Interactions between radiation therapy and immunotherapy: the best of two worlds?
D. H. Thamm
The Animal Cancer Center
Department of Clinical Sciences
College of Veterinary Medicine and Biomedical Sciences
Colorado State University
300 West Drake Road
Contrary to earlier thought, there is now accumulating evidence that local radiation therapy and various form of immunotherapy may be complementary. This may occur as a result of radiation-induced changes to the tumour cells themselves, the tumour microenvironment, the tumour vasculature and modulation of antigen-presenting and effector cell function. This study discusses recent findings and potential applications of this apparent synergy.
Despite recent advances in the treatment of neoplasia, locoregional tumour control and the control of distant metastasis are still difficult to achieve in many patients with cancer. Combined modality therapy has the potential both to improve local tumour control and to decrease the incidence and/or progression of metastasis. When treatment modalities with different but complementary antitumour mechanisms and non-overlapping side-effects are combined, the potential for effectiveness is increased while toxicity remains acceptable.
Combinations of radiation therapy (RT) and immunotherapy (IT) meet the criteria listed above, as their mechanisms of action are different but potentially synergistic. RT is a primarily local modality, while IT has the potential to exert its effect both locally and systemically. A very small increase in antitumour efficacy of IT has the potential to translate into a meaningful therapeutic difference when applied over a fractionated RT protocol. In addition, an increase in metastatic frequency when local RT is applied has been documented in certain experimental and spontaneous tumours1,2, and concurrent treatments that might enhance systemic immunity have the potential to help control the development of metastasis in the face of RT. Thus, combined modality cancer treatment with RT and IT deserves to be thoroughly investigated in an attempt to better control regional and metastatic disease.
RT and immune function
Teicher et al.3 made the following statement concerning the effects of RT on tumour immunity:
Local radiation, like other tumor cell cytotoxic therapies, facilitates T cell-mediated tumor regression. The exact mechanisms for the response are unknown, but they could involve decreased synthesis of immune suppressant factors, increased expression of immunogenic epitopes by tumor cells, or the generation of immune-enhancing cytokines as a result of inflammation secondary to tumor cell killing.
This was convincingly shown in a murine study, in which an immunogenic, established murine sarcoma was rendered significantly less radioresponsive when implanted in an immunosuppressed host. The authors were able to document an immunity-related 20-fold cell kill enhancement in immunocompetent mice when compared with immunosuppressed mice4. Increases in cytotoxic T-cell infiltration and in situ CD8 T-cell proliferation have been seen in irradiated tumour tissue when compared with unirradiated tumour tissue5–7, and the degree of inflammatory infiltrate after RT has been positively correlated to treatment outcome in humans with cervical carcinoma8,9. A study using coarsely fractionated RT before surgery in patients with carcinomas of the breast and stomach showed a significant increase in lymphoplasmacytic infiltrate compared with those patients receiving surgery alone10, showing the potential immunomodulatory effects of RT in a relevant human patient population.
Local RT can alter beneficially the characteristics of tumour cells themselves. Observed changes include increased expression of major histocompatibility complex (MHC) molecules11–18, increases in expression of tumour-associated antigens (TAA)11,18–20 or the induction of the expression of novel TAA17 and increases in cytokine gene expression21. Upregulation of important costimulatory signals such as B7.1 and intracellular adhesion molecule-1 (ICAM-1) has also been seen14,16,18,22,23, as has upregulation of heat shock protein chaperone molecules, which are responsible for antigen presentation and capable of serving as maturation factors for dendritic cells (DCs)24,25. Additionally, increased susceptibility to both natural killer (NK)-mediated lysis and specific immune effector cell killing has been observed26–28, in part mediated through upregulation of Fas expression by the tumour cells, which leads to enhanced Fas-dependent cytotoxic T lymphocyte killing18,29,30 and upregulation of ligands for the NKG2D receptor present on NK and certain T cells31,32.
Local RT can also have effects on systemic immune function. RT can increase granulocyte–macrophage colony-stimulating factor secretion from human peripheral blood monocytes, increase monocyte tumour cytotoxicity33 and increase peripheral blood NK activity28,34. In vitro, as little as 10 Gy (1000 rad) of RT is capable of priming macrophages (Mφ) for tumour cell killing and increasing Mφ MHC I expression35, and in some circumstances, irradiated Mφ exhibit an increased capacity for cytokine secretion and antigen presentation to T cells36. A recent study showed that while RT can diminish the capacity of DCs to present endogenous antigens, it appears to enhance their capacity for presentation of exogenous antigens37. Additionally, it appears that antigen-primed T cells are more radioresistant than unprimed, quiescent T cells36,38. Thus, RT may select for activated, antigen-specific T cells while eliminating other T-cell clones. It has also been shown that Th1 and Th2 subsets appear to have differing radiosensitivities: RT preferentially inhibits the activity of Th2 clones, thus potentially generating effector cells capable of inducing more efficacious antitumour responses39.
