A non-interventional biomarker study in patients with adenocarcinoma of the lung treated with nintedanib plus docetaxel following progression on chemotherapy and/or immunotherapy: LUME-BioNIS

Introduction

Immune checkpoint inhibitors are key frontline agents for the treatment of advanced non-small cell lung cancer (NSCLC) (1,2). Immunotherapy alone or in combination with platinum doublet chemotherapy represents first-line standard of care for patients with advanced NSCLC without targetable mutations (3,4). While the most appropriate second-line treatment for patients progressing on immunochemotherapy is yet to be established, antiangiogenic agents represent potential options (5-7). Indeed, there is biological rationale for assessing regimens that incorporate both immunotherapy and anti-angiogenics due to angiogenesis and immunosuppression being interconnected biological processes (8,9). Nintedanib is an oral triple angiokinase inhibitor that inhibits fibroblast growth factor receptors 1–3, platelet-derived growth factor receptors α/β, and vascular endothelial growth factor (VEGF) receptors 1–3 (10,11).

Following the Phase III LUME-Lung 1 study (12), nintedanib combined with docetaxel was approved in Europe and some other countries for the treatment of adult patients with locally advanced, metastatic or locally recurrent NSCLC after first-line chemotherapy (13). In this study, nintedanib plus docetaxel significantly improved progression-free survival (PFS) vs. docetaxel alone [median 3.4 vs. 2.7 months; hazard ratio (HR): 0.79; 95% confidence interval (CI): 0.68–0.92]. The combination also significantly improved overall survival (OS) vs. docetaxel in patients with adenocarcinoma, especially those who progressed within 9 months of the start of first-line therapy (12). As LUME-Lung 1 was undertaken prior to the era of first-line immunotherapy, there are few prospective data assessing the effectiveness of nintedanib plus docetaxel following prior immunotherapy. However, several recent observational studies indicate that nintedanib plus docetaxel is effective in both second- and third-line settings following immunotherapy. In the non-interventional VARGADO study, the combination conferred a median PFS of 4.7 months and an objective response rate (ORR) of 35% following first-line immunochemotherapy in 164 patients with advanced NSCLC (14). In a third-line setting, the combination was associated with a median PFS of 6.4 months and an ORR of 50% following sequential chemotherapy and immunotherapy (15). Other smaller studies have also demonstrated promising activity of nintedanib plus docetaxel after immunotherapy (16-20). These data indicate that nintedanib plus docetaxel could play an important role in contemporary treatment regimens for NSCLC following immunotherapy. However, at present, there is an unmet need for clinical biomarkers that predict response to anti-angiogenic drugs, including nintedanib (21,22).

LUME-BioNIS was a prospective, non-interventional biomarker study that aimed to determine whether genetic/genomic markers (alone, or combined with clinical covariates), could be used to predict OS in patients with NSCLC (adenocarcinoma) receiving nintedanib plus docetaxel in a real-world clinical setting. The analysis was undertaken both in the overall dataset and in a subgroup of patients who had previously received immunotherapy (as this reflects current standard of care for most patients). We present this article in accordance with the TREND reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-326/rc).

Methods

Trial design and patients

LUME-BioNIS (NCT02671422) was a prospective, European, multi-center, non‑interventional study of patients with advanced, metastatic, or locally recurrent NSCLC (adenocarcinoma) who received nintedanib plus docetaxel according to the approved European Union label in routine clinical practice. All patients (or their legally accepted representatives) provided written informed consent according to the regulatory and legal requirements of the participating country.

For each patient, formalin-fixed paraffin-embedded (FFPE) tumor tissue, obtained at diagnosis and/or at rebiopsy before first-line treatment had to be available. Key exclusion criteria were: any contraindication to nintedanib or docetaxel as specified in their respective labels, initiation of nintedanib >7 days prior to inclusion in the study, and simultaneous participation in another clinical trial.

The study conformed to the provisions of the Declaration of Helsinki (as revised in 2013). The study was approved by the ethics committee/review boards of the Schleswig-Holstein Medical Association, Germany (No. 015/16 II); the General Hospital of Chest Diseases of Athens “Sotiria”; Greece (No. 10315); the Lithuanian Bioethics Committee (No. 1201423); the Independent Review Board Nijmegen, the Netherlands (No. 2015017); and the Ethics and Drug Research Committee of Parc Taulí, Spain (No. EPA-LA 1199.223).

Treatment and study endpoints

Patients received nintedanib (200 mg twice daily orally on days 2–21 of three-week cycles) plus docetaxel (75 mg/m² intravenous on day 1 of every cycle). Data were collected via electronic case report forms. Patients were to be followed until they died, were lost to follow-up, withdrew consent, or until the required number of OS events had occurred, whichever occurred first.

The primary outcome was OS in relation to exploratory biomarker assessment, including gene expression and genomic alterations. Further objectives were: to investigate time since the start of first-line chemotherapy until the start of nintedanib therapy [time since start of first-line therapy (TSFLT)] in relation to gene expression patterns; the identification of potential biomarkers to characterize the survival outcome potential of patients with late progression (≥9 months) vs. early progression (<9 months); and effectiveness outcomes: OS, PFS, and tumor response. Adverse events (AEs) and adverse drug reactions were recorded in the patient file throughout the study. AEs were categorized as mild [awareness of sign(s) or symptom(s), which were easily tolerated], moderate (enough discomfort to cause interference with usual activity), or severe (incapacitating or causing inability to work or perform usual activities).

