| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics |
Department of Immunology, The Scripps Research Institute, La Jolla, California 92037 [C. L., H. H., A. E., T. S. E.]; Attenuon, LLC, San Diego, California 92121 [F. D.]; Applied Molecular Evolution, San Diego, California 92121 [C. D.]; Wayne State School of Medicine, Detroit, Michigan 48201 [R. S.]; and Department of Urology, University of Washington, Seattle, Washington 98185 [R. V.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
or 
linkages (1
, 4)
. This protein also has been described in brain as a neuropeptidase (1
, 5)
, as well as in small intestine as folate hydrolase (1
, 4)
. In normal prostatic epithelium, a cytosolic form of PSMA is found, whereas in CaP, there is a nearly 100-fold increase of the membrane form (6)
. There is evidence that indicates PSMA expression is up-regulated not only in tumor cells but also is found associated with local, what appear to be, microvascular lining cells in CaP and other tumors (1
, 7
, 8)
. We here explore whether PSMA-positive cells constitute, in part, intratumoral vasculature structures and what have been thought to be PSMA-expressing endothelial cells may be tumor cells adapted to vasculogenic mimicry (9)
. The inappropriate expression of a novel gene product on the lumenal surface of intratumoral microvascular lining cells provides a potential target to localize and assemble molecules for imaging or therapy. We have previously demonstrated the feasibility of localizing TF to tumor microvasculature to induce local tumor vasculature thrombosis (10) . This strategy has successfully induced selective infarctive necrosis of tumors and frequent complete eradication in a proof of principle murine tumor model and without undesirable effects (10, 11, 12) . We have characterized PSMA expression in both the human LuCap tumor model and the rat Mat Lu prostate tumor model. Using a STVT incorporating a PSMA catalytic site inhibitor as the selective targeting element, i.v. administration induced selective local infarction of Mat Lu tumors. Combined therapy with doxorubicin significantly enhanced tumor eradication and prolonged the tumor-free status.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Immunohistochemistry.
Immunohistochemical analysis was performed on formalin fixed as well as fresh frozen 5-µm tissue sections mounted on polylysine-coated slides. For endothelial identification, biotinylated murine antirat CD-31 monoclonal antibody (TLD-3A12) or biotinylated rat antimouse CD-31 monoclonal antibody (MEC 13.3) was used at 1 µg/ml as first antibody, then the reaction was developed with fluorescein-conjugated strepavidin. For identification of PSMA in frozen sections, reaction of mouse monoclonal antibody J591 was followed by biotinylated rabbit antimouse IgG, and the reaction was visualized with Texas-red-conjugated strepavidin. Staining of PSMA in formalin-fixed tissue was performed with biotinylated 7E11C-5 antibody. The tissue sections were analyzed with the aid of laser scanning confocal microscopy (Bio-Rad, Hercules, CA).
Strep-TF Fusion Protein.
TF cDNA containing amino acids 3–311 was obtained by PCR of a human cDNA library (Clontech, Palo Alto, CA) with primers BM21: 5'-ACTACAAATACTGTGGCAGCA-3' and BM33: 5'-TTTaagcttTCACGTGCCCATACACTCTACCGG-3'. The resulting 639-bp fragment was isolated by gel electrophoresis and subjected to a second PCR with BM33 and BM51: 5'-AAATggatccTGGTGCCTAGGGGCCCGGGACTACAAATACTGTGGCAGCA-3'. The resulting 670-bp fragment was digested with BamHI and HindIII and ligated into the BamHI and HindIII sites of the vector pTrcHisC (Invitrogen, Carlsbad, CA). The BM51 oligo also encodes a thrombin cleavage site (Val-Pro-Arg-Gly-Ser) for selective proteolytic deletion of the His tag from the expressed protein. This plasmid (NuV120) was further modified to contain a linker sequence with three repeats of Gly4Ser between the thrombin cleavage sequences and those of TF. The following overlapping oligos were annealed and inserted into the BamHI and AvaI sites of NuV120: nuv20–1: 5'-GATCTTGGTCCCTAGGGGATCCGCAGAACCAATGCCT-3';nuv20–2: 5'-PO4-CACTCGCTAAACTTCAGTCAATACCTCTGGTATACT-3';nuv20–3: 5'-PO4-GGTACCGGAGGAGGCGGTTCAGGTGGTGGAGGTTCA-3';nuv20–4: 5'-PO4-GGAGGTGGAGGTTCTC-3';nuv20–5: 5'-PO4-TCTGCGGATCCCCTAGGGACCAA-3';nuv20–6: 5'-PO4-AGGTATTGACTGAAGTTTAGCGAGTGAGGCATTGGT-3';nuv20–7:5'-PO4-CCACCTGAACCGCCTCCTCCGGTACCAGTATACCAG-3'; and nuv20–8: 5'-CCGGGAGAACCTCCACCTCCTGAACCTCCA-3'.