When compared with normal, quiescent endothelium, tumour endothelium is known to express decreased levels of vascular adhesion molecules such as ICAM-1(CD54), partially as a result of basic fibroblast growth factor-mediated suppression. This has the potential to impair the infiltration of leucocytes into tumours, thus protecting the tumour from immune destruction40–42. RT can increase ICAM-1 expression in tumour and normal vessel endothelium43 and tumour cells themselves14,44, which may contribute to an immune component of RT-induced tumour regression45.
Table 1 summarizes the potential immunomodulatory effects that RT can have on the tumour or effector cells. Enhancement of immune function using IT with RT has the potential to capitalize on these effects, thus creating a more robust local and systemic antitumour response.
Table 1. Possible immunomodulatory effects of radiation therapy in cancer
| Increased MHC expression|
| Increased NKG2D ligand expression|
| Increased heat shock protein production|
| Increased costimulatory molecules (B7.1, ICAM-1)|
| Increases or alterations in tumour-associated antigen expression|
| Increase in Fas expression|
| Increased cytokine expression|
| Upregulation of ICAM-1|
| Enhanced leucocyte trafficking|
| Activation of macrophages|
| Activation of DCs (enhanced presentation of exogenous antigens)|
| Selective destruction of Th2 lymphocytes over Th1|
RT combined with systemic nonspecific biological response modifiers has been proven additive or synergistic in several rodent models of neoplasia46–49. Human clinical trials investigating this approach have also been performed in patients with various malignancies50–54. Mixed results have been obtained, with some studies showing increased response rate or median survival, while others showing no significant difference.
Several recent studies have sought to increase DC/macrophage activation or number in conjunction with RT. Systemic or local treatment with cytosine–phosphate–guanosine (CpG) oligonucleotides, potent activators of Toll-like receptor 9, was shown to be additive or synergistic with RT in several mouse models55–57, as has the direct intratumoural injection of DC58–60. Interestingly, the systemic (subcutaneous or intravenous) administration of DC was similarly shown to be synergistic with RT, and increased trafficking of DC into irradiated tumour tissue was observed61. Additionally, a recent study showed that CpG exposure protects T cells and macrophages from RT-mediated cytotoxicity62. Furthermore, lipid A stimulation of macrophages was shown to dramatically increase their production of nitric oxide, which was capable of increasing the radiation sensitivity of the tumour cells63.
Specific IT consisting of tumour cell vaccines64–66 or adoptive IT with T cells67,68 has also been combined with RT in murine models. Both increased local tumour response and decreased metastasis have been observed. One prominent effect of RT appears to be in the promotion of lymphocyte infiltration into the irradiated tissue23,67, although some studies have failed to show this phenomenon68. Two clinical trials evaluating specific IT in the form of intralesional DC administration or a peptide vaccine combined with RT have been evaluated in early-stage human clinical trials69,70. Both studies showed the safety of this approach; however, evaluation of efficacy was not possible in these preliminary studies.
A relatively large body of literature exists investigating the effects of systemic therapy with recombinant cytokines and local RT. RT combined with interleukin (IL)-6 and macrophage colony-stimulating factor enhanced local tumour control and, in some cases, decreased metastasis in mouse models71. RT prior to systemic IL-2 therapy has been shown to increase the efficacy of the IL-2 in preclinical murine models of neoplasia72. Both in a spontaneous and in an experimental metastatic model of renal cell carcinoma (RCC), the application of RT enhanced local tumour control and decreased the development of metastasis. Radiation alone caused an increase in Mφ infiltration into tumours, and IL-2 caused an influx of T cells. Combined therapy resulted in a massive influx of T cells, Mφ and NK cells, suggesting a combined effect on tumour control15,73.
A decrease in the incidence of hepatic metastases from colon carcinoma was documented when a combination of systemic IL-2, tumour-infiltrating lymphocyte administration and either hepatic RT or total body irradiation was used, but this effect was lost if the liver was shielded from radiation. This suggested that inclusion of the tumour-bearing organ was necessary for optimal response74. In another experimental metastasis model, combined systemic IL-2 therapy and half-lung RT decreased murine RCC metastasis formation in both the irradiated lung and the opposite lung, suggesting that RT was capable of potentiating the systemic immune response15.