Overview of biomarker strategy

As well as clinical variables, comprehensive analysis of molecular factors from the tumor and blood was undertaken based on: genomic biomarkers (somatic mutations of individual genes); RNA expression levels, categorized according to continuous pathways, cell phenotype, and individual angiogenesis genes; immunohistochemical markers [programmed death-ligand 1 (PD-L1) and Ki67 expression]; and Fms-related receptor tyrosine kinase 1 (FLT1) genotype (Table S1). Multivariate and univariate analysis was performed to determine whether any of these potential biomarkers served as predictors of OS in patients treated with nintedanib plus docetaxel. Biomarker analysis was performed by Boehringer Ingelheim or a qualified central laboratory.

DNA and RNA extraction from samples

FFPE tumor tissues (blocks or slides) were processed for genomic and transcriptomic analysis. Processing, histopathological analyses, including quality control, and DNA/RNA extraction was performed by a central laboratory (BARC Lab, now Cerba Research). Nucleic acid extraction was performed using the AllPrep DNA/RNA FFPE Kit (QIAGEN, Hilden, Germany). In addition, germline DNA was extracted from ethylenediamine tetraacetic acid (EDTA) blood samples using Chemagic DNA Blood Kit (PerkinElmer Chemagen Technologie GmbH, Baesweiler, Germany) with EB buffer (QIAGEN).

Targeted gene panel for genomic alteration detection using SeqCap EZ DNA sequencing

Targeted next-generation sequencing (NGS) of extracted tumor DNA from archival FFPE tissue, as well as matching germline DNA from EDTA blood samples, was undertaken using a custom-designed gene panel (180903_HG38_1199_0223_aLC_EZ_HX3; Roche Diagnostics GmbH, Mannheim, Germany) for detection of somatic alterations in 78 genes (Table S1). The panel was designed to detect several types of somatic alterations such as single nucleotide polymorphisms (SNPs), short indels, copy number variations, and gene rearrangements. Sequencing was performed on a NextSeq System (Illumina, San Diego, CA, USA) using the NextSeq^®500/550 High Output Kit v2 and the instrument specific software was used for all sequencing runs.

RNA analysis using NanoString IO panel

Analysis of gene expression from tumor RNA was performed using the PanCancer IO 360^TM panel (NanoString Technologies Inc., Seattle, WA, USA). The PanCancer IO 360^TM panel is a 770 gene CodeSet that is designed for profiling tumor biopsies and characterizing gene expression patterns associated with the tumor, the immune response, and the microenvironment, which shape tumor-immune interactions. The panel enables inference of the abundance of different cell types as well as scores for signaling pathway signatures, which describe key aspects of tumor biology and immuno-oncology to aid in sample characterization. Analyzed genes are shown in Table S1. A NanoString nCounter^® Prep Station and NanoString nCounter^® Digital Analyzer (NanoString Technologies Inc.) were used according to the manufacturers’ instructions. Analysis of the Nanostring gene expression data was performed using the nSolver 4.0 Analysis Software (NanoString Technologies Inc.).

Genotyping by TaqMan PCR technology

Genotyping of the FLT1 SNP, rs9582036 from germline DNA was also undertaken using a commercial TaqMan^®-based assay (C_1910658_10; Applied Biosystems, ThermoFisher Scientific, Waltham, MA, USA). The functional FLT1 variant rs9582036 has been identified as a prognostic determinant of recurrence in stage I–III NSCLC (23).

Immunohistochemistry (IHC)

Immunohistochemical analysis of tumor and immune cells was undertaken with validated assays by Targos Molecular Pathology GmbH to assess PD-L1 and Ki67 expression. The following parameters were assessed: percentage of PD-L1 positive tumor cells (continuous and categorical) measured by Dako IHC 28-8 pharmDX (Agilent Technologies Inc., Santa Clara, CA, USA); PD-L1 H-Score of tumor cells (continuous and categorical); Ki67 positive tumor cells (continuous and categorical); percentage of PD-L1 positive immune cells (continuous and categorical); and location of PD-L1 positive immune cells (peritumoral vs. intratumoral).

Statistical analysis

Target enrollment was up to 375 patients, assuming that ≥300 would be evaluable. Planned analysis of OS was to be conducted when 250 patients had died. However, as this planned number of OS events could not be reached, analysis was performed earlier (after 257 patients had been treated) in agreement with the European Medicines Agency.

