© 2000 Nature America, Inc. 0929-1903/00/$15.00/ϩ0 www.nature.com/cgt Progress in cancer gene therapy David T. Curiel,1 Winald R. Gerritsen,2 and Mark R. L. Krul3 1Division of Human Gene Therapy, Departments of Medicine, Pathology, and Surgery, Gene Therapy Center, University of Alabama, Birmingham, Alabama 35294; 2Academic Hospital Vrije Universiteit Amsterdam, Department of Medical Oncology, Amsterdam, The Netherlands; and 3NDDO Oncology, Section of Drug Development Strategy, Amsterdam, The Netherlands. The “First International Symposium on Genetic Anticancer Agents,” which took place in Amsterdam on March 8–9, 2000, served as a forum to review the results of preclinical and clinical gene therapy studies for cancer endeavored to date. Despite the fact that gene therapy was initially conceptualized as an approach for inherited genetic disease, it is currently finding its widest employ for treating neoplastic disorders. In this regard, more than 70% of patients treated to date in human clinical gene therapy protocols have been in the context of anticancer regimens.1 Of note, the application of gene therapy for cancer has proceeded from the same rational basis as was originally conceptualized for inherited genetic disorders. Specifically, the molecular basis of these disorders is increasingly being understood, therapeutic genes are available, and alternative therapies are often lacking. Most recently, the field of gene therapy has enjoyed the realization of the first incontrovertible evidence of clinical benefit, for hemophilia and cardiovascular disease, in its first 15 years of human application.2 This recent recognition of the potential power of gene therapy, and the current lack of realizing such ends for neoplastic disease, has led to a reassessment of the field. Such a critical analysis is a necessary step in defining the means to progress the technology toward achieving the potential benefits of gene therapy for cancer. Cancer Gene Therapy (2000) 7, 1197–1199 Key words: Gene therapy; preclinical research; clinical trials; vectors; targeting; replication. hereas human clinical gene therapy trials for can- an analysis of this phenomenon whereby, in some in- Wcer have not yielded clear benefit to date, these stances, the basis of tumor cell resistance can be defined. studies have served to highlight the deficiencies of For instance, one of the major vector systems employed current approaches. In this regard, a number of cancer for cancer gene therapy is the recombinant adenovirus gene therapy approaches have been developed based on (Ad).3 Analysis of a variety of primary tumors has direct correction of cancer-relevant genetic lesions. demonstrated profound deficiencies in the primary re- These “mutation compensation” approaches represent a ceptor for Ad, Coxsackie and Ad receptor (CAR), which diversity of specific interventions, all designed to rectify serves to explain the observed tumor cell vector resis- the molecular lesions etiologic of neoplastic transforma- tance in human clinical trials. tion/progression, and have demonstrated efficacy in nu- merous reports in the context of in vitro and in vivo Gene delivery model systems. Application of these approaches for human disease requires direct, in situ delivery of these The recognition that CAR deficiency is a nearly univer- anticancer genes. Furthermore, this must be achieved in sal feature of epithelial neoplasms has predicated the the context of a plurality of the neoplastic cells. This need for adenoviral vectors capable of “CAR-indepen- requirement predicates the need for gene delivery vec- dent” gene delivery.4 Strategies that have been endeav- tors capable of efficient and selective gene delivery to ored to achieve this employ retargeting complexes as tumor cells in the context of in vivo gene delivery well as genetic capsid modifications. The ability to schemas. In this regard, one of the major findings of achieve CAR-independent gene delivery has allowed human clinical gene therapy trials for cancer has been a effective infection of otherwise refractory tumor cells. disappointingly low level of achieved gene transfer. This has allowed direct therapeutic gain in the context of Thus, despite demonstrated efficacy in model systems, murine models of human neoplastic disease. These new the primary tumor has proved to be relatively refractory vector paradigms are being evaluated currently in the to the same vector systems. These results have provoked context of tumor clinical trials. Of note, it was the earlier generation of human clinical trials that first revealed the biological limits of current generation vectors. On this Address correspondence and reprint requests to Dr. Mark R. L. Krul, basis, vector design modifications were endeavored that NDDO Oncology, Section