The Impact of Aging on Cancer Vaccination

Abstract

Cancer vaccination is less effective at old than at young age, due to T cell unresponsiveness, caused by various age-related changes of the immune system. This includes lack of naïve T cells, defects in activation pathways of T cells and antigen-presenting cells (APC), and age-related changes in the tumor microenvironment. Although evidence exists that also natural killer (NK) and natural killer T (NKT) cells of the innate immune system change with age, comparison of various studies involving adaptive and innate immune responses in elderly and cancer patients, as well as cancer vaccination at young and old age in this review, indicates that also innate immune responses should be tested as a potential candidate to improve immunotherapy against cancer at older age.

Introduction

Cancer is a disease of the elderly. With the increase in the elderly population, we can expect an increase in the number of cancer patients and mortality. For most cancers, primary tumors can be removed by surgery, followed by radiation, chemo- or adjuvant therapy. However until today, these therapies are ineffective against metastases [1]. In contrast, cancer vaccination shows great promise against metastases but is hampered by the fact that vaccines are less effective at old than at young age, due to T cell unresponsiveness [2–4]. Analysis of various vaccine studies in preclinical cancer models at young and old age, showed that vigorous anti-tumor responses could be obtained by tailoring vaccination to older age, but in most cases T cell responses were hardly detectable. Therefore, we questioned the feasibility of T cell activation in the tumor microenvironment by vaccination at older age, and whether activation of innate immune responses against cancer could be a more feasible approach since innate immune responses seems less affected by aging than adaptive immune responses. To answer these questions, we reviewed adaptive and innate immune responses in elderly and cancer patients, and comparison of vaccine studies in preclinical models at young and old age. Based on these studies, new approaches to improve adaptive and innate immune responses against cancer through vaccination or immunotherapy, respectively, at older age will be proposed.

Impaired immune responses in elderly

Adaptive Immune Responses

One of the most important changes in the immune system at older age is the decline in responsiveness of T cells to new antigens. This is mainly caused by strong decrease in the number of naive T cells (capable of reacting to new antigens) and an increase in the number of memory T cells (capable of reacting to previously exposed antigens) at old compared to young age [5,6]. However, also other possible causes for decreased T cell responses in aged humans and mice have been described, such as defects in T cell receptor (TCR)/CD3-mediated phosphorylation events or aberrant regulation of tyrosine kinases associated with the TCR [7], and an age-related decrease in the αβ repertoire of the human TCR [8]. The TCR is expressed by T cells, and is required for recognition of foreign antigens in association with self-major-histocompatibility complex (MHC) molecules, presented by APC to the immune system, and for subsequent activation of T cells. In addition, an age-related decrease in the expression of CD28 on the cell membrane of T cells, which provides a secondary signal for T cell activation when ligated to the B7 molecule on APC has been reported (9–11). Decreased production of interleukin (IL)-2 or interferon (IFN)γ at old compared to young age in individuals vaccinated with influenza virus or in vitro upon stimulation with influenza virus has been shown as well [11, 12].

Innate Immune Responses

Cumulative evidence indicates that ageing exerts significant effects on all cells of the innate immune system [13]. This includes natural killer (NK) cells, natural killer T (NKT) cells, dendritic cells (DC), macrophages, and neutrophils. NK cells are the most well known cells of the innate immune system. NK cell function has been extensively studied in relation to aging in mice and humans. Although in 25-month-old mice NK cell number and function, such as the production of IFNγ, IL-2 of perforin, is decreased at old compared to 8 week-old mice, it has been reported that in human healthy centenarians NK cytotoxicity by activation with IL-12, IFNα, and IFNγ is well preserved, but somewhat decreased in less healthy elderly [13–15]. In our studies we found that the production of IFNγ by NK cells induced by vaccination with an attenuated Listeria monocytogenes was almost as good in old as in young mice (unpublished data).

NKT cells are considered to be a member from the innate immune system because of their early response against infection and perhaps against cancer. They represent a heterogeneous T cell population that shares some functional and phenotypical characteristics with NK cells [16]. It has been reported that the number of NKT cells increases with age [21], while their Th1 cytokines decreases with age [20]. However, liver NKT cells bearing TCRγδ are not only strongly increased in number but also their functions are well preserved in very old mice and humans [17]. It has been reported that NKT cells communicate with NK cells when activated by α-galactosylceramide (GalCer) [19].

