Immunological Implications of Androgen Ablation

Author: Yan Lin

Introduction

There is a great need for novel therapeutic approaches to prostate cancer, the second
leading cause of cancer related deaths in American men (1). In this hypothesis paper we
evaluate immunological implications of hormonal manipulations in several models
systems and try to extrapolate the findings to the immunological impact of androgen
ablation in metastatic prostate cancer patients. The essence of our hypothesis lies in the
possibility that standard androgen ablation therapy may posses an unrecognized
immunomodulatory effect. Should such an effect exist, it may provide a critical window
of opportunity to further increase immune response with the hope of eradicating the
cancer.

Does the Immune System Recognize Prostate Cancer?

Innate Responses

One of the most important components of the innate response against cancer is the natural
killer (NK) cell. These cells can kill a variety of tumors in vitro and in vivo via Fas
ligand and/or perforin dependent mechanisms (9). NK cell-mediated killing of the
human androgen resistant cell lines DU-145 and PC-3 has been demonstrated in vitro (15,
16), while killing in vivo was shown in a mouse model using the RM-1 cell line (17).
One explanation for the ability of NK cells to specifically destroy prostate cancer cells is
the observation that in contrast to normal prostatic tissue, cancerous tissue possesses low
or absent MHC-1 molecules (18). MHC-1 molecules are needed to activate cytotoxic T
cells, therefore the lack of this molecule may be induced by selective pressure of the
cytotoxic T cell. However, NK cells preferentially lyse cells lacking MHC-1. This is
postulate to occur in part by a negative signaling receptor which inhibits NK cells from
being activated when they are in contact with MHC-1 expressing cells (19).
Another cell which possesses innate ability to recognize and destroy tumors is the γδ-T
cell. These cells have been shown to possess cytolytic activity as well as the ability to
secrete cytokines which have the potential to recruit other immune cells. Prostate cancer
cells as well as cells of breast and lung cancer have been shown to basally possess the
stress-induced, MHC-I-like antigens MICA and MICB. γδ-T cell can specifically be
activated by this antigens alone or when expressed on the respective cancer cells (6).

Specific Responses
Antibody responses to PSA, PAP, HER-2/neu and p53 have been reported in prostate
cancer patients but not in healthy volunteers (2). Titers of antibodies to PSA and HER-
2/new were higher in patients with androgen independent disease. This could be the
result of an increased antigen concentration in the host, due to advanced disease. Another
possibility is that the higher titers of antibodies are due to initiation of Th2-antibody
producing responses in these patients in contrast to the protective cell mediated Th1
response. In fact, serum levels of antibody-promoting Th2 cytokines have been shown in
higher concentrations in androgen-independent patients (3). Regardless of the reason for
elevated antibody titers, this observation implies B cell recognition of prostate antigens is
occurring in patients.

Both CD4 helper cells and CD8 cytotoxic cells can recognize and proliferate in response
to peptide determinants from PSA and PSMA (4). This implies that these two proteins
can be processed by host antigen presenting cells in such a manner that T cells that have
not been thymically-deleted can respond to. This was a concern since both PSA and
PSMA are found in non-malignant tissue and therefore a possibility existed that the host
immune response could not be generated.

In vitro generation of cytotoxic T cells specific to PAP using dendritic cells pulsed with
antigen was recently accomplished by Peshwa et al (5). These T cells were able to lyse in
vitro prostate cancer cells from patients with the same HLA as the effector cells. Such a
system demonstrates that not only can antigen-specific responses be generated, but these
responses are effective against primary cells.

The in vivo relevance of such antigen-specific T cells is demonstrated in patients who
have lymphocytes infiltrating the tumor. Tumor infiltrating lymphocytes are seen in
patients with prostate cancer and lack of them correlates with poor prognosis (8).
Evidence that these T cells can mount an anticancer response is implied from findings
that in some patients the infiltrating cells can secrete IFN-γ a cytokine with direct and
immune-mediated anticancer activity (7).

