Live Viruses in Cancer Treatment

November 1, 2002

Dr. Nemunaitis gives a scholarly and informative historical review of antineoplastic viral therapy using recombinant DNA biotechnologies. The field predates the polymerase chain reaction and restriction enzymes; it has its roots in observations by Jenner and experiments that are over 100 years old.

Dr. Nemunaitis gives a scholarly and informativehistorical review of antineoplastic viral therapy using recombinant DNAbiotechnologies. The field predates the polymerase chain reaction andrestriction enzymes; it has its roots in observations by Jenner and experimentsthat are over 100 years old.

The fundamental assumption underlying the use of viruses to treat cancer isthat the genes they employ in their own life cycle can be redirected towardkilling cancer cells. Viruses have evolved in our biosphere to be highlyeffective at killing host cells. Viruses express the genes involved in thatprocess in a robust and precise way, making them powerful immunologiccostimulatory molecules as antigens. Our species’ only countermeasure has beento develop vigorous immune responses upon infection. Thus, using viral genes as"associated antigens" for presenting human tumor-associated antigensto the immune system is a pharmacologic exploitation of the power thatattenuated viruses possess in vaccination.

Nemunaitis reviews two overall therapeutic approaches: (1) the use of virusesas platforms for inducing immune responses against infected cells (often livevirus vaccines containing tumor associate antigens), and (2) the use ofoncolytic viruses that infect and kill cancer cells with a selective cytoxicityfavoring the killing of tumor cells rather than adjacent normal cells.

Potential Advantages

The attraction of viral therapy for experimental therapeutics derives fromtwo properties of viruses that are seen regardless of whether they are directlyinfecting and killing tumor cells or generating T cells and antibodies againstthe tumor after a vaccination approach.

The first rationale comes from the fact that either strategy can kill tumorcells independent of where they are in the cell cycle. Unlike the majority ofFood and Drug Administration (FDA)-approved antineoplastic drugs currently usedby clinicians, viruses can kill tumor cells anywhere in the cell cycle. Manycancers have developed genetic mechanisms to escape apoptosis but have a lowpercentage of cells at any given time in DNA replication. The "kineticresistance" of solid tumors—unlike leukemias and lymphomas—is not aproblem for current strategies in the clinical development of viraltherapies.

The second rationale derives from multiple laboratory observations (cited byNemunaitis) that historically chemotherapy-refractory tumor types are sensitiveto the cell-killing effects of certain classes of viruses. For example,selective "oncolytic" herpesviruses can kill chemotherapy-refractorybrain tumor cells, and "oncolytic" adenoviral vectors can replicateand kill chemotherapy-resistant head and neck tumor cells and hormone-refractoryprostate cancer cells.[1-5] Importantly, the permutations of genetic-engineeringpossibilities for viruses that confer highly selective replication in tumorcells but not normal cells continues to expand. For example, the uniqueexpression of the prostate-specific antigen (PSA) gene and its upstreamregulatory DNA sequences have permitted the construction of adenoviral vectorsthat selectively replicate and kill PSA-expressing cells.[3,4]

Several groups have convincingly shown proof of concept in the intratumoralinjection of tumor-selective viral vectors in the clinic. The ONYX-015 virus hasclear clinical activity in recurrent head and neck cancer.[2,3] Recently, in thefirst application of intratumoral viral "brachytherapy" in patientswith local recurrence of prostate cancer following external-beam radiotherapy,DeWeese and colleagues reported evidence of > 50% PSA responses, some ofwhich were durable.[5] Equally as important, these investigators confirmedseveral rounds of tumor-killing viral replication with very tolerable patientside effects of mild prostatitis and fever. The Hopkins group also observed thatpreexisting levels of antibodies to adenovirus type 5 apparently did not blockreplication of the virus when applied intratumorally under stereotacticultrasound guidance—although, after treatment, patients clearly developed hightiters of antibodies to the virus.[5]

Approaches to brachytherapy in head and neck and prostate cancer arecontinuing along paths toward phase III registration trials. New tumor target-specificviruses are also being created on the basis of sophisticated, molecularproof-of-concept, phase I trials, which have demonstrated not just safety, butevidence of viral replication, killing by viral lysis, and objective measurableclinical cancer burden reductions. A bladder cancer-specific (uroplakin IIpromoter) replication-selective adenoviral vector has been developed that mightbe used under cystoscopy, for example.[6]

