| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Otolaryngology — Head & Neck Surgery (B.W.O.,
D.A.S., D.L., L.D), Johns Hopkins University, Baltimore, Maryland
21203,
Department of Cell Biology (S.L.C.W., S.-H.C.),
Baylor College of Medicine, Houston, Texas 77030,
Department of Pathology (K.-i.K.), Osaka University Medical
Center, Osaka 565, Japan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
IL-2 is produced by stimulated T lymphocytes (4) and is a known T cell growth factor as well as T cell activation factor. IL-2 also appears to be a potent growth and activation factor for natural killer (NK) cells (5, 6, 7). In experimental animal tumor studies, IL-2 has been shown to augment the effect of concurrently administered cytotoxic T cells and has restored normal proliferative responses in patients with cancer and acquired immunodeficiency syndrome (8). Phase I clinical trials using systemically administered IL-2 have demonstrated some success in tumor regression (2, 3). A major limitation of systemic IL-2, however, is the severe toxicity, which includes fever, chills, headaches, and capillary leak syndrome (2, 3, 9).
In addition to the toxicity, the systemic administration of cytokines to stimulate an immunological response bypasses a critical principle in lymphokine physiology. This principle is that lymphokines act in a paracrine fashion to generate and maintain the specificity of the immunological response (10). Under physiological circumstances, specific cytokines are produced locally in high concentrations at the antigen site. These cytokines act locally in concert with antigen-driven signals to generate effector responses. Pharmacological doses of cytokines administered systemically, however, result in high concentrations in the vasculature at sites distant from the antigen but often in suboptimal levels in tissues at the site of antigen.
Considering both the toxicity and the lack of paracrine function, local delivery of cytokines appears to be a safer and more physiological approach to cytokine-based cancer therapy. Important to this concept are studies that have shown that local cytokine delivery can produce dramatic inflammatory effects without significant systemic toxicity (1, 11, 12).
A pilot clinical trial for inoperable squamous cell cancer of the head and neck was reported by Forni and co-workers (13). In this study, patients received daily local injections of recombinant IL-2 around the regional draining lymph nodes for 10 days, which were repeated on a monthly basis for 1 yr. Twenty five percent of the patients demonstrated partial or complete tumor regression after local IL-2 therapy. Although disease-free survival was increased compared with nontreated controls, the responding patients developed a recurrence 3–5 months after treatment. These and other promising results of local cytokine delivery have led to the interest in developing cancer vaccine strategies.
A notable limitation to local delivery of the recombinant protein is the quick clearance of the protein and the need for multiple closely spaced injections to maintain an antitumor response. A solution to this limitation may lie in the application of gene transfer of cytokines such as IL-2 to provide local sustained release of the therapeutic protein. The replication-defective adenovirus is a widely studied vector for gene transfer and has many important features that are useful for cancer therapy strategies. Adenoviral vectors can carry therapeutic genes at titers of up to 1011 plaque-forming units (pfu)/ml, which is significantly greater than retroviral vectors (14). The adenoviral isolate can also be injected alone and directly into tumors with resulting effective gene expression. Furthermore, the adenovirus genome remains episomal rather than integrating into the chromosome as occurs with retroviruses. The adenovirus is also being used in multiple safety studies and human clinical trials for various human diseases (15, 16, 17, 18).
We have developed a head and neck cancer model to study gene transfer strategies and have demonstrated antitumor efficacy in both single and combination gene therapy treatments (19, 20). Our investigations center on combining a popular "suicide gene", the herpes virus thymidine kinase gene (tk) with the gene for IL-2 for the treatment of squamous cell carcinoma of the head and neck. Delivery of the tk gene coupled to systemic administration of the nucleoside analog ganciclovir (GCV) results in necrosis and direct cytotoxicity to dividing cells. When local IL-2 expression is combined at the site of tumor necrosis, we propose that a synergistic antitumor response results (20, 21). With respect to the safety of this strategy, we have not detected any local or distant pathological effects either from a direct adenovirus vector toxicity or from a secondary inflammatory response in the brain, liver, or floor of mouth and neck (19, 20, 22). The immune response after gene transfer of IL-2 in the head and neck cancer model, however, has yet to be evaluated. The following study investigates the immunological role IL-2 plays in single or combination gene therapy and provides insight into the limitations of adenoviral-mediated cytokine delivery to head and neck cancer.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Previous in vitro studies have
demonstrated no toxic effects of tk or ß-gal control adenovirus
vectors (19, 20).
|
). Functional transduction was confirmed by assaying
the tumor cell culture supernatant for IL-2. Using an IL-2-specific
enzyme-linked immunosorbent assay kit, IL-2 protein was detected in
supernatants at both 24 and 72 h after transduction (950
pg/liter x 106 cells).
