In Vivo Studies of Adenovirus-Based p53 Gene Therapy for Ovarian Cancer 1

Transcription

1 GYNECOLOGIC ONCOLOGY 69, (1998) ARTICLE NO. GO In Vivo Studies of Adenovirus-Based p53 Gene Therapy for Ovarian Cancer 1 Vivian E. von Gruenigen, M.D.,*, Joseph T. Santoso, M.D.,*, Robert L. Coleman, M.D.,*, Carolyn Y. Muller, M.D.,*, David Scott Miller, M.D.,*,,2 and J. Michael Mathis, Ph.D.*,, Division of Gynecologic Oncology, *Department of Obstetrics and Gynecology, and Department of Biochemistry, Hamon Center for Therapeutic Oncologic Research, University of Texas Southwestern Medical Center, Dallas, Texas Received August 18, 1997 Objectives. To test the safety, efficacy, and toxicity of gene therapy using wild-type p53-expressing adenovirus (Ad-CMVp53) in a nude mouse model with intraperitoneal (ip) 2774 human ovarian cancer cell line that contains a p53 mutation. Study design. An initial study of adenovirus tolerance was determined in nude mice by a single ip injection of increasing doses of Ad-CMV-p53. Nude mice were implanted with an LD 100 dose of cells. To study the efficacy and specificity of Ad-CMV-p53 treatment, the mice received treatment with different adenovirus constructs. One group received Ad-CMV-p53 and another group received a control adenovirus construct, Ad-CMV- gal. To study the treatment response to Ad-CMV-p53, the mice were divided into groups and received various treatment schedules of pfu of Ad-CMV-p53. Results. The mice tolerated Ad-CMV-p53 without adverse effects at doses of pfu. The response to Ad-CMV-p53 showed significant survival duration in each dose regimen, with a survival time greater than that of untreated animals (P ). However, no statistically significant survival advantage was observed between Ad-CMV-p53- and Ad-CMV- gal-treated mice. Conclusions. These studies show that at the adenovirus dose and administration regimen used, there is effective but not specific 2774 tumor growth inhibition in vivo. Efficient introduction of biologically active genes into tumor cells would greatly facilitate cancer therapy. Thus, although promising, these results caution that much effort will be required to realize the potential for clinical application of adenovirus-based ovarian cancer gene therapy Academic Press INTRODUCTION Ovarian cancer is the most lethal of the gynecologic malignancies. Most women have advanced disease at diagnosis and require extensive debulking surgery and aggressive chemotherapy. Unfortunately, despite good initial response rates, the majority of patients will develop drug resistance, recur, and die 1 Presented at the 27th Annual Meeting of the Society of Gynecologic Oncologists, New Orleans, LA, February 10 14, To whom correspondence should be addressed at University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX Fax: (214) of their malignancy. Despite advances in operative management and chemotherapeutic agents in the past 20 years, the 5-year survival rate has remained unchanged [1]. New strategies in the management of ovarian cancer therapy are based on our emerging understanding of the molecular genetics involved in this disease [2]. Ovarian cancer arises from the accumulation of mutations in multiple combinations of genes [3]. These changes involve inherited or acquired activation of cellular proto-oncogenes and somatic or germline inactivation of tumor suppressor genes. The most extensively studied tumor suppressor gene in solid tumors is p53. The p53 gene product plays a role in normal cellular proliferation by regulating gene transcription, cell cycle control, and apoptosis [4]. It has been described in malignancies of the ovary, breast, colon, lung, esophagus, head and neck, and hematopoieic system [5]. Mutations of the p53 gene have been identified in 30 to 79% of epithelial ovarian cancers [6, 7]. In spite of cancer s polygenic nature, there is strong evidence that correction of only one of the genetic defects in cancer cells in vitro can lead to the inhibition of cell growth and reversal of tumorigenicity [8, 9]. The strategy of human gene therapy to alter the biology of human cancers is an area of investigation that shows potential for improving survival [10, 11]. The adenovirus-based vector has emerged as a leading candidate for gene therapy. Adenoviruses can be produced in high titers, can infect most cell types, and are efficient. The expression of the transferred gene is limited because the gene remains extrachromosomal and is therefore not propagated during cell division. The immunodeficient mouse is an established animal model system to study intraperitoneal dissemination in ovarian cancer [12]. Certain ovarian cancer cell lines can grow in the peritoneal cavity of nude mice and induce ascites and intraabdominal carcinomatosis. The purpose of this study is to determine the safety, efficacy, and toxicity of adenovirus-based p53 gene therapy in the treatment of ovarian cancer using a microscopic intraperitoneal disease animal model system. The results of this study, although promising, caution that much effort will be required to /98 $25.00 Copyright 1998 by Academic Press All rights of reproduction in any form reserved.

