Lung cancer remains the leading cause of cancer death, with 80-85% of the total number of lung cancer cases being non-small cell lung cancer (NSCLC) (1). Unfortunately, the vast majority of NSCLC patients are diagnosed late with local progression or evidence of metastasis (2). Bevacizumab combined with chemotherapy was initially approved by the United States Food and Drug Administration (US FDA) in 2006 as a first-line treatment for locally unresectable advanced, recurrent, metastatic, nonsquamous NSCLC due to evidence of a significant survival benefit and acceptable safety (3). However, not every patient benefits from this combination and so predictive markers are increasingly being used to identify NSCLC subpopulations who may benefit from receiving this combined intervention. Unfortunately, to date, research has been largely unsuccessful in identifying even a single response predictor for combined antiangiogenic treatments (4). Therefore, it remains necessary to identify predictive factors to ensure that this treatment is provided to those who will benefit most.

Integrin αvβ3, which forms complexes with vascular endothelial growth factor (VEGF) signaling pathways, is highly expressed on newly formed vessels (5, 6). The tripeptide sequence arginine-glycine-aspartate (RGD) also has a high affinity with and can therefore bind to integrin αvβ3. As such, ¹⁸F-AlF-NOTA-PRGD2 positron emission tomography/computed tomography (¹⁸F-RGD PET/CT) has been recommended for noninvasive angiogenesis imaging (7). ¹⁸F-RGD is a novel tracer that one-step labels integrin αvβ3, which targets the PET probe, and has proven relatively safe (8). In a pilot clinical study, patients with lesions broader than ¹⁸F-RGD PET/CT parameters appeared to respond better to antiangiogenic drugs alone (9). These initial studies validated some of the underlying theories that are emerging within this field and further suggest the need to explore the predictive capacity of ¹⁸F-RGD PET/CT for combining antiangiogenic and chemotherapeutic interventions.

Serum inflammatory cells have a complex relationship with angiogenesis and can be used to reflect the immune system status by providing insight into the tumor microenvironment (10). In particular, neutrophil and monocyte subpopulations have been found to contribute to angiogenesis (11, 12). However, activated T cells appear to inhibit neoangiogenesis by releasing the immune-related cytokine IFN-γ (4). Antiangiogenic therapies with bevacizumab may therefore increase the infiltration of immune effector cells and convert an intrinsically immunosuppressive tumor microenvironment into an immunosupportive microenvironment (13). Here, we propose that systemic immune factors, including the pretreatment neutrophil-to-lymphocyte ratio (PreNLR), pretreatment platelet-to-lymphocyte ratio (PrePLR), and pretreatment lymphocyte-to-monocyte ratio (PreLMR), may be useful predictors for combined antiangiogenic treatments. As such, we investigated whether these ¹⁸F-RGD PET/CT parameters with systemic immune factors could be used as predictive markers for combined antiangiogenic treatments.

This prospective study provided the first evidence that tumor ¹⁸F-RGD uptake parameters and systemic biomarkers for inflammation at baseline can be used to predict outcomes in patients receiving combined antiangiogenic therapy for advanced NSCLC. The tumor SUVmax obtained from ¹⁸F-RGD PET/CT was significantly higher in responders than nonresponders. Patients with tumors with higher SUVpeak or higher PreLMR had comparatively longer PFS. Although without a linear correlation, the combination of SUVpeak and PreLMR may increase the prognostic value and enable the identification of a subgroup of patients with an intermediate PFS. The exclusion of the seven patients did not change the prognostic value of these parameters since the included patients received a standard treatment.

A large proportion of the included patients had received two prior combined antiangiogenic treatments (26.09%). This may be the reason why we observed a median PFS slightly lower than 6.2 months in the bevacizumab combined treatment arm in the ECOG4599 trial. For tumors with high microvessel density (MVD), anti-angiogenic drugs are more likely to induce blood vessel normalization, thereby increasing the effect of anti-tumor treatment, which in turn makes the tumor insensitive to treatment. Therefore, the baseline MVD level has the potential to predict the efficacy of anti-angiogenic therapy. However, immunohistochemistry of tumor is an invasive test, and the heterogeneity of the tumor is likely to cause unstable results. Therefore, molecular imaging and hematology markers are expected to become potential indicators.

In a previous preclinical study, we found that the degree of tumor responses to bevacizumab alone, apatinib alone, or bevacizumab combined with chemotherapy or radiotherapy was positively associated with ¹⁸F−RGD tumor uptake prior to treatment (15). ¹⁸F-RGD uptake decreased more in high-uptake tumors than in low-uptake tumors. Several other preclinical studies have found that RGD uptake from pre- to post-antiangiogenic monotherapy decreased more than that in the control groups, and this occurs much earlier than the fluorodeoxyglucose (FDG) metabolic response (16). From a preclinical perspective, tumor RGD uptake at baseline or RGD uptake changes in the early stage seems to be a strong predictor for antiangiogenic therapies.