Characterization of Immune Infiltrates after Adenovirus Treatment
in Vivo
The purpose of this experiment was to evaluate the role of
designated lymphocytes in the therapeutic effect of adenovirus gene
transfer. Twenty mice received floor of mouth injections with the
squamous carcinoma cells as described. At the time of adenovirus
injection, tumor sizes ranged from 70–100 mm3. There were
no significant differences in pretreatment tumor sizes between
experimental groups. In the four experimental groups (tk+IL-2, tk
alone, IL-2+ ß-gal, and ß-gal control), each tumor was treated with
a total of 1 x 109 pfu in a total volume of 50 µl.
For animals receiving combined therapy, 2 x 108 pfu
of ADV/RSV-mIL-2 was delivered. All animals subsequently received
intraperitoneal administration of GCV at 25 mg/kg twice daily for six
days and were killed on day 7. Consistent with our previous findings,
only the groups treated with tk + IL-2 and tk alone demonstrated
significant tumor regression compared with the ß-gal alone control
animals (P = 0.0004; Mann Whitney analysis). Also
consistent was the finding that the tk + IL-2-treated animals
demonstrated significant regression as compared with the tk
alone-treated animals (P = 0.0006).
Immunohistochemical analysis was performed on all tumor specimens, and
positive staining cells were counted per ten high-powered fields (Fig. 2). There were no statistical differences between groups
for CD4 staining, but CD8 staining in the tk+IL-2 group was
statistically greater than in the tk or ß-gal alone group. Although
CD8 staining in the tk+IL-2 group was higher than in the IL-2+ß-gal
group, significant differences were not evident. Notable was the fact
that all samples showed a minimum staining for both CD4 and CD8
lymphocytes, whereas the residual tumors treated with IL-2 alone or in
combination with tk showed an average of 2 to 5 times more positive CD8
cells. The enhanced CD8 lymphocyte tumor infiltration appears to be a
direct result of adenoviral gene transfer of IL-2. Despite the
increased CD8 lymphocytes in the IL-2+ß-gal group, however, no
therapeutic benefit was seen. This finding indicates a lack of
significant tumor recognition by these cytotoxic T lymphocytes in the
groups not treated with tk.
|
). Although the IL-2+ß-gal group had
the widest range of posttreatment tumor sizes, the mean value was very
close to the ß-gal alone group, and there was no significant
difference from the control (P = 0.7). Based on
previous experience, tumors less than 10 mm3 do not project
from the floor of mouth enough to allow identification and caliper
measurement. Therefore, tumors less than 10 mm3 could have
been present in the tk+IL-2 and tk animals that received a tumor size
calculation of 0 mm3.
|
). All tumors at second sites were greater than 100
mm3 in size. The protective systemic antitumor immunity is
dependent on the combination of tk+IL-2 as this group was significantly
different from all other groups (P = 0.003; Fisher
Exact analysis).
|
). Microscopic
evaluation of all residual tumors was performed after routine
hematoxylin and eosin staining. A minimal inflammatory infiltrate was
detected equally in each group, and no tumor necrosis or evidence of
cytotoxic effect was seen (data not shown). These results suggest that
adenovirus delivery of IL-2 alone is ineffective as a single injection
in the established tumors.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Tumor regression was significant in the tk alone group, but the combination tk+IL-2 therapy was more effective. Inasmuch as CD8 infiltration was 2 to 5 times more intense in the tk+IL-2 group, local IL-2 production appears to be a critical component in increasing tumor-specific lymphocytes and inducing synergistic effects on tumor regression when combined with tk treatment. It is possible that the CD8 cell infiltrate may also reflect, to some degree, an immune response against tk or ß-gal proteins. However, the CD8 values are significantly lower than the IL-2-containing groups. In ongoing work with these vectors, we have never seen a significant antitumor response to tumors treated with tk or tk+IL-2 without GCV or to tumors treated with ß-gal with or without GCV (Refs. 19, 20 and our unpublished data). Furthermore, when established tumors are treated with ADV/RSV-mIL-2 alone (without ß-gal), there is still an increased CD8 infiltrate vs. those tumors treated with ADV/RSV-ß-gal alone (our unpublished data). These findings collectively support the importance of the IL-2 and not tk or ß-gal on both the extent and tumor specificity of inflammatory infiltrate.