Thus there appears to be immune recognition of prostate cancer. The question arises,
why is the immune response ineffective at destroying the cancer? Part of the answer is
that prostate cancer possesses many mechanisms of suppressing initiation and effector
functions of immune response.

Is the Immune Response Altered in Prostate Cancer Patients?

Dendritic Cells

Dendritic cells (DC) act as sentinals of the immune system, constantly gathering antigens
from the periphery and bringing them to the lymph nodes where they can activate T cells
and therefore initiate the immune cascade. Evidence of the potent immune initiating
abilities of DC comes from observations that these are the only cells which can activate
naive T cells. DC can act as potent adjuvants to antigens and have been employed in
numerous protocols to stimulate antigen-specific immunity (20).
In melanoma, part of the antigen-specific suppression of immune responses to tumor
antigens is believed to be due to “programming” of the DC progenitors by the tumor.

Intra-melanoma DC do not initiate immune response but instead induce an antigenspecific state of non-responsiveness in the T cells called anergy (21). Part of the reasonfor the existence of these altered DC is believed to be tumor-secreted VEGF, a cytokine which induces angiogenesis but also blocks maturation and expression of immunestimulatory molecules on DC (22). Although such immune-suppressive DC have not been found in prostate cancer, the cytokine VEGF is highly expressed and positively
correlates with Gleason score (23).

A recently identified abnormality of DC in respect to prostate cancer is that these cells
seem to be in present in very low numbers or outright lacking. This is in contrast to nonmalignant prostate tissue which contains DC and macrophages (24,25). The prostate
cancer cell lines PC-3, DU-145 and RM-1 have been shown to directly induce apoptosis
in DC through a Fas-independent mechanism (26). The relevance of DC in immune
response to prostate cancer is illustrated by the potent antitumor effects reported by
Shurin’s group (27). These investigators transfected DC with the anti-apoptotic gene
Bcl-XL and transferred them into mice bearing RM-1 prostate tumors. Injection of Bcl-
XL transfected DC but not marker transfected resulted in potent regression of tumors.

Natural Killer Cells

Decreased number and cytotoxic activity of NK cells has been observed in patients with
prostate cancer. This study also found a positive correlation between stage of cancer and
decreased activity (10). Several other investigators have reported suppression of NK
activity in a variety of cancers. Although no common mechanism for cancer-induced
suppression has been established, an interesting explanation is that an important signal
transduction molecule, the T cell receptor zeta chain, is absent in the T cells and NK cells
of prostate cancer patients (11). One explanation for the absence of this protein is that
Fas ligand expressed by cancer cells initiates a Fas-dependent activation of caspases
which cleave the zeta chain.

T Cells

In prostate cancer patients there is a deficiency of T cell proliferation in response to
mitogenic stimulation. Such a deficiency could be the result of soluble inhibitory factors
secreted by the cancer, or abnormal signal transduction molecules in the T cells of tumorbearing patients. As is the case with NK cell activity, lack of T cell proliferation
positively correlates with disease stage (12,13). Proliferation of T cells in response to
mitogen has the possibility of being a non-specific assay since T cells in some situations
may actually promote progression of the tumor (14). In order to determine whether
cancer-inhibitory T cells are deficient in patients a more sensitive assay of T cell function
was used, which detects the cytokines secreted (15). This method revealed the T cells of
prostate cancer patients secrete less interleukin-2 (IL-2) and more interleukin 10 (IL-10),
than T cells from controls. IL-2 is a cytokine which activates a variety of anti-tumor
immune functions including NK and cytotoxic T cells. IL-10 suppresses anti-tumor
immunity by inhibiting activation of NK and cytotoxic T cells.