DC in blood or Langerhans cells in skin, play a central role in T cell activation, but the results reported so far are variable. For instance, it has been demonstrated that blood DC from old individuals can still function as powerful antigen-presenting cells when exposed to purified protein derivate (PPD) of Mycobacterium tuberculosis or influenza vaccine [20,21], while others have shown that DCs from aged individuals are more mature and have impaired ability to produce IL-12 [22], or that secretion of tumor necrosis factor (TNF)α and IL-6 significantly increased upon stimulation with lipopolysaccharide (LPS) and ssRNA in DC of aged compared to young individuals [23]. Aged macrophages and neutrophils have impaired respiratory burst and reactive nitrogen as a result of altered intracellular signaling [24].

Impaired immune responses in cancer patients

Adaptive Immune Responses

In cancer patients cytotoxic T lymphocytes (CTL), recognizing tumor-associated antigens (TAA) in association with major histo-compatibility complex (MHC) molecules on the tumor cells through their T cell receptor, and expected to destroy tumor cells when exposed simultaneously to both TAA/self-MHC complexes and co-stimulatory molecules, are often found at the site of the tumor, but have evidently been unable to destroy the tumor cells [25]. Multiple possible causes have been described for this unresponsiveness of the CTL in cancer patients [for a review see 4]. This includes decreased expression of MHC, TAA, or co-stimulatory molecules by tumor cells, and immune suppression induced by the primary tumors. In humans and mice, many tumors secrete lymphokines or factors that inhibit vaccine-induced T cell and NK cell responses. Examples are transforming growth factor (TGF)β, IL-6, IL-10, cyclooxygenase-2 (COX-2) and its products prostaglandine E2 (PGE₂), PD1-ligand, or indolamine 2,3-dioxygenase (IDO).

Also immune cells in the tumor microenvironment attracted and activated by the primary tumor such as myeloid-derived suppressor cells (MDSC), suppress T cell and NK cell responses by the production of IL-6, IL-10, TGFβ, reactive oxygen species (ROS), inducible nitric oxide synthase (iNOS) or arginase [26•,27•], or contributes to the expansion of regulatory T cells (T_regs) in the tumor microenvironment [28•], while tumor associated macrophages (TAM) and M2 macrophages strongly suppress T cell responses through the production of IL-6, IL-10, TGFβ in the tumor microenvironment [29]. Interestingly, it has been reported that the tumor microenvironment changes with age, i.e. it appeared that the number of MDSC increases with age, and that this contributed to the T cell unresponsiveness at older age [30]. So far, little research has been performed on MDSC and T cell unresponsiveness in relation to aging. Inducible T_regs play an important role in suppression of the immune system in cancer patients, through the production of soluble factors such as IL-10 and TGFβ or through direct cell-cell contact, resulting in the inhibition of T cell and NK cell responses [31–33]. Moreover, evidence exists that the number of T_regs increases with age [34]

Innate Immune Responses

In vivo depletion of NK cells leads to a poor control of tumor growth in various cancer models, indicating the importance of NK cells in anti-tumor responses and tumor surveillance [35–38]. Evidence exists from mice and humans that NK cells alter with age, but that they still function at older age. However, the effect of aging on NK cells against cancer has been far less extensively studied than T cells. A few reports describe that NK cells of elderly had a lower ability to respond to IL-2, lower spontaneous cytolytic activity towards tumors than young adults [24]. However, NK cells can also be used to kill tumor cells through other pathways then perforin-mediated tumor cell destruction. For instance, a clinical trial is ongoing with bortezomib which sensitizes tumor cells for TRAIL- and FasL-mediated destruction by NK cells in cancer patients between 20–70 years (NCT00720785) [39]. We found NK cell responses (producing IFNγ) in vivo in old mice with metastatic breast cancer after vaccination with pcDNA3.1-Mage-b [40•].

Also NKT cells have anti-tumor activity in mice, including lung and hepatic cancer metastases when activated by αGalCer, by secreting large amounts of IFNγ and IL-4, resulting in activation of other cells of the immune system including NK cells [41,42]. In a phase I clinical trial with αGalCer in patients with solid tumors, the effect was dependent on the high number of NKT cells present pretreatment [43]. Since the number of NKT cells increases with age, αGalCer could be a potential candidate to activate NKT cells against cancer at older age.

Cancer vaccination at older age in preclinical models

More than 50% of all cancer patients are 65 years or older [44]. The vaccine studies discussed below show that cancer vaccination is less effective at old than at young age, but that tailoring cancer vaccination to older age is feasible. Moreover, also innate immune responses may be a potential target for immunotherapy against cancer.