Drug Development Strategy, Amstelveenseweg allowed vector improvement with a direct gain in model 641, 1081 JD Amsterdam, The Netherlands. systems of cancer.4 Clearly, future human clinical trials Cancer Gene Therapy, Vol 7, No 8, 2000: pp 1197–1199 1197 1198 CURIEL, GERRITSEN, KRUL, ET AL: PROGRESS IN CANCER GENE THERAPY with the advanced-generation vectors will provide the cal results have been achieved, confirming the proof of ultimate test of efficacy. In any event, the role of direct principle, the induction of clinical responses in humans clinical translation in allowing rational advancement of still has to be confirmed. cancer gene therapies is clearly illustrated. Immunotherapy Replicative vectors So far however, most cancer gene therapy trials involve The recognition that advanced-generation vectors are immunotherapeutic approaches. Based on the increas- limited in their in vivo transduction efficacy has also led ing knowledge regarding tumor-associated antigens to the development of novel anticancer gene therapies (Ags) and Ag presentation, new approaches for gene embodying the principles of posttransductional amplifi- immunotherapy are being designed. Research on den- cation. In this scheme, an initial transduction event dritic cells is aimed at specific transduction of these provokes a biological cascade whereby anticancer ef- Ag-presenting cells with integrins or CD40 as target Ags, fects, over and above the transduced cells, are achieved. to further improve the functioning of these cells. The One such approach employs replicating viral agents. A first clinical results with cDNA vaccination against car- variety of viruses have been employed to this end, cinoembryonic Ag, with dendritic cells loaded with including parvovirus,5 herpes viruses,6 reovirus,7 and mRNA or with tumor cells transduced with the granu- Ad.3 The unifying feature of these agents is that the locyte-macrophage colony-stimulating factor gene, dem- antitumoral effect is achieved directly by virtue of the onstrated limited clinical benefit in patients with end- replicative cycle of the virus. Such antitumoral “oncoly- stage disease. In a melanoma trial with autologous, sis” is ideally achieved in a fashion whereby the virus irradiated tumor cells transduced with granulocyte-mac- replicates selectively in tumor cells. A conceptually rophage colony-stimulating factor, 20% of patients who received three vaccinations were still alive after a fol- attractive aspect of these systems is their intrinsic ability 11 to achieve an amplified effect via the capacity of infec- low-up of 3–5 years. tious viruses to lateralize within the context of solid Clinical trial design tumors. On this basis, replicative viral agents have been recognized as agents of exceptional promise and have In the paradigms presented, information gained in hu- been rapidly translated into human clinical gene therapy man clinical trials has allowed rational steps to improve trials. As noted previously in the context of the issue of the design of gene-based anticancer agents. Whereas vector efficacy, clinical trials carried out to date with this recognition argues for stringent clinical evaluation replicative viral agents have revealed limitations in these of new agents, it also mandates that maximal informa- systems that were not fully noted in model systems. In tion be derived from these trials. Heretofore, the real- the first instance, tumor cell resistance to viral infection ization of such key information has not been fully based on viral receptor deficiency presents a barrier to feasible based on the crude and invasive methods of viral lateralization. Thus, vector modifications will be gene transfer/expression employed. A number of groups required to address this limit of effective amplification. have recently developed noninvasive imaging methods In addition, the achievement of precise replicative spec- as a means to monitor heterologous vector-mediated ificity has been more difficult to achieve than initially gene expression.12 These systems offer a number of envisioned. It must therefore be recognized that the advantages for this type of analysis and potentially rapid translation of these agents into clinical trials will provide a truly dynamic method to achieve the key goal mainly serve to highlight key functional areas warranting of interval data, short of direct therapeutic benefit. The address. For advanced-generation vectors, it may be evaluation of these systems in human trials remains to be anticipated that such design reconfigurations will yield a endeavored. Nonetheless, such systems may prove piv- direct gain in terms of therapeutic results. otal in guiding the empiric development of future gen- erations of anticancer agents. Prodrug therapy Another approach to overcome the