Multivariate analysis assessed associations of biomarker signatures with OS. Clinical and molecular biomarkers were organized into eight sets. An overview of the multivariate analysis is shown in Table S2. All analyses were exploratory with the purpose of hypothesis generation. Sets 1 and 2 were subjected to integrative least absolute shrinkage and selection operator (LASSO) with penalty factors (IPF LASSO) (24), sets 3–7 were subjected to standard LASSO, and set 8 was subjected to Random Forest. Details of biomarkers included in sets 1–8 are shown in Table S3 (overall dataset analysis) and Table S4 (patients who had received previous immunotherapy). Quantitative and categorized variable observations were excluded if values were missing in at least 10 cases (in at least two categories for categorized variable). IPF LASSO (main biomarker analysis) was applied with two modalities (Table S2): Modality 1 assessed clinical parameters, IHC, and genotyping; Modality 2 assessed RNA markers, genetic alterations by gene, and tumor mutational burden. As many missing values for IHC were expected, IPF LASSO was applied twice (with and without IHC). Other variables with many missing variables (≥50%) could also be excluded. Standard LASSO was carried out for the following sets of markers separately: DNA (genomic characterization of tumor tissue); RNA (continuous pathways and cell types plus selected individual angiogenesis genes); DNA, RNA, and clinical covariates; IHC (percentage of PD-L1 positive tumor and immune cells, PD-L1 H-Score, and Ki67); and IHC, genotyping DNA, RNA, and clinical covariates. Random Forest was applied once to all variables together [IHC, genotyping (genomic evaluation of blood samples), DNA, RNA, and clinical covariates]. Random Forest was implemented by the R packages party/partykit. The marker sets were applied for all four endpoints analyzed [OS, TSFLT <9 and ≥9 months, TSFLT (continuous), and best response to first-line therapy].

Univariate analyses were performed for gene expression-based pathways and cell types, expression of selected individual angiogenesis genes, DNA alterations by gene, putative tumor mutational burden, IHC, genotyping, and the clinical covariates for OS and PFS.

Results

Patient disposition and baseline characteristics

The study was conducted at 71 oncology/pulmonology sites in 13 countries: Austria, Belgium, Denmark, Germany, Greece, Hungary, Italy, Lithuania, Luxembourg, Netherlands, Spain, Sweden, and the UK. Between March 14, 2016 and October 19, 2018, a total 295 patients were screened and 257 were treated. At data cut-off, all patients had discontinued treatment; 156 (60.7%) due to progressive disease (PD), 68 (26.5%) due to AEs, 1 (0.4%) lost to follow-up, and 32 (12.5%) for other reasons. The last patient left the study on September 03, 2019.

Four analysis sets were defined: the entered set (ES; n=260) that included all patients who entered the study (including patients who did not receive nintedanib); the biomarker evaluation set (n=239 including two patients who did not receive nintedanib) that included all patients from the ES with a valid per-protocol tissue sample; the treated set (TS; n=257) that included all patients who were documented to have received at least one dose of nintedanib; and the per-protocol set (n=213) that included all patients from the TS without any protocol deviation that led to exclusion.

At the end of the study, all patients in the TS had discontinued nintedanib and docetaxel, mostly due to PD (Table S5). A total of 129 patients (49.6%) received at least one line of subsequent therapy after discontinuing nintedanib plus docetaxel. Post-progression therapy included immunotherapy (n=84, 32.3%), chemotherapy (n=58, 22.3%), investigational therapies (n=2, 0.8%), bevacizumab (n=1, 0.4%), and others (n=5, 1.9%). Baseline characteristics and oncological history for the ES, and the subgroup of patients who received prior immunotherapy (n=67), are shown in Table 1.

The ES had a relatively high rate of poor prognostic features compared with the LUME-Lung 1 patient population, reflecting the ‘real-world’ setting of the study. In total, 91.9% of patients were current or ex-smokers, 59.6% had Eastern Cooperative Oncology Group performance status (ECOG PS) ≥1 and 30.8% of patients had received more than two lines of therapy prior to nintedanib plus docetaxel (Table 1, Table S6). Patient characteristics were similar in the ES and in patients who received prior immunotherapy (Table 1).

Most patients received nintedanib plus docetaxel in the second-line (n=179, 68.8%) or third-line (n=67, 25.8%; Table S6). In the ES, median duration of treatment was 12.9 weeks (range, 0–111 weeks). In patients who had received prior immunotherapy, median duration of treatment was 16.9 weeks (range, 1–111 weeks).

Efficacy

Overall efficacy

After a median follow-up of 19.7 months, median OS was 8.1 months (95% CI: 7.1–9.5). One- and two-year survival rates were 34.2% and 11.6%, respectively (Figure 1A). Univariate subgroup analysis indicated that the following clinical parameters had potential prognostic significance for OS: presence of liver metastases (yes vs. no); presence of adrenal metastases (yes vs. no); ECOG PS (≥1 vs. 0); TSFLT (≥9 vs. <9 months); response to first-line therapy (PD vs. non-PD; Figure 1B). In IPF LASSO multivariate analysis, the presence of brain/liver metastases and the presence of >2 metastatic sites at baseline were associated with OS (Figure 1C).

Figure 1 Overall survival in the overall population. (A) KM curve of OS in the overall population (n=260). (B) Forest plot of clinical parameters with potential prognostic significance for OS based on univariate analysis. (C) Forest plot of clinical parameters that were associated with OS based on IPF LASSO multivariate analysis. AJCC, American Joint Committee on Cancer; CI, confidence interval; diag, diagnosis; diff, differentiated; ECOG PS, Eastern Cooperative Oncology Group performance status; FDR, false discovery rate; IPF LASSO, integrative LASSO with penalty factors; KM, Kaplan-Meier; LCL, lower confidence limit; OS, overall survival; PD, progressive disease; PD-L1, programmed death-ligand 1; pos, positive; PR, partial response; SD, stable disease; UCL, upper confidence limit; UICC, Union for International Cancer Control.