Scientific and Clinical Challenges

There are several problems and unaddressed issues in the clinical developmentof oncolytic viruses. The first one has to do with the biology of adenovirusreplication. The receptor involved in viral entry—the coxsackie-adenovirusreceptor (CAR)—is downregulated or absent in many tumor types. This may confera resistance mechanism at the cell entry point. It could be difficult to getenough viral particles into certain tumor types to create a tumor-lyticinfection. Tumor resistance may emerge from the growth of CAR-negative clones.Some groups are attempting to reengineer these viruses to dock and enterreceptors that are overexpressed by certain tumor types. These geneticcountermeasures to viral therapy resistance await phase I/II proof-of-principaltrials.

The multidimensionality of genetically altering these viruses for humanclinical trials has not been lost upon Federal regulatory agencies in apost-9/11 world. Clinical development in human gene therapy after the tragicdeath of Jessie Gelsinger (in a clinical trial of adenoviral vector genetherapy)—and in a state of bioterrorism awareness—is likely to put theburden on academic investigators and biotechnology companies to carefully andrepeatedly explain the scientific rationale of approaches and to definesafeguards in both patient informed consent and clinical trial conduct.

Lastly, it remains unclear whether oncolytic viruses can be given safely athigh enough doses intravenously to "get into" metastatic sites andinfect and destroy metastases through rounds of replication. Overcoming thishurdle is clearly the great unmet clinical need. In preclinical models ofrodents, tail vein injections of oncolytic viruses can destroy xenografts atdistant sites. But in the clinic, these viruses must pass through the humanliver and the reticuloendotheilal system complexed with neutralizing antibodiesafter initial infusion. Both clinicians and researchers await with greatinterest the clinical results of dose-escalation trials assessing PSA-selectiveoncolytic adenoviruses given intravenously in patients with hormone-refractoryprostate cancer.

Future Directions

A major question for the field of oncolytic viruses concerns whether theywill be useful in brachytherapy, to improve the therapeutic index ofradiotherapy. FDA registration trials will find approval in better local controland delays in time to progression that confer survival and quality-of-lifeimprovement. The other fundamental question is whether the next generations ofgenetically engineered viruses can ultimately be programmed with"countermeasures" to be exquisitely selective for replication in tumorcells only, and be given effectively as intravenous oncologic drugs.

Some optimism appears warranted. In vivo clinical responses to intratumortreatment with early generations of these viruses have been seen in wellconducted phase I and phase II trials.[1-3] That is a big step in a young field.The approach warrants further investigation in selected areas of medicaloncology. Positive clinical trial results in viral therapy have been seen inpatients for whom the entire 20th century of chemotherapeutic drug developmentoffers no more promise. Despite the global political electricity currentlysurrounding viruses, the power of genetically engineered, tumor-selective viralvectors to kill chemotherapy- and radiotherapy-resistant clones is anaesthetically appealing and scientifically sound direction for cancer research.


1. Bischoff JR, Kirn DH, Williams A, et al: An adenovirus mutant thatreplicates selectively in p53-deficient human tumor cells. Science 274:342-343,1996.

2. Khuri FR, Nemunaitis J, Ganly I, et al: A controlled trial of intratumoralONYX-015, a selectively-replicating adenovirus, in combination with cisplatinand 5-fluorouracil in patients with recurrent head and neck cancer. Nat Med6:862-863, 2000.

3. Nemunaitis J, Khuri F, Ganly I, et al: A phase II trial of intratumoralinjection of ONYX 015 in patients with refractory head and neck cancer. J ClinOncol 19:289-298, 2001.

4. Rodriguez R, Schuur ER, Lim HY, et al: Prostate attenuated replicationcompetent adenovirus (ARCA) CN706: A selective cytotoxic for prostate-specificantigen-positive prostate cancer cells. Cancer Res 57:2559-2563, 1997.

5. DeWeese TL, van der Poel H, Li S, et al: A phase I trial of CV706, areplication-competent, PSA selective oncolytic adenovirus, for the treatment oflocally recurrent prostate cancer following radiation therapy. Cancer Res61:7464-7472, 2001.

6. Zhang J, Ramesh N, Chen Y, et al: Identification of human uroplakin IIpromoter and its use in the construction of CG8840, a urothelium-specificadenovirus variant that eliminates established bladder tumors in combinationwith docetaxel. Cancer Res 62:3743-3750, 2002.