High expression of PD-L1 Predicts Worse Overall Survival in the Cavitary Lung Adenocarcinoma

Objective Solitary cavitary lung cancer is one of the rare types of lung cancer. Generally, the relationship between cavitary lung adenocarcinoma and immunotherapy remains unknown. We aimed to assess programmed cell death ligand-1(PD-L1) expression and CD8-positive (CD8 + ) tumor inltrating lymphocytes (TILs) density, and evaluate their prognostic signicance of patients with cavitary lung adenocarcinoma (LUAD).


Introduction
In both sexes combined, lung cancer is the most common type of cancer diagnosed and the leading cause of cancer death worldwide in 185 countries [1]. Among all types of lung cancer, cavitary lung cancer is particularly unique and seldom reported [2], which occurs in 8% of all lung cancers [3], while other researchers reported the incidence rate of 1.00-2.07% [2]. Cavitation in a tumor nodule is previously thought to be more prevalent in patients with lung squamous cell carcinoma [4]. Following with LUAD increases, cavitary LUAD has also been reported, with an incidence of 5.7 to 14.9% in patients with LUAD [5]. As a rare type of lung cancer, cavitary lung cancer is not easily diagnosed by radiological measures, and has a worse prognosis because of high TNM stage, compared with noncavitary lung cancer [6][7][8][9]. Moreover, the biological features of the underlying walled cavity are still poorly understood.
Immunotherapy is regarded as a novel choice in the treatment of a variety of cancers with poor prognosis [10]. The development of immune checkpoint inhibitors has changed the treatment of non-small cell lung cancer (NSCLC) [11]. As one of the typical checkpoint inhibitors, programmed cell death 1 (PD-1) is an inhibitory cell-surface receptor that is expressed on activated T-cells and other immune cells. Anticancer immunotherapy targeting immune checkpoints with antibodies to PD-1 and its ligand PD-L1 is an established treatment modality for NSCLC [12,13]. One of the important mechanisms that anti-PD-L1 monoclonal antibodies restraint the lymphocyte inhibition by binding to the PD-1 receptor, which preventing the PD-1 binding with its ligands 1 or 2 (PD-L1 or PD-L2), and permit T cells to maintain their tumor cell killing function [14,15]. PD-L1-positive patients had a higher chance of achieving an objective response when treated with anti-PD-L1 monoclonal antibodies [16,17].
Tumor in ltrating lymphocytes (TILs) also play a vital role in predicting tumor progression in different kinds of cancers [18]. As the most studied component of tumor-associated immune response, the cytotoxic CD8-positive (CD8+) T cells expression could predict better prognosis in patients with breast or ovarian cancers [19,20]. However, the relationship between solitary thin-walled cavity lung cancer and immune checkpoint remains unknown. Therefore, in this study, we take advantage of human specimens to assess PD-L1 expression and CD8 + TIL density, and investigate its prognostic signi cance in cavitary LUAD.

Material And Methods
Samples collection 65 patients who were diagnosed as the cavitary lung cancer in the period from September 2005 and October 2015 in the General Hospital of Central Theater Command Hospital, PLA were included in the research, and they had no neoadjuvant therapy before surgical resection. All patients underwent the 64row spiral computed tomography with slice thickness of 1.25 mm, 1.5 mm or HRCT. Two radiologists (J Liu and Y Xue) examined the imaging features of these samples independently. According to previous studies, the tumor cavitation was de ned as an air-lled space with a shortest diameter of ≥ 5 mm within a tumor [5,21,22] (Fig. 1).
Two pathologists (Q Wang and Y Ren) recon rmed the histopathologic features of each sample independently. Another 30 cases of general LUAD patients who didn't present as a solitary cavity and excluded post-neoadjuvant therapy were collected as controls.
Clinicopathological data were retrieved from clinical records and histopathology reports. The follow-up began on the date of surgery or biopsy and ended in October 2018. The median follow-up was 45 months (1-115 months). Overall survival was de ned as the period from diagnosis to death or the end of followup. LUAD specimens were classi ed according to the 8th edition of TNM classi cation by Union for International Cancer Control/ American Joint Committee on Cancer (2017) [23].

Evaluation of immunohistochemistry
Immunostaining intensity was observed using light microscopy (Olympus BX-53 with CCD DP73). Results were scored by two pathologists (Q Wang and Y Ren) who were independent and blinded to the clinicopathological characteristics of the research.
The immunohistochemistry characteristics and cut-offs of PD-L1 for being regarded as positive varies in different research. Here, the cut-off of PD-L1 protein expression in tumor cells were de ned as 5%. PD-L1 ≥ 5% was regarded as high expression, which was consistent with many types of cancers [24,25].
For the CD8 evaluation, CD8 + TILs were counted in each slide at × 200 magni cation. The mean of the three counts was calculated for each case and the cutoff point of high or low expression was determined on the median number of total scores [20,26].