2 198 VON GRUENIGEN ET AL. realize the potential for clinical application of adenovirusbased ovarian cancer gene therapy. Animals MATERIALS AND METHODS Female BALB/c athymic nu/nu mice, 5 6 weeks old (B&K Universal Inc., Fremont, CA), were housed in microisolator cages under sterile conditions in a laminar flow unit and handled according to aseptic techniques. Tumor Cell Lines The 2774 human ovarian carcinoma cell line, derived from ascitic fluid of a patient with an untreated ovarian carcinoma of endometriod type [13] was maintained in DME/F12 media containing 10% fetal bovine serum (FBS). The adenovirus stocks were propagated in the human transformed embryonal kidney 293 cell line as described previously [9, 14]. Adenovirus Preparation As described by Santoso et al. [9], the replication-defective human adenovirus vectors were produced via homologous recombination between two transfected plasmids overlapping at the E1a-flanking region. An expression cassette driven by cytomegalovirus (CMV) was inserted in place of the E1a deletion. The titer of viral concentrations were determined using a plaque-formation assay. -Galactosidase-Staining Assay To determine the efficiency of the adenovirus vector to infect tissues, the nude mice were inoculated intraperitoneal (ip) with cells. After 3 days, the mice were treated with three daily injections of Ad-CMV- gal. One day after the final treatment, the animals were sacrificed, and peritoneal tissues were harvested and Formalin-fixed. The tissues were evaluated for the presence of Escherichia coli -galactosidase activity by staining in a solution of 5 mm K 4 Fe (CN) 6, 5 mm K 3 Fe (CN) 6, 2 mm MgCl 2, and 200 g/ml X-gal overnight at 37 C. Stained tissues were paraffin embedded, sectioned, and counterstained in eosin [15]. Tumor Growth in Vivo Freshly harvested 2774 cells were washed and resuspended in DME/F12 medium supplemented with 10% FBS. To determine the LD 100, the cell suspension was injected ip into female athymic nude mice at densities of ,1 10 6,5 10 6, or of tumor cells per 1.0 ml. For adenovirus efficacy and specificity studies, the animals were divided into three treatment arms: Ad-CMV-p53 treatment, Ad-CMV- gal treatment (serving as a control adenovirus construct), or PBS treatment (serving as a mock control treatment). The mice were monitored daily for clinical signs of tumor in the peritoneal cavity, and were sacrificed when they exhibited heavy tumor burden according to the guidelines of the Institutional Animal Care and Research Advisory Committee (IACRAC). Survival times were determined by the Kaplan Meier method, and comparisons between treated and untreated groups were made by the log-rank test. Statistical significance is based on an of 0.05 adjusted for multiple group comparisons, where appropriate, by the method of Dunn. RESULTS Tumor Growth in Vivo Escalating doses of 2774 cells were implanted ip into nude mice in order to determine the minimum concentration at which 100% of the animals acquired tumors. As shown in Fig. 1, five animals were implanted with 2774 cells at each dose, ranging from to cells per 1.0 ml. At a dose of cells, all five of the inoculated animals were dead of disease at 51 days. In the lower concentrations of 2774 cells implanted, however, each dose lost only one animal to disease. Thus, the dose of cells was selected for subsequent adenovirus treatment regimens. In mice inoculated with cells ip, microscopic implants were present upon histologic examination of the peritoneal organs and tissues (Fig. 2). After 4 to 6 weeks, abdominal distention and ascites were grossly apparent and the animals died of tumor burden. The gross intraperitoneal tumors retrieved at the time of necropsy showed a pattern of distribution similar to that found in women with ovarian cancer. These tumors were composed of nodules of neoplastic cells disoriented to one another with variation in size, shape, and mitoses. There was diffuse peritoneal studding with tumor implants, with involvement on all peritoneal surfaces. Histologic sections confirmed the widespread peritoneal tumor distribution; no metasatic spread into the pleural cavity, however, was detected. As shown in Figs. 2A, 2B, and 2C, the tumor cells covered the peritoneal surfaces of the liver, diaphragm, and ovary. Figure 2D illustrates the diffuse infiltration of tumor into the pelvic fat pad. Ad-CMV- gal Distribution in Vivo To evaluate adenovirus-mediated gene transfer to the peritoneum