Median PFS based on investigator assessment was 3.7 months (95% CI: 2.9–4.6; Figure 2A). Univariate subgroup analysis identified the following clinical parameters as having potential prognostic significance for PFS: line of treatment (third vs. second); prior immunotherapy (yes vs. no); presence of liver metastases (yes vs. no); ECOG PS (≥1 vs. 0); TSFLT (≥9 vs. <9 months; Figure 2B). IPF LASSO multivariate analysis of PFS was undertaken with the clinical parameters identified as being associated with OS only (Figure 2C).

Figure 2 Progression-free survival in the overall population. (A) KM curve of PFS in the overall population (n=260). (B) Forest plot of parameters with potential prognostic significance for PFS based on univariate analysis. (C) Multivariate analysis of PFS for the clinical parameters that were associated with OS based on IPF LASSO multivariate analysis. AJCC, American Joint Committee on Cancer; CI, confidence interval; diag, diagnosis; diff, differentiated; ECOG PS, Eastern Cooperative Oncology Group performance status; FDR, false discovery rate; KM, Kaplan-Meier; LCL, lower confidence limit; PD, progressive disease; PD-L1, programmed death-ligand 1; PFS, progression-free survival; pos, positive; PR, partial response; SD, stable disease; UCL, upper confidence limit; UICC, Union for International Cancer Control.

The investigator assessed ORR was 14.6% (all partial responses). The disease control rate (DCR) was 58.7%; 44.1% of patients showed stable disease.

Efficacy in patients who received prior immunotherapy

In patients who had received prior immunotherapy (either alone or combined with chemotherapy), median OS was 8.8 months (Figure 3A), compared with 7.8 months in patients who had not received immunotherapy (HR: 0.95; 95% CI: 0.69–1.30).

Figure 3 Overall survival in patients who received prior immunotherapy. (A) KM curve of OS in patients who received prior immunotherapy (n=67). (B) Forest plot of parameters with potential prognostic significance for OS in patients who received prior immunotherapy. AJCC, American Joint Committee on Cancer; CI, confidence interval; CR, complete response; diag, diagnosis; diff, differentiated; ECOG PS, Eastern Cooperative Oncology Group performance status; FDR, false discovery rate; KM, Kaplan-Meier; LCL, lower confidence limit; OS, overall survival; PD, progressive disease; PR, partial response; SD, stable disease; UCL, upper confidence limit; UICC, Union for International Cancer Control; n.c., not calculated.

The following clinical factors had potential prognostic significance for OS in patients who had received prior immunotherapy: line of treatment (third vs. second); presence of adrenal metastases (yes vs. no); number of metastatic sites at baseline (>2 vs. ≤2); TSFLT (≥9 vs. <9 months; Figure 3B). Of note, only a small number of patients were assessed for PD-L1 expression levels, precluding meaningful analysis.

PFS was significantly higher in patients who had received prior immunotherapy compared with those who had not (median 4.6 vs. 3.2 months; HR: 0.64; 95% CI: 0.47–0.86; Figure 4A). The only clinical parameter that showed potential prognostic significance for PFS was TSFLT (≥9 vs. <9 months; Figure 4B).

Figure 4 Progression-free survival in patients who received prior immunotherapy. (A) KM curve of PFS in patients who received prior immunotherapy (n=67). (B) Forest plot of parameters with potential prognostic significance for PFS in patients who received prior immunotherapy. AJCC, American Joint Committee on Cancer; CI, confidence interval; CR, complete response; diag, diagnosis; diff, differentiated; FDR, false discovery rate; KM, Kaplan-Meier; LCL, lower confidence limit; PD, progressive disease; PFS, progression-free survival; PR, partial response; SD, stable disease; UCL, upper confidence limit; UICC, Union for International Cancer Control; n.c., not calculated.

The DCR was higher in patients who had received prior immunotherapy compared with those who had not (78.2% vs. 51.9%).

Biomarker analysis

Biomarker analysis of the overall dataset

An overview of the biomarker analyses undertaken is shown in Figure S1. Overall, valid results from NGS, Nanostring gene expression analysis, FLT1 genotyping, PD-L1 IHC, and Ki67 IHC were obtained for 99, 135, 220, 137, and 150 patients, respectively. The top 15 most frequently altered genes were: KRAS (35% of patients with valid NGS results), TP53 (20%), KMT2D (18%), FLT4 (13%), LRP1B (13%), NOTCH1 (12%), ARID1A (11%), FAT1 (11%), CDKN2A (10%), ERBB2 (10%), STK11 (10%), ATM (9%), epidermal growth factor receptor (EGFR) (8%), KEAP1 (8%), and FLT3 (7%). Angiogenesis pathway analysis based on gene expression data from tumor tissue showed that the median angiogenesis pathway expression score was 0.36 (range, −8.76, 5.78). FLT1 SNP rs9582036 genotypes were: ‘AA’, 55.0%; ‘AC’, 39.5%; and ‘CC’, 5.5%.