The resulting plasmid (NuV127) encodes a His-tag, a thrombin cleavage site, three repeats of the spacer Gly4Ser, and TF residues 3–211. This vector can be used to create expression vectors for various STVT molecules by inserting a cDNA sequence encoding the derived amino acids into the unique BamHI and KpnI sites. The streptavidin gene was amplified by PCR with Pfu polymerase (Stratagene) and oligonucleotides strep1: 5'-ACCACGGTCTCGATTACGGC-3' and strep2: 5'-ACTACTGCTGAACGGCGTCG-3'resulting in a 514-bp fragment. The 514-bp fragment was purified and used as template for a second PCR amplification, this time with the oligonucleotides strep3: 5'-CACACAGGATCCGCCGCCGAGGCCGGCATCAC-3' and strep4: 5'-CACACAGGTACCCTGCTGAACGGCGTCGAGCG-3'. BamHI and KpnI sites, respectively, are underlined, and extra nucleotides, in italics, were added for efficient enzyme digestion. The resulting DNA fragment of 486 bp was purified, digested with BamHI and KpnI, and cloned into the BamHI and KpnI sites of NuV127. The resulting plasmid NuV159 expresses a protein shown diagrammatically in Fig. 2A
. BL21 transformed with NuV159 plasmid were grown in Super Medium (25 g of tryptone, 15 g of yeast extract, and 5 g of NaCl/liter) supplemented with biotin. Cells were induced with 1 mM isopropyl-1-thio-ß-D-galactopyranoside when the A600 nm reached 0.6 and were cultured for 24 h at 37°C. The protein (Strep-TF) accumulated in inclusion bodies, which were isolated as described previously (11)
and solubilized in 6 M GuHCl (pH 8.0). The protein construct was partially purified in a Ni-NTA (Qiagen) column equilibrated and washed with 6 M GuHCl (pH 8.0) and eluted with 250 mM imidazole and 6 M GuHCl (pH 8.0). Protein folding was performed by dilution of GuHCl-solubilized Strep-TF in 20 mM Tris, 300 mM NaCl, 0.8 M GuHCl (pH 8.0), and glutathione redox buffer. After 16 h at 4°C, the sample was concentrated with a Pellicon XL concentrator (MWCO, 10,000). The His tag is removed by thrombin digestion, and the protein construct purified in two steps with a Source 15Q 16/10 column followed by a Sephacryl S-200 gel filtration.
|
Cell Culture.
The LnCap cells were cultivated in RPMI 1640 supplemented with 10% FCS, glutamine (2 mM), HEPES (10 mM), sodium pyruvate (1 mM), and glucose (4.5 g/liter). Mat Lu cells were cultured in RPMI 1640 with 10% FCS, glutamine (2 mM), and 250 nM dexamethasone.
Factor Xa Generation Assay.
Factor Xa generation assays were performed as described previously (12)
with modification provided for association of the STVT constructs to PSMA-expressing LnCap cells. Cells were plated at 8 x 104/well in 96-well plates and allowed to attach for 4 h in medium above. Medium was replaced with HBSA buffer [150 mM NaCl, 5 mM CaCl2, 0.5% BSA, and 20 mM HEPES (pH 7.4)], and serial concentrations of DßE:strep-TF:VIIa or Strep-TF:VIIa complex were add to the wells. After 5 min of incubation, factor X was then added to a final concentration of 1 µM. After 5 min at 37°C, the limited proteolytic conversion of factor X to factor Xa was arrested with 100 mM EDTA. Factor Xa amidolytic chromogenic substrate Spectrozyme Xa (American Diagnostica, Greenwich, CT) was added to a final concentration of 200 µM, and substrate hydrolysis was determined kinetically at A405 nm in a spectrophotometric plate reader (Molecular Devices, Sunnyvale, CA).
Animal Models.