in vivo, nude mice were administered three daily ip injections of the Ad-CMV- gal construct, the control adenovirus vector. At 24 h after the last injection, the animals were sacrificed, and the peritoneal tissues were harvested, fixed, and stained for -galactosidase activity. The peritoneal tissues showed extensive blue staining of cells on the surface of the mesothelial lining of the peritoneal cavity, including the surface of the intestines and mesenteries (data not shown). Thus, a diffuse widespread pattern of gene transfer was obtained upon ip inoculation of adenovirus. Adenovirus Gene Therapy In order to study dose scheduling of Ad-CMV-p53, 28 nude mice were injected ip with cells. After 3 days, the

3 p53 GENE THERAPY FOR OVARIAN CANCER 199 FIG. 1. Survival time of nude mice inoculated with increasing doses of 2774 cells. Five to six-week-old nude mice were injected ip with 1.0 ml of 2774 cells containing a concentration of cells ( ), cells (Œ), cells ( ), and cells (F). Five animals were inoculated at each concentration and animal survival was followed until termination of the experiment at 112 days. animals were divided into four groups, and three dosing strategies were studied utilizing Ad-CMV-p53 at a concentration of pfu (Fig. 3). The schedules included single injection, seven daily injections and seven weekly injections. The fourth (control) group received 7 weekly ip injections of PBS. The survival of animals in each group was followed for 200 days. There was a statistically significant difference in survival (P ) among all treatment arms versus the control (PBStreated) group. However, no significant difference in survival was detected between the three adenovirus treatment regimens. To study the efficacy and specificity of Ad-CMV-p53 treatment, 30 nude mice were inoculated ip with cells. After 3 days they were divided into three groups. One treatment group received seven weekly ip injections of Ad- CMV-p53 at a concentration of pfu. A second treatment group received seven weekly ip injections of a control adenovirus construct (Ad-CMV- gal) at a concentration of pfu. The third (control) group received seven weekly ip injections of 1.0 ml PBS. As shown in Fig. 4, the survival of animals in each group was followed for 105 days. The response to Ad-CMV-p53 treatment showed a survival time that was significantly greater than that of the PBS control animals (P 0.005). Likewise, the survival in Ad-CMV- gal-treated mice compared to PBS-treated controls was also significant (P ). However, no statistically significant survival advantage was observed between treatment with the Ad-CMV-p53 versus treatment with the Ad-CMV- gal construct. To determine whether an alternate treatment regimen would show Ad-CMV-p53 specificity, 30 nude mice were implanted ip with cells, and after 3 days, were divided into three groups receiving three daily ip injections (Fig. 5A). One group received Ad-CMV-p53 at a concentration of pfu. A second treatment group received Ad-CMV- gal at a concentration of pfu. The third (control treatment) group received three daily ip injections of 1.0 ml PBS. The survival of animals in each group was followed for 105 days. Similar to the results shown in Fig. 4, the response to adenovirus treatment showed a survival time that was significantly greater than the PBS-treated control animals (P 0.041). However, there was no significant difference between treatment regimens when the experiment was terminated at 200 days. To determine whether the interval between ip implantation with 2774 cells and initiation of adenovirus therapy had an effect on survival, this interval was increased from 3 to 7 days (Fig. 5B). Seven days after tumor cell inoculation, the animals were divided into three groups receiving three daily ip injections (Fig. 5B). One group received Ad-CMV-p53 at a concentration of pfu. A second treatment group received Ad-CMV- gal at a concentration of pfu. The third (control treatment) group received three daily ip injections of 1.0 ml PBS. The survival of animals in each group was followed for 105 days. Although the response to adenovirus treatment showed a survival time that was significantly greater

4 200 VON GRUENIGEN ET AL. FIG. 2. Histopathologic sections of peritoneal organs and tissues harvested from a nude mouse 40 days postimplantation with cells. Nodules of tumor implants are shown on: (A) liver, (B) diaphragm, (C) ovary, fallopian tube, and uterus, and (D) pelvic fat pad.

