Abstract
Our preclinical work suggests that appropriate angiogenesis inhibition could potentiate PD-1/PD-L1 blockade via alleviating hypoxia, increasing infiltration of CD8+ T cells and reducing recruitment of tumor-associated macrophages. We hereby conducted a clinical trial to evaluate this combination in pretreated patients with advanced non–small cell lung cancer (NSCLC).
The study included phase Ib apatinib dose-escalation and phase II expansion cohorts. Patients received apatinib at doses of 250–500 mg orally once daily, in combination with camrelizumab 200 mg intravenously every 2 weeks.
From March 2017 to October 2018, 105 chemotherapy-pretreated patients with nonsquamous NSCLC were enrolled and received apatinib 250 mg (recommended phase II dose) and camrelizumab. Among them, one (1.0%) complete response, 28 (26.7%) partial responses, and 48 (45.7%) stable diseases were observed. In the efficacy-evaluable population (n = 94), objective response rate (ORR) was 30.9% [95% confidence interval (CI), 21.7–41.2]. The median progression-free survival was 5.7 months (95% CI, 4.5–8.8) and overall survival was 15.5 months (95% CI, 10.9–24.5). Efficacy of combination therapy was evident across all PD-L1 and tumor mutation burden subgroups, and appeared to be improved in patients with STK11/KEAP1 mutation (mutant vs. wild-type, ORR: 42.9% vs. 28.1%; 1-year survival rate: 85.1% vs. 53.1%). No unexpected adverse events were observed.
Combined apatinib and camrelizumab showed encouraging antitumor activity and acceptable toxicity in chemotherapy-pretreated patients with advanced nonsquamous NSCLC. Patients with STK11/KEAP1 mutation might derive more benefits from this combination. We will validate these results in an ongoing phase III trial (NCT04203485).
PD-1/PD-L1 blockade immunotherapies have profoundly altered the treatment algorithm of non–small-cell lung cancer (NSCLC). Nevertheless, only a small subset of patients benefit from anti-PD-1/PD-L1 monotherapy. Our preclinical work has demonstrated that angiogenesis inhibition could alleviate hypoxia, increase infiltration of CD8+ T cells, and reduce recruitment of tumor-associated macrophages and therefore may potentiate cancer immunotherapy. We hereby conducted a proof-of-concept clinical trial to investigate whether such combination represented a potential therapeutic strategy. The current analysis included 105 chemotherapy-pretreated and immunotherapy-naïve patients with nonsquamous NSCLC who received apatinib 250 mg (every day) and camrelizumab (once every 2 weeks). We observed an objective response rate of 30.9% and a median progression-free survival of 5.7 months, with no unexpected adverse events. Further biomarker analysis showed that STK11/KEAP1 mutation might be associated with improved treatment efficacy. A phase III confirmative trial is currently underway to evaluate apatinib and camrelizumab combination therapy as a first-line treatment for advanced NSCLC (NCT04203485).
Introduction
Immunotherapy has revolutionized the management of advanced/metastatic non–small cell lung cancer (NSCLC). Immune checkpoint inhibitors (ICI) targeting the programmed death 1 (PD-1)/PD ligand 1 (PD-L1) pathway, such as nivolumab, pembrolizumab, and atezolizumab, are the standard-of-care second-line treatment for advanced NSCLC (1–4). In addition, pembrolizumab or atezolizumab monotherapy is now a preferred first-line choice in patients with advanced/metastatic NSCLC with high PD-L1 expression (5–8). However, the majority of patients do not benefit from single-agent anti-PD-1/PD-L1 therapy, highlighting an unmet need to develop novel combination strategies.
Antiangiogenic therapies targeting the VEGF or VEGFR-2 have been found to increase the trafficking of cytotoxic T cells into tumors, reduce immunosuppressive cytokine levels, and inhibit regulatory T-cell proliferation, and therefore might have synergistic or additive effects with ICIs (9–11). Two phase III studies in metastatic renal cell carcinoma showed that the combination of angiogenesis inhibitor axitinib with either pembrolizumab or avelumab improved progression-free survival (PFS) and objective response rate (ORR) compared with sunitinib alone (12, 13). In advanced NSCLC, bevacizumab was found to enhance atezolizumab in combination with chemotherapy in patients with metastatic nonsquamous NSCLC in the first-line setting (14), whereas ramucirumab or lenvatinib in combination with pembrolizumab showed promise as a later-line therapy in two recent phase I studies (15, 16).