The results of the second tumor challenge experiment support the hypothesis that combination therapy creates a tumor-specific immune response. Effective tumor regression was seen in both tk+IL-2 and tk alone groups; however, only the combination therapy provided a tumor-specific immunity. After the various adenoviral treatments to floor of mouth tumors, SCC VII cells injected in the flanks grew as tumors in all groups except those treated with combination tk+ IL-2. The specificity of this antitumor immunity was further illustrated by the significant growth of a heterologous sarcoma cell line (RIF-1) injected concurrently in the opposite flank. Therefore, the systemic immunity in the tk+IL-2-treated animals was specific for the parental SCC VII tumor cells.
The in vitro experiments reveal a lack of direct cytotoxicity ADV/RSV-mIL-2 up to MOI values well above those achievable in vivo and demonstrate effective IL-2 expression in transduced cells. To determine whether our lack of efficacy for IL-2 alone was simply a matter of concentration of adenovirus delivered to the floor of mouth tumors, the dose-response experiment was performed. Adenovirus containing a range of IL-2 up to the maximum possible dose for 50 µl (5 x 108 pfu) was delivered to established tumors. No significant tumor regression was seen as compared with controls, and microscopic examination revealed no necrosis or other pathological effects. The combination therapy data coupled to the dose-response results strongly support the adjuvant role of IL-2 in conjunction with adenovirus tk therapy.
Based on the above results, it appears that direct tumor killing from tk (and GCV administration) provides a medium that enables enhanced local IL-2 expression to generate tumor-specific immune responses. We hypothesize that the necrosis and cellular debris from tk’s direct tumor killing results in release or concentration of tumor-specific antigens. Local antigen-presenting cells present the tumor antigens to CD8 cells and stimulate a tumor-specific immune response. The increased local production of IL-2 after intratumor adenovirus delivery enhances the complete immune response, providing both effective local tumor regression and systemic antitumor immunity. Although the data are consistent with such a conclusion, more extensive immune studies are needed to provide stronger support of this hypothesis in the head and neck tumor model.
The notable weakness of this system is a lack of complete cure despite effective tumor regression, immune system stimulation, and increased survival. We have thus far been unable to consistently prevent tumor recurrence with this strategy. The majority of animals develop recurrent floor of mouth tumors within 1–4 weeks after the response (20). As in most animal tumor models, recurrence signifies persistence of the original tumor. The persistence of tumor may be a result of incomplete tumor transduction with adenovirus vector or possibly immunoselection. Future studies will address these issues and will include both repeat adenovirus injection and the addition of other cytokines, such as granulocyte-macrophage colony-stimulating factor, which may enhance antigen presentation as well as long-term antitumor effects. The adenovirus gene therapy strategy is still in its infancy, but the findings of effective tumor regression and tumor-specific immune stimulation support the need for continued work toward future clinical application in the treatment of head and neck cancer.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In Vitro Experiments
The squamous carcinoma cell line SCC VII was used in all
experiments. Originally, the SCC VII squamous cell carcinoma arose
spontaneously in C3H/HeJ mice and has subsequently been propagated
in vivo (23). The cells were cultured in T-75 tissue culture
flasks (Corning Glass Works, Corning, NY) containing 30 cc RPMI 1640
media (Sigma Chemical Co., St. Louis, MO), 12% bovine calf serum, 1%
penicillin-streptomycin, and 1% L-glutamine. Cells were
maintained in 5% CO2 incubators. The recombinant
adenovirus containing IL-2 was added at MOI values ranging from 0–50
to the different wells just after the cells were plated. Seventy-two
hours after plating, the cells were trypsinized and counted with a
hemacytometer.
PCR reaction was performed to confirm the presence of the vector construct within the cells. Cells were digested with 1% SDS-proteinase K at 42 C for 12 h. DNA was extracted by ethanol precipitation. PCR amplification was performed with the primers RSV 270A (GACTCCTAACCGCGTACA) and ADV 3205 (GTGTTACTCATAGCGCGTAA), which are specific for the ADV/RSV-mIL-2 vector, with the following PCR protocol: 95 C for 1 min, 5x for 1 min, and 7x for 1 min for 35 cycles, and then 7x for 5 min. Five microliters of the 50-µl reaction was run on a 1% agarose gel.
In Vivo Experiments
All animal experiments were performed on C3H/HeJ mice (Jackson
Laboratories, Bar Harbor, ME) using sterile technique under a laminar
flow hood in accordance with the Johns Hopkins Animal Care and Use
Committee regulations. Mice 6–10 weeks old were anesthetized using the
inhalational agent Metophane, and a 0.1 cc suspension of 5 x
105 SCC VII cells in HBSS was injected directly into the
floor of the mouth. The animals were then maintained in standard
housing conditions.