Potentiation of radiation response has also been shown with systemic IL-12 administration. Systemic IL-12 administration plus local RT induced a tumour growth delay of 5.4–10.6 days in a Lewis lung carcinoma model compared with RT alone, conferring a dose-modifying factor of 1.1–2.8 depending on the dose of IL-123,75.
There have been two clinical trials in humans exploring the combination of local RT and systemic IL-2 therapy. In both studies, there appeared to be no additive toxicity from the combined treatment, but there was little evidence of an additive or synergistic effect with regard to local tumour response or the development of metastasis76,77. However, a study comparing systemic IL-2 and lymphokine-activated killer therapy following chemotherapy, RT with chemotherapy, or RT alone in patients with lung cancer showed a significant survival advantage for the patients treated with IT78.
Local cytokine administration circumvents some of the potentially serious toxicities encountered with the systemic administration of cytokines such as IL-1279,80 and IL-281,82 and provides sufficient local cytokine levels for immune cell activation. Low-dose local IL-2 administration plus RT is capable of enhancing local control and inducing regression in distant, nonirradiated tumour without enhancing radiation toxicity83–85. This approach was translated into a study of humans with head and neck squamous cell carcinoma, in which patients were randomized to receive either perilymphatic IL-2 administration combined with surgery and/or RT or surgery/RT alone. This study showed a significant benefit in terms of survival and disease-free interval in those patients treated with IT and a significant increase in tumour inflammatory infiltrate86. A summary of clinical investigations of RT/IT combinations performed to date is given in Table 2.
Table 2. Published RT/IT clinical trials in humans
|RT plus nonspecific immunomodulation (BCG, OK-432, levamisole, sizofiran, Corynebacterium parvum)||Cervix, HNSCC, SCLC, rectal CA||Randomized phase III||5 of 7 studies showed no difference between groups. One study showed reduced recurrence and one showed improved 5-year survival in the IT arm||DiSaia et al. 50, Kimura et al. 52, Padmanabhan et al. 54, Cheng et al.87, O’Connell et al.88, Jackson et al.89 and Miyazaki et al.90|
|RT plus interferon-β||NSCLC||Phase I||Evidence of biologic modulation: MTD established||Byhardt et al. 91|
|RT plus systemic IL-2 ± TIL||RCC, multiple tumours||Single-arm phase II||No increased toxicity compared with IT alone: no evidence of increased response rate versus IT alone||Lange et al.76 and Redman et al.77|
|RT ± perilymphatic IL-2||HNSCC||Randomized phase III||Addition of IL-2 significantly increased DFI and OS||De Stefani et al. 86|
|RT ± allogeneic viral oncolysate vaccine||Cervix||Randomized phase III||No difference in DFI or OS between groups||Freedman et al. 92|
|RT ± autologous vaccine||Rectal CA||Randomized phase III||No difference in DFI or OS between groups||Hoover et al. 93|
|RT ± peptide vaccine, GM-CSF, IL-2||Prostate||Randomized phase II||Well tolerated. Majority of vaccinated patients developed specific T-cell responses||Gulley et al. 70|
|RT plus intralesional DC injection||Hepatic CA||Phase I||Well tolerated. Majority of patients developed specific T-cell responses. Only 50% showed increased NK activity||Chi et al. 69|
These examples provide compelling proof of principle that combined modality RT/IT is capable of potentiating radiation response and antitumour immunity. However, the disappointing results in the few human trials conducted to date and the significant toxicity associated with systemic cytokine administration suggest that additional studies must be performed to refine this methodology. Current studies from our group are evaluating several different RT/IT combinations for various malignancies. These include neoadjuvant RT plus the liposomal immunomodulator liposome muramyl tripeptide phosphatidylethanolamine (L-MTP-PE) for canine osteosarcoma; coarsely fractionated RT, L-MTP-PE and allogeneic tumour cell vaccination for canine oral melanoma and half-lung RT plus intraperitoneal cationic liposome–DNA complexes for osteosarcoma lung metastasis.
The author wishes to acknowledge Varian Medical Systems, Inc., for financial support of the radiation therapy/immunotherapy clinical trials ongoing at the Colorado State University Animal Cancer Center and Drs S. Formenti, C. Guha, E. Lord, W. McBride, L. Milas, D. Panical, J. Schlom, H. Streicher, R. Sutherland and D. Vail for their helpful discussions.