Peripheral Blood Biomarkers Provide Prediction of Survival and Immune-Related Pneumonitis in Stage III/IV Non-Small Cell Lung Cancer Treated with Immune Checkpoint Inhibitors

Backgroud: Immune checkpoint inhibitors (ICIs) have been considered as a standard therapy for non-small cell lung cancer (NSCLC). This study aimed to explore associations between peripheral blood biomarkers with survival and immune-related pneumonitis (IRP) in patients with stage III/IV NSCLC treated with ICIs, and establish a novel risk stratication to predict outcomes. Methods: This study enrolled 95 patients with stage III/IV NSCLC receiving ICIs. Univariate and multivariate cox regression analyses were used to determine prognostic factors affecting overall survival (OS) and progression-free survival (PFS). Logistic regression analysis was applied to explore the correlation between peripheral blood markers and IRP. Results: Multivariate Cox analysis revealed that pretreatment absolute lymphocyte count (ALC) (HR: 0.113, 95%CI: 0.113–0.863, P=0.025), and post-treatment systemic inammation response index (SIRI) (HR: 4.199, 95%CI: 1.058–16.662, P=0.041) were independent prognostic factors for OS. Similarly, pretreatment ALC (HR: 0.333, 95%CI: 0.129–0.857, P=0.023) and post-treatment SIRI (HR: 3.160, 95%CI: 1.046–9.551, P=0.041) were also independent predictors of PFS. Based on these biomarkers, patients were classied into 3 categories: low-risk group: ALC >1.5 and SIRI ≤ 1.69 (n=28); medium-risk group: ALC >1.5 or SIRI ≤ 1.69 (n=35); high-risk group ALC ≤ 1.5 and SIRI >1.69 (n=32). The 1-year OS rate were 69.2%, 63.6%, 27.1%, respectively (P=0.001) and 1-year PFS rate were 51.8%, 44.8%, 21.1%, respectively (P=0.007). Furthermore, we found patients with post-treatment NLR >3 were associated with a signicantly higher risk for occurring IRP than that NLR ≤ 3 (HR: 2.917, 95%CI: 1.037-8.206, P=0.043), and the time to IRP onset in post-treatment NLR>3 group was signicantly shorter than that in NLR ≤ 3 (P=0.018). absolute lymphocyte count; AMC, absolute monocyte count; LMR, lymphocyte-to-monocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte; PNI, prognostic nutritional index; SII, systemic inammation index; SIRI, systemic inammation response index. Abbreviations: IRP, immune-related pneumonitis; KPS, karnofsky performance status; PD-L1, programmed cell death-Ligand 1; IQR, interquartile range; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; AMC, absolute monocyte count; LMR, lymphocyte-to-monocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte; PNI, prognostic nutritional index; SII, systemic inammation index; SIRI, systemic inammation response index.


Backgroud
Lung cancer is one of the most prevalent cancers globally and the incidence increases 26% annually.
Non-small cell lung cancer (NSCLC), which is comprised of approximately 85% of all lung cancer patients. [1] And almost 75% of the patients are diagnosed at stage III/IV NSCLC with the poor prognosis. [2] Recently, immune-checkpoint inhibitors (ICIs), anti-PD-1 (programmed cell death-1) and anti-PD-L1 (programmed cell death-Ligand 1) have demonstrated to improve long-term survival and considered as a novel standard treatment for advance non-small-cell lung cancer (NSCLC). [3][4][5][6][7][8] However, it was only approximately 20% patients who had positively response to ICIs, and a minority of patients who could not bene t from ICIs. [9] Furthermore, the use of ICIs also commonly emerged immune-related adverse effects (irAEs), [10] one particular irAE is the development of immune-related pneumonitis (IRP), which is a rare but potentially fatal. [11][12][13][14] Therefore, it is important to select eligible patients who may bene t the ICIs therapy and identify factors associated with the onset of IRP.
In ammatory response and immune surveillance have been deemed crucial signatures regarding cancer development and treatment outcomes. [15] A variety of evidence indicated that peripheral blood In ammatory biomarkers such as neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), prognostic nutritional index (PNI), systemic in ammation index (SII), and systemic in ammation response index (SIRI) could facilitate the prediction of treatment outcomes for various solid carcinomas, including melanoma, prostate cancer, colorectal cancer, esophageal squamous cell carcinoma, and NSCLC. [16][17][18][19][20] These In ammatory biomarkers are easy to obtain in clinical practice and directly re ect the condition of the immune system, and may potentially impact on the response to ICIs. [20,21] However, in currently, there were few reported studies on the predictive value of peripheral blood biomarkers for NSCLC treated with ICIs. Therefore, we performed a retrospective analysis to investigate the usefulness of In ammatory biomarkers including ANC, ALC, AMC, LMR, NLR, PLR, PNI, SII, and SIRI as predictors of outcomes and the development of IRP in stage III/IV NSCLC treated with ICIs therapy.