Five days after cell implantation, mice were anesthetized with 0.5 cc avertin at a concentration of 20 mg/ml with the depth of anesthesia determined by toe pinch. A skin incision was made in the lower neck, and surgical dissection revealed the established floor of mouth tumors. Tumors were measured in three dimensions with calipers. Using a 100-µl syringe (Hamilton, Reno, NV) and 26-gauge needle, 1.0 x 109 total pfu of either ADV/RSV-tk, ADV/RSV-ß-gal control, ADV/RSV-tk+ADV/RSV-mIL-2 (2.0 x 108), or ADV/RSV-mIL-2 (2.0 x 108) + ADV/RSV-ß-gal in 50 µl solution were injected directly into the tumors. Neck incisions were closed with 4–0 silk suture (Ethicon, Somerville, NJ). Eighteen hours after adenoviral injection, the mice were administered GCV ip at a regimen of 25 mg/kg twice daily for 6 days.
For the second tumor challenge experiments, tumor sizes were assessed on the seventh day after adenoviral treatment by external caliper measurements. The right and left flanks were then injected separately with tumorigenic doses of either SCC VII or the syngeneic fibrosarcoma cell line RIF-1. Tumor growth was evaluated 1 and 2 weeks after injection.
Immunohistochemistry
For the immunohistochemistry studies, mice were killed 1
week after adenoviral treatment of the floor of mouth tumors. Tumor
sizes were measured, and the fluorescein anti-fluorescein system was
used to identify infiltrating inflammatory cells in residual tumor
masses. Frozen tissues were sectioned at 4 µm and placed on
silane-coated slides. Endogenous peroxidase activity in the tissue was
blocked by H2O2 treatment. Nonspecific binding was blocked
with PBS containing 0.3% BSA. Fluorescein-conjugated primary
monoclonal antibodies used in the assay were as follows: rat anti-mouse
CD4 (L3T4) (GIBCO BRL, Grand Island, NY), rat anti-mouse CD8a (Ly-2)
(GIBCO, BRL), mouse anti-mouse NK (5E6) (Pharmingen, San Diego, CA).
After reaction with primary antibodies, the sections were rinsed and
incubated with peroxidase-conjugated rabbit anti-fluorescein
isothiocyanate (DAKO, Carpinteria, CA) for 2 h at room
temperature. After rinsing, the slides were incubated in chromogen
solution (diaminobenzidine, 3 mg; PBS, 10 ml; 8% NiCl, 50 µl; 30%
H2O2, 1 µl) for 10 min. The reaction was stopped in
running distilled water for 1 min, and the slides were counterstained
with Nuclear Fast Red for 5 min.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication January 29, 1997. Accepted for publication March 21, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Minemura, T. Takeda, K. Minemura, T. Nagasawa, R. Zhang, R. Leopardi, and L. J. DeGroot Cell-Specific Induction of Sensitivity to Ganciclovir in Medullary Thyroid Carcinoma Cells by Adenovirus-Mediated Gene Transfer of Herpes Simplex Virus Thymidine Kinase Endocrinology, May 1, 2000; 141(5): 1814 - 1822. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Li, W. Jiang, J. S. Bishop, R. Ralston, and B. W. O'Malley Jr. Combination Surgery and Nonviral Interleukin 2 Gene Therapy for Head and Neck Cancer Clin. Cancer Res., June 1, 1999; 5(6): 1551 - 1556. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Zhang, F. H. Straus, and L. J. DeGroot Effective Genetic Therapy of Established Medullary Thyroid Carcinomas with Murine Interleukin-2: Dissemination and Cytotoxicity Studies in a Rat Tumor Model Endocrinology, May 1, 1999; 140(5): 2152 - 2158. [Abstract] [Full Text]
|
|
|
|
 |
|
 X. Danthinne, K. Aoki, A. L. Kurachi, G. J. Nabel, and E. G. Nabel Combination Gene Delivery of the Cell Cycle Inhibitor p27 with Thymidine Kinase Enhances Prodrug Cytotoxicity J. Virol., November 1, 1998; 72(11): 9201 - 9207. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Zhang, K. Minemura, and L. J. De Groot Immunotherapy for Medullary Thyroid Carcinoma by a Replication-Defective Adenovirus Transducing Murine Interleukin-2 Endocrinology, February 1, 1998; 139(2): 601 - 608. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