The median percentage of Ki67-positive tumor cells, a marker for cell proliferation, was 45% (range, 0–99%). Of 126 patients with known tumor cell PD-L1 status, positivity <1% and ≥1% was observed in 68 (54.0%) and 58 (46.0%) patients, respectively. The median PD-L1 H‑Score was 0.5 (range, 0–270) overall. Among patients with prior immunotherapy with known PD-L1 status (n=28), positivity <1% and ≥1% was observed in 13 (46.4%) and 15 (53.6%) patients, respectively; median PD-L1 H-Score was 1.5 (range, 0–225).

In the IPF LASSO analysis of biomarker set 1, the optimal penalty factor for the model was 1:2. Only three variables were obtained with the penalized IPF LASSO for the optimal penalty factor 1:2 and OS: brain metastases (yes vs. no or missing); liver metastases sites (yes vs. no or missing); and number of metastatic sites at baseline (≤2 or missing vs. >2). No other factors were associated with OS across the eight biomarker sets.

For biomarker sets 1–5, variables with ≥30% relative frequency of IPF LASSO stability selection for TSFLT (dichotomized) are shown in Figure S2. For set 6, no variable showed ≥30% relative frequency of IPF LASSO stability selection. For set 7, variables with ≥30% relative frequency of IPF LASSO stability selection were nintedanib plus docetaxel by treatment line, and gene alteration in FAT1.

The random forest analysis (set 8), in support of the findings of the primary multivariate analysis, showed the highest variable importance (standardized variable importance ≥ P30) for two variables: the line of treatment in which nintedanib plus docetaxel was received, and prior immunotherapy (yes vs. no) (Figure S3).

Comprehensive univariate analysis identified the following molecular factors as having potential prognostic significance for OS: abundance of macrophages; mutations in the ATM gene; and percentage of PD-L1-positive tumor cells (Figure 5A). Abundance of macrophages and mutations in the ATM gene were also associated with PFS. In addition, mutations in the MET gene and gene expression levels of VEGFA demonstrated a possible association with PFS (Figure 5B). No other factors were associated with OS or PFS.

Figure 5 Molecular markers that demonstrated potential prognostic significance based on univariate analysis of the overall dataset. (A) Molecular markers that demonstrated potential prognostic significance for OS. (B) Molecular markers that demonstrated potential prognostic significance for PFS. OS, overall survival; ATM, ataxia-telangiectasia mutated; cat, category; CI, confidence interval; FDR, false discovery rate; MET, mesenchymal epithelial transition; P, percentile; PD-L1, programmed death-ligand 1; VEGFA, vascular endothelial growth factor A.

Biomarker analysis of patients who had previously received immunotherapy

Biomarker analysis was repeated in the 67 patients who had received previous immunotherapy. Several factors had potential prognostic significance for OS based on univariate analysis: RNA expression levels of genes involved in cell proliferation, DNA damage repair and interferon signaling, and abundance of neutrophils (Figure S4A). RNA expression levels of genes involved in DNA damage repair and interferon signaling, and abundance of neutrophils, were also associated with PFS. Other factors associated with PFS were RNA expression levels of genes involved in matrix remodeling and metastasis, and TGF-beta signaling and RNA expression levels of VEGFA (Figure S4B).

Safety

Of 257 patients, AEs were reported in 205 (79.8%) patients, of which 122 (47.5%) were severe (Table S7). The most common AEs (any intensity) were diarrhea (29.6%), nausea (10.5%), and fatigue (7.8%). The most common severe AEs (besides malignant neoplasm progression) were febrile neutropenia (3.5%), diarrhea (3.1%), pneumonia (3.1%), asthenia (1.6%), and neutropenia (1.6%). AEs leading to dose reduction were reported in 22 (8.6%) patients, most commonly diarrhea (n=12, 4.7%), nausea (n=4, 1.6%), vomiting, alanine aminotransferase increased, aspartate aminotransferase increased, gamma-glutamyltransferase increased, transaminases increased, and decreased appetite (n=2, 0.8% each). AEs leading to treatment discontinuation were reported in 110 (42.8%) patients. The most common AEs leading to discontinuation were malignant neoplasm progression (11.7%), diarrhea (9.3%), and nausea (2.7%). Eighty-nine (34.6%) patients died during treatment. The most frequent fatal AEs were malignant neoplasm progression (24.9%), pneumonia (2.3%), performance status decreased, and respiratory failure (1.2% each). Around half (50.6%) of patients had drug-related AEs and one fatal drug-related AE (ischemic stroke) was reported.

Discussion

In this trial, nintedanib plus docetaxel demonstrated encouraging efficacy in patients with NSCLC, especially given that 31% of the patients were treated in third-line or later settings and/or had other poor prognosis features. In patients who received prior immunotherapy, median OS was 8.8 months and median PFS was 4.6 months, vs. 7.8 and 3.2 months in those without. Moreover, no new safety signals were observed in this real-world population. To date, no biomarkers have been identified that predict effectiveness of anti-angiogenic treatments. Despite our comprehensive biomarker investigations, the primary multivariate analysis only identified three covariates that were predictive of OS. These factors—brain metastases, liver metastases, and the number of metastatic sites—were already known as prognostic biomarkers (25-27). Accordingly, biomarkers to predict response to nintedanib plus docetaxel unfortunately remain elusive, including in patients previously treated with contemporary immunotherapy-based regimens.

