The explosive increase in the apparent incidence of prostate cancer
in the United States (which is due, in large measure, to wider efforts
at early detection) has been accompanied by a dramatic stage migration,
which can also be attributed to the increased use of prostate-specific
antigen (PSA). Testing with PSA is used not only for the detection of prostate
cancer but also as a measure of both response to therapy and progression
after therapy. As a consequence, a lead-time bias in the identification
of progressive disease has been introduced, which affects all treatment
modalities, including surgery, radiation therapy, and systemic therapy,
such as androgen deprivation. Consequently, a large group of patients have
been identified who have early prostate cancer or in whom the only evidence
of progressive disease is a rising PSA. This group of patients present
a unique opportunity for the development of novel therapeutic approaches
such a gene therapy that are likely to be more efficacious in patients
with minimal tumor burden.
The rationale for the development of gene therapy in prostate cancer,
potential situations in which it may be utilized, and its challenges and
promises are carefully laid out in the article by Drs. Harrison and Glode.
The authors outline the various components required for successful therapy
involving the introduction of genetic material into patients' cells. These
include construction of the appropriate vector for delivery of the gene
of interest, selection of an appropriate promoter to ensure that adequate
expression of this gene occurs, and, most importantly, identification of
the gene of interest.
Although numerous investigators are exploring nonviral means of gene
transfer, it is fair to say that most approaches have taken advantage of
the capacity of viruses to introduce genetic material into a cell. The
authors correctly state that most viruses in current use are replication-deficient.
However, it must be pointed out that a precedent now exists for mutant
viruses that maintain the capacity for replication but retain specificity
for targeted malignant cells. The first example of such a virus was recently
reported. Bischoff et al showed that an E1B-deficient adenovirus can
infect and lyse only cells that harbor a mutant p53 gene.
Although, technically, the use of this virus would not constitute gene
therapy, since novel genetic material is not being introduced, the net
effect, that of genetic selection of a target cell population, is the same.
Contrary to the replication-deficient adenovirus used in gene transfer
studies, the utility of the E1B-deficient virus would be that more than
a limited number of cells could be infected and potentially destroyed.
The authors point out that one advantage of retroviral vectors is the
possibility of selective targeting to the dividing tumor cell. However
it is not entirely clear that overexpression (eg, of a wild-type tumor-suppressor
gene) in normal tissue is necessarily deleterious. Nevertheless, the authors
are correct in pointing out that one of the safety concerns in gene therapy
is the possibility of inadvertently targeting normal as opposed to neoplastic
cells. Extensive preclinical and phase I testing will be required to determine
whether or not such an event would necessarily be detrimental.
The authors suggest that one of the unique aspects of prostate cancer
is that, with the use of prostate-specific promoters, only prostatic tissue
(both benign and malignant) could be targeted. Given that an innocent bystander
effect involving normal prostatic tissue is not likely to have significant
clinical repercussions and that patients may have already received definitive
therapy for localized disease, targeting of both benign and malignant prostatic
tissue would be acceptable.
Candidate Genes in Prostate Cancer
The authors have carefully listed features unique to gene therapy in
prostate cancer, including unique clinical features of prostate cancer,
as well as an overview of candidate genes. Regarding the utility of inserting
wild-type tumor-suppressor genes, the authors appropriately note that consistent
deletions of chromosomal material in patients with localized prostate cancer
are not well identified. Nevertheless, some candidate genes do exist. In
particular, p53 mutations have been found in anywhere from 20% to 75% of
prostate cancer specimens, albeit at the lower end of the spectrum in patients
with localized disease.
Furthermore, novel techniques, including comparative genomic hybridization,
show great promise for identifying regions of the chromosome with frequent
amplifications or deletions. More recently, the area distal to the BRCA-1
(breast-ovarian cancer susceptibility) gene has been shown to be deleted
in 70% of patients with localized prostate carcinoma.
Other Potential Targets
In addition to the insertion of wild-type tumor-suppressor genes, efforts
to inactivate specific genes or gene products, in particular, oncogenes,
must be considered a form of gene therapy. In this regard, the utility
of antisense or ribozyme technology to destroy or inactivate specific genetic
sequences must be considered. In prostate cancer, potential targets include
oncogenes, such as ERB-B2 and EGFR, although their exact role in the pathogenesis
of prostate cancer is not well understood. Humanized antibodies to these
oncogene products exist as well, and are the basis for ongoing trials.
Also discussed in the article are direct toxins targeted at specific
cells. These approaches frequently capitalize on a bystander effect. In
this regard, the E1B-deficient adenovirus noted above should be included
as an agent capable of infecting and affecting more than a single cell.
The authors also describe, in general terms, the potential efficacy
of targeting local growth factors, such as basic fibroblastic growth factor.
