Cancer immunotherapy: how low-level ionizing radiation can play a key role

The concept that in vertebrates, elements of the immune system specifically recognize and eliminate incipient neoplastic cells and protect thereby against the development of overt malignancy dates back to late 1950s [2, 3]. In accordance with this “cancer immunosurveillance” hypothesis, it was demonstrated that both immuno-compromised human patients and experimental animals are at increased risk of developing various neoplasms (reviewed in [107]). However, investigations by Stutman showed that chemically induced sarcomas or adenomas do not develop more often in athymic, T-cell-deficient, nude mice than in their wide-type, immunocompetent counterparts [108]. This observation seriously challenged the cancer immunosurveillance model and almost led to its abandonment [7]. Yet, evidence accumulated in recent years has helped to explain what was wrong with the original cancer immunosurveillance hypothesis and why some neoplasms progress to their clinical stage. Thus, it was found that innate immunity initially senses the presence of transformed cells and exercises the first line of anti-cancer defense. Soon after the activation, elements of the innate immune system promote induction of adaptive (specific) anti-tumor responses. However, owing to genetic and epigenetic changes in the developing neoplastic cells, tumors may become “invisible” to immune effectors through loss or aberrant expression of the MHC class I antigens (reviewed in [109, 110]) or of other molecules on cancer cells involved in triggering of the innate and/or adaptive immune responses [111, 112]. For example, a change in hydrophobicity of tumor cells may lead to suppressed expression of the “damage-associated molecular pattern” (DAMP) molecules necessary to alert the innate immune system to a “danger” incurred by the presence of aberrant cells [113]. Notably, even the “danger signals”, such as high-mobility group box 1 protein (HMGB1), can actually support cancer growth through stimulation of myeloid-derived suppressor cells (MDSCs) [114] or nurse-like cells [115] that create conditions favorable for cancer progression. Furthermore, tumor-associated specific antigens may assume forms similar to those expressed on normal cells and evade recognition as “non-self” by the immune system (reviewed in [116]).

Developing tumors create microenvironments that not only support neoplastic growth and metastasis, but also significantly reduce the effectiveness and corrupt the functions of both the innate and adaptive arms of anti-cancer immunity [10]. Among the immunosuppressive components of tumor microenvironments are various soluble factors such as IL-10, TGF-β, vascular endothelial growth factor (VEGF), prostaglandin E2 (PGE₂), HMGB1, indoleamine-2,3-dioxygenase (IDO), as well as soluble forms of phosphatidylserine, Fas receptors, and MHC class I-related chain A proteins (reviewed in [117, 118]). Another recently recognized immunosuppressive mechanism involves the activation of the so-called immune checkpoints whose function is to prevent overstimulation of the immune system (reviewed in [119, 120]). The two most important immune checkpoint co-inhibitory molecules likely to play a role in induction and maintenance of the immunosuppressive state within tumors are members of the immunoglobulin gene superfamily, the cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) whose expression on T helper cells suppresses the activity of cytolytic CD8⁺ T lymphocytes, and the programmed death 1 (PD-1; CD279) receptor primarily expressed on tumor-infiltrating lymphocytes (TILs) and monocytes which, upon combining with its respective ligands (PDL-1 and PDL-2), negatively regulates the anti-neoplastic function of T cells [121, 122]; in addition, the PD-1:PDL-1 interaction may promote the development and function of regulatory T (Treg) cells [123].

Active immunosuppression is also exerted by many non-specific and specific cellular effectors residing in or attracted to neoplastic tissue. Many different cells capable of inhibiting anti-cancer immunity and promoting cancer growth have now been identified. These include Treg lymphocytes [125, 126], MDSCs [127‐130], macrophages [128, 131‐133], natural killer T (NKT) [134‐136], Th17 [137‐139] and B lymphocytes [140‐143], but also neutrophils [131, 144‐147], dendritic cells (DCs) [148‐151], mast cells (MCs) [152], and mesenchymal stem cells [153‐155].

It has been finally well established that persistent activation of pro-inflammatory immunity facilitates cellular transformation and promotes tumor advancement. Unlike acute transient inflammatory responses which attract and activate elements of the innate immune system, chronic inflammation not only supports cancer progression, but also prevents the host from mounting effective immune defenses against it [129, 156‐162]. An intermediate role in this process of the inflammation-driven type 2 immune response is played by MDSCs which are attracted to inflammatory sites and facilitate tumor growth [163, 164]. Chronic inflammation, as a powerful driver of carcinogenesis, is associated with aberrant signaling mediated by the nucleotide-binding oligomerization domain (NOD)-like receptors expressed on DCs, macrophages, and lymphocytes [165, 166]. Critical immunosuppressive mechanisms operating in the tumor microenvironment during the advanced stages of carcinogenesis are outlined in Fig. 1.

