| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Endocrine-Hypertension Division Department of Medicine Brigham and Women’s Hospital and Harvard Medical School Boston, Massachusetts 02115
Address all correspondence and requests for reprints to: Ursula B. Kaiser, M.D., Endocrine-Hypertension Division, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, 221 Longwood Avenue, Boston, Massachusetts 02115. E-mail: ukaiser{at}partners.org
 |
Introduction |
|---|
 Top Introduction References |
|---|
Pituitary adenomas are typically benign tumors, representing 10–15% of primary intracranial neoplasms (7). They are very common, recognized in up to 20% of the population in autopsy series. While the majority of these remain clinically silent, others can cause severe and progressive disease. Clinical manifestations include mass effects related to the displacement of normal surrounding brain structures (e.g. compression of the optic nerves resulting in visual field deficits), hormone overproduction, and/or pituitary insufficiency, manifested by the lack of adequate production of any of the pituitary hormones. Pituitary adenomas are caused by clonal expansion of one of the five major anterior pituitary cell types and include prolactinomas that hypersecrete PRL resulting in reproductive dysfunction, GH-secreting tumors that cause the syndromes of acromegaly or gigantism, ACTH-producing tumors that cause Cushing’s disease, TSH-producing tumors that result in hyperthyroidism, and gonadotrope-derived tumors that may hypersecrete LH and/or FSH.
The goals of therapy of pituitary tumors include: 1) reduction or elimination of tumor mass; 2) limitation and prevention of damage to surrounding structures; 3) normalization of hormone hypersecretion; 4) correction of hormonal deficits; and 5) preservation of remaining pituitary function. Current therapeutic options for pituitary tumors include surgical resection, radiotherapy, and pharmacological approaches (8). Transsphenoidal surgery is the treatment of choice for most pituitary tumors. For microadenomas (< 1 cm), transsphenoidal surgery is successful in 70–90% of cases, in the hands of an experienced neurosurgeon. However, larger macroadenomas are rarely cured by surgery alone. Pharmacotherapies are available for selected types of pituitary tumors. For example, dopamine agonist therapy for prolactinomas effectively reduces hormone secretion and tumor size in most patients, and is the treatment of choice for this pituitary tumor type. Somatostatin analogs can reduce GH and TSH secretion and, in some cases, tumor size in patients with acromegaly or TSH-secreting adenomas. However, pharmacologic therapy does not cure the tumors, and they recur if the medications are discontinued. Furthermore, some patients fail to respond to medical therapy or cannot tolerate associated side effects. Radiotherapy is rarely used as primary treatment; more commonly, it is used as an adjunctive treatment after surgery. Radiotherapy has variable efficacy in reducing tumor size and hormone levels and several years are usually required before it has full effect. Furthermore, it may have adverse effects such as hypopituitarism, neurologic deficits, and secondary central nervous system tumors. As a result, despite recent advances in the current therapeutic modalities, it is not always possible to achieve cure, particularly in patients with tumors that are large, locally invasive, hormone hypersecreting, or unresponsive to pharmacologic therapies.
The limitations in our current therapeutic options, and the characteristics of the pituitary gland that render it a suitable model for gene targeting, have led several investigators to hypothesize that gene therapy strategies may be useful for the treatment of pituitary adenomas (6, 9). These strategies take advantage of pituitary cell-specific activity of selected gene promoters to target therapeutic transgenes. Pituitary tumors represent a feasible target for gene therapy not only because of the ability to target specific cells. They are also mostly benign, slow growing, and localized. Reductions in tumor size and function, even if not completely curative, are of clinical benefit and are unlikely to be associated with rapid recurrence or re-growth. Several in vitro and in vivo studies have been performed over the past 3 yr, culminating in the current study using a large animal model, as reported in this issue (1).