Endpoints
The primary endpoints included overall survival (OS) and progression-free survival (PFS). OS was de ned as the time between initial ICIs treatment and death for any reason or the last follow-up. PFS was measured from the rst time of ICIs treatment to the disease progression or death due to any cause. The secondary endpoint was the development of IRP. The time to onset of IRP was ranged from the start of ICIs therapy to the development of IRP. The IRP was reviewed by the National Cancer Institute's Common Terminology Criteria for Adverse Events (CTCAE), version 5.0. All patients were followed up until death or the last follow-up. The median follow-up period was 10 months (range 2-38 months).

Statistical analysis
A Chi-squared test or Fisher exact test was used to compare categorical variables and a Mann-Whitney U test was used to compare continuous variables. Survival curves were estimated by the Kaplan-Meier method and compared by the log-rank test. Variables with a P-value ≤ 0.15 in the univariate analysis was included in the multivariable analyses. Multivariate analysis was carried out to identify independent prognostic factors related to OS and PFS. Logistic regression analysis was applied to explore the correlation between peripheral blood markers and the onset of IRP. All analyses were considered as statistically signi cant with a two-sided P-value of < 0.05, and were performed using SPSS 24.0 (IBM, Armonk, NY, USA).

Univariate and multivariate cox analysis for PFS and OS
Univariate and multivariate cox analysis for OS and PFS were detailed in Table 2 and Table 3, respectively. As shown in Table 2 (Table 3).

Immune prognostic factors for risk strati cation
Pretreatment ALC and post-treatment SIRI were considered as signi cant favorable prognostic factors for both OS and PFS. The Kaplan-Meier curves for OS and PFS were presented in Fig. 1. The median OS and PFS among patients who had post-treatment SIRI > 1.69 were signi cantly shorter than that posttreatment SIRI ≤ 1.69 (OS: 5 months vs 11 months, P = 0.013; PFS: 3 months vs 8 months, P = 0.031).
As showed in Table 4, patients with IRP had signi cantly higher post-treatment NLR than those without IRP (the medium 4.23 versus 3.07, P = 0.045). Logistic regression analysis (  all patient developed grade 2 ICI-pneumonitis. But it was no signi cant difference in the distribution of grade between post-treatment NLR > 3 group and post-treatment NLR ≤ 3 group (P = 0.280) (Fig. 3A). In addition, we observed that most of patients (18/24) developed IRP within the rst 3 months after the start of ICIs treatment, and the time to IRP onset in post-treatment NLR > 3 group was shorter than in posttreatment NLR ≤ 3 group (Fig. 3B). The median onset time of IRP was 2 months (range 1-18 months), and the median IRP onset time in post-treatment NLR > 3 group was 1.5 months (range 1-16 months), and 6.5 months (range 1-18 months) in post-treatment NLR ≤ 3 group. Median OS and PFS were similar in patients who developed IRP compared with those who did not, however, patients with grade ≥ 3 IRP were signi cantly associated with shorter OS (8 months versus 4 months, P = 0.04) and PFS (6.5 months versus 3 months, P = 0.054) (Fig. 4).