In a series of our own experiments carried out in the relatively radiosensitive BALB/c mice and the relatively radioresistant C57BL/6 mice, both single and multiple WBI with X-rays at total doses ranging from 0.05 to 0.2 Gy reproducibly suppressed development of the induced neoplastic colonies in the lungs. Since the mice were whole-body irradiated before inoculation of the syngeneic tumor cells, the obvious suggestion was that the low-level X-ray exposures stimulated systemic innate anti-neoplastic reactions. Although we were not able to directly estimate the activities of immune cells in the lungs, a significant stimulation of the cytotoxic activities of NK cells and LPS- and IFN-γ-stimulated macrophages obtained from the spleen and peritoneal cavity, respectively, was detected in the X-ray-exposed mice from both strains. Interestingly, no elevation of the activities of these cells was detected after their in vitro irradiation at the same doses of X-rays indicating that enhancing of the NK- and macrophage-mediated cytolytic functions by LLR depends on the presence of factors occurring in in vivo but not the in vitro conditions [81‐90, 176‐179].

The above-described epidemiological and experimental observations of anti-neoplastic and immunomodulatory effects of LLR exposures provide grounds for clinical trials with WBI or HBI of oncological patients [101, 180]. Even before the aforementioned evidence gained significance, a few LLR-based therapy trials had been performed. In 1965, Holder reported on positive therapeutic effects of the low-level total-body irradiation of patients with multiple myeloma [181]. In 1975, Kazem described curative effects of WBI (0.15 Gy of γ-rays daily for the first 5 days and thereafter at 0.1–0.15 Gy every other day or at longer intervals to the total doses of 2.0–2.65 Gy applied over 5–12 weeks) of patients with disseminated stage III lymphomas [182]. Likewise, Chaffey et al. obtained complete remissions in 32 out of 40 patients with advanced lymphocytic lymphoma after repeated WBI (0.15 Gy twice a week to a total dose of 1.5 Gy) as an initial and only primary therapy [102]. Very promising results of low-level total-body exposures to γ-rays of patients with non-Hodgkin’s lymphoma (NHL) were also reported by Qasim [183] and Choi et al. [103]. In one of the later trials, 24 out of 26 patients with stage IV low grade NHL were in complete remission after two courses of low-dose, total-body irradiation at 0.75 Gy given in five fractions; when the initially pathological lymph node areas of these patients were 1 month later treated with the conventional RT (total dose of 40 Gy applied in 20 fractions), the disease remitted in yet another patient [184]. Similarly, Safwat et al. who used low-level total-body exposures (0.1–0.25 Gy several times a week to the total dose of 1.5–2.0 Gy) obtained complete remissions in 11 out of 35 patients and 2-year progression-free survival in 12 patients with relapsed and/or chemo-resistant NHL; in 14 patients, a significant increase in the percentage of CD4⁺ T cells in the blood was noted [105]. In addition, as demonstrated by Sakamoto et al., low-dose HBI with X-rays (0.1–0.15 Gy two times a week for 5 weeks) combined with local RT (2 Gy five times a week for 6 weeks) resulted in the 5-year survival of 84% of patients with stage I and II NHL as compared to 65% survival of patients treated solely with local RT (the difference significant at p < 0.05); in these patients, percentages of peripheral blood CD4⁺ T helper lymphocytes were significantly elevated [100].

While more clinical trials employing WBI or HBI with LLR are needed, they are hampered by radiation safety regulations based on the linear, no threshold (LNT) model of the dose–effect relationship assuming that any absorbed dose of radiation causes a finite increase in cancer risk. There is a growing consensus that the LNT hypothesis lacks a solid experimental foundation and is based largely on ideology rather than science [25, 169, 185‐193]. Hopefully, the many recent appeals from radiobiologists, physicians, and health physicists to various regulatory bodies and authorities to base the radiation protection system on scientific data indicating that there are quantitative and qualitative differences between the effects of low doses delivered at low dose rates and high doses delivered at high-dose rates [171, 187, 188, 190, 192, 194, 195] will lead to a revision of current radiation protection regulations, so that WBI with LLR can be tested in clinical trials.

