We scanned the medical data of patients with LC treated with anlotinib who were admitted at the West China Hospital from June 2018 to January 2021. Patients treated with anlotinib after the failure of previous antiangiogenic therapy (bevacizumab or endostatin), were assigned to group A (83 patients). By contrast, patients treated with anlotinib without any previous antiangiogenic therapy, were assigned to group B (77 patients).

The initial dose of anlotinib was 12mg or 10mg orally once daily on days 1-14, every 3 weeks. The dose for some patients was reduced to 10mg/day or 8mg/day when they became intolerant to adverse reactions. Anlotinib was continued until disease progression or when patients remained intolerant to adverse reactions. The anlotinib regimens include:(1)anlotinib monotherapy,(2)anlotinib combined with chemotherapy (no restriction on chemotherapy regimen), (3) anlotinib combined with immune checkpoint inhibitors (anti-PD-1/L1 antibodies), (4) anlotinib combined with targeted agents (including epidermal growth factor receptor TKI and ALK/ROS inhibitors), and (5) anlotinib combined with local radiotherapy (no limitation on the radiotherapy site and dose).

Radiographic examinations were performed to evaluate the efficacy two cycles after initiation, and then every two cycles or periodically according to clinical conditions. Efficacy evaluation was directly performed by the researchers using the Picture Archiving and Communication Systems in the hospital according to the Response Evaluation Criteria In Solid Tumor 1.1, with reference to imaging reports and clinical practice. The best response evaluation was the best response record from the beginning to the end of anlotinib treatment. Progression-free survival (PFS) was defined as the time from anlotinib initiation to the presence of objective evidence of disease progression (or death for any reason). Overall survival (OS) was defined as the time from anlotinib initiation until death, or loss to follow-up or reaching the study observation deadline. Adverse events were documented according to the electronic medical records of our hospital, and the incidence of adverse events was lower than the actual incidence.

Statistical analyses were performed using R version 4.0.5. Categorical variables were compared using Pearson’s chi-square test or Fisher’s exact test. Survival curves were created using the Kaplan–Meier method. The log-rank test was used for univariate analysis of PFS and OS. Propensity score matching (PSM) was performed using 1:1 nearest neighbor matching (caliper 0.2). The matching factors were sex, age (< 60 years old, ≥ 60 years old), histological subtype (adenocarcinoma, squamous cell carcinoma, others), clinical-stage (stage IV), and the number of previous treatment lines (<3 lines, ≥3 lines). Statistical significance was set at P<0.05 (both sides).

A total of 172 patients were treated with anlotinib. Three patients were lost to follow-up, two cases were discontinued anlotinib after several days, one of which was due to dizziness, and the other was due to fatigue, hypertension, and proteinuria. Efficacy evaluation could not be performed in seven patients due to imaging inaccessibility, therefore they were excluded. A total of 160 patients were included in the analysis, 83 of whom had received prior antiangiogenic therapy (bevacizumab or endostatin) (group A) and 77 of whom did not receive antiangiogenic therapy (group B).The research flowchart is shown in Figure 1 , and the baseline patient characteristics are shown in Table 1 . As seen in the table, the two patients groups were not balanced in terms of sex (P=0.020), age (P=0.000), pathological type (P=0.000), clinical-stage (P=0.015), PD-L1 expression level (P=0.019), smoking history (P=0.02), number of previous treatment lines (P=0.000), history of targeted therapy (P=0.024), and therapeutic schemes (P=0.014) ( Table 1 ). PSM was used to control the influence of confounding factors between the two groups. After matching, the baseline characteristics of the two groups became comparable ( Table 2 ).

The best response in group A was partial response (PR) (n =8), stable disease (SD) (n = 51) and progressive disease (PD) (n =24). In group B, the best response PR (n =8), SD (n = 55) and PD (n =14). The objective response rate (ORR) and disease control rate (DCR) of the two groups were 9.6% in group A vs. 10.4% in group B (P=0.874), and 71.1% in group A vs. 80.5% in group B (P=0.165), respectively ( Table 3 ).

