Retinoblastoma is the most common primary eye cancer in children and an important clinical problem facing ocular oncologists. Retinoblastoma usually manifests before 3 years of age, and the tumors either grow locally within the eye, extended outside the globe migrating down the optic nerve into the central nervous system and the cerebrospinal fluid, or spread through the vasculature to form metastasis in other parts of the body, particularly the bone. Left untreated, retinoblastoma is uniformly fatal. Approximately 40% of retinoblastoma cases are hereditary, in which the child receives a single germline mutation in the retinoblastoma (Rb) gene from one parent and acquire a spontaneous Rb mutation in the other allele during retinal development. The remaining 60% of cases are sporadic, in which the child acquires spontaneous mutations in both Rb alleles during development of the retina. In hereditary retinoblastoma, a majority of the children have cancer in both eyes, and the tumors have a tendency to form earlier and grow faster. These children also have a much higher risk for developing secondary malignancies later in life, especially those of the bone (osteosarcoma). Reports indicate the rate of secondary malignancies in these patients is increased by radiation treatment and possibly by chemotherapy.
The Rb gene was the first tumor suppressor gene identified and has been studied extensively over many decades for its role in controlling cell cycle progression. Tyler Jacks, PhD, however, observed the surprising result that targeted deletion of the Rb gene in the retina of mice failed to result in tumor formation, indicating that more genes are required.1 This issue was resolved recently in studies by Michael A. Dyer, PhD, and coworkers,2 who produced the first gene knockout model of Rb in mice by the inducible deletion of Rb, p53, and, p107. These data suggest that in humans, a similar cascade of gene mutations is likely to be required during malignant transformation in the developing retina.
WHY IS A NEW THERAPY NEEDED?
Advances in the management of retinoblastoma have
improved prognosis dramatically over the past 30 years,
and the overall survival of retinoblastoma patients in the
United States is currently greater than 90%. Originally, the
only treatment option was enucleation of the tumorcontaining
eye. Today, enucleation is used only in
patients with exceptionally large tumors and patients
who fail to respond to treatment. Current treatments,
such as radiotherapy or combining chemotherapy and
radiotherapy, are successful at controlling the growth of
small tumors in unilateral or bilateral retinoblastoma,
respectively, and at preserving functional vision in the
treated eye. The recent use of intraarterial chemotherapy
has provided a new method to deliver chemotherapy
directly to the tumor-containing eye. This technique has
yielded some intriguing results among the early groups of patients receiving this treatment.3 (See “The Evolution of
Treatments for Retinoblastoma” on page 56). Despite
these advances, there are still numerous cases of
retinoblastoma, particularly larger ones that are difficultto-
treat tumors. Moreover, there can be significant side
effects related to use of radiotherapy or chemotherapy in
children, some of which can be quite severe.
WHY IS AN IMMUNOTHERAPY
APPROPRIATE?
Historically, it has been thought that the immune
response was capable of providing the ideal protection
against tumor progression because of its ability to
distinguish between malignant and normal cells.
Lymphocytes, specifically T cells, display a high level of
specificity conveyed by receptors (TcR-T-cell receptors)
that are capable of identifying protein fragments
derived from mutated genes expressed only in tumor
cells and not in normal cells. Although enthusiasm for
cancer immunotherapies has waxed and waned over
the years, there has been generally steady progress in
this field over the past decade, particularly in the treatment
of metastatic skin melanoma.
While the molecular biology of the Rb gene has been studied extensively, experiments that examine the induction and expression of tumor-specific T cells against retinoblastoma have not been conducted. This lack of information exists despite circumstantial evidence that indicates retinoblastoma may be highly immunogenic. The rate of spontaneous regression of Rb tumors has been predicted to be higher than other tumors (as high as 1.0%). Although the mechanism of Rb tumor regression is unknown, it may result from 1) ischemic necrosis, 2) formation of a benign retinoma, or 3) immune-mediated rejection. Although there is no direct evidence indicating that immune-mediated rejection of Rb occurs in patients, a local inflammatory response has been reported to accompany spontaneous regression of retinoblastoma in a few cases.
GENERAL CONCEPT OF A TUMOR
CELL VACCINE
Our laboratory has investigated the creation of tumor
cell vaccines. Tumor tissue is recovered from enucleated eyes, and primary explant cultures are established in
the laboratory (Figure 1). Cultures that yield tumor cell
lines are analyzed for Rb gene mutations and expression
of Rb protein. These data are compared with sections
from the original tumor specimen to validate the
tumor cell line. Tumor cells are genetically modified to
express two genes (CD80 and MHC class II, as described
below) using lentiviral vectors. The tumor cells are tested
in the laboratory to demonstrate that they effectively
activate tumor-specific T cells and are inactivated to
prevent proliferation. The vaccine can be used via either
direct immunization of the patient or in vitro activation
of the patients, in which lymphocytes are then adoptively
transferred back into the patient. Activated T cells
then migrate systemically and eliminate malignant cells.
