Incomplete Freunds adjuvant hampers cancer vaccine efficacy in mice

Of 98 cancer vaccines formulated with incomplete Freund’s adjuvant (IFA) – a mineral oil-based adjuvant included in many vaccines to stimulate the immune response and increase vaccine longevity – and tested in U.S. government-approved clinical trials, none have been FDA-approved (

Why so many failures? A research team led by Dr. Willem Overwijk of the M.D. Anderson's Department of Melanoma Medical Oncology at the University of Texas suggests that IFA is at least partly to blame. Overwijk and his team found that, in mice, IFA-based vaccines induce cancer-specific CD8+ T cells to accumulate at vaccination sites instead of at tumors. There, they are incapacitated and eventually die. In contrast, non-IFA based vaccines significantly ameliorate this problem, suggesting that they may be more effective in fighting human cancers (Anderson Cancer Center 2013; Hailemichael et al. 2013).

Incomplete Freund’s adjuvant correlates to reduced cancer-specific CD8+ T cell life spans

Overwijk and his team wanted to know why cancer vaccines so often increase the number of circulating tumor-specific CD8+ T cells without eradicating the cancer. Based on their and others’ research in mice and humans, they wondered if IFA might be to blame. They tested their hypothesis in a variety of mice.

First, the Overwijk team used mice with traceable CD8+ T cells: namely, pmel-1 trangenic mice - B6.Cg-Thy1a/Cy Tg(TcraTcrb)8Rest/J (005023) - which carry a melanoma-specific T cell receptor transgene and the T lymphocyte-specific Thy1a (Thy1.1) allele. By using appropriate antibodies and flow cytometry, they could track the fate of melanoma-specific CD8+ T cells originating from these mice. They then observed that if these mice are vaccinated with an incomplete Freund’s adjuvant-based vaccine containing a melanoma peptide, only a short-term increase in melanoma-specific CD8+ T cells and negligible memory results, and that T cells are unresponsive to a booster shot. In contrast, the CD8+ T cell increase persists if the mice are vaccinated with a non IFA-based vaccine (but not if it is administered along with the IFA-based vaccine or if the IFA-based vaccine is administered 30 days later).

The Overwijk team observed similar results on ovalbumin-specific CD8+ T cells when OT-I and OT-II TCR transgenic mice – C57BL/6-Tg(TcraTcrb)1100Mjb/J (003831) and B6.Cg-Tg(TcraTcrb)425Cbn/J (004194), respectively – were administered IFA-based, ovalbumin peptide vaccines. They also observed that a range of vaccine doses produces similar results.

These preliminary results suggested that the immune hyporesponsiveness associated with IFA-based vaccines is not peculiar to the antigens in the vaccine, the T cell specificity to those antigens or the vaccine dose.

Overwijk and his colleagues also found that neither B nor CD4+ cells appear to play a role in the effects of IFA-based vaccines: neither B cell-deficient mice (B6.129S2-Ighmtm1Cgn/J (002288)) nor CD4 knockout mice (B6.129S2-Cd4tm1Mak/J (002663)) are protected from the IFA-based vaccine effects.

IFA-based vaccines attract T cells to vaccination site and lymph nodes

Overwijk and his colleagues wondered what the fate of pmel-1 effector T cells would be if they were transferred to mice with B16 melanoma. They found that whereas these cells multiply robustly and accumulate in tumors if transferred to mice vaccinated with a non IFA-based vaccine, they accumulate at the vaccination site in mice vaccinated with both non IFA- and IFA-based vaccines. The IFA-based vaccination sites induce the expansion of and attract pmel-1 T cells for as long as 96 days post-vaccination.

The researchers also observed that an antigen in a commonly used IFA-based, HLA-A0201-binding, melanoma peptide vaccine is persistently presented by antigen-presenting cells in the vaccine-draining lymph nodes (VdLNs) of HLA-A0201 transgenic (C57BL/6-Tg(HLA-A2.1)1Enge/J (003475) ) mice. These results indicated that IFA-mediated antigen persistence sequesters T cells to the vaccination site and VdLNs.

IFA-induced T cell hyporesponsiveness is mediated by interferon-g and Fas ligand

The B6.129S2-Cd4tm1Mak/J (002663) mouse is CD4 cell-deficient.

The Overwijk team observed that the death receptor encoded by the Fas gene is highly expressed on the surfaces of pmel-1 T cells at IFA-based vaccination sites. Knowing that interferon-g (IFNG) induces the Fas and Fasl (Fas ligand; FasL) genes and T cell apoptosis, they explored the roles of IFNG and FasL in IFA-mediated T cell hyporesponsiveness. They found that pmel-1 cell apoptosis at IFA-based vaccination sites is considerably reduced in FasL-deficient B6Smn.C3-Faslgld/J (001021) mice and in Ifng-deficient B6.129S7-Ifngtm1Ts/J (002287) mice.

Additionally, IFA-based vaccination sites become populated with immunosuppressive cells that show IFN-γ-driven upregulation of FasL and the immunosuppressive PD-1 ligand, PD-L1.   Collectively, the experiments led the Overwijk team to conclude that IFA-based vaccines allow constituent antigens to persist at VdLNs and vaccination sites. There, they induce an immunosuppressive, T cell-hostile environment, where antigen-specific T cells are sequestered, primed, and undergo IFNG/FasL-mediated apoptosis.

Non IFA-based melanoma vaccines improve anti-tumor efficacy

The Overwijk team wondered if vaccine efficacy would be improved if incomplete Freund’s adjuvant were replaced with a non-persistent adjuvant. Indeed, they found that a vaccine containing a combination of saline and three immunostimulatory molecules – a CD40-specific antibody, a toll-like receptor 7 (TLR7) agonist, and interleukin 2 (IL2), collectively called covax – is short-lived, induces a pro-T cell survival environment, mitigates T cell apoptosis and favors strong pmel-1 T cell expansion that can be boosted. When administered to pmel-1 transgenic mice, such a vaccine specific for melanoma induces T cell priming and accumulation at melanoma tumors rather than at vaccination sites and suppresses the growth of even established, palpable tumors. It also efficiently boosts melanoma-specific T cell numbers in pmel-1 mice vaccinated with a different non IFA-based vaccine.

In summary, Dr. Overwijk and his colleagues demonstrated that, in mice, incomplete Freund’s adjuvant markedly protects a cancer vaccine's constituent antigen from degrading at the vaccination site. The persistent antigen initiates a chemokine-driven feedback loop in which cancer antigen-specific T cells are primed and expand in the vaccine-draining lymph nodes (VdLNs) and are recruited to the inflamed vaccination site. There, they are incapacitated, and, along with nearby vaccination site tissue, are destroyed. In contrast, a short-lived vaccine induces long-lived, functional anti-tumor T cells that preferentially localize to tumor sites. These findings suggest that CD8+ T cell-inducing cancer vaccines would be more effective if based on rapidly biodegradable adjuvants.