The efficacy of adding hyperthermia to the treatment of advanced NSCLC patients based on the states of EGFR

Background: The study aims to explore the efficacy of adding hyperthermia to the treatment of advanced NSCLC patients based on the states of epidermal growth factor receptor (EGFR). Patients and methods: We included 205 advanced NSCLC patients who were received hyperthermia plus other treatment (hyperthermia group) or non- hyperthermia and other treatments (non- hyperthermia group). The OS and progression free survival (PFS) were retrospectively estimated. Using Kaplan-Meier and the log-rank test compare the OS and PFS between the groups. Results: The median follow-up was 22 months. The Univariate analysis have shown that 1-year OS and PFSfirst rates in the hyperthermia group and non- hyperthermia group were 83.3% vs 71.5% (P=0.010) and 62.0% vs 42.7% (P=0.001). The subgroup analyses revealed that patients didn’t have EGFR mutant who received hyperthermia had significantly higher 1 year OS and PFSfirst rates than those treated with non- hyperthermia (OS: 79.1% vs 65.2% P=0.037, PFS: 64.2% vs 36.5%, P=0.001). For patients with EGFR mutation, there was no significant difference between the two groups. The PFSfirst in first-line and PFSpost in posterior-line was no significant difference between the groups. Conclusions: This retrospective study revealed that adding hyperthermia to the treatment of NSCLC patients without EGFR mutation had better prognosis than those who did not adding hyperthermia to the regimen. Moreover, adding hyperthermia in first-line or in posterior-line treatment was no significant difference. However, these results need more prospective studies to confirm the conclusions.

have shown that IO not only could prolong the PFS, but also could improve the OS [8][9][10][11][12]. However, patients with EGFR mutation who treated with IO had limited efficacy. Previous meta-analysis study showed that IO did not benefit patients with EGFR mutation, but increased adverse reactions [13]. And the latest phase II single arm trial had been ended early because of lacking significant clinical efficacy [14]. The efficacy of adding IO to the treatment of patients with EGFR mutation in advanced NSCLC is still controversial.
Whether IO should be added to the regimen of advanced NSCLC with EGFR mutated remains unclear. It is also unclear why adding IO to advanced NSCLC with EGFR mutated decreases the efficacy of treatment. It may be related to the immune microenvironment. Hyperthermia is an effective measure to overcome tumor hypoxic cells, and it mainly kills the S phase cells, which were resistant to radiotherapy. Moreover, it can reduced the production of immunosuppressive cells. Guo et al. [15] found that hyperthermia could transform Treg cells into Th17 cells by increasing IL-6 levels. Wendt MK [16] found hyperthermia could also indirectly reduce TGF-β by up-regulating the expression of miR-10b to inhibit breast cancer cell proliferation. Guo et al. [17] found hyperthermia could not only activate Cytotoxic T lymphocyte (CTL) by increasing the expression of B7 molecules and ICAM-1, but it could also increase CD+4 /CD+8 ratio. Therefore, hyperthermia can improve the tumor microenvironment (TME), and promote the efficacy of IO in advanced NSCLC.
On the basis of these theories, we hypothesis that patients with EGFR mutation may not benefit from hyperthermia. Our study aims to explore the prognosis of adding hyperthermia to the treatment of advanced NSCLC with different EGFR mutated states and evaluate the efficacy of adding hyperthermia as the firstline treatment method or as a posterior-line method.

Characteristics of patients
The median follow-up was 22 months. A total of 205 patients were included the study. Among the 205 III-IV NSCLC patients, 77 patients in the hyperthermia group and 128 patients in the non-hyperthermia group. The baseline characteristics of the two groups were shown in Table 1 Table 2).

Subgroups analysis
The subgroups analyses have shown that the 1-year OS rate in hyperthermia group and in non-hyperthermia group was 79.1% versus 65.2% (P=0.037, Figure 2A) in the non-EGFR mutant subset, while the 1-year PFSfirst rate in hyperthermia group and non-hyperthermia group were 64.2% versus 36.5% (P=0.001, Figure 2B). The outcomes had significant difference between the two groups.

Prognostic factors
Multivariate Cox regression analysis demonstrated that smoke, clinical stage and hyperthermia were the independent prognostic factors for PFSfirst and OS (Tables 3, 4).

