**1. Introduction**

Non-small cell lung cancer (NSCLC) is the leading cause of cancer-related deaths worldwide. Recent studies have proven that anti-programmed death-1 (PD-1)/programmed death ligand-1 (PD-L1) antibodies provide significant survival benefits in patients with advanced NSCLC, when compared with standard chemotherapy [1–4]. Although the efficacy of the anti-PD-1 antibody varies according to the immunohistochemical degree of PD-L1 expression within tumor cells, there are no established biomarkers to predict the outcome after the administration of the anti-PD-1 antibody and the expression of PD-L1. If a useful biomarker is obtained from common modalities, this discovery can be easily adopted into daily practice.

Notably, 2-deoxy-2-[18F]fluoro-d-glucose (18F-FDG) positron emission tomography (PET) is a distinguished radiological modality to distinguish benign lesions from malignant tumors [5]. The uptake of 18F-FDG is observed in non-malignant lesions such as sarcoidosis, granuloma, pneumonia, and tuberculosis, but it can closely resemble the metabolic activity of malignant tumors [5]. Previous reports demonstrated that the accumulation of 18F-FDG within tumor cells was significantly linked to the presence of glucose transporter 1 (Glut1), hypoxia-inducible factor 1α (HIF-1α), and vascular endothelial growth factor (VEGF)b [6]. Several researchers have described that the accumulation of 18F-FDG exhibited a significant correlation with the expression of PD-L1 in patients with NSCLC [7–9]. Although little is known about the close association between immune environment and glucose metabolism, it is important to discover how metabolic tumor activity can affect the upregulation of PD-L1 expression. Moreover, it has been reported that 18F-FDG-PET is useful for assessing the therapeutic monitoring of anti-PD-1 antibody with regard to the prognosis and overall response rate [10]. Little is known as to whether the uptake of 18F-FDG before the administration of anti-PD-1 antibody can predict the efficacy and prognosis of immune checkpoint inhibitors (ICI) in patients with advanced NSCLC.

The maximum standardized uptake value (SUVmax) is extensively used to evaluate the metabolic degree of 18F-FDG within tumor cells. Since SUVmax reflects the maximal point of glucose metabolism within tumor specimens, it remains unclear whether it can indicate the total metabolic tumor volume (MTV). Recently, total lesion glycolysis (TLG) and MTV—indicators of 18F-FDG accumulation within tumor cells—have been identified as significant prognostic markers for predicting the treatment outcome in patients with NSCLC. A meta-analysis described TLG and MTV as better predictive markers than SUVmax [11]. Moreover, an exploratory study documented that the tumor metabolic activity assessed by TLG and MTV is better than SUVmax in evaluating the therapeutic monitoring of anti-PD-1 antibody in patients with previously treated NSCLC [10]. Recent reports suggested that baseline tumor size could predict adverse outcomes in patients with NSCLC who received ICI [12–14]. However, the detailed mechanism behind tumor burden (TB), determined by baseline tumor size and causing poor efficacy of ICI, is unclear. Tumors possibly suppress the immune response by different mechanisms other than the PD-1/PD-L1 pathway. Tumor hypoxia, determined by HIF-1, induces VEGF, which induces immunosuppressive T-lymphocytes such as regulatory T-cells (Tregs) and myeloid-derived suppressor cells [15,16]. These evidences suggest that the high HIF-1 expression caused by an increased tumor size creates an immunosuppressive environment regardless of ICI treatment and contributes to the poor outcome for patients with NSCLC [15,16]. However, whether morphological assessment, based on baseline tumor size, can accurately reflect the volume of tumor hypoxia or tumor metabolic activity is debatable. Therefore, 18F-FDG uptake is expected to assess tumor activity more accurately than TB on computed tomography (CT).

We conducted this study to investigate whether the degree of 18F-FDG uptake before the administration of anti-PD-1 antibody can predict the prognosis in patients with previously treated advanced NSCLC, by evaluating the correlation between TB and SUVmax, TLG, and MTV.
