Studies have shown that radiotherapy, especially stereotactic body radiation therapy (SBRT) or hypofractionated radiotherapy (HFRT), can cause DNA damage, which leads to tumor cell death, induces release of pro-inflammatory factors, and enhances tumor immune stimulation cells and cytokines to remodel the tumor immune microenvironment (TIME) (Formenti and Demaria, 2013; Demaria et al., 2015). More importantly, radiotherapy can also promote immune cell infiltration and transform “cold” tumors into “hot” tumors, a status suitable to immune checkpoint inhibitors (ICIs) (Ostrand-Rosenberg and Sinha, 2009). Therefore, the combination of radiotherapy and ICI therapy has gained more and more attention and is considered a promising treatment for cancer (Formenti et al., 2018; Chicas-Sett et al., 2019). Dewan et al. (Dewan et al., 2009) found that the combination of HFRT and ICI therapy could induce an abscopal effect in a mouse model of breast cancer. Besides, Verbrugge et al. (Verbrugge et al., 2012) verified that ICI therapy enhanced the curative capacity of radiotherapy in established breast malignancy.

ICI treatment was given at different timepoints after radiotherapy in many studies; therefore the optimal time window remains elusive (Dovedi et al., 2015; Schapira et al., 2018). The PACIFIC study showed that interventional immunotherapy within 14 days after radiotherapy had the longer progression-free survival (PFS) and overall survival (OS) in patients with locally advanced NSCLC (Antonia et al., 2018). The KEYNOTE-001 study found that radiotherapy followed by immunotherapy had better PFS (4.4 months vs. 2.1 months) and OS (10.7 months vs. 5.3 months) in patients with advanced NSCLC (Shaverdian et al., 2017). The Pembro-RT study verified that pembrolizumab within 1 week after SBRT doubled the objective response rate (ORR), and prolonged PFS (6.6 months vs. 1.9 months) and OS (15.9 months vs. 7.6 months) in patients with advanced NSCLC (Theelen et al., 2019). Bauml’s study revealed that pembrolizumab in 4–12 weeks after local ablations had a PFS of 19.1 months in patients with metastatic NSCLC, tripling the previous PFS of 6.6 months (Aggarwal et al., 2019). However, Wegner (Wegner et al., 2019) showed that immunotherapy at least 3 weeks after radiotherapy would exhibit longer OS in a retrospective study. Therefore, to explore the right timing for ICI therapy intervention after radiotherapy has great significance in clinical treatment.

What is more, different fractionations also have different effects on TIME. Lugade (Lugade et al., 2008) found that a single 15 Gy was more effective than 3 Gy × 5 F in activating DC cells in lymph nodes in the B16 melanoma model. However, Schaue (Schaue et al., 2012) found that the 7.5 Gy × 2 F was better than a single dose of 15 Gy in inducing T cell initiation in another melanoma model. An appropriate fractionation could enhance an immunoreactive effect, but an extra high dose would cause damage to lymphocyte subsets and produce an immunosuppressive effect and immune dysfunction (Zitvogel and Kroemer, 2015). A single high-dose radiotherapy could cause damage and collapse of the tumor vasculature, which was not conducive to the infiltration of T cells into the tumor (Timke et al., 2008). It would cause radioresistance of tumor cells due to hypoxia caused by destruction of the vascular system (Barker et al., 2015). Radiation produces two-way immune effects like the “seesaw,” including positive and negative responses. The appropriate fractionation could push the immune effects into the positive response. Previous studies have shown that HFRT or SBRT was more capable of mobilizing local and systemic immune responses than conventional fractionation (Schaue and Mcbride, 2015). Since there are many choices in clinical practice, it is a conundrum as to which fractionation is appropriate and optimal.

As the intervention timepoints of ICI therapy after radiotherapy and the fractionations are various and controversial in previous studies, this study aimed at exploring the dynamic changes of TIME at different timepoints and fractionation doses of HFRT in NSCLC and providing an experimental basis for the optimal intervention timing and fractionation dose for the combination of radiotherapy and ICI therapy.