Posterior-line analysis
Of patients who occurred progress and had not received hyperthermia in first-line treatment, twenty-nine patients were treated with hyperthermia as a posteriorline treatment method. Of the 29 patients, 12(41.4%) relapsed, 2 (6.9%) had distant metastasis and 6(20.7%) both had recurrences and distant metastasis. The median PFS were no significant difference between the first-line treatment and posterior-line treatment (24m VS 21m, P = 0.225, Figure 4).

DISCUSSION
The relationship between tumor stroma cells is very important and that certain cytokines play an important role in the relationship between tumor hosts [18]. This study compared the efficacy of adding hyperthermia to the current regimens or without hyperthermia in advanced NSCLC. Our results showed that patients who received hyperthermia had better PFSfirst and OS than those without hyperthermia. A retrospective study [19] showed that hyperthermia combined with RT had a better prognosis and did not increase side effects. The author found the objective response rate (ORR) and 3year OS rates between two groups were 97% versus 70% and 37% versus 6.7%, retrospectively (P<0.05). Italian scholars [20] were the first to report that hyperthermia combined with Chemoradiotherapy in LC in 1998, the results demonstrated that stage IV patients who received hyperthermia had longer OS (13.2m vs 8.4m, P<0.01). However, the simple size was small and these studies were retrospective studies. A review showed hyperthermia could exert synergistic effect with other treatments [21], Which were similar to our study. The mechanisms of hyperthermia enhance the anticancer effects of radiotherapy are as below: 1. DNA doublestrand breaks (DSB) are the most critical radiationinduced damage, which is the main cause of cell death. It can inhibit the repair of DSB in NSCLC cell line, promotes apoptosis and inhibits the invasion of tumor cells when radiotherapy combined with hyperthermia [22]. 2. Hyperthermia can ameliorate the TME such as low PH, hypoxia, and vascular immaturity through increases blood flow and perfusion [23]. 3. Cells in the S phase which are insensitive to radiotherapy, are sensitive to hyperthermia. Hyperthermia increases the effectiveness of chemotherapy by increasing the concentration of chemotherapy drugs, inhibiting DNA repair, increasing the release of free radicals, reversing the drug resistance of chemotherapy drugs and so on [24].     efficacy of adding hyperthermia for NSCLC patients with different EGFR states and our outcomes are consistent with the results of the above studies.
The reason why patients with EGFR mutation could not benefit from hyperthermia now is still unclear. We suppose hyperthermia had a positive effect on immune microenvironment as we described previously, and the special immune microenvironment in patients with EGFR mutation. Dong et al. [30] performed a poolanalysis of 3283 patients from 15 studies to systematically assessed the association between EGFR mutation and PD-L1 expression. They found that EGFR wild-type tumors were more likely to be PD-L1-positive than EGFR mutant tumors (OR: 1.79; 95% CI 1.10-2.93; P = 0.02). They analyzed the protein and mRNA profiles of PD-L1 in the repository (The Cancer Genome Atlas; TCGA) and internal (Guangdong Lung Cancer Institute; GLCI) databases and performed IHC detection of PD-L1 in resected NSCLC tissues found same results. What's more, they explored the correlation between EGFR status and CD8+ T-cell infiltration. Analysis of mRNA profiles in the GLCI cohort indicated that patients with EGFR mutations had a lower CD8A expression than those with EGFR wild-type (P = 0.031). IHC analysis of CD8+ TILs in the 255 resected NSCLC specimens confirmed that EGFR mutant tumors showed lesser T-cell infiltration than EGFR wild-type ones (P = 0.003). They also found patients with sensitive EGFR mutation display low immunogenicity and show impaired response to PD-1 blockade. Eri Sugiyama et al. [31] have shown that tumor mutation burden (TMB) was decreased and regulatory T cells (Treg) was increased in EGFR-mutated lung adenocarcinomas (LUADs). Thus, we suspected low TMB, low tumor infiltrating lymphocytes (TILs) and high Treg in patients with EGFR mutation may resulted in poor efficacy of IO of them. Hyperthermia can promote the efficacy of IO in advanced NSCLC by improving the TME, which may lead to patients with EGFR mutation do not benefit from hyperthermia. But more basic researches should be carried out to clarify the mechanisms.
Moreover, our study also investigated the best time to add hyperthermia. The outcomes demonstrated that the PFSfirst in first-line and PFSpost in posterior-line was no significant difference between the entire groups. Takayuki Ohguri et al. [32] carried out a study to assess the efficacy of re-irradiation plus regional hyperthermia for 33 recurrent NSCLC patients, the results showed that the median disease PFS after reirradiation were 6.7 months. In contrast, our median PFSpost was 21 months.
However, our study was a retrospective clinical study. The sample size was small and the follow-up time was short. Some patients did not detect EGFR or combined with other gene mutations may impact on the results. It need more prospectively, large sample size, randomized controlled trails to confirm the results.