5 p53 GENE THERAPY FOR OVARIAN CANCER 201 FIG. 3. Survival time of nude mice treated with various doses of Ad-CMV-p53. Five- to six-week-old nude mice were injected ip with cells. Three days after tumor implantation, the animals were treated with a single injection of pfu Ad-CMV-p53 ( ), with seven daily injections of pfu Ad-CMV-p53 (Œ), with seven weekly injections of pfu Ad-CMV-p53 ( ), or with seven weekly injections of PBS (F). Seven animals were included in each treatment arm and survival differences were followed until termination of the experiment at 200 days. than the PBS-treated control animals (P ), there was no significant difference between Ad-CMV-p53 and Ad-CMV- gal treatment. When the study was terminated at 200 days, the adenovirus treatment continued to show a survival time that was significantly greater than the PBS-treated control animals (P 0.045). DISCUSSION Early clinical trials of virus-mediated gene therapy have begun in head and neck squamous carcinomas as well as non-small cell lung carcinomas [16]. Results of human clinical trials directed at inherited disease such as ADA deficiency, familial hypercholesterolemia, and cystic fibrosis, although promising, have yielded inconclusive results [5]. Many obstacles must still be overcome before the clinical potential of human gene transfer strategy can be fully achieved. These obstacles include inconsistent results between experimental animal models and early human trials, production problems in the generation of clinical grade materials for clinical trials, delivery of the therapeutic gene selectively to target tumor cells, and unpredicted toxicities in some early human studies. Other development problems facing the clinical application of gene therapy are specific to each vector chosen, the gene(s) to be delivered, the dosing strategy used, and the route of vector administration. The adenovirus vector delivering the expression cassette containing the wild-type p53 gene driven by a CMV promoter has been administered in vivo including intratracheal [17], intravenous [18], intramuscular [19], intratumoral [20], portal venous injection [21], and stereotactic inoculation in the brain [22]. The role of adenovirus-based p53 gene therapy in the treatment of ovarian cancer was addressed using an in vivo mouse model system. The intraperitoneal route of administration of the adenovirus vector in this study is a novel approach for the treatment of human ovarian cancer metastasis, since iv administration is much less efficient [23]. The targeted role for adenovirus-based p53 gene therapy in this study is treatment of advanced invasive ovarian carcinoma. Patients with minimal residual disease after primary surgery demonstrate the best response rates to conventional chemotherapy but survival remains poor. Since bulky tumor models have demonstrated poor vector penetration and suboptimal tumor responses, the microscopic disease ovarian cancer model should hold the most promise for clinical applicability. The immunodeficient mouse is an established animal model system to study intraperitoneal dissemination in ovarian cancer [12]. Certain ovarian cancer cell lines grow in the peritoneal cavity of nude mice and induce ascites and intraabdominal carcinomatosis. In the present study, the PBS control mice sacrificed 3 and 7 days later showed microscopic serosal im-

6 202 VON GRUENIGEN ET AL. FIG. 4. Survival time of nude mice treated with Ad-CMV-p53 or with Ad-CMV- gal. Nude mice were injected ip with Three days after tumor implantation, the animals were treated with seven weekly injections (shown as arrows) with pfu Ad-CMV-p53 ( ), with seven weekly injections of pfu Ad-CMV- gal (Œ), or with seven weekly injections of PBS (F). Ten animals were included in each treatment arm and survival differences were followed until termination of the experiment at 105 days. plants within the peritoneum. The nonsacrificed control mice continued to develop clinical tumor burden, observed as ascites fluid with gross carcinomatosis throughout the peritoneal cavity. Therefore, to model microscopic cancer, a 3-day postinoculation time period was established before any treatment modality was begun. Toxicity can differ depending on the route of administration of any biologic or therapeutic agent and must therefore be examined. Our results indicate that intraperitoneal administration of Ad-CMV-p53 or the control Ad-CMV- gal construct is nontoxic to the animal. Doses had no clinical ill effects on the animals as demonstrated by 