The efficacy data in LUME-BioNIS were largely consistent with other observational studies that have assessed nintedanib plus docetaxel following immunotherapy. VARGADO Cohort B assessed the combination in a third-line setting after sequential chemotherapy and immunotherapy and reported median OS and PFS of 12.1 and 6.4 months, respectively (15). Additionally, a named patient use program undertaken in Spain reported median PFS of 3.2 months (16). Median OS and PFS were also similar to that observed in VARGADO Cohort C in patients who received second-line nintedanib plus docetaxel following first-line chemotherapy plus immunotherapy (median OS: 8.1 months; median PFS: 4.7 months) (14). Of note, given that immunotherapy with or without chemotherapy is now first-line standard of care, OS with nintedanib plus docetaxel was similar in patients who received prior immunotherapy and those who received chemotherapy only (8.8 vs. 7.8 months). Subgroup analysis indicated that OS may vary according to the treatment line in which immunotherapy was received, but this likely reflects low patient numbers in the analysis or rapid changes in first-line standard of care during the period that the trial was undertaken. Of interest, PFS with nintedanib plus docetaxel in LUME-BioNIS was longer in the subgroup of patients who received prior immunotherapy than those who received chemotherapy only. This may suggest an interaction between anti-angiogenic and pro-immunogenic treatment modalities. Indeed, preclinical investigations have suggested a hypothetical ‘angio-immunogenic switch’, in which anti-angiogenic therapy restores an immunosupportive tumor microenvironment (28). However, further clinical validation of this hypothesis is required. The encouraging efficacy in the present study, along with the VARGADO trial (5,15), indicates that nintedanib plus docetaxel may be a treatment option in a post-immunotherapy setting.

The failure of LUME-BioNIS to identify any new biomarkers by multivariate analysis illustrates the wider challenge of developing clinically applicable biomarkers for anti-angiogenic treatments (22). For example, ANSELMA, a recent meta-analysis in over 8,000 patients across 16 trials, compared OS and PFS benefit in patients receiving second-line anti-angiogenic therapy plus standard second-line treatment vs. standard second-line treatment alone (29). In this study, no novel predictors of OS benefit were identified from the clinical parameters examined other than age and TSFLT <9 months. Despite the negative results of multivariate analysis, univariate analysis identified some potentially interesting clinical and molecular covariates that warrant further investigation. For example, consistent with previous studies (5,12), outcomes were poorer in patients with TSFLT of <9 months, reflecting a patient population at risk of rapid progression. Of note, in LUME-Lung 1, a more pronounced difference in efficacy was observed with nintedanib plus docetaxel than with docetaxel alone in patients with TSFLT <9 months than in the overall population (12). Molecular biomarkers of potential interest were macrophage abundance, ATM gene mutations, and percentage of PD-L1-positive tumor cells (macrophage abundance and ATM mutations were also associated with PFS). MET gene mutations and gene expression levels of VEGFA demonstrated a possible association with PFS. Ultimately, the identification of biomarkers for anti-angiogenics in the post-immunotherapy setting is compounded by a lack of clinical data and biological understanding of how and why patients become resistant to immunotherapy. More mechanistic data relating to progression on immunotherapy are required so that potential biomarkers can be tested in patient subgroups based on their biology.

This study had a number of limitations. The non-interventional, non-comparative study design may have impacted the completeness of data collection. Survival may not be the ideal endpoint in this type of analysis as survival may be affected by factors not related to therapeutic efficacy (22). Furthermore, use of radiologic criteria was not mandatory for response evaluation. The majority of patients treated in the trial received first-line chemotherapy only, which no longer reflects standard of care for advanced NSCLC without targetable mutations. While subgroup analysis was undertaken in patients who received prior immunotherapy, this analysis was post hoc and exploratory in nature, with relatively few patients in the subgroups compared. Additionally, biomarker sample quality was variable. Non-experienced sites and the use of archived samples led to a low rate of analyzable samples. Finally, despite comprehensive biomarker analysis, all clinical and molecular parameters were based on baseline measurements. There was no analysis of dynamic changes in parameters and their possible correlation with OS.

Conclusions

The comprehensive biomarker assessment in the present study did not identify new actionable biomarkers predictive of OS in real-world patients with advanced NSCLC who received nintedanib plus docetaxel following failure of first-line chemotherapy and/or immunotherapy. There remains an unmet need for clinical biomarkers to help identify NSCLC patients who might benefit from anti-angiogenic agents.

Acknowledgments

The authors thank the patients and their caregivers for making these studies possible. Medical writing support for the development of this manuscript, under the direction of the authors, was provided by Lynn Pritchard, DPhil, and Jim Sinclair, PhD, of Ashfield MedComms, an Inizio company, and funded by Boehringer Ingelheim.

Funding: This study was funded by Boehringer Ingelheim. Boehringer Ingelheim was given the opportunity to review the manuscript for medical and scientific accuracy as well as intellectual property considerations.