TGF- β

TGF-β is a pleiotropic cytokine which can actually inhibit growth of some types of
cancers. In advance prostate cancer, an excess amount of this cytokine is produced, but
this does not result in autocrine inhibition of proliferation (28). TGF-b is postulated to
have an immunosuppressive function since non-immunogenic prostate tumors whose
TGF-b production is blocked by treatment with antisense oligonucleotides have a reduced
ability to grow in immune competent rats but not in immune deficient animals (29).
TFG-b suppresses immune activation through several mechanisms these include: 1)
Suppression of antigen presentation so the antigen presenting cells do not properly
instruct the T cell to induce the appropriate response (30). 2) Induction of T cell
unresponsiveness so that T cells presented with antigen will not proliferated (31). 3)
Suppression of Th1 memory cell activation, this does not allow memory cells which
secrete anticancer cytokines to be reactivated (32).

The multifaceted mechanisms by which prostate cancer suppresses immune response are
still being elucidated, and the above was just a general overview. It is the hope of this
paper to put androgen ablation therapy in the context of the myriad of protective
mechanisms prostate cancer uses to protect its self.

Androgen Ablation

Currently, the only effective therapy for metastasized prostate cancer is depletion of
androgens. The rational for this treatment is that since the prostate gland is dependent on
circulating androgens for it’s survival, decreasing the androgens in the serum will result
in death of cancers derived from prostatic tissue. Huggins and Hodges brought androgen
ablation therapy into the mainstream in 1941 when they demonstrated that decreasing
serum androgen levels by either bilateral orchidectomy or administration of estrogens
resulted in regression of metastisized prostate cancer (33). Androgen ablation protocols
have subsequently evolved, however the main methods used today are 1) bilateral
orchidectomy, 2) LHRH agonists, which decrease pituitary production of leutenizing
hormone, 3) androgen receptor agonists and 4) administration of estrogens (34).
Unfortunately, all of these methods are effective in controlling the tumor for an average
of 18 months, after which androgen resistance occurs and little therapeutic options are
available.

A possible explanation for the inability to completely abolish the tumor was provided by
Geller et al who demonstrated that intraprostatic DHT concentrations are 10-15% intact
after androgen ablation (37). These residual levels of androgen can still maintain the
viability and proliferation of several prostate cancer cell lines. In order to reduce residual
androgens, Labrie proposed inducing a maximal androgen blockade (MAB) by
combining castration (chemical or surgical) with androgen receptor antagonists. In a
double-blind clinical trial with 603 patients castration with leuprolide combined with the
androgen receptor antagonist flutamide was compared with leuprolide plus placebo.
Median progression free survival was 16.5 months for the combined treatment and 13.9
months for leuprolide plus placebo. Survival was 35.6 months for the combination group
and 28.3 for the leuprolide alone group (38). The benefit of MAB compared with
monotherapy was assessed in other clinical trials with conflicting results (35). Overall,
there does appear to be a slight increase in progression-free survival in patients with
MAB, however due to cost and side effects, the overall benefit of this therapy is under
discussion.

Another modification of androgen ablation therapy is intermittent therapy. Due to the
undesirable effects of androgen ablation, and the possibility that such a treatment selects
for cancer cells which are androgen-resistant, a clinical trial was conducted to assess the
possibility of intermittently reducing androgen concentrations. This treatment involved
suppressing androgen concentrations until the tumor regressed, then allowing patient
androgen concentrations to rise until a predetermined peak, after which androgen ablation
was reintroduced. This treatment allowed for a better quality of life, including restoration
of sexual activity when the patient’s androgen levels were unmanipulated. Presently, no
strong data exists for a delayed androgen-independence using intermittent therapy,
however the median survival times appear to be similar (36). An interesting observation
is that a wide heterogeneity exists in specific time until tumors regress after the first
androgen depletion session. In light of the postulated immunological mechanisms of
androgen ablation therapy, it would be interesting to compare immune parameters in
patients receiving intermittent therapy.