For the overall population, the median OS and PFS were was 14.6 months (95%CI 11.1-18.1) and 3.8 months (95%CI 3.1-4.5), respectively. The median OS was 11.8 months (95%CI 6.7-16.9) in group A vs. 16.1 months (95%CI 12.6-19.6) in group B (P=0.121) ( Figure 2 ). Median PFS was 3.1 months vs. 4.7 months (95%CI 3.9-5.5),(P=0.009; Figure 3 ).

After matching, the ORR of the two groups was 13.0% vs. 10.9% (P=0.748). The DCR of two groups was 78.3% vs. 82.6% (P=0.599), respectively ( Table 3 ). The median OS of the two groups was 14.6 months (95%CI 10.1-19.1) in group A vs. 16.2 months (95%CI 9.2-23.2) in group B (P=0.320; Figure 4 ). Median PFS was 3.5 months (95%CI 2.7-4.3) and 4.5 months (95%CI 3.7-5.3) (P=0.040; Figure 5 ).

After anlotinib progression, 57 patients (68.7%) in group A and 61 patients (79.2%) in group B received subsequent therapy. However, there was no statistically significant difference in subsequent line therapy between the two groups before and after matching. The most common posterior line treatments in both groups were targeted therapy, immunotherapy, and radiotherapy, whereas the other relatively uncommon treatments included chemotherapy and liver interventional therapy. However, data regarding the posterior line treatment of a few patients were unavailable ( Table 4 ).

The effect of previous antiangiogenic treatment on anlotinib in group A was further analyzed, the Kaplan-Meier analysis showed that the PFS for anlotinib in the bevacizumab and endostatin groups were 3.2 and 2.1 months, respectively (P =0.973). The PFS for anlotinib in the PR, SD, and PD groups that received prior antiangiogenic therapy were 2.7 months, 3.1 months and 3.8 months (P =0.918), respectively. Similarly, after grouping according to the PFS achieved from previous antiangiogenic treatment, the PFS for anlotinib in the shorter (≤6 months) and longer (>6 months) groups were 2.9 and 3.3 months (P=0.592), respectively. The OS data for the anlotinib group after the progression of previous antiangiogenic therapy are detailed in Supplementary Table 1 .

The univariate analysis showed that female patients (P=0.021), patients with ECOG scores of 0-2 (P=0.001), and patients with brain metastases (P=0.015) had longer PFS ( Table 5 ). Factors with P<0.05 in the univariate analysis and those clinically associated with prognostic factors (liver metastasis) were included in the multivariate analysis. ECOG score (P=0.002), brain metastasis (P=0.006), and liver metastasis (P=0.006) were factors found to independently influence PFS ( Table 6 ). In terms of OS, the univariate analysis showed that patients with an ECOG score of 0-2 had significantly longer OS than those with an ECOG score of 3, whereas other factors did not affect the OS of the patients ( Table 5 ).

No life-threatening adverse events associated with anlotinib were documented in this study. However, six patients were spontaneously discontinued in group A due to being intolerant to adverse reactions, including one patient who experienced vaginal bleeding and fatigue, one who experienced hematuria, three who experienced skin symptoms, and one who experienced nausea and vomiting. In group B, four patients were discontinued due to adverse reaction intolerance, including one who experienced nausea and vomiting, one who had hand and foot syndrome, one who had an acute stroke, and one who had upper gastrointestinal bleeding after an esophageal cancer diagnosis.

There was no statistically significant difference in the incidence of adverse events between the two groups. The most common adverse events in group A were skin symptoms (24.1%), including rash, pruritus, hand and foot syndrome, followed by hypertension (8.4%) and fatigue (6.0%). The other relatively rare adverse events of concern included vaginal bleeding (3.6%), nosebleeds (2.4%), and hematuria (1.2%). In group B, the most common adverse events were skin symptoms (14.3%), hypertension (10.4%), and anorexia (10.4%), followed by diarrhea (7.8%) and fatigue (7.8%). The other relatively uncommon adverse events of concern included hemoptysis (2.6%), upper gastrointestinal bleeding (1.3%), and stroke (1.3%) ( Table 7 ).

Our study has several limitations. The results of this study were based on the experience of a single-center and a small sample size. The baseline characteristics of the patients in the two groups were influenced by confounding factors before matching. After PSM, the number of patients available for analysis was halved, which reduced the representativeness of our cohort, and could also have introduced certain limitations in the results. The variety of drugs used in combination with anlotinib might have also affected the outcomes.