Footnote

Reporting Checklist: The authors have completed the TREND reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-326/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-326/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-326/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-24-326/coif). M.R. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; consulting fees, payments, honoraria, meeting and travel support from Amgen, AstraZeneca, BeiGene, Boehringer Ingelheim, Bristol Myers Squibb, Eli Lilly, GSK, Merck, Mirati Therapeutics, MSD, Novartis, Pfizer, Regeneron, and Roche; participates on data safety monitoring boards or advisory boards for Daiichi Sankyo and Sanofi; and receives medical writing support funding (Articulate Science, LLC) from Daiichi Sankyo. K.S. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; payment and honoraria from AstraZeneca, Bristol Myers Squibb, and MSD; and meeting and travel support from AstraZeneca. S.M. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim. S.C.v.W. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim. B.M. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; travel grants from AstraZeneca, Bristol Myers Squibb, MSD, and Roche; and personal fees from BeiGene and Takeda. H.B. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; consulting fees, payments and honoraria from Boehringer Ingelheim. A.V.S. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; and employment by Boehringer Ingelheim. R.M.L. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; consulting fees and employment by Boehringer Ingelheim. A.M.E. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; and employment by Boehringer Ingelheim. T.K. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; and employment by Boehringer Ingelheim and Merck Healthcare KGaA. K.K. reports funding, article processing charges and medical writing support funding (Ashfield MedComms) for this study from Boehringer Ingelheim; consulting fees from AbbVie, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol Myers Squibb, Daiichi Sankyo, Debiopharm, Diaceutics, Eli Lilly, Merck Serono, MSD, Novartis, Pfizer, Regeneron, Roche, Roche Diagnostics, Ventana, and Sanofi; payment and honoraria from AstraZeneca, Amgen, Boehringer Ingelheim, Bristol Myers Squibb, Eli Lilly, Merck Serono, MSD, Novartis, Pfizer, Roche, Roche Diagnostics, Ventana, Medscape, and Prime Oncology; and leadership roles on the IASLC Pathology Committee. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study conformed to the provisions of the Declaration of Helsinki (as revised in 2013). The study was approved by the ethics committee/review boards of the Schleswig-Holstein Medical Association, Germany (No. 015/16 II); the General Hospital of Chest Diseases of Athens “Sotiria”; Greece (No. 10315); the Lithuanian Bioethics Committee (No.1201423); the Independent Review Board Nijmegen, the Netherlands (No. 2015017); and the Ethics and Drug Research Committee of Parc Taulí, Spain (No. EPA-LA 1199.223). Informed consent was taken from all the patients.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Wang F, Xia T, Li Z, et al. Current status of clinical trial research and application of immune checkpoint inhibitors for non-small cell lung cancer. Front Oncol 2023;13:1213297. [Crossref] [PubMed]
2. Olivares-Hernández A, González Del Portillo E, Tamayo-Velasco Á, et al. Immune checkpoint inhibitors in non-small cell lung cancer: from current perspectives to future treatments-a systematic review. Ann Transl Med 2023;11:354. [Crossref] [PubMed]
3. Hendriks LE, Kerr KM, Menis J, et al. Non-oncogene-addicted metastatic non-small-cell lung cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol 2023;34:358-76. [Crossref] [PubMed]
4. Jaiyesimi IA, Leighl NB, Ismaila N, et al. Therapy for Stage IV Non-Small Cell Lung Cancer Without Driver Alterations: ASCO Living Guideline, Version 2023.3. J Clin Oncol 2024;42:e23-43. [Crossref] [PubMed]
5. Grohé C, Wehler T, Dechow T, et al. Nintedanib plus docetaxel after progression on first-line immunochemotherapy in patients with lung adenocarcinoma: Cohort C of the non-interventional study, VARGADO. Transl Lung Cancer Res 2022;11:2010-21. [Crossref] [PubMed]
6. Brueckl WM, Reck M, Rittmeyer A, et al. Efficacy of docetaxel plus ramucirumab as palliative second-line therapy following first-line chemotherapy plus immune-checkpoint-inhibitor combination treatment in patients with non-small cell lung cancer (NSCLC) UICC stage IV. Transl Lung Cancer Res 2021;10:3093-105. [Crossref] [PubMed]
7. Reck M, Popat S, Grohé C, et al. Anti-angiogenic agents for NSCLC following first-line immunotherapy: Rationale, recent updates, and future perspectives. Lung Cancer 2023;179:107173. [Crossref] [PubMed]
8. Fukumura D, Kloepper J, Amoozgar Z, et al. Enhancing cancer immunotherapy using antiangiogenics: opportunities and challenges. Nat Rev Clin Oncol 2018;15:325-40. [Crossref] [PubMed]