Radiotherapy can achieve a synergistic effect with immunotherapy by recruiting T cells to the irradiated tumor area and increasing the vulnerability of tumor cells to T cells (Galluzzi et al., 2017). Nevertheless, how to determine the optimal combination strategy of radiotherapy and immunotherapy remains an unsolved problem in clinical practice.

As regarded, timing between radiotherapy and immunotherapy as well as the fractionations are important considerations. Studies have shown that the timing of radiotherapy combined with ICIs depends on different types of tumors and ICIs (Young et al., 2016; Lesueur et al., 2018). Therefore, there is controversy about the optimal timepoints of ICIs and radiotherapy. Most studies have revealed that the comprehensive immune effect is positive and CD8⁺ T cells play a vital role in HFRT and SBRT (Reits et al., 2006; Lee et al., 2009). CD8⁺ T cells residing in tumors are mainly a group of high proliferation capability and exhausted function, which cannot effectively kill tumor cells (Li et al., 2019; Sanmamed et al., 2021). After HFRT, CD8⁺ T cells in peripheral blood migrate to tumors, kill tumor cells, and further activate DCs, thereby turning “cold” tumors into “hot” tumors. The effect of PD-1 inhibitors depends on the intratumoral infiltration of CD8⁺ T cells derived from peripheral blood (Huang et al., 2017; Yost et al., 2019). In this study, we found that immune cells, especially CD8⁺ T cells, in mice tumor tissues and peripheral blood showed a time-dependent dynamic change after radiotherapy. CD8⁺ T counts increased from 48 h to 3 weeks in 4.6 Gy × 3 F and 6.2 Gy × 2 F, and CD8⁺ PD-1 shortly increased from 48 h to 2 weeks in 6.2 Gy × 2 F and 10 Gy × 1 F. These results indicate that 48 h after HFRT may be a critical timepoint for immune response. Then we selected the timepoint of 48 h to verify in mouse tumor tissue and found that CD8⁺ T cells were not increased yet, indicating that the immune system has been activated in blood within 48 h after radiotherapy, earlier than in tumor tissue, which was consistent with the research by Matsumura (Matsumura et al., 2008). Besides, in peripheral blood of NSCLC patients, CD8⁺ T cells and CD8⁺ PD-1 were increased significantly from 24 to 96 h after radiotherapy. Furthermore, DC cells began to be significantly increased and continued to rise until 3 weeks in mice, and there was a similar trend in NSCLC patients before 96 h. DC PD-L1 showed a high level from 24 to 96 h in patients. On the other hand, MDSCs, which were regarded as suppressors of the immune cell response, increased from 2 weeks and continued to rise until 3 weeks in mouse peripheral blood. Besides, we also compared the ratio of CD4⁺/Treg and CD8⁺/Treg, respectively. All the ratios were consistent with the cell counts.

Collectively, the rule of changes in our study suggests that intratumoral infiltrating T cells being induced after local radiotherapy may be derived from the mobilization and chemotaxis of the systemic immune system. We speculate that CD8⁺ T cells are mobilized fully in the peripheral blood from 48 h to 2 weeks after radiotherapy, preferentially recruited and activated into tumor tissue; in that period, the efficacy of radiotherapy will be enhanced when combined with ICIs. This result is consistent with the PACIFIC study in which the timing of intervention ICIs is within 14 days. Most studies have shown that early interventional immunotherapy after radiotherapy has better efficacy (Shaverdian et al., 2017; Aggarwal et al., 2019; Theelen et al., 2019), and only a retrospective study reported at the 2019 ASCO meeting showed that immunotherapy given at 3 weeks after SBRT has better OS (Wegner et al., 2019). Using evidence-based medicine, the subgroup analysis of randomized controlled studies is more reliable than that of retrospective analysis. Therefore, it is possible that a shorter interval of immunotherapy after radiotherapy leads to better effect. However, due to the high single dose of SBRT or HFRT, synchronization immunotherapy or premature use of ICIs after radiotherapy may cause an increase in side effects. Our results are generally consistent with previous research, but differ slightly in the specific timepoints. We speculate that the difference may be due to race, number of cases, type of ICIs, and fractionation dose. In future, the optimal timepoints for combining radiotherapy with immunotherapy need further study.