The earliest studies of pituitary gene targeting were performed in vitro, using primary rat anterior pituitary cell cultures and immortalized pituitary cell lines (2). Replication-deficient adenoviral vectors were used, which have the advantage that they are episomal and not integrated into the host genome (10, 11). As a result, these vectors are able to successfully transfer genes into nondividing or slowly dividing, differentiated cells. This also minimizes concerns about the possibility of random integration events causing inadvertent activation of oncogenes or disruption of tumor suppressor genes. On the other hand, a drawback of adenoviral vectors is the limited duration of transgene expression after delivery. Castro et al. (2) demonstrated that replication-deficient adenoviral vectors encoding ß-galactosidase, driven by Rous sarcoma virus or cytomegalovirus promoters (AdRSVGal or AdCMVGal), were able to infect and express ß-galactosidase at high levels within all cell types of primary rat anterior pituitary cultures, as well as in the corticotropic AtT20 and somatolactotropic GH3 endocrine tumor cell lines. There were no detrimental effects on cell survival. ACTH and LH secretion were impaired after infection, but only at a multiplicity of infection (MOI) of 20 or greater. In primary pituitary cell cultures, transgene expression was shown to persist for at least 21 days. However, the decline may occur at a faster rate in vivo, if the adenovirus were to induce an immune response. Nonetheless, these studies demonstrate that high efficiency gene transfer into all anterior pituitary cell types can be achieved.
The efficient gene transfer into pituitary cells using adenoviral vectors led to the hypothesis that gene therapy might provide a feasible therapeutic approach for the treatment of pituitary adenomas. "Suicide" gene therapy involves the transfer of genes encoding nontoxic proteins, which can cause cell death via the conversion of the nontoxic prodrug into a toxic product. A well characterized conditional cytotoxic approach uses the herpes simplex virus-thymidine kinase (TK) gene, which converts nucleoside analogs, such as ganciclovir (GCV), into their phosphorylated metabolites, which then act as competitive inhibitors of endogenous nucleotides and cause termination of DNA synthesis, resulting in death of proliferating cells (10, 12). Infection of GH3 or AtT20 cells with adenovirus containing the TK gene driven by the cytomegalovirus promoter (AdCMVTK) resulted in the specific induction of apoptosis by GCV (4). In contrast, no apoptosis was detected in primary rat anterior pituitary cells following infection with AdCMVTK and exposure to GCV, indicating that this treatment is not deleterious for normal pituitary cells in primary culture. An important advantage of the use of the TK gene is that the cytotoxicity induced by GCV is limited to cells that are dividing. This may provide selectivity for pituitary adenomas, with minimal toxicity to normal pituitary cells.
While the use of viral promoters has the advantage of high levels of
expression, its utility is limited by its widespread expression
indiscriminate of cell type, leading to potential toxic or adverse
effects in normal tissues. This problem may be circumvented if gene
expression could be restricted by using cell-specific promoters. This
rationale led to studies using adenoviruses containing the GH or
-subunit gene promoters to target transgene expression
(3). Using the GH promoter to drive ß-galactosidase gene
expression (AdGHGal), 95–100% of GH3 cells stained blue
by 4–5 days after adenoviral infection. Similar results were obtained
using the -subunit promoter-driven adenovirus (AdGal) in the
gonadotrope-derived T3 tumor cell line. Peak expression occurred
5–7 days after infection, thereafter declining but persisting until at
least 21 days after infection. These same promoters were also used to
drive expression of TK. AdGHTK and AdTK were able to confer
dose-dependent GCV sensitivity, cytotoxicity, and growth inhibition to
GH3 and T3 cell lines. To develop an in vivo
model for assessing the effects of the recombinant adenoviruses,
GH3 cells were injected into nude mice, resulting in the
development of sc tumors. These GH3 cell tumors, in turn,
were injected with AdGHTK virus. Treatment with GCV caused marked
regression of the tumors. An additional benefit of the use of TK is its
ability to generate a "bystander" effect (13).
Complete regression of tumors has been observed even when only 10% of
tumor cells expressed TK, possibly reflecting the transfer of
triphosphate nucleoside analogs from one cell to another through gap
junctions, or the uptake of apoptotic bodies by nontransduced cells.
These studies indicate that using pituitary-specific promoters,
adenoviral vectors can efficiently target expression of genes to
pituitary tumor cells in vitro and in vivo, and
growth inhibitory and cytotoxic effects are observed upon transferring
"suicide" genes by this approach.