As reviewed above, both acute and chronic exposures to LLR stimulate various anti-neoplastic immune reactions that are stifled or corrupted within the tumor microenvironment, especially during the later stages of carcinogenesis. Based on evidence indicating that tumor-inhibiting effects of LLR have been observed in both humans and experimental animals exposed in many different ways to single, multiple, and chronic irradiation with LLR, it may be argued that many, if not all, of the above-reviewed tumor-promoting immune mechanisms are likely to be blocked and/or reversed by such exposures (Fig. 2). Indeed, data indicating that LLR exposures may reverse the tumor-associated immune suppression has recently begun to emerge, even though many underlying LLR-induced mechanisms remain to be clarified. Based on the current evidence it may be postulated that, in addition to the direct activation of NK lymphocytes [83, 196, 197] and possibly other anti-tumor cytotoxic cells, LLR exposures enhance the “visibility” and/or susceptibility of cancer cells to immune assaults through stimulation of the expression by neoplastic and immune cells of molecules and ligands (e.g., CD2, B7, CD28, NKG2D) necessary for triggering of cytotoxic reactions [198‐200] and/or turning on “danger signals” in the neoplastic tissue [201, 202]. Furthermore, low-level radiation exposures are likely to alleviate or reverse the tumor-associated immune degeneracy through elimination or inhibition of the multiple cells, cytokines, and other factors associated with immunosuppressive loops induced by the tumor [175, 203‐207]. This could result in: (a) shifting of the immune response in favor of the anti-neoplastic phenotypes such as Th1 in the case of CD4⁺ T cells [97, 208], M1 in the case of macrophages [209, 210], and N1 in the case of neutrophils [211], (b) targeting the Treg-Th17 and Th17-DC interactions conducive to tumor regression [212‐214], (c) activation of the Toll-like receptor-mediated signaling in phagocytes and antigen-presenting cells [215‐217], (d) attenuation of the chronic inflammation pertinent to cancer initiation, promotion, and progression [94, 95, 170, 218, 219], and/or (e) down-regulation of the immune checkpoint molecules such as the CTLA-4, PD-1, and/or PD-L1 on T cells [198, 220‐222]. Indeed, one of the recent reports indicates that hypofractionated γ-ray irradiation of tumors induced in C57BL/6 mice combined with blockade of the PD-1 checkpoint stimulated accumulation of TILs associated with complete eradication of very large neoplasms [222]. In addition, there are numerous non-immune mechanisms triggered by LLR that positively affect normal, but not malignant cells [29]. These include: (a) increased cell proliferation, (b) stimulation of anti-oxidant reactions associated with the reduction of tissue injury, (c) improved repair of the DNA damage, and (d) metabolic shift from oxidative phosphorylation to aerobic glycolysis resulting in increased radioresistance of healthy tissues. Such outcomes are of primary importance for the combination of the LLR-based immunotherapy with classic forms of cancer therapy (i.e., high-dose RT and chemotherapy) that are lethal to normal cells and tissues and promote the formation of reactive oxygen species and inflammation. It is expected that other LLR-triggered reactions and mechanisms will be detected providing additional grounds for the use of the truly low-level exposures to IR in the treatment of cancer and, possibly, other diseases.

Cancer immunotherapy has matured from the application of several therapeutic agents, including tumor cell- and dendritic cell-based vaccines, anti-cytokine antibodies, checkpoint inhibitors, and genetically engineered T cells and stem cells, which collectively act to reverse immune suppression in the tumor environment and/or immune resistance of tumor cells (reviewed in [208]). There are also clinical trials combining such agents with local irradiation of tumors at moderate doses (i.e., >0.5–1.0 Gy per fraction) currently used in RT [16]. The recently acknowledged capacity of locally applied moderate or high (radiotherapeutic) doses of radiation to induce immunogenic death of cancer cells and local inflammatory reactions associated with stimulation of dendritic cells and enhancing the suppressed anti-cancer immunity has been employed as an adjuvant to improve the efficacy of existing immunotherapy protocols (reviewed in [11‐19, 21]). However, such exposures can also cause persistent inflammation and multiple cell death in normal tissues, impede various immune and other physiological functions, and increase the risk of secondary primary cancers. In contrast, LLR exposures do not kill or impair and actually support functions of normal cells and tissues, selectively eliminate precancerous and transformed cells, attenuate rather than induce chronic inflammation, stimulate various anti-neoplastic reactions of the immune system, and are not associated with the development of secondary malignancies [21, 29, 94, 95, 170]. Finally, as indicated by the above-reviewed results of experimental and epidemiological studies as well as several clinical trials, WBI or HBI with LLR are not likely to induce any untoward side effects and can thus be used in treatment of patients with systemic or metastatic cancer.

It is, therefore, time to employ whole- or half-body exposures to LLR (alone or as an adjuvant to conventional therapeutics) to restore the efficacy of systemic anti-cancer functions of the immune system, the most potent guardian against neoplasia. This approach is expected to mediate improved clinical responses in cancer patients, as well as protect normal tissues from the well-known adverse effects associated with standard chemo- and radiotherapy used in contemporary cancer therapeutics.