9. Lee WS, Yang H, Chon HJ, et al. Combination of anti-angiogenic therapy and immune checkpoint blockade normalizes vascular-immune crosstalk to potentiate cancer immunity. Exp Mol Med 2020;52:1475-85. [Crossref] [PubMed]
10. Hilberg F, Roth GJ, Krssak M, et al. BIBF 1120: triple angiokinase inhibitor with sustained receptor blockade and good antitumor efficacy. Cancer Res 2008;68:4774-82. [Crossref] [PubMed]
11. Hilberg F, Tontsch-Grunt U, Baum A, et al. Triple Angiokinase Inhibitor Nintedanib Directly Inhibits Tumor Cell Growth and Induces Tumor Shrinkage via Blocking Oncogenic Receptor Tyrosine Kinases. J Pharmacol Exp Ther 2018;364:494-503. [Crossref] [PubMed]
12. Reck M, Kaiser R, Mellemgaard A, et al. Docetaxel plus nintedanib versus docetaxel plus placebo in patients with previously treated non-small-cell lung cancer (LUME-Lung 1): a phase 3, double-blind, randomised controlled trial. Lancet Oncol 2014;15:143-55. [Crossref] [PubMed]
13. European Medicines Agency Committee for Medicinal Products. Summary of product characteristics: vargatef. 2015, last updated 07 December 2021. Available online: https://www.ema.europa.eu/en/documents/product-information/vargatef-epar-product-information_en.pdf
14. Grohé C, Wehler T, Henschke S, et al. Effect of performance status (ECOG PS) on treatment outcome with second-line (2L) nintedanib (NIN) + docetaxel (DOC) for patients (pts) with lung adenocarcinoma after first-line (1L) immune checkpoint inhibitor (ICI) combination therapy. J Clin Oncol 2022;40:9092.
15. Grohé C, Blau W, Gleiber W, et al. Real-World Efficacy of Nintedanib Plus Docetaxel After Progression on Immune Checkpoint Inhibitors: Results From the Ongoing, Non-interventional VARGADO Study. Clin Oncol (R Coll Radiol) 2022;34:459-68. [Crossref] [PubMed]
16. Corral J, Majem M, Rodríguez-Abreu D, et al. Efficacy of nintedanib and docetaxel in patients with advanced lung adenocarcinoma treated with first-line chemotherapy and second-line immunotherapy in the nintedanib NPU program. Clin Transl Oncol 2019;21:1270-9. [Crossref] [PubMed]
17. Reck M, Syrigos K, Miliauskas S, et al. Non-interventional LUME-BioNIS study of nintedanib plus docetaxel after chemotherapy in adenocarcinoma non-small cell lung cancer: A subgroup analysis in patients with prior immunotherapy. Lung Cancer 2020;148:159-65. [Crossref] [PubMed]
18. Metzenmacher M, Rizzo F, Kambartel K, et al. Real-world efficacy of docetaxel plus nintedanib after chemo-immunotherapy failure in advanced pulmonary adenocarcinoma. Future Oncol 2021;17:3965-76. [Crossref] [PubMed]
19. Ljubicic L, Janzic U, Unk M, et al. Efficacy and safety of nintedanib and docetaxel in patients with previously treated lung non-squamous non-small cell lung cancer: a multicenter retrospective real-world analysis. Radiol Oncol 2023;57:397-404. [Crossref] [PubMed]
20. Hochmair MJ, Kolb R, Wurm R, et al. Nintedanib plus Docetaxel after Immune Checkpoint Inhibitor Failure in Patients with Advanced Non-Small-Cell Lung Cancer: A Case Series. Case Rep Oncol 2022;15:138-48. [Crossref] [PubMed]
21. Wehland M, Bauer J, Magnusson NE, et al. Biomarkers for anti-angiogenic therapy in cancer. Int J Mol Sci 2013;14:9338-64. [Crossref] [PubMed]
22. Kurzrock R, Stewart DJ. Exploring the Benefit/Risk Associated with Antiangiogenic Agents for the Treatment of Non-Small Cell Lung Cancer Patients. Clin Cancer Res 2017;23:1137-48. [Crossref] [PubMed]
23. Glubb DM, Paré-Brunet L, Jantus-Lewintre E, et al. Functional FLT1 Genetic Variation is a Prognostic Factor for Recurrence in Stage I-III Non-Small-Cell Lung Cancer. J Thorac Oncol 2015;10:1067-75. [Crossref] [PubMed]
24. Boulesteix AL, De Bin R, Jiang X, et al. IPF-LASSO: Integrative L(1)-Penalized Regression with Penalty Factors for Prediction Based on Multi-Omics Data. Comput Math Methods Med 2017;2017:7691937. [Crossref] [PubMed]
25. Ali A, Goffin JR, Arnold A, et al. Survival of patients with non-small-cell lung cancer after a diagnosis of brain metastases. Curr Oncol 2013;20:e300-6. [Crossref] [PubMed]
26. Riihimäki M, Hemminki A, Fallah M, et al. Metastatic sites and survival in lung cancer. Lung Cancer 2014;86:78-84. [Crossref] [PubMed]
27. Oh Y, Taylor S, Bekele BN, et al. Number of metastatic sites is a strong predictor of survival in patients with nonsmall cell lung cancer with or without brain metastases. Cancer 2009;115:2930-8. [Crossref] [PubMed]
28. Popat S, Curioni-Fontecedro A, Dafni U, et al. A multicentre randomised phase III trial comparing pembrolizumab versus single-agent chemotherapy for advanced pre-treated malignant pleural mesothelioma: the European Thoracic Oncology Platform (ETOP 9-15) PROMISE-meso trial. Ann Oncol 2020;31:1734-45. [Crossref] [PubMed]
29. Remon J, Lacas B, Herbst R, et al. ANtiangiogenic Second-line Lung cancer Meta-Analysis on individual patient data in non-small cell lung cancer: ANSELMA. Eur J Cancer 2022;166:112-25. [Crossref] [PubMed]