The research group of Provinciali reported that immunization with a highly engineered mammary adenocarcinoma TS/A-IL-2, protected both young and old mice from TS/A challenge which was not possible without IL-2 [45]. CD4 and CD8 T cells were present in tumors of young but hardly detectable in tumors of old mice, while macrophages and neutrophils were detected at both ages. However, protective memory responses that could reject tumor cells upon re-challenge of tumor-free mice was only obtained in young mice. Another study of the group of Provinciali showed that vaccination with pCMV-neuNT against Her2/neu-expressing breast tumor cells (TUBO) completely protected young mice but only 60% of the old mice from TUBO challenge, and correlated with proliferation of spleen cells of young compared to old mice, in vitro upon re-stimulation with the Her/2 neu antigen [46].

Also the group of Lustgarten found that cancer vaccination was less effective at old than at young age. They showed that young but not old mice developed long-lasting memory responses to a pre-B-cell lymphoma (BM-185). However, inclusion of CD80 to the BM-185 cell line (BM-185-CD80) plus agonist anti-OX-40 or anti-4-1BB (receptor for co-stimulation on T cells) mAb induced equally strong long-lasting memory responses at young and old age, suggesting the involvement of T cell responses [47]. Also in another study they found that adding anti-OX40 or anti-4-1BB mAb to a DC vaccine, resulted in vigorous anti-tumor responses in a syngeneic TRAMP-C2 model at young and old age, while without anti-OX40 or anti-4-1BB, protection was significantly better in young than in old mice [48]. Moreover, immunization of young and old mice with DC-TRAMP-C2 vaccine plus anti-OX40 or anti-4-1BB mAb resulted in improved CTL responses to apoptotic TRAMP-C2 cells in vitro upon re-stimulation, compared to the same vaccination without OX40 or anti-4-1BB mAb at old age, but the CTL responses were less vigorous compared to the same immunizations at young age.

Grolleau-Julius et al [49] showed that vaccination with a DC-OVA vaccine derived from young mice was less effective against B16-OVA melanoma tumors in old than in young mice, indicating the altered tumor microenvironment at older age and its effect on vaccination. Also the group of Zhang found that the tumor microenvironment was altered at old compared to young age. They demonstrated that the number of myeloid-derived suppressor cells (MDSC) increased in the tumor environment of old compared to young mice, and that this contributed to the age-related T cell unresponsiveness [48].

In our laboratory, we developed a DNA vaccine of Mage-b (pcDNA3.1-Mage-b) and tested this vaccine at young and old age in two syngeneic metastatic mouse breast tumor models, 4TO7cg and 4T1, both overexpressing Mage-b in metastases and primary tumors [49]. Vaccination of both models with Mage-b was highly effective against metastases and young age but not at old age, and this correlated with strong Mage-b-specific T cell responses in vitro and in vivo at young but not at old age [40]. Interestingly, we found that Mage-b vaccination activated macrophages and NK cells (producing IFNγ) in old mice [40]. In another more recent vaccine study with Mage-b delivered through a highly attenuated Listeria monocytogenes we found a dramatic effect on the metastases in the 4T1 model at young age [50•]. However, we discovered that this was not solely due to Mage-b, but rather to the direct infection and kill of tumor cells by Listeria [50•]. We concluded that this approach could be highly interesting for old age since T cells were not required. Indeed, the Listeria-based vaccine was equally effective against metastatic breast cancer at young and old age, while Mage-b-specific T cells responses were strong at young but almost undetectable at old age (unpublished results). Moreover, also NK cell responses were strongly activated by Listeria at young and old age, and may have contributed to the reduced growth of metastases at both ages as well. Since Listeria-infected tumor cells highly express Listeria proteins, the tumor cells become a highly sensitive target for NK cells and Listeria-specific CTL [50•].

Conclusions and future prospects

The main conclusion from the studies analyzed here is that also the innate immune system should be considered for testing as a potential candidate for immunotherapy at older age. This is based on the following findings. While the effect of cancer vaccination on growth of tumors and metastases could be strongly improved by tailoring the vaccine to older age, as shown in the preclinical studies analyzed here, in most cases improvement was not the result of T cell activation but rather the result of other immune cells stimulated by the vaccine. Although various functions of NK and NKT cells are decreased at old age, it is far less dramatic than the age-related decline in T cell function, and both cells play an important role in anti-tumor responses. However, also improvement of T cell activation against cancer through vaccination at older age should be further optimized. Below, new strategies to improve adaptive and innate immune responses against cancer at older age through vaccination or immunotherapy, respectively, are proposed below and summarized in Table I.

