb RH (1994). Localization of immune effector cells to tumor metastases. In Tumor Immunology and Cancer Therapy R Goldfarb and T Whiteside ( Eds ). Marcel Dekker, New York. [17] Whiteside TL, and Goldfarb RH (1994). Antitumor effector cells: extravasation and control of metastasis. In Tumor Immunology and Cancer Therapy R Goldfarb and T Whiteside ( Eds ). Marcel Dekker, New York. [18] Topalian SL, and Rosenberg SA (1991). Adoptive cellular therapy: basic principles. In Biological Therapy of Cancer V DeVita, Jr, S Hellman and S Rosenberg ( Eds ). JB Lippincott, Philadelphia. [19] Greenberg PD (1991). Adoptive T cell therapy of tumors: mechanisms operative in the recognition and elimination of tumor cells. Adv Immunol 49, 281 – 355. [20] De Boer RJ, Hogeweg P, Dullen HF, De Weger RA, and Den Otter W (1985). Macrophage T lymphocyte interaction in the anti - tumor immune response: a mathematical model. J Immunol 134, 2748 – 58. [21] De Boer RJ, and Hogeweg P (1986). Interaction between macrophages and T lymphocytes: tumor sneaking through intrinsic to helper T cell dynamics. J Theor Biol 120, 331 – 51. [22] Kuznetsov VA, Makalkin IA, Taylor MA, and Perelson AS (1994). Nonlinear dynamics of immunogenic tumors: parameter estimation and global bifurcation analysis. Bull Math Biol 56, 295 – 321. [23] Jain RK (1998). The next frontier of molecular medicine: delivery of therapeutics. Nat Med 4, 655 – 657. [24] Schoepf U, Marecos E, Melder R, Jain R, and Weissleder R (1998). Intracellular magnetic labeling of lymphocytes for in vivo trafficking studies. BioTechniques 24, 642 – 646. [25] Lewin M, Carlesso N, Tung C, Tang X, Cory D, Scadden D, and Weissleder R (2000). Tat peptide derivatized magnetic nanoparticles allow in vivo tracking and recovery of progenitor cells. Nat Biotechnol 18. [26] Zhu H, Melder RJ, Baxter LT, and Jain RK (1996). Physiologically based kinetic model of effector cell biodistribution in mammals: implications for adoptive immunotherapy. Cancer Res 56, 3771 – 81. [27] Kjaergaard J, and Shu S (1999). Tumor infiltration by adoptively transferred T cells is independent of immunologic specificity but requires down - regulation of L - selectin expression. J Immunol 163, 751 – 759. [28] Sasaki A, Melder RJ, Whiteside TL, Herberman RB, and Jain RK (1991). Preferential localization of human adherent lymphokine activated killer cells in tumor microcirculation. J Natl Cancer Inst 83, 433 – 437.
8
Localization of Lymphocytes
Melder et al.
[29] Kuzu I, Bicknell R, Fletcher CD, and Gatter KC (1993). Expression of adhesion molecules on the endothelium of normal tissue vessels and vascular tumors. Lab Invest 69, 322 – 328. [30] Jain RK, Koenig GC, Dellian M, Fukumura D, Munn LL, and Melder RJ (1996). Leukocyte - endothelial adhesion and angiogenesis in tumors. Cancer Metastasis Rev 15, 195 – 204. [31] Fox SB, Turner GDH, Gatter KC, and Harris AL (1995). The increased expression of adhesion molecules ICAM - 3, E - and P - selectins on breast cancer endothelium. J Pathol 177, 369 – 376. [32] Piali L, Fichtel A, Terpe H, Imhof BA, and Gisler RH (1995). Endothelial vascular cell adhesion molecule 1 expression is suppressed by melanoma and carcinoma. J Exp Med 181, 811 – 816. [33] Melder RJ, Elmaleh D, Brownell AL, Brownell GL, and Jain RK (1994).
A method for labeling cells for positron emission tomography ( PET ) studies. J Immunol Methods 175, 79 – 87. [34] Melder RJ, Brownell AL, Shoup TM, Brownell GL, and Jain RK (1993). Imaging of activated natural killer cells in mice by positron emission tomography: preferential uptake in tumors. Cancer Res 53, 5867 – 5871. [35] Melder RJ, Kristensen CA, Munn LL, and Jain, RK. Modulation of A - NK cell rigidity: in vitro characterization and in vivo implications for cell delivery. Biorheology 38, 151 – 159. [36] Melder RJ, Koenig GC, Witwer B, Safabakhsh N, Munn LL, and Jain RK (1996). During angiogenesis, vascular endothelial growth factor and basic fibroblast growth factor regulate natural killer cell adhesion to tumor endothelium. Nat Med 2, 992 – 997.
Neoplasia . Vol. 4, No. 1, 2002