The LuCap human prostate tumor was carried as a xenograft in WEHI nude mice (The Scripps Research Institute Breeding Facility) (13)
. The tumors were passaged by implantation of 
2-mm3 fragments in the s.c. tissue of the back of the mice. The rat Mat Lu prostate carcinoma, carried in male Copenhagen rats ages 4–6 weeks (Harlan Sprague Dawley, Germantown, NY), was inoculated with 5 x 105 Mat Lu cells s.c./site in the back of the rats. Treatment was initiated once tumors reached 200 mm3 through bolus i.v. injection of the STVT or control protein (0.1 mg/kg based on strep-TF protein) and repeated twice at 2-day intervals. For combination therapy, liposomal doxorubicin (Doxil) at 2 mg/kg was separately injected i.v. Tumor growth and other physical signs were monitored daily, including gross evidence of tumor necrosis, local tumor ulceration, as well as evidence of toxicity, including mobility, response to stimulus, eating, and weight of each animal. The studies have been reviewed and approved by the Institutional Animal Care and Use Committee of The Scripps Research Institute. The work was conducted in the Scripps Research Institute facilities, which are accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care. The Scripps Research Institute maintains an assurance with the U.S. Public Health Service and is registered with the United States Department of Agriculture and is in compliance with all regulations relating to animal care and welfare.
Statistical Analysis.
Statistical significance was determined by the two-tailed Student’s t test, except for statistical significance of survival curves, which used the logrank test using GraphPad Prism version 3.00 (GraphPad Software, San Diego, CA).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The lumens of these channels are formed by tumor cells, and PSMA expression is more intense on the aspect of tumor cell membranes that constitute the lumenal surface of the channels (Fig. 1B)
. Double staining of the LuCap tumor with antimouse CD31 antibody and PSMA-specific antibody suggests these PSMA-delineated microchannels are distinct and mutually exclusive of microvascular channels lined by CD31-positive cells (Fig. 1C)
. Similar microchannels lined by PSMA-positive cells were also observed in syngeneic rat Mat Lu tumors (Fig. 1D)
.
|
2 min after infusion, rapidly frozen, and sectioned. Double immunostaining for PSMA and bacteriophage revealed that the PSMA-positive cell-lined microchannels contained phage (Fig. 1D)
indicative of immediate functional continuity with the general vasculature.
The STVT Functionally Associates with PSMA-positive Cells.
To confer cell surface assembly of the designed STVT to PSMA positive cells, a known suicide inhibitor of PSMA glutamyl carboxypeptidase, namely DßE, was incorporated (14, 15, 16)
. The biotin-GSGSDßE inhibitor structure was coupled to Strep-TF protein (Fig. 2, A and B)
through the high affinity binding of biotin to the streptavidin domain. However, targeting alone is not sufficient for function because the STVT must also align properly on an anionic cell membrane surface and associate with factor X substrate that has localized to the same locus. The specific activity of the assembled DßE:strep-TF:VIIa complex on PSMA-expressing LnCap cells (Fig. 2C)
was analyzed in a factor Xa generation assay that requires the functional assembly. Unlike most tumor cells, LnCap cells do not express TF as indicated by the coagulation assays and examined by Western blot (data not shown) and do not directly activate substrate factor X to the active product factor Xa and thereby drive the thrombogenic cascade (Fig. 2D)
. The dose dependent increase of factor Xa generation in the presence of LnCap cells was striking in comparison to the control Strep-TF lacking the PSMA-targeting element, indicating that the PSMA-directed STVT functionally assembles on the cell surface via binding of DßE to PSMA (Fig. 2D)
and functionally initiates the thrombogenic cascade.
Tumor Infarction in Vivo.
The control Strep-TF protein was not toxic in rats over a wide range of concentrations, thereby permitting evaluation of the potential for selective tumor thrombosis and infarctive necrosis in tumor-bearing rats. i.v. administration of the PSMA-directed STVT was associated with a rapid wave of microthrombosis and resultant infarction of Mat Lu tumors (Fig. 3A)
with significant retardation in tumor growth (Fig. 3B)
. An average 70% reduction in tumor mass was observed compared with controls (Fig. 3C)
. The center of the tumors in the experimental group showed gross signs of ischemic necrosis. In contrast, there was no microthrombosis or areas of necrosis in these highly cellular tumors from the control group (Fig. 4A)
. Occluded tumor microvessels were widespread in the experimental group (Fig. 4B)
, with platelet aggregates, packed erythrocytes, and fibrin (Fig. 4C)
. The tumor interstitium that commonly contained a few erythrocytes was infiltrated with inflammatory cells (Fig. 4D)
. After the standard three infusions at 2-day intervals, tumors showed very extensive necrosis with liquefaction of the entire central region of the tumors. However, at the growth edge of tumors from the treated animals, a rim of viable tumor tissue remained.
|
|
. This same combination therapy had a significant beneficial effect on survival of the tumor-bearing animal hosts (P < 0.001). The prolongation of survival of rats treated with STVT alone was modest but was significant. Therapy with low-dose liposomal doxorubicin alone had no measurable benefit (Fig. 6)
.
|
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
or linkages to release glutamate (1
, 4)
. Whereas normal prostate epithelial cells produce a cytosolic form of PSMA, transformation results in a nearly 100-fold increase of the membrane isoform (6)
. Some recent studies have suggested that PSMA expression is up-regulated not only in prostate carcinomas but also is associated with the local endothelial cells in prostate carcinomas and even in association with other tumors (1
, 7
, 8) .