When radiotherapy is combined with ICIs, the fractionation dose is another key factor for optimal outcome. Dewan et al. (Dewan et al., 2009) used the TSA breast cancer cell model and the MCA38 colorectal cancer model and found that when combined with CLTA-4 inhibitors, 8 Gy × 3F was significantly better than 6 Gy × 5 F or 12 Gy × 1 F regardless of local tumor control or abscopal immune response, suggesting that different fractionations also have different effects on the cancer therapy when combined with ICIs. In this study, the tumor volume was significantly reduced after radiotherapy with 4 different fractionations of the same BED. Moreover, the longer survival time appeared after radiotherapy of 6.2 Gy × 2 F. A clinical trial in the United Kingdom demonstrated the similar results that different fractionation doses with similar BED could cause different efficacy and prognosis of tumors (The et al., 2008). On the other hand, we found that CD8⁺ T cells were significantly increased in 6.2 Gy × 2 F, along with CD8⁺ PD-1 being increased in 6.2 Gy × 2 F fractionation in the peripheral blood of mice. In peripheral blood of NSCLC patients, CD8⁺ T cells and CD8⁺ PD-1 maintained a high level in 6.2 Gy × 4 F, and CD8⁺ T cells also increased in 10 Gy × 2 F. Besides, DC cells began to be significantly increased after radiotherapy in all fractionations in mice. In peripheral blood of NSCLC patients, we discovered that DC cells significantly increased and hit a small peak in 6.2 Gy × 4 F and 10 Gy × 2 F and then gradually returned to the baseline level, being accompanied by a high expression of DC PD-L1 in 6.2 Gy × 4 F and 10 Gy × 2 F.

Chen et al. (Welsh et al., 2020a) conducted a phase I/II randomized clinical trial comparing the efficacy of paprizumab alone versus combination with conventional fractionated radiotherapy (45Gy/15F) or SBRT (50Gy/4F) in the treatment of advanced NSCLC. The mPFS of the paprizumab + SBRT group was significantly better than that of the paprizumab + conventionally fractionated radiotherapy group (9.1 months vs. 5.1 months). In addition, Baas et al. (Welsh et al., 2020b) analyzed pembro-RT and MDACC studies and found that the combination of pabulizumab with ablative radiotherapy (24Gy/3F and 50Gy/4F) had a better ORR than non-ablative radiotherapy (45Gy/15F) and the pabulizumab alone group. Therefore, in terms of fractionation dose, studies tend to support SBRT or HFRT combined with immunotherapy, which can enhance the antitumor effect more than conventional fractionated radiotherapy. Here, we demonstrate that 6.2 and 10 Gy would be better doses of HFRT, which will produce an optimal immunoactivated status from 48 h after radiotherapy and the immune synergistic effect may be maximized when combined with ICIs. At these doses, both numbers and PD-1 expression of positive immune cells increase in TIME, which will be beneficial for ICIs in NSCLC. There is a reminder that the best timing of immune intervention could be 96 h later under 10 Gy fractionation, especially in two but not single doses.