Immunogical Effects of Androgen Ablation

Several studies have shown immune responses are generally higher in females than in
males, partly due to the immunoregulatory effects of testosterone (39,40). Since during
androgen ablation the hormonal patterns of the male are altered towards a more feminine
state, we ask the question of whether the immune response may be altered in this manner
as well?

The possibility of immunological alterations during androgen ablation has been
demonstrated in a rat model where following treatment the prostate is infiltrated by
macrophages and CD8 T cells. Interestingly, this same study was able to show a
decrease in tumor regression in rats which received the immune suppressive drug FK-506
during androgen ablation (41). The question with this study remains, does the androgen
ablation activate immune responses, or does the death of tumor cells induced by the
treatment recruit immune cells which further augment the tumorolytic process?

Suppression of VEGF Production

A mechanism by which androgen ablation may augment anti-tumor immunity is by
decreasing production of VEGF. It has been demonstrated, as stated above, that prostate
cancer secretes high levels of the immune suppressing cytokine VEGF (23). Previous
studies have demonstrated tumor suppressing synergy between antibody therapy to
VEGF and immune stimulators (42). It has been shown that VEGF secretion by prostate
cancer cells is dependent on androgen (43). Further, patients treated with androgen
ablation exhibit decrease production of this cytokine (23). Thus suppression of tumor
VEGF production may be an immunomodulatory effect of androgen ablation.
Augmentation of Th1 Responses by Androgen Ablation
A murine model for human inflammatory autoimmune disease is the SJL mouse which
develops a multiple sclerosis-like syndrome termed experimental allergic
encephalomyelitis (EAE) when immunized with the autoantigen myelin basic protein
(MBP). Development of autoimmunity is dependent on a Th1 immune response, with
cytotoxic T cells mediating tissue-specific damage. It is this type of immune response
which would be ideal for eradication of prostate cancer. Interestingly, autoimmunity can
only be induced in the female mouse. This has been demonstrated to be dependent on the
lack of gonadal hormones in the females since castrated male mice can develop
autoimmunity subsequent to autoantigen injection (44). Further evidence for antigenspecific immune modulation by androgens arises from observations that MBP-specific T cell clones derived from female SJL mice secrete high levels of IFN-g, low levels of IL- 10 and can cause autoimmunity when transferred to a recipient. When these clones are grown in the presence of androgen at concentrations found in male serum, the T cells lose pathogenic potential and possess a typical Th2 cytokine profile in that they secrete highlevels of IL-10 and low levels of IFN-g (45). The ability of androgens to inhibit immune activation against a self antigen in the case of SJL mice may be an explanation for the difficulty in achieving immunity using prostate cancer vaccines.
Another model of autoimmunity is the non-diabetic obese (NOD) mouse which develops
diabetes due to infiltration and destruction of pancreatic beta cells by cytotoxic and
helper T cells. Unlike the SJL mouse, autoimmunity in the NOD mouse is not induced
by immunization but is spontaneous. Interestingly, diabetes occurs primarily in the
female with only 10-20% of males developing disease. The possibility of androgens
protecting against development of diabetes in was evaluated by administration of
testosterone to prediabetic female mice, which resulted in protection from disease (46).
In another study, the in vitro culture of female T cells with DHT demonstrated that DHT
can increase the Th2 cytokine IL-4, as well as inducing a Th2-like population in mice
treated with DHT. This data suggests androgens may play a role in controlling proinflammatory Th1 responses (47).

Conclusion

The immune response seems to recognize prostate cancer, although it is ineffective at
inducing eradication. Several immunological interventions are being presently attempted
in order to systemically boost immune response to prostatic and prostate tumor antigens.
A reason for the mediocre performance of immunotherapy may be the state of immune
suppression which exists in the patient. We propose that androgen ablation may induce a
transient alleviation of the prostate cancer-induced immune suppression in the patient.
Therefore, the optimal time for immune therapies, especially antigen-specific
interventions, would be during or immediately after androgen ablation.