We have demonstrated in this study that PSMA-positive cells are found lining intratumoral microchannels that are not lined by conventional CD31-positive endothelial cells. Tumor cell surface expression of PSMA is more intense on that aspect of tumor cell membranes that delineate the lumenal surface of these tumor cell-lined channels. These PSMA-delineated microchannels are continuous with the general vasculature based on the very rapid entry of bacteriophage into these channels after infusion by the tail vein. Additional studies demonstrate that extensive thrombotic infarction occurs after administration of a PSMA-localizing STVT. It is reasonable to hypothesize that the PSMA-positive cell delineated channels, which lack the usual endothelial marker CD31, are likely tumor cells adapted to a pseudoendothelial phenotype. Such adaptation has been described as vasculogenic mimicry (17, 18, 19, 20, 21, 22, 23)
wherein tumor cells, rather than endothelial cells, adapt and line intratumoral microvascular channels. A recent study showed heterogeneous-invasive prostate carcinoma cell lines have the potential to form perfusable vasculogenic-like networks in culture (9)
. Existence of such networks in aggressive rat and human tumors in vivo, similar to this study, were observed (9)
. Although a current topic of some interest and debate (24, 25, 26, 27)
, a greater degree of elucidation of the intrinsic cell biology and vasculogenic characteristics remain to be developed (28, 29, 30)
. However, in addition to the vasculogenic mimicry hypothesis, tumor cell surface molecules may be directly accessed by molecules in blood, including therapeutic agents, through direct transmigration of tumor cells through the microvascular lining cells to the lumen and subsequent detachment into the circulation during metastasis. These tumor cells initially localize to endothelial cells locally before releasing into circulation (29)
. In one study, it was estimated that 15% of perfused vessels of a colon carcinoma xenograft were mosaic vessels with focal regions where tumor cells appeared to contact the microvessel lumen (30)
. Tumor cells accounted for 4% of the total vascular surface area in these colon carcinomas. Similar numbers of mosaic vessels were found in human colon carcinoma biopsies (30)
, underscoring the complexity of intratumoral microvasculature and differences from normal vasculogenic rules.
A selective tumor microvascular thrombotic infarction strategy was used to determine whether tumor cell surface-expressed molecules, in this case PSMA, could both localize and properly assemble the cell surface TF:VIIa:X:membrane complex to initiate the thrombogenic cascade in vivo. Using this PSMA-directed STVT, we observed robust and highly selective tumor microvascular thrombosis and infarctive necrosis of syngeneic prostate tumors in the rats. The gross and histopathological changes observed were similar to those previously described by us and others for selective infarctive therapy of tumors (10 , 31) . Signs of tumor vasculature thrombosis occurred immediately after initial infusion of the STVT followed by infarction and necrosis. The adopted three-dose protocol was without any general adverse effect on the rats. However, despite the rapid tumor destruction, there remained viable tumor cells at the tumor periphery. However, host survival was significantly extended.
Combination therapy of the PSMA-directed STVT with low-dose liposomal doxorubicin was far more effective. Doxorubicin alone had no observable effect on tumor growth or survival. However, separate bolus infusions of doxorubicin at the time of STVT infusion greatly potentiated the effect on the tumors. Abrupt infarctive features of the tumors were observed. Although not directly addressed, the doxorubicin appears to facilitate eradication of tumor cells at the peripheral edge of the tumors where microthrombosis and necrosis appears to be less effective. However, the more attractive interpretation is that because doxorubicin is known to induce endothelial cell apoptosis (32) , it may also have injured the tumor microvascular endothelium and thereby increasing exposure of the tumor cells to plasma proteins to potentiate the local thrombotic activity of the PSMA-directed STVT.
These experiments underscore the potential importance of tumor cell exposure and even participation in intratumoral microvasculature. Such exposure to the blood supports the therapeutic potential to target large molecules such as proteins to tumor cell surface molecules. The success of targeting the present STVT to a tumor cell surface-specific molecule expands the possible application of this approach to include other tumor cell surface molecules as the facilitators and targets. Dual therapy of a functional STVT and a cytotoxic agent shows promise in enhancing the infarctive eradication of tumors. Infarctive tumor eradication as been reported by us and others (10 , 31) has the potential to develop as an effective therapeutic tactic.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work is supported by NIH Grants P01 HL 16411 and T32 CA 75924. 