Discussion
The role of immunotherapy in the treatment of NSCLC has been increasingly emphasized, and its application in clinical practice largely changes the treatment and prognosis of NSCLC patients. [4][5][6]8] In this study, a total of 95 stage III/IV NSCLC patients treated with ICIs therapy was retrospectively analyzed.
We found that both pretreatment ALC and post-treatment SIRI were useful predictors of outcomes in NSCLC patients receiving ICIs. Then we established a novel risk strati cation based on these prognostic factors. To our knowledge, this is the rst study to combine pretreatment ALC with post-treatment SIRI for predicting outcomes in patients treated with ICIs. Furthermore, our results showed that post-treatment NLR was a risk factor for the occurrence of IRP.
Currently, a growing body of studies illustrated that the clinical value of peripheral blood biomarkers which could be helpful for predicting treatment outcomes in different solid tumors, including NSCLC. [16][17][18][19][20] Lymphocyte is one of peripheral blood biomarkers which is a vital indicator of the immune system, re ecting the immune system activation, and play a fundamentally important role in tumor proliferation and migration. [23][24][25][26] Previous studies have reported that the lymphopenia correlated with worse outcomes in cancer patients. [27][28][29] Ryoko et.al. found that pretreatment ALC less than 1.5 associated with adverse survival.
[28] Another analysis of patients with advanced breast cancer, sarcoma, and lymphoma, suggested that lymphopenia was an independent predictor of survival. [29] In our study, we found that pretreatment ALC less than 1.5 predicted adverse OS and PFS for patients treated with ICIs.
The mechanism of cancer immunotherapy is to promote the activity of cytotoxic T lymphocytes (CTLs) and assist in the activation of tumor-speci c CTLs in lymphoid organs, thus establishing an e cient and durable anti-tumor immune microenvironment for anti-tumor therapy. [30] Therefore, it is not di cult to understand that a higher pretreatment ALC is bene cial to improve the survival of patients receiving ICIs.
Another peripheral blood biomarker is the systemic in ammation response index (SIRI), which combined ANC, ALC, and AMC, has been proved to be an effective prognostic biomarkers in different cancers such as esophageal squamous cell carcinoma, non-small cell lung cancer, pancreatic cancer, gastric adenocarcinoma, clear cell renal cell carcinoma. [20,[31][32][33][34] Geng et al. suggested that SIRI was an independent prognostic indicator for ESCC patients after the radical surgery. [33] and Hu et al. proposed pretreatment SIRI was an independent predictor of outcomes in stage III NSCLC patients who undergoing chemoradiotherapy. [20] But few studies accessed the effect of SIRI for patients receiving immunotherapy.
As far as we know, the present study was the rst to demonstrate predictive roles of SIRI for prognosis of III/IV stage NSCLC patients treated with ICIs and revealed that high post-treatment SIRI was signi cantly associated with poor OS and PFS. Then, we established a novel risk strati cation based on pretreatment ALC and post-treatment SIRI. Patients were classi ed into 3 categories: low-risk group: ALC > 1.5 and SIRI ≤ 1.69; medium-risk group: ALC > 1.5 or SIRI ≤ 1.69; high-risk group: ALC ≤ 1.5 and SIRI > 1.69. Survival analysis showed 1-year OS rates of 69.2%, 63.6%, 27.1% in low-risk, medium-risk, and high-risk group respectively (P = 0.001). In addition, PFS of high-risk group were also signi cantly shorter than that patients in low-risk or medium-risk group (P = 0.007). Given that ALC and SIRI are easily obtainable, a simple immune risk strati cation based on these factors may easily predict survival in III/IV NSCLC treated with ICIs in clinical practice.
Although immunotherapy has brought hope to patients with advanced tumors in recent years, the immune-related adverse effects (irAEs) it brings have also drawn attention. [10] Immune-related pneumonitis (IRP) is a rare but potential fatal irAE that is related to poor outcomes. [11][12][13][14]  In addition, we also explored the correction between NLR and the time to IRP onset, and found that most of patients developed IRP within the rst 3 months after the start of ICIs treatment, and the onset time in post-treatment NLR > 3 group was shorter than in post-treatment NLR ≤ 3 group. The median onset time in post-treatment NLR > 3 group was 1.5 months (range 1-16 months), and 6.5 months (range 1-18 months) in post-treatment NLR ≤ 3 group. Therefore, for those patients with post-treatment NLR > 3, we should be alert to the development of IRP in the rst 3 months after ICIs therapy and adopt a timely treatment strategy.
The present study has several limitations. Firstly, this is a retrospective study, so there may be selective bias in our study. Secondly, there are small numbers of eligible patients, and all patients come from a single institution. Accordingly, larger prospective studies are needed to con rm our results. Third, our current study only explored parameters that are commonly used and easily accessible in clinical practice, but other relevant variables involving genomics and radiomics may provide more valuable information to improve the predictive accuracy of IRP as well as the prognosis of patients treated with ICIs.

Conclusion
Our data indicated that pretreatment ALC and post-treatment SIRI appeared to be biomarkers of outcome in stage III/IV NSCLC patients treated with ICIs. A novel risk strati cation was established basing on these prognostic factors. Patients were classi ed into 3 categories: low-risk group: ALC > 1.5 and SIRI ≤ 1.69; medium-risk group: ALC > 1.5 or SIRI ≤ 1.69; high-risk group: ALC ≤ 1.5 and SIRI > 1.69. Furthermore, our study also shows that post-treatment NLR > 3 were likely to develop IRP of patients who receiving ICIs. A prospective study to validate the signi cance of peripheral blood biomarkers is necessitated.

Declarations
Ethics Approval and consent to participate This retrospective study was approved by the ethics committee of the Fujian Province Cancer Hospital (YK2021-009-01). All patients provided written informed consent prior to treatment, and all information was anonymized prior to analysis.

Guideline statement
All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication
Not applicable.

Availability of data and materials
The data that support the ndings of this study are available from the corresponding author upon reasonable request.

Competing interests
The authors declare the submitted work was not carried out in the presence of any personal, professional or nancial relationships that could potentially be construed as a con ict of interest.