100% survival up to 60 days. No animal developed any clinically relevant symptoms, and all demonstrated normal growth and development. Our necropsy slides were reviewed by a pathologist and there were no histological signs of peritonitis. Likewise, repeated injections of the Ad-CMV-p53 did not induce non-tumor-related deaths. Zhang et al. [17] and Clayman et al. [24] have used the Ad-CMV-p53 vector in mice to simulate lung cancer and head and neck squamous carcinomas treatment. Encouragingly, the results of these two studies show significant adenovirus inhibition of tumor growth. The result of our study using Ad- CMV-p53 to inhibit tumorigenesis in the microscopic intraperitoneal ovarian cancer mouse model also shows promise. Dosing strategies were first addressed in Fig. 3. Initial adenovirus doses, using a single injection of Ad-CMV-p53 ( pfu), seven daily Ad-CMV-p53 injections, or seven weekly Ad-CMV-p53 injections, showed no statistical difference in survival among the treatments. There was, as expected, a significant difference in survival between the adenovirustreated animals and the control (PBS-treated) animals. These data suggest that the maximal effect of Ad-CMV-p53 may be achieved on the initial dose and subsequent exposure to the adenovirus construct is without additional benefit. Unexpectedly, the results of our study show inhibition of tumorigenesis of the 2774 ovarian cancer cell line in both the Ad-CMV-p53-treated animals and the Ad-CMV- gal-treated animals compared to the control (PBS-treated) animals. This was a reproducible result and was not dependent on the number of treatments or dosing strategy used. Interestingly, other in vivo studies have not reported Ad-CMV- gal as a control adenovirus vector [17, 20, 24 26]. This may be an important observation, since the expression cassette and vector are the same in both adenovirus constructs used this study, and only differ in the gene which is expressed. Our data suggest that suppression of tumorigenesis may be independent of the expressed gene, either p53 or -galactosidase in this microscopic disease model. The data presented by Clayman et al. [24] indirectly support this hypothesis, since inhibition of tumorigenesis in the cell lines used was independent of the mutation

7 p53 GENE THERAPY FOR OVARIAN CANCER 203 FIG. 5. Survival time of nude mice treated with Ad-CMV-p53 or with Ad-CMV- gal. (A) Nude mice were injected ip with cells. Three days after tumor implantation, the animals were treated with three daily injections (shown as arrows) with pfu Ad-CMV-p53 ( ), with pfu Ad-CMV- gal (Œ), or with three daily injections of PBS (F). Ten animals were included in each treatment arm and survival differences were followed until termination of the experiment at 105 days. (B) Seven days after tumor implantation, the animals were treated with three daily injections (shown as arrows) with pfu Ad-CMV-p53 ( ), with pfu Ad-CMV- gal (Œ), or with three daily injections of PBS (F). Ten animals were included in each treatment arm and survival differences were followed until termination of the experiment at 105 days.

8 204 VON GRUENIGEN ET AL. status of the endogenous p53 gene. An alternative explanation for these findings is the possibility that an inflammatory response to adenovirus injection caused a nonspecific inhibition of tumor cells implantation and growth. It is also possible that overexpression of a foreign protein (either human p53 or E. coli -galactosidase) by the adenovirus vector may have induced an inflammatory response that inhibited residual tumor formation. These results indicate that the in vivo immune response to standard titers of adenovirus must explored in future studies before adenovirus-based gene therapy may be clinically applicable in the intraperitoneal treatment of microscopic residual ovarian carcinoma. The results observed in this study have raised additional issues that must be addressed. Different human cancer cell lines and different p53 mutations may not be created equal and may respond differently to adenovirus-based p53 gene therapy. This may be due to different p53 mutations and the fact that the 2774 cell line has a mismatch repair defect. The adenovirus constructs prepared by our laboratory and those prepared by others may not be directly comparable. It is possible that differences in titers or purity of the adenovirus constructs used may be responsible for the inconsistencies seen among each of the published reports. Finally, the issue of local immune response to the