2 To whom requests for reprints should be addressed, at E-mail: chengliu{at}scripps.edu and E-mail: tse{at}scripps.edu
3 The abbreviations used are: PSMA, prostate-specific membrane antigen; CaP, carcinoma of the prostate; TF, tissue factor; STVT, selective tumor vascular thrombogen; GuHCl, guanidinium hydrochloride; DßE, aspartyl-ß-linked glutamate.
Received 2/22/02. Accepted 7/30/02.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-linked acidic dipeptidase activity by N-acetyl-L-aspartyl-ß-linked L-glutamate. J. Neurochem., 55: 39-46, 1990.[Medline]
-linked acidic dipeptidase. J. Med. Chem., 39: 619-622, 1996.[Medline]
This article has been cited by other articles:
|
|
 |
|
  X. Huang, W.-Q. Ding, J. L. Vaught, R. F. Wolf, J. H. Morrissey, R. G. Harrison, and S. E. Lind A soluble tissue factor-annexin V chimeric protein has both procoagulant and anticoagulant properties Blood, February 1, 2006; 107(3): 980 - 986. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Wu, Y. Luo, C. Sun, Y. Liu, P. Kuo, J. Varga, R. Xiang, R. Reisfeld, K. D. Janda, T. S. Edgington, et al. Targeting Cell-Impermeable Prodrug Activation to Tumor Microenvironment Eradicates Multiple Drug-Resistant Neoplasms Cancer Res., January 15, 2006; 66(2): 970 - 980. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Akiyama, K. A. Mohamedali, R. L. e Silva, S. Kachi, J. Shen, C. Hatara, N. Umeda, S. F. Hackett, S. Aslam, M. Krause, et al. Vascular Targeting of Ocular Neovascularization with a Vascular Endothelial Growth Factor121/Gelonin Chimeric Protein Mol. Pharmacol., December 1, 2005; 68(6): 1543 - 1550. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kessler, R. Bieker, T. Padro, C. Schwoppe, T. Persigehl, C. Bremer, M. Kreuter, W. E. Berdel, and R. M. Mesters Inhibition of Tumor Growth by RGD Peptide-Directed Delivery of Truncated Tissue Factor to the Tumor Vasculature Clin. Cancer Res., September 1, 2005; 11(17): 6317 - 6324. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Dienst, A. Grunow, M. Unruh, B. Rabausch, J. E. Nor, J. W. U. Fries, and C. Gottstein Specific Occlusion of Murine and Human Tumor Vasculature by VCAM-1-Targeted Recombinant Fusion Proteins J Natl Cancer Inst, May 18, 2005; 97(10): 733 - 747. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Liu, C. Dickinson, J. Shobe, F. Donate, W. Ruf, and T. Edgington A hybrid fibronectin motif protein as an integrin targeting selective tumor vascular thrombogen Mol. Cancer Ther., July 1, 2004; 3(7): 793 - 801. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. E. Thorpe Vascular Targeting Agents as Cancer Therapeutics Clin. Cancer Res., January 15, 2004; 10(2): 415 - 427. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Palumbo, K. E. Talmage, H. Liu, C. M. La Jeunesse, D. P. Witte, and J. L. Degen Plasminogen supports tumor growth through a fibrinogen-dependent mechanism linked to vascular patency Blood, October 15, 2003; 102(8): 2819 - 2827. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Ruf, E. A. Seftor, R. J. Petrovan, R. M. Weiss, L. M. Gruman, N. V. Margaryan, R. E. B. Seftor, Y. Miyagi, and M. J. C. Hendrix Differential Role of Tissue Factor Pathway Inhibitors 1 and 2 in Melanoma Vasculogenic Mimicry Cancer Res., September 1, 2003; 63(17): 5381 - 5389. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Hess, E. A. Seftor, R. E. B. Seftor, and M. J. C. Hendrix Phosphoinositide 3-Kinase Regulates Membrane Type 1-Matrix Metalloproteinase (MMP) and MMP-2 Activity during Melanoma Cell Vasculogenic Mimicry Cancer Res., August 15, 2003; 63(16): 4757 - 4762. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
, Strep-TF; 
, STVT.
, n = 8; PSMA-directed STVT treated group 
, n = 8). Tumor volume was calculated as D x d2 where D and d are the tumor length and width, respectively. Each point represents the mean ± SE of eight rats (P < 0.001). The experiment was reproducible with comparable results. Tumor size does not represent viable tumor because most of the treated tumors are necrotic. C, day 7 tumor mass comparison where tumor weight of the treated group (n = 8, ) is compared with that of the control group (n = 8, 