Due to the limit of tail vein blood in each mouse, we detected the immune cells in tumor tissues instead of cytokines. Likewise, we detected cytokines in NSCLC patients instead of immune cells in tumor tissues. Interleukin-10 (IL-10) was discovered as an anti-inflammatory factor. However, increasing evidence revealed IL-10 can induce antitumor effects in an immune-dependent manner, which indicates that it also plays a bidirectional role in immune regulation of tumors (Sato et al., 2011). Studies have demonstrated that effector T cells were the main source of IL-10, and IL-10 can promote CD8⁺ T cell responses by binding to IL-10 receptor (Fujii et al., 2001; Wilke et al., 2011). Besides, Qiao (Qiao et al., 2019) generated a Cetuximab-based IL-10 fusion protein (CmAb-(IL10)2) and found that it could prevent CD8⁺ tumor-infiltrating lymphocyte apoptosis, further revealing that IL-10 could potentiate CD8⁺ T cell-mediated antitumor immunity. In this study, IL-10 was increased significantly from 24 to 96 h, when the number of CD8⁺ T cells and CD8⁺ PD-1 expression were also increased after 6.2G×4F radiotherapy in human peripheral blood. The result revealed that HFRT can induce CD8⁺ T cells to produce IL-10, which enhances the proliferation, differentiation, activity, and function of CD8⁺ T cells, especially in 6.2 Gy. In the TGF family, TGF-β1 is the most widely distributed in the immune system, and with the development of tumors, it can continuously promote the invasion, metastasis, and deterioration. Marie confirmed that TGF-β1 mainly plays a role in Tregs induction by regulating Foxp3 positive expression in Tregs (Marie et al., 2005). Furthermore, convincing evidence has verified that IL-4 directly acts on CD8⁺ T cells, reduces or even eliminates its cytotoxicity, and promotes the infiltration of Treg into tumors, thus establishing the immunosuppressive state and promoting tumor growth by avoiding the recognition of the immune system (Wynn, 2003). Here, we found that TGF-β1 decreased significantly in 3.7 Gy × 8 F, 4.6 Gy × 6 F, and 6.2 Gy × 4 F and that IL-4 decreased in 10 Gy × 2 F after radiotherapy, revealing that HFRT also achieves antitumor immunity by reducing TGF-β1 and IL-4, thus contributing to regulating Tregs and CD8⁺ T cell function in TIME. Besides, IL-2 and IL-5, which are regarded as positive immune regulators, are also found to increase significantly in 4.6 Gy × 6 F. Therefore, we find different cytokines change in different fractionations and speculate that the immune effect of different fractionation radiotherapies may be related to different target cytokines and induce an immune activation state after HFRT.

In the part of clinical data, we provided the bOR, mPFS, and mOS. However, there are many impact factors in this clinical practice, such as the baseline of patients, the number of treatment lines, and the different duration of immunotherapy. We do not think the efficacy and survival for the patients with or without immunotherapy are important results. Moreover, most of the clinical cases who received immunotherapy were not in the time window of 48 h to 2 weeks because we did not have any conclusion before this work.

To the best of our knowledge, this study is the first to prove the positive comprehensive immune effect on the timing and fractionation dose of HFRT. CD8⁺ T cells are the most important indicator of TIME, so we took CD8⁺ T cells as the main effector cells to judge the immune response. CD8⁺ T increased after HFRT both in peripheral blood of NSCLC patients and mouse models, revealing that HFRT can induce a positive immune response which may be beneficial for ICIs. Comprehensively considering the results of CD8⁺ T cells, DC cells, and cytokines, 3.7 and 4.6Gy are not recommended as the preferred fractionations compared with 6.2 and 10Gy. However, this is just our preliminary result. We are designing another animal experiment and prospective study based on our current findings and carry out clinical trials with larger samples for verification in the future.

In spite of the major strength of validating the changes of TIME after HFRT in both animal and human experiments, this study still has several limitations. First, taking into account clinical treatment efficacy, we adopted different fractionation doses between mouse and human. Second, we only choose LLC cells to establish the implanted mice model because it is not suitable to use xenografts derived from human cancer cells in immunodeficient mice to explore the immune microenvironment. Third, there were some other combination treatments during radiotherapy in different groups in patients. Finally, we did not perform cell isolation to detect cytokine secretion in tumor tissue and cell killing experiments in vitro.

In summary, HFRT induces the increase in CD8⁺ T cells and positive immune cytokine response in specific periods and fractionation doses. It was the optimal time window from 48 h to 2 weeks for immune response, especially in 6.2 Gy fractionation. The best immune response was 96 h later in 10 Gy fractionation, delivered twice instead of a single dose. During this time window, the intervention of immunotherapy may achieve a better effect. Future work should include exploration of the relationship between radiotherapy and TIME deeply.