Subsequent testing turned to in vivo rodent models in an
attempt to target adenoviral vectors to the pituitary gland
(4, 5, 6). Intravenous or intracarotid arterial injection of
AdCMVGal in adult rats yielded a high level of ß-galactosidase
expression in the liver, but virtually no ß-galactosidase activity
was detected in the pituitary (6). These data indicate
that the transfer efficiency of the adenoviral vector to the pituitary
gland is very low after intravascular administration. No positive
pituitary cells were detected following injection of AdGHGal or
AdGal, although the absence of expression of these vectors in the
liver did indicate that pituitary-specific promoters effectively
restricted transgene expression, an important consideration to avoid
potential adenoviral hepatotoxicity. In contrast, direct stereotactic
injection into the pituitary gland yielded positive ß-galactosidase
activity around the injection site, using either AdCMVGal, AdGHGal, or
AdGal. Furthermore, the use of AdGHGal or AdGal targeted
pituitary cell types with high specificity, with AdGHGal producing
ß-galactosidase activity almost exclusively in somatotropes, and
AdGal resulting in staining of gonadotropes and thyrotropes. Similar
studies were performed in rats with sulpiride- and estrogen-induced
lactotrope hyperplasia. Stereotactic intrapituitary injection of
AdCMVTK followed by treatment with GCV for 7 days caused a 60%
reduction in circulating PRL levels and a 26% reduction in pituitary
weight (4). No effects on circulating GH, ACTH, and TSH
levels were observed, indicating that the toxic effects were largely
limited to the more actively proliferating lactotropes, with sparing of
the other hormone-producing cell populations within the anterior
pituitary gland. On the other hand, when AdPRLTK was used in the same
experimental paradigm, treatment was not effective in either reducing
the weight of the pituitary gland, the number of lactotropes, or
circulating PRL levels (5, 8). Of note, in the study using
AdGHGal and AdGal, ß-galactosidase staining was less intense than
with AdCMVGal (6). It may be that AdPRLTK is able to
target transgene expression to lactotropes, but the levels of TK
expressed under the control of the PRL promoter are insufficient to
induce cell death by GCV. Studies of gene therapy for posterior
pituitary disease have also been performed; stereotactic injection of
AdRSVGal directly into mouse neurohypohysis resulted in uptake by nerve
terminals and ß-galactosidase expression in magnocellular neurons
(14).
This brings us to the most recent study, in this issue of Endocrinology (1), applying to sheep, similar experimental approaches as those used previously in rodents. The aims of this study were to evaluate adenoviral gene transfer in a large animal pituitary model, in an attempt to provide further validation of a gene therapy strategy for potential human therapy. Ewes were anesthetized and a burr-hole was made in the skull to allow stereotactic transcranial injection of adenoviral vectors encoding the ß-galactosidase gene driven by either the cytomegalovirus promoter (RAd-CMV-ßgal) or by -4429/+14 of the human PRL promoter (RAd-hPRL-ßgal) into the pituitary gland. Following adenoviral injection, blood samples were taken daily until the animals were killed 4–7 days later and pituitary glands were harvested for further analysis. Intense staining indicative of ß-galactosidase activity was observed along needle tracks and at the base of the pituitary, with stronger expression in animals injected with RAd-CMV-ßgal than in those injected with RAd-hPRL-ßgal. Dual immunofluorescence demonstrated expression of the transgene in all endocrine cell types as well as in folliculostellate cells, in animals injected with RAd-CMV-ßgal. In contrast, in animals injected with RAd-hPRL-ßgal, 93% of the cells staining for ß-galactosidase coexpressed PRL, with small numbers of gonadotropes and other endocrine cell types staining, indicating high cell-type specificity. Measurement of plasma hormones showed no abnormal responses with the exception of transient rises in cortisol and PRL immediately following surgery, likely a stress response to anesthesia, although no sham-operated animals were included as controls. This study extends the previous reports using in vitro cell cultures or in vivo rodent models. The results indicate that stereotactic adenoviral transgene delivery is reliable, accurate, and safe in a large animal model, with no noted adverse effects on endocrine function, supporting a possible role as a therapeutic modality for pituitary adenoma therapy. Cell-type-specific targeting was achieved using a cell-specific hormone promoter. One potential shortcoming is the limited distribution of the transgene in rather localized areas rather than throughout the pituitary. This may limit the effectiveness of cytotoxic therapies, although they may benefit from the previously noted "bystander" effect (although the "bystander" effect is not cell-type specific). The next step is clearly the evaluation of the safety and efficacy of delivery of cytotoxic genes in large animal models to test the potential for ablative therapy.