Although our initial studies examined the T-cell response to specific tumor cell lines, we do not plan to make individual tumor cell vaccines for each patient. This would not be feasible for many reasons. Most important, it is not possible to obtain tumor samples unless the patient's eye is enucleated. Moreover, even if we were able to obtain tumor samples, Rb cell lines are produced from only a small percentage of samples. Therefore, an individualized tumor cell vaccine is absolutely impossible for Rb patients.
In order to solve this problem, we are producing a group of stably transfected Rb cell lines that express CD80 plus different Class II alleles. How we will use these cell lines to vaccinate Rb patients is illustrated in Figure 2. The available vaccines will contain a mixture of different Rb cell lines that together will express the widest possible array of tumor antigens. Therefore, when a potential Rb patient is considered for vaccination, we would customize a combined vaccine from a number of tumor cell lines that each match the class II alleles of the patient. Because a majority of people in the general population expresses class II (either DR1, DR2, DR4, DR7, or DR15), it is very likely that we will have the appropriate matching class II vaccines already produced.
ACTIVATION OF A TUMOR-SPECIFIC
IMMUNE RESPONSE
There are two major types of specific T cells (Figure
3): Cytotoxic T cells express TcR (T cell receptors)
that recognize tumor antigens presented by class I and
are identified by the expression of CD8; by contrast,
T-helper cells express TcR that recognize tumor antigens presented by class II and are identified by the
expression of CD4. The function of these two populations
of lymphocytes is distinct; cytotoxic T cells eliminate
tumor cells by lysing the target cells, and T-helper
cells release a variety of lymphokines that activate
cytotoxic T cells and macrophages. Successful activation
of T-helper cells is normally triggered by antigenpresenting
cells (APC), but this occurs only when APCs
express two signals: 1) an antigen presented by class II,
and 2) a costimulatory signal provided by CD80 (Figure
4). The TcR on the T-helper cell recognizes the antigen,
and a second receptor (CD28) recognizes CD80. When
the T cell receives both of these signals, it is activated
to release a variety of lymphokines including interferon
gamma (IFN-γ) and interleukin-2 (IL-2). This pathway of
activating T-helper cells in cancer patients is blocked by
factors secreted by tumor cells that inhibit the two signals
provided by APCs. To activate T-helper cells
against tumor antigens, our approach is to convert
tumor cells into APCs by genetically altering the tumor
cells to express CD80 and Class II (Figure 5). This allows
tumor cells to express the two signals required to activate
tumor-specific T-helper cells.
T-helper cells that are activated by the tumor cell vaccine provide the necessary “helper” lymphokines that are needed to activate CD8 cytotoxic T cells (Figure 6) capable of eliminating the tumor cells. Once activated, cytotoxic lymphocytes no longer require two signals in order to trigger lysis of target cells. They require only one signal: the expression of the tumor-specific antigen. This allows the vaccine to activate specific T cells that circulate systemically throughout the body searching for malignant cells. One critical factor in activating protective antitumor immunity is the development of a sustained T cell response. Past attempts at tumor cell vaccines focused on activation of cytotoxic T cells in the absence of Thelper cells. These vaccines were capable of triggering a T-cell-mediated cytotoxic T-cell response in patients, but the duration of the response was short-lived and unable to sustain protective immunity. We predict that activation of T-helper cells is critical for a sustained immune response against the tumor.
RISKS/BENEFITS OF A TUMOR CELL VACCINE FOR RB
The most significant and common side effect of a
protective anti-tumor immune response is the development
of autoimmunity. For example, the most recent
immunotherapy for metastatic skin melanoma uses a
Gp-100 (tumor antigen) vaccine combined with an antibody
(ipilimumab or tremelimumab) that promotes
T-cell activation. A recently completed randomized
double-blind phase 3 trial demonstrated a benefit in
overall survival in the treated population.4,5 However,
patients who responded displayed a variety of secondary
inflammatory complications including colitis/diarrhea,
dermatitis, hepatitis, endocrinopathy, nephritis,
and uveitis. Although these complications were successfully
treated in most patients, their manifestation illustrates
the close connection between protective antitumor
immunity and potentially damaging autoimmunity.
The potential benefit of immunotherapy, however,
is also illustrated by the limited success of the immunotherapy
clinical trials. Currently, the only documented
cure for metastatic melanoma is restricted to therapies
that stimulate anti-tumor immunity.
Bruce Ksander, PhD, is an Associate Scientist with the Schepens Eye Research Institute and an Assistant Professor of Ophthalmology, Harvard Medical School in Cambridge, MA. Dr. Ksander has a US patent pending for “Animal models of adnexal tumors.” He can be contacted at +1 617 912 7443; fax: +1 617 912 0113; or via e-mail at bruce.ksander@schepens.harvard.edu.
SHARE YOUR FEEDB
Would you like to comment on an author's article?
Do you have an article topic to suggest?
Do you wish to tell us how valuable
Retina Today is to your practice?
We would love to hear from you. Please e-mail us at
letters@bmctoday.com with any thoughts or
questions you have regarding this publication.
_1773249222.png?auto=compress,format&w=75)