In a previous clinical study with bevacizumab-containing therapy in ovarian and cervical cancers, larger decreases in the SUVmean related to RGD uptake were found in two patients with an early objective response compared to one patient with disease progression (17). However, this initial study provided only preliminary findings, which need to be confirmed in larger validation studies. In another clinical trial, researchers found that a higher ¹⁸F-RGD uptake in pretreated tumors predicted better responses to apatinib alone (9). The results of this study were, however, not entirely consistent with previous research. For example, higher RGD uptake values at baseline have been shown to predict poorer outcomes after concurrent chemoradiotherapy in both NSCLC and glioblastoma (18, 19). The primary role of antiangiogenic therapy alone, for example, apatinib, endostar or bevacizumab, may be to prevent new vessels from forming and maintaining tumor vessels in an inactive state, although with different mechanisms of action (20). Integrin αvβ3, which is highly expressed on newly formed vessels, should be reduced by these drugs. It seems therefore that higher RGD uptake can represent a high density of effective target receptors, which increases the sensitivity to antiangiogenic treatments alone and therefore also increases the predictive power.

However, dilated, tortuous, and hyperpermeable angiogenesis represented by RGD uptake values increased tumor interstitial fluid pressure and led to poor blood flow and severe hypoxia, which increased tumor resistance to chemoradiotherapy (19, 21). In this trial, the predominant role of the antiangiogenic drug was to prune tumor vessels and improve the function of the remaining vessels to induce chemopotentiation by enhanced delivery of chemotherapeutic agents according to the vascular normalization theory. In this instance, high RGD uptake not only predicted improved outcomes with antiangiogenic agents but also suggests that the penetrative capacity of chemotherapy is enhanced and therefore can kill tumors with a higher sensitivity.

Previous studies have provided evidence that serum inflammatory factors are associated with immunological status within tumor microenvironments (10). In a retrospective study of patients receiving bevacizumab combined with chemotherapy, researchers found that NLR and PLR in sera decreased in patients with CR/PR but increased in patients with PD, while LMR was elevated in patients with CR/PR and reduced in those with evidence of PD (22). Additionally, low levels of baseline LMR, an increased NLR and a decreased LMR were all independent risk factors for death. Our findings further validate the predictive value of inflammatory factors, confirming that LMR pretreatment is an independent prognostic factor for PFS. This phenomenon may be explained simply because more lymphocytes (and fewer monocytes) migrate into the tumor microenvironment and therefore improve immunological responses, which help to prevent tumor growth. Lymphocytes are thought to play a critical role in immunology by inhibiting tumor cell proliferation and migration (23). Additionally, neutrophils and macrophages, derived from monocytic precursors within tumors, promote angiogenesis. By releasing proangiogenic growth factors such as VEGF and triggering VEGF-independent angiogenesis, this process ultimately leads to antiangiogenic agent resistance (24, 25). Therefore, inflammatory factors certainly appear to be useful predictors of combined antiangiogenic responses.

Although there was no significant correlation between the SUVpeak and PreLMR, the combination of the two parameters may provide a more accurate prognostic value and enable the identification of a subgroup of patients with a high SUVpeak with a low PreLMR and patients with a low SUVpeak with a high PreLMR who encountered intermediate survival. While this study may advance shared decision-making processes allowing patients to plan their futures, it also suggests that intercalating parameters may further increase predictive power.

To the best of our knowledge, this is the first clinical study to apply ¹⁸F-RGD uptake on PET/CT and the combined biomarker with serum inflammatory factors for selecting optimal advanced NSCLC patients to receive combined antiangiogenic treatment. As such, we ought to reflect on the limitations of this study design and our testing. First and foremost, the sample size was relatively small, and the credibility of the results need to be further verified. Therefore, larger studies are needed, not only to verify the results but also to further develop and validate biomarkers. Additionally, predictors of ¹⁸F-RGD PET/CT for short-term efficacy and survival were inconsistent. The earlier the effective treatment is determined, the greater the benefit for patients (26); however, the objective response rate did not appear to have a predictive function for PFS in this study. This may have led to inconsistent predictions, and studies with a larger sample size are needed. Second, inflammatory factors may be influenced by other causes, such as comorbid infections and the use of steroids. Even though patients with autoimmune diseases and active infections were excluded from this study, some of our participants may have been in the early stages of developing other conditions.

In conclusion, pretreatment ¹⁸F-RGD uptake on PET/CT imaging and systemic inflammatory biomarker may predict outcomes of combined antiangiogenic treatment and chemotherapy, and a higher ¹⁸F-RGD uptake or higher PreLMR may better predict short-term responses and PFS. Combined biomarkers based on molecular imaging with ¹⁸F-RGD PET/CT and serum inflammatory biomarkers may increase the prognostic value and provide a basis for risk stratification.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Shandong cancer hospital. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin. Written informed consent was obtained from the individual(s), and minor(s)’ legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

JL, LW, and ZL collected the data. JL wrote the manuscript. JNL and ZG helped to collect literature and participated in discussions. JY and XM designed and verified the study. SS helped to write and then edited this report for publication. All authors contributed to the article and approved the submitted version.

This study was supported by National Natural Science Foundation of China (81972864); Science and Technology Support Plan for Youth Innovation Teams of Universities in Shandong Province (2019KJL001); Science and Technology Plan of Jinan (201907113); The National Key Research and Development Projects of China (2018YFC1312201); Radiation Oncology Innovate Unit, Chinese Academy of Medical Sciences (2019RU071); the Academic Promotion Program of Shandong First Medical University (2019ZL002) and the foundation of National Natural Science Foundation of China (81972863, 8162790 and 82030082).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.