Although there is no question that proangiogenic molecules, such as fibroblast
growth factor (FGF) and perhaps vascular endothelial growth factor (VEGF),
are involved in the progression of prostate cancer, it is not at all clear
that the technology currently exists to target the gene sequences controlling
the generation of these molecules. Other approaches, including the use
of suramin (which is known to bind FGF), may have a similar effect. Other
antiangiogenic molecules under development may be easier, less expensive,
and perhaps as efficacious or more efficacious than gene therapy.
The authors discuss the utility of promoting the expression of various
cytokines as a means of soliciting or enhancing an immunologic response.
Granulocyte-macrophage colony-stimulating factor (GM-CSF) is felt to be
especially promising, and is the basis for one therapeutic gene trial in
prostate cancer, referenced by the authors.
Preclinical animal data available on GM-CSF should be tempered by in
vitro data suggesting that it appears to stimulate, rather than inhibit,
several prostate cancer cell lines. Clearly, bench-top observations
made in cell lines and tissue culture cannot be directly extrapolated to
the clinical setting. Nevertheless, these data must be taken into account
as therapeutic trials are designed.
It should also be pointed out that there are other ways of obtaining
an immune response without the expense and difficulty of transfecting cells
with a particular gene of interest. Numerous vaccine trials utilizing purified
antigens, prostate cancer cells, and/or immune effector cells are currently
underway and merit evaluation as well.
Guarding Against Unrealistic Expectations
The authors have carefully delineated the advantages and unique aspects
of prostate carcinoma that may allow for the successful administration
of gene therapy, but they appropriately warn about harboring unrealistic
expectations about the early phases of these trials. They point out that
perhaps the most appropriate setting in which to study immune therapy is
in patients with minimal or microscopic disease. On the one hand, they
argue that the long latency of prostate cancer offers a large window of
opportunity for the use of gene therapy. However, if patients with minimal
disease are to be treated, then trials will, by necessity, be large and
will require lengthy maturation time before any determination can be made
regarding an impact on end points.
Certainly, in this relatively early phase of gene therapy, it is prudent
to test gene therapy in localized disease prior to undertaking systemic
therapy. The authors hold that the anatomic considerations of prostate
cancer make it an ideal tumor to treat. However, accessibility of tissue
should not be confused with ease of administration. Although local therapy,
such as brachytherapy and cryotherapy, have been successfully used in the
context of real-time imaging of the prostate with either ultrasound or
CT, the fact remains that these modalities can target a reasonable volume
of tissue. By contrast, injections of nonreplicating virus into the prostate
cannot be hoped to have significantly more activity than in the immediate
microenvironment. Furthermore, if nonreplicating viruses are used, repeated
injections may be required, which is not necessarily a pleasant prospect
The authors correctly conclude that gene therapy approaches are less
likely to work in patients with overtly metastatic disease. It is not entirely
clear that the reasonably restricted metastatic pattern involving the lymph
nodes and lungs would be of a particular benefit in designing gene therapy,
until the mechanism of these tropisms (eg, the overexpression of specific
adhesion molecules) are better understood.
The authors' final note of caution regarding the overselling of expectations
about these early clinical trials cannot be overstated. Particularly for
gene therapy trials that exploit immunologic manipulations, it is likely
that large randomized phase III adjuvant trials with lengthy follow-up
will be required before any conclusions can be made about the utility of
such an approach. Although the concept of gene therapy may be topical and
splashy, this approach is clearly in its infancy, and it is unlikely that
anything other than safety data will be derived from the early activated
phase I trials.
Since it is indeed true that prostate cancer patients frequently perceive
that they have few, if any, alternatives, every effort should be made to
guard against promoting unrealistic expectations about these trials. On
the other hand, the explosion in understanding of the molecular basis for
the development and progression of prostate carcinoma makes the prospect
of molecular medicine or gene therapy both a realistic and important goal
1. Bischoff JR, Kirn DH, Williams A, et al: An adenovirus mutant that
replicates selectivity in p53 deficient human tumor cells. Science 274:373-376,
2. Heidenberg HB, Sesterhenn I, Gaddipati J, et al: Alteration of the
tumor suppressor gene p53 in a high fraction of treatment resistant prostate
cancer. J Urol 154:414-421, 1995.
3. Cher M, Bova GS, Moore DH, et al: Genetic alterations in untreated
metastases and androgen-independent prostate cancer detected by comparative
genomic hybridization and allelotyping. Cancer Res 56:3091-3102, 1996.
4. Williams B, Jones E, Zhu XL, et al: Evidence for a tumor suppressor
gene distal to BRCA1 in prostate cancer. J Urol 155:720-725, 1996.
5. Savarese DM, Valinski H, Quesenberry P, et al: Expression and function
of colony-stimulating factors and their receptors in human prostate carcinoma
cell lines. The Prostate, 1997 (in press).