CONCLUSIONS
This retrospective study revealed that adding hyperthermia to the treatment of NSCLC patients without EGFR mutation had better prognosis than those who didn't adding hyperthermia to the regimen. Moreover, adding hyperthermia in first-line or in posterior-line treatment was no significant difference. However, these results need more prospective studies to confirm the conclusions.

Patients
We retrospectively collected patients from the Second Affiliated Hospital of Guangxi Medical University and Liuzhou Worker's Hospital between January 2018 and December 2019. All patients had pathologically confirmed adenocarcinoma (AC) and squamous cell carcinoma (SCC) of lung and stage III-IV disease by the staging criteria of the 8 th edition of the American Joint Committee on Cancer Staging System (AJCC) and the International Union Against Cancer (UICC) [33]. The inclusion criteria were as follows: 1) Patients aged 18 to 70 years old, 2) ECOG score ≤ 2, 3) Patients with complete clinical data, 4) Stage III -IV disease restaged by the 8 th edition of AJCC/UICC TNM staging system, 5) Patients previously didn't received any treatment, 6) Patients didn't have any malignancies before. Exclusion criteria were: 1) ECOG score ≥3, 2) Patients with severe hepatic, renal and pulmonary dysfunction, 3) The pathological type was not adenocarcinoma or squamous cell carcinoma, 4) Cognitive dysfunction. A total of 205 patients were included in this study based on the inclusion and exclusion criteria. On the basis of adding hyperthermia or not to the regimen in the advanced NSCLC before first progression, we divided all the eligible patients into hyperthermia group (N=77) and nonhyperthermia group (N = 128). There were 25(32.5%) patients had EGFR mutation in hyperthermia group while 44(34.4%) patients in non-hyperthermia group among the groups. The patients' base-line characteristics were shown in Table 1.

Surgery
Lobectomy or partial or total pneumonectomy± lymph node dissection through thoracotomy or thoracoscopy. AGING

Radiotherapy (RT)
All included patients were received 6 MV-X ray liner accelerator intensity modulated radiation therapy (IMRT), and the treatment steps were as follows: (1

Hyperthermia
One hour before other treatment, Jilin Maida medical radiofrequency hyperthermia system NRL-004 was used to patients for 60 minutes, the temperature was controlled at 42° C, twice a week, with an average of 4-8 times.

Endpoints
The primary endpoint of this study was 1-year PFS. PFSfirst defined as the time from the diagnosis of the disease to the first progression or death from any cause. PFSpost in the Posterior line therapy adding hyperthermia was defined as from disease first progression to the last progression or death from any cause. The secondary endpoint was 1-year OS, which defined as the time from disease diagnosis to death or the last follow-up.

Statistical analysis
SPSS 22.0 software (IBM) and graphpad prism 8 software were used to analyze the data. Chi-square test was used for continuous variables, and t-test was used for categorical variables. Kaplan-Meier method was conducted for the analysis of the time-to-event endpoints, and the log-rank test was used to compare the differences between the two groups. Multivariate analysis were performed in the detection of prognostic factors related to the endpoints. Using graphpad prism8 to draw the survival curve. All statistical tests were two-sided, and P<0.05 was considered statistically significant.

Ethics approval
The study was approved by the ethics committees of The Second Affiliated Hospital of Guangxi Medical University, and The Liuzhou Worker's Hospital.

CONFLICTS OF INTEREST
All the authors have declared no conflicts of interest.

Editorial note
& This corresponding author has a verified history of publications using a personal email address for correspondence.