This study takes us one step closer to realizing the potential for gene therapy in pituitary disease. However, many questions remain to be addressed before this goal can become a reality. Will pituitary cell-type specific promoters be able to direct cell-specific expression of transgenes at levels sufficient for growth inhibitory or lethal effects on pituitary cells? To date, no studies using cell-specific promoters to target TK to pituitary cells have demonstrated toxic effects on the pituitary gland in vivo. Will the transient nature of expression of adenovirally mediated gene delivery limit the effectiveness of such therapy on pituitary disorders? Studies have indicated that the transgene is expressed for at least 21 days. This may be sufficient for the therapy of pituitary adenomas, where long-term expression is not necessary. However, for other pituitary diseases, such as pituitary insufficiency, the goal of hormone replacement by gene therapy presents a greater challenge because chronic expression of the transgene would be required for effective therapy. Modifications of adenoviral vectors or the use of other types of vectors may allow more long-term transgene expression. For example, the use of so-called "gutless" adenovirus, containing no viral genes, shows promise for prolonged transgene expression (9). Nonetheless, in many genetic causes of pituitary insufficiency, the target cell type may be absent from the pituitary due to developmental defects, making physiologically regulated expression of a hormone a difficult goal to attain. A more realistic goal may be to target hormone-encoding transgenes to other tissues, using other promoters, as has been done for GH in muscle cells, although this approach does not allow for physiological regulation of hormone production and release (15, 16). Will pituitary gene therapy cause immune responses or inflammatory changes in the pituitary gland? To date, no inflammatory changes have been observed in the injected pituitary glands in animal models at the time points studied (5–7 days). However, immune reactions may develop at later time points. It is conceivable that inflammatory and immune reactions may be of benefit in achieving toxic effects on pituitary tumors, although on the other hand such changes may theoretically have adverse effects on the function of surrounding normal pituitary gland.
At present, the ability to target adenoviral vectors to the pituitary requires direct pituitary injection. As such, gene therapy will be limited to adjunctive therapy in conjunction with neurosurgery. Further experimentation with varying amounts of virus, with additional viral vectors, or with modification of adenoviral vectors to target the pituitary gland may lead to the development of successful gene delivery to the pituitary by intravascular injection. Modification of the adenoviral vector is a possible approach to restrict viral tropism—for example, incorporation into the viral capsid of a ligand for a pituitary cell-type specific receptor may help to target the adenoviral vector to specific pituitary cell types.
Alternative cytotoxic genes can also be tested and may have improved efficacy. The use of angiogenesis inhibitors represents a possible strategy. In mice, heterologous disruption of the Rb gene leads to POMC-expressing pituitary tumors originating from the intermediate lobe of the pituitary gland; therapy with adenovirus encoding the Rb cDNA led to inhibition of tumor cell proliferation, inhibition of tumor growth, and prolongation of the life spans of affected animals (17, 18). One of the earliest studies of gene therapy for pituitary tumors used adenovirus encoding tyrosine hydroxylase, the rate-limiting enzyme in the biosynthesis of dopamine, to infect cultured human lactotrope adenoma cells in vitro. This therapy led to increased production of dopamine, resulting in reduced PRL secretion (19).
The field of gene therapy for pituitary diseases remains in its infancy and many hurdles remain before successful application to human pituitary disorders can be achieved. The need for targeted gene delivery, and for pituitary cell type specific gene expression, points to the importance of continued studies and further understanding of the mechanisms underlying regulation of cell-specific gene expression. Despite the long road ahead, expectations for gene therapy in general, and in pituitary disease in particular, remain high.
Received November 20, 2000.
|
References |
|---|
|
TopIntroductionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