Multiplex immuno uorescence staining
Manual multiplex immuno uorescence (mIF) staining was performed in 4-µm sections obtained from FFPE lung cancer blocks by using the Opal 4-Color IHC Kit (PerkinElmer, Waltham, MA) [11]. The stained slides were scanned by a Vectra multispectral microscope (Akoya Biosciences, USA). The immuno uorescence markers were consisted of PD-L1 (E1L3N, dilution 1:200; Cell Signaling Technology, USA), CK (AE1/AE3) and CD8(C8/144B) are ready-to-use antibodies from Agilent/DAKO, California, USA. Primary antibody was visualized by using tyramide signal ampli cation linked to a speci c uorochrome from the multiplex IHC Kit for each primary antibody. A stripping procedure, based on the Meidi microwave (Meidi, China), was performed for each consecutive antibody staining. Human tonsil FFPE tissues were also used with and without primary antibodies as positive and negative (auto uorescence) controls, respectively. The mIF-stained slides were scanned with a Vectra 2.3 microscope system (Akoya Biosciences, USA) under uorescent illumination. From each slide, Vectra automatically captured the uorescent spectra from 420 nm to 720 nm at 20-nm intervals with the same exposure time and then combined the captured images to create a single stack image that retained the particulate spectral signature of all IF markers.

Statistical analysis
Data were expressed as frequencies for categorical variables and mean ± SD for numerical variables. SPSS 21.0 software (Chicago, IL, USA) was used to perform all statistical analyses. χ 2 test or Fisher exact test were carried out to evaluate the correlations of PD-L1 expression and clinicopathological parameters of cavitary LUAD patients. We explored the relationship between PD-L1 and CD8 + TILs using Spearman correlation analysis. The survival analysis was assessed using the Kaplan-Meier curve and log-rank test to the statistical difference survival data. Cox proportion hazard regression model was conducted to evaluate univariate and multivariate analysis of survival as well as the independent prognostic values. P values < 0.05 were considered statistically signi cant.

Patient characteristics
As showed in Table 1, among the 65 cavitary LUAD patients, 36 (55.4%) were male and 29 (44.6%) were female, with the mean age of 58 years old (range 48-71). 31 (47.7%) patients were alive and 34 (52.3%) died at the end of follow-up. The data of T, N, M and TNM stage as well as the clinicopathological parameters of noncavitary LUAD patients also showed in Table 1.

PD-L1 expression and clinicopathological parameters in cavitary LUAD patients
The relationship between tumor PD-L1 expression and clinicopathologic variables in the cavitary LUAD was investigated by χ 2 test. As listed in Table 3, high expression of PD-L1 protein was signi cant correlated with the lymph node metastasis (N) (P=0.001) and TNM stage (P=0.024). Negative correlation was observed between PD-L1 and CD8 TIL status (rs= -0.272, P=0.025). However, PD-L1 expression was not signi cantly associated with age, gender, tumor size in the cavitary LUAD patients.
Prognostic value of PD-L1 expression in cavitary LUAD patients Survival analysis determined by Kaplan-Meier curve and log-rank test was determined to investigate the prognostic value of PD-L1 expression and CD8 TIL status in cavitary LUAD patients. High expression of PD-L1 group predicted poorer survival and high mortality rate in cavitary LUAD patients (Fig.5A, P=0.004).
In univariate analysis, both PD-L1 and CD8 expression levels were found to be signi cantly related to the overall survival (OS) of LUAD patients (HR and 95% CI=2.670, 1.334-5.345, P=0.006; HR and 95% CI=0.995, 0.991-0.998, P=0.002, respectively. Table 4). Moreover, the phenotype of high PD-L1 expression and low CD8+ TIL had higher risk than the other phenotype (HR and 95% CI=2.999, 1.518-5.923, P=0.002). Simultaneously, TNM stage was signi cantly correlated with the OS of LUAD (Table 4). In order to analyze whether the above univariate was an independent prognostic factor, a multivariate COX proportional hazard model on OS was performed. The clinicopathological characteristics and the two protein expressions were added in the multivariate analysis model. The results indicated that only CD8 expression was an independent prognosis parameter in OS of LUAD patients (HR and 95% CI=0.005, 0.007-0.401, P =0.004, Table 4).