adenovirus construct must be further examined, particularly in the intraperitoneal cancer model. As our understanding of the molecular genetics of human cancer progresses, we must also proceed to undertake the challenges presented by these initial in vivo gene transfer experiments prior to obtaining the reality of clinically useful human gene therapy strategies. REFERENCES 1. Miller DS, Spirtos NM, Ballon SC, Cox RS, Soriero OM, Teng NNH: Critical reassessment of second-look exploratory laparotomy for epithelial ovarian carcinoma. Cancer 69: , Mathis JM, Muller CY, von Gruenigen VE, Miller DS: Biologic Strategies in the Gynecologic Cancers, in Langdon S, Miller W, Berchuck A (eds): Biology of Female Cancers. Boca Raton, FL, CRC Press, 1997, in press 3. Bast RC Jr., Boyer CM, Jacobs I, Xu FJ, Wu S, Wiener J, Kohler M, Berchuck A: Cell growth regulation in epithelial ovarian cancer. Cancer 71: , Harris CC, Hollstein M: Clinical implications of the p53 tumor-suppressor gene. N Engl J Med 329: , Crystal RG: Transfer of genes to humans: early lessons and obstacles to success. Science 270: , Kohler MF, Marks JR, Wiseman RW, Davidoff AM, Clarke-Pearson DL, Soper JT, Bast RC Jr, Berchuck A: Spectrum of mutation and frequency of allelic deletion of the p53 gene in ovarian cancer. J Natl Cancer Inst 85: , Kupryjnczyk J, Thor AD, Bauchamp R, Merritt V, Edgerton SM, Bell DA, Yandell DW: p53 gene mutations and protein accumulation in human ovarian cancer. Proc Natl Acad Sci USA 90: , Fujiwara T, Grimm EA, Roth JA: Gene therapeutics and gene therapy for cancer. Curr Opin Oncol 6:96 105, Santoso JT, Tang DC, Lane SB, Hung J, Reed DJ, Muller CY, Carbone DP, Lucci JA, Miller DS, Mathis JM: Adenovirus-based p53 gene therapy in ovarian cancer. Gynecol Oncol 59: , Mulligan RC: The basic science of gene therapy. Science 260: , von Gruenigen VE, Muller CY, Miller DS, Mathis JM: Gene therapy in ovarian cancer. Contemp Obstet Gynecol 42(2): , Christman JE, Miller DS, Coward P, Smith LH, Teng NNH: Study of selective cytotoxic properties of cationic, lipophilic mitochondrial-specific compounds in gynecologic malignancies. Gynecol Oncol 39:72 79, Freedman R, Pihl E, Kusyk C, Gallager HS, Rutledge F: Characterization of an ovarian carcinoma cell line. Cancer 42: , Graham FL, Smiley J, Russell WS, Nair R: Characteristics of a human cell line transformed by DNA from human adenovirus type 5. J Gen Virol 36:59 74, Alam J, Cook JL: Reporter genes: application to the study of mammalian gene transcription. Anal Biochem 188: , Roth JA, Nguyen D, Lawrence DD, Kemp Bl, Carrasci CH, Ferson DZ, Hong WK, et al: Retroviral-mediated wildtype p53 gene transfer to tumors of patients with lung cancer. Nat Med 2(9): , Zhang WW, Fang X, Mazur W, French BA, Georges RN, Roth JA: High-efficiency gene transfer and high-level expression of wild-type p53 in human lung cancer cells mediated by recombinant adenovirus. Cancer Gene Ther 1:5 13, Lemarchand P, Jones M, Yamada I, Crystal RG: In vivo gene transfer and expression in normal uninjured blood vessels using replication-deficient recombinant adenovirus vectors. Circ Res 72: , Ragot T, Vincent N, Chafey P, Vigne E, et al: Efficient adenovirusmediated transfer of a human minidystophin gene to skeletal muscle of mdx mice. Nature 361: , Liu TJ, Zhang WW, Taylor DL, Roth JA, Goepfert H, Clayman GL: Growth suppression of human head and neck cancer cells by the introduction of a wild-type p53 gene via a recombinant adenovirus. Cancer Res 54: , Jaffe HA, Danel C, Longenecker G, Metzger M, et al: Adenovirusmediated gene transfer and expression in normal rat liver. Nat Genet 1: , LeGal La Salle G, Robert JJ, Bernard S, Ridoux V, et al: Defective and non-defective adenovirus vectors for expressing foreign genes in vitro and in vivo. Gene 101: , Tang DC, Johnston SA, Carbone DP: Butyrate-inducible and tumorrestricted gene expression by adenovirus vectors. Cancer Gene Ther 1:15 20, Clayman GL, El-Naggar AK, Roth JA, Zhang WW, Goepfert H, Taylor DL, Liu TJ: In vivo molecular therapy with p53 adenovirus for microscopic residual head and neck squamous carcinoma. Cancer Res 55:1 6, Fujiwara T, Grimm EA, Mukhopadhyay TEA, De WC, Owen-Shaub LB, Roth JA: A retroviral wild-type p53 expression vector penetrates human lung cancer spheroids and inhibits growth by inducing apoptosis. Cancer Res 53: , Yang C, Cirielli C, Capogrossi MC, Passaniti A: Adenovirus-mediated wild-type p53 expression induces apoptosis and suppresses tumorigenesis of prostatic tumor cells. Cancer Res 55: , 1995