Our study demonstrated the promising effects of G-CSF combined with radiation by analyzing 1312 patients with NSCLC. Because this study is not a randomized study, the propensity score (PS) matching method was used to eliminate selection biases. We found that concurrent administration of rhG-CSF improved radiation response significantly and generated abscopal effects in NSCLC.
Currently, rhG-CSF is widely used in clinical to stimulate the production of neutrophils in patients who suffer from chemotherapy or radiotherapy-induced neutropenia. However, the anticancer effects of G-CSF are still unclear. Several studies have demonstrated the promising effects of G-CSF with chemotherapy for acute myeloid leukemia (35, 36). The underline mechanisms were that G-CSF drove leukemic cells into the cell cycle, which was more sensitive to chemotherapy. However, rare clinical evidence showed for solid tumors. A study that included 7 patients indicated the synergistic effect of G-CSF and chemotherapy for metastatic melanoma (37). On the contrary, some researchers have described the potential adverse effects of G-CSF (38–40). Shojaei et al. found that G-CSF mediated tumor refractoriness to anti-VEGF therapy by stimulating myeloid cell mobilization and angiogenesis in mouse models (41). Another preclinical study also found that G-CSF diminished the antitumor activity of chemotherapy by contributing to tumor revascularization (38). These controversial results may be related to that G-CSF can increase or activate the activity and recruitment of not only antitumor TANs but also protumor TANs and other myeloid-derived suppressor cells (MDSCs)(42). In addition, different types of neutrophils and other MDSCs have variable effects depending on the therapeutic regimen used and tumor growth stage.
Recent studies have revealed that radiation could enhance immune response that mediates tumor regression (43–45), but the immunologic effects of radiation has not been fully explored. TANs play important roles in tumor immunity and tumor development. Traditionally, TANs were mostly described for their ability to promote tumor progression through different mechanisms, including accessing tumor cell proliferation, suppressing the function of CD8+T cells (8, 46). Recently, accumulating data suggested that TANs also play roles in antitumor immunity. TANs could induce tumor cell apoptosis by secreting cytotoxic molecules such as ROS (32), or promoting the detachment of tumor cells leading to tumor cell death (32). Therefore, as the main stimulator of neutrophils, G-CSF had reversible effects when combined with radiation. Takeshima et al observed that radiation recruited tumor-associated neutrophils (RT-Ns) and induced apoptosis of tumor cells by increasing the production of reactive oxygen species. Also, they found the concurrent administration of G-CSF enhanced RT-mediated antitumor activity by activating RT-Ns (32). However, another study conducted by Kim et al. found that administration of G-CSF with radiotherapy promoted tumor growth by stimulating vascularization in tumor-bearing mice (47). Most of these data stay at the preclinical level. As to our knowledge, our study was first to demonstrate the synergistic effect of G-CSF to combine with radiation in patients with NSCLC. G-CSF significantly promoted radiation response and increased abscopal effects. In our results, low baseline neutrophils (<3.59) were found correlated with better response in univariate analysis, but it was not an independent factor. We believe the predictive effect of baseline neutrophils was dependent on its correlation with the administration of G-CSF. Because G-CSF was mostly given to patients with neutropenia in the study. Further studies are needed to find out the predictive effect on neutrophils, emphatically on TANs. Besides, KPS score >80, squamous carcinoma, negative N stage, M stage and high BED were found favorable indicators for PFS, which was consistent with previous studies (48, 49).
To better understand the underlying factors for the abscopal response, we compared the clinicopathological factors between the abscopal group and non-abscopal group. Radiation site, BED, fractionation regimens, radiation intention and synchronous chemotherapy were not associated with abscopal response in our results, but further studies are needed. The underlying mechanism of abscopal response has not been fully explored. The in-situ vaccination effect of radiation was widely accepted as the potential mechanism of abscopal response (50). Radiation released tumor associated antigens and could modulate immunological pathways, increasing tumor antigen presentation, activating tumor-specific cytotoxic T cells, as well as enhancing T-cell homing and function in tumors. Previous study reported that extracranial radiation showed more advantage in activating antitumor immune response (51). Also, enough BED of radiation is needed to induce the immunologic effect of radiation (52). Camphausen et al. found that the abscopal antitumor response was significantly more potent when mice were treated with greater BED. Tang et al evaluated multiple SBRT dose fractionation regimens (BED ranging from 96 to >100 for liver and lung metastases and found abscopal responses in 10% of patients (53). To be noted, these enhanced abscopal responses cannot be simply attributed to greater BED, since different fractionation regimens were used in those studies. Hypofractionation was reported more favorable in immunostimulants with a marked increase in T cell priming than conventional fractionation. Further studies are needed to figure out the best radiation regimen to maximize antitumor immunity.
The major limitation of our study was the retrospective study design, which failed to complete the elimination of bias in the selection of patients, even by using the PS matching method. The only way to eliminate all of the confounding factors and to verify the antitumor effects of G-CSF combined with radiation is to conduct a prospective randomized clinical trial (RCT). Our results provided strong evidence for clinical usage and to conduct the RCT subsequently. Despite the indicative effect of chemotherapy was not significant in the study, the contribution of systemic therapy to abscopal response reported herein cannot be defined, including chemotherapy, targeted therapy and immunotherapy.
In conclusion, our results showed that concurrent administration of G-CSF improved radiation response significantly and generated abscopal effects in NSCLC. These findings suggested that the combination of radiotherapy and G-CSF is potential a new strategy and should be further evaluated in prospective clinical trials.