Discussion
To our knowledge, the current study was the rst to investigate that high expression of PD-L1 protein in the cavitary LUAD tissues, compared to the noncavitary LUAD tissues. PD-L1 expression level was signi cantly correlated to lymph node metastasis, TNM stage and CD8 TIL status. Interestingly, our study showed that high PD-L1 expression and low CD8 + TIL can predict poorer overall survival of the patients with cavitary LUAD. CD8 + TIL had an independent predictor of LUAD prognosis.
The examinations of chest radiography and CT are common used for the clinical diagnosis of lung cancer. Radiographic features of cavities that suggest malignancy include multiple holes, a nodular illde ned inner or outer wall, and an eccentric excavation with irregular margins [27]. Cavitary lung cancers are more prevalent in patients with worse survival than that for noncavitary NSCLC patients [5,22], because of advanced tumor stage, and vascular, lymphatic, or pleural invasion of cavitary LUAD [5]. Currently, there is a lack of cognition concerning its onset and progression for solitary cavitary LUAD, thus, this rare type of cancer is subject to misdiagnosis and missed diagnosis [2]. Necrosis may cause solitary cavity because of primary cancer overgrowth. The lesion may originate from the distal part of lung or a preexisting cystic lesion. Tumor growth leads to bronchial obstruction and vascular invasion, which provides an environment of ischemia and hypoxia, and then resulting in the tumor necrosis. Furthermore, the autophagy of neoplastic cell can also induce the cavitary lesion [6,9].
Immune checkpoint inhibitors targeting the PD-1/PD-L1axis have shown promising results in patients with NSCLC. Overexpression of PD-L1 is associated with poor recurrence-free survival and overall survival [28]. Pembrolizumab has been approved as rst-line treatment for advanced PD-L1 positive NSCLC patients [29]. In our study, we found that high expression of PD-L1 was detected in 44.6% (29/65) cases of cavitary LUAD, signi cantly increased, compared to 20% (6/30) cases of noncavitary lung cancer, which demonstrated tumor necrosis correlates with higher PD-L1 expression in the LUAD [30]. Moreover, high expression of PD-L1 was correlated with high TNM stage, and could predict poor prognosis of patients with cavitary LUAD. The above results con rmed that high expression of PD-L1 may promote malignant progression, and regard as one of cancer immunotherapy targets of the cavitary LUAD. However, recent paper have reported high expression of PD-L1 is also positively asscioated with mutations in KRAS, TP53, and MET, however, negatively associated with mutations of EGFR and STK11of LUAD[31]. Thus, we had to further the speci c molecular features of cavitary LUAD to analyze the mechanism of malignant progression and poor prognosis.
In ammation is one of the notable features of cancer, and can lead to tumor progression [32]. Cytokines also has an anti-tumor immune effect, IFN-α, IFN-γ, and TNF-α can increase expression of PD-L1 in a variety of cancers [33]. In addition, CD8 + T cells also correlate with PD-L1 expression and participated in the in ammation of anti-tumor immune response [34,35]. Our results also con rmed that negative correlation was observed between high expression of PD-L1 and CD8 + TIL in cavitary LUAD. Moreover, this phenotype with high PD-L1 expression and low CD8 + TIL could predict poorer overall survival of the patients with cavitary LUAD, compared to the other phenotypes (such as low PD-L1 and high CD8 + TIL). Moreover, CD8 + TIL was an independent marker to predict prognosis of LUAD. In a comprehensive view, it suggested that in ammation or necrosis induced by the dysregulation of tumor growth, and then activates the cytokines secretion as well as leads to the formation of thin-walled cavity lesion in the LUAD development. Subsequently, cytokines increase the expressions and activities of PD-L1, which contributes tumor cells escaping from immune surveillance, further promote malignant progression of cavitary LUAD.
CD8 + TILs involved in in ammation and immune processes described above. Further studies are required to elucidate the mechanisms of PD-L1 overexpression and its relationship with in ammation or necrosis in solitary cavity LUAD cases.
In conclusion, we rstly demonstrated that PD-L1 expression is upregulated in the cavitary LUAD patients, and high expression of PD-L1 negatively correlates with CD8 T cell in ltrating status. High PD-L1 expression and low CD8 + TIL can predict poorer overall survival of the patients with cavitary LUAD. Our results illustrate that PD-L1 is a critical immune checkpoint and improve our mechanistic understanding of cavitary LUAD.

Declarations
Authors' contributions JYL and WCH initiated and designed the work, supervised the data collection. JYL and MMG prepared the manuscript and performed the experiments. YX and YR contributed to the acquisition of patients and tissues specimens and to the analysis and interpretation of data. All authors read and approved the nal manuscript.

Funding
This work was supported by grants from the Natural Science Foundation of Hubei Province General Project (grant number: 2020BCB059).

Availability of data and materials
The data sets and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This research was approved by the Ethics Committee of General Hospital of Central Theater Command, PLA, Wuhan (2018-002-1). All patient specimens and clinical data involved in this study complied with the Declaration of Helsinki.

Consent for publication
Written informed consent was obtained from a legally authorized representative(s) for anonymized patient information to be published in this article.

Competing interests
The authors state that there are no con icts of interest to disclose.   soft tissue density shadow, which was lobulated, and pulled by the adjacent pleura, with eccentric cavity.    Correlation of PD-L1 and CD8 expression level and overall survival in the patients with cavitary LUAD. A: Patients with low PD-L1 expression showed better overall survival than that with High PD-L1 expression (P<0.001). B: Patients with high CD8 TIL showed better overall survival than that with low CD8 TIL