Table I.

Improvement of cancer vaccination or immunotherapy at older age

Adaptive Immune Responses -Generation of memory T cell responses at young age, when sufficient naïve T cells are available, and reactivation of memory T cells at old age -Recruitment of naïve T cells by IL-7 -Activation of co-stimulatory molecules by OX40 or 1-4BB1 -Activation of Th1 type immune responses by CpG-DNA -Elimination or polarizing MDSC, TAM or M2 macrophages -Elimination of Tregs -Reduction of lymphokines or factors that inhibits T cell activation

Innate Immune Responses -Activation of NK and NKT cells by bacterial products -Activation of NKT and NK cells by α-GalCer -Activation of NK and NKT cells by CpG-DNA -Sensitizing tumor cells for NK cell-mediated destruction by Bortezomib -Elimination or polarizing MDSC

Other Approaches -Delivery of genes into tumor cells through bacterial vectors that leads to tumor cell destruction: potential candidates are attenuated Listeria monocytogenes or other nonpathogenic bacteria such as Lactococcus lactis or Escherichia coli.

α-GalCer=α-galactosylceramide, MDSC=myeloid-derived suppressor cells, NK=natural killer, NKT=natural killer T, Th1=T helper 1, TAM=tumor-associated macrophages.

As mentioned above, innate immune responses should be considered as a potential target for improvement of immunotherapy against cancer at older age. For instance, NK cells and TCRγδ NKT cells could be activated by attenuated Listeria or αGalCer, since both cell types are present in sufficient numbers at older age, could be activated at older age, and exhibit anti-tumor activity. Not only αGalCer has an effect on MDSC, but also CpG ODN, vitA, and several chemotherapeutica may decrease the number or polarize MDSC into an immune-stimulating phenotype [51•]. It has been shown that CpG seems especially good at enhancing cellular and humoral immunity and promoting Th1-type responses in old mice [52]. Improved innate immune responses may also lead to improved adaptive immune responses. Since the number of MDSC increases with age and contributes to T cell unresponsiveness at older age, and since the number of TCRγδ NKT cells also increases with age, vaccination in combination with αGalCer or CpG, could enhance vaccine-induced T cell responses at older age. Also elimination of immune suppressing TAM and M2 macrophages may lead to improved T cell activation in the tumor microenvironment at young and old age.

T cell could also be activated through other strategies. For instance, the problem of lack of naïve T cells, one of the most important changes at older age, could be avoided by immunizing at young age when sufficient naïve T cells are present, followed by recall at old age to reactivate memory T cells. Such an approach has been successfully used for improving antibody production at older age [53]. Also, naïve T cells could be recruited by IL-7 [54]. However lack of naïve T cells is not the only hurdle to overcome. TAA are weakly immunogenic and T cells need help to become activated against TAA expressed by cancer cells. As shown in the studies discussed here, just adding IL-2 to TS/A tumor cells will improve anti-tumor responses but not memory responses to the tumor at old age. The best results so far has been shown by the group of Lustgarten by activating T cells against cancer through vaccination plus co-stimulation using anti-OX40 or 1-4BB mAb at young and old age. Also, elimination of T_regs could improve T cell activation at older age.

Finally, we have shown that an attenuated Listeria monocytogenes can be used to deliver genes directly and selectively in tumor cells in vivo [50•]. Also other nonpathogenic bacteria are currently under investigation for the delivery of genes selectively into tumor cells such as Lactococcus lactis and E.coli [55•]. Our results suggest that such an approach could be effective at young and old age.

In summary, despite all the obstacles that need to be overcome, vaccination against cancer is potentially the most promising approach. While cancer vaccination has limited success against late stage tumor development, they can be particularly effective where almost all other therapies struggle, i.e. against metastases and recurrence of cancer. The vaccine studies analyzed here show that improvement of vaccine efficacy at older age is possible, but that in addition to activation of T cells, the innate immune system also should be considered as a possible target for immunotherapy against cancer at older age. The advantage of activating adaptive immune responses by vaccination is its prophylactic and therapeutic application, while activating innate immune responses by immunotherapy can only be applied therapeutically. Finally, the results of these studies demonstrate the need of testing and tailoring cancer vaccines to older age in preclinical models before entering the clinic.