Camrelizumab (SHR-1210), a novel IgG4-κ PD-1 mAb, has been approved as a third-line treatment for classical Hodgkin lymphoma in China. In addition, the combination of camrelizumab with chemotherapy has been shown to improve ORR and PFS as first-line therapy for advanced nonsquamous NSCLC (17, 18). Apatinib, a VEGFR-2 tyrosine kinase inhibitor (TKI), has been approved as a third-line therapy for advanced gastric cancer in China (19). In our preclinical study, we observed that apatinib could modulate tumor immune microenvironment (TIME) via alleviated hypoxia, increased tumoral infiltration of CD8+ T cells, and reduced recruitment of tumor-associated macrophages (11). This is in agreement with several other studies showing angiogenesis inhibitors, when given at appropriate dose, could reverse the immunosuppressive tumor microenvironment and potentiate cancer immunotherapy (20–23).
We hereby conducted a phase Ib/II study of camrelizumab plus apatinib in patients with advanced NSCLC. Here, we report the results for a subgroup cohort of 105 chemotherapy-pretreated and immunotherapy-naïve patients with advanced nonsquamous NSCLC.
Patients and Methods
Study design and patients
This open-label, multicenter, multicohort, phase Ib/II study of camrelizumab in combination with apatinib in patients with NSCLC was conducted at 26 sites in China (ClinicalTrials.gov identifier: NCT03083041). The study comprised of two phases: the dose-escalation Ib phase was designed to assess the tolerability, safety, and pharmacokinetics of camrelizumab in combination with apatinib and to establish the recommended phase II dose (RP2D) for apatinib; the dose-expansion phase served to further assess the efficacy and safety of the combination at the RP2D (Supplementary Fig. S1). In this article, only patients with advanced nonsquamous NSCLC without targetable driver mutations who received camrelizumab and apatinib at the RP2D were reported (including patients from phase Ib and from phase II cohort 1).
Patients were eligible if they were 18–70 years old; had histologically or cytologically confirmed recurrent or metastatic NSCLC, measurable disease, and disease progression after at least one prior platinum-based chemotherapy regimen (one for cohort 1 and ≥2 for phase Ib cohort). Other key eligibility criteria included no EGFR or ALK targetable mutations (except for pharmacokinetic study), Eastern Cooperative Oncology Group performance status 0 or 1, adequate organ function, a life expectancy of at least 3 months and no prior treatment with immunotherapy. All participants were required to provide a fresh or archival tissue sample for the analysis of PD-L1, tumor mutation burden (TMB) and cancer genomics.
The trial was carried out in accordance with the International Conference on Good Clinical Practice Standards and the Declaration of Helsinki. The Institutional Review Boards or independent ethics committees of all participating centers approved the protocol and amendments, and all patients provided written informed consent.
Procedures
In the dose-escalation phase Ib study, the dose of camrelizumab was fixed (200 mg intravenously every 2 weeks) and the starting dose of apatinib was 250 mg (orally once daily) and could escalate up to a maximum of 500 mg (Supplementary Fig. S2). All participants in phase II cohort 1 received the RP2D regimen of camrelizumab plus 250 mg once-daily apatinib established from phase Ib study (24). We considered every 4 weeks of treatment as a cycle. In the phase II part, patients were treated with combination therapy until disease progression, intolerable toxicity, or withdrawal by investigator or patient's decision. Tumor response was assessed by radiographic imaging by the investigators according to the RECIST version 1.1 every 8 weeks for the first six cycles and every 12 weeks thereafter. Treatment response was required to be confirmed by a subsequent radiographic assessment after no less than 4 weeks. Adverse events (AE) were evaluated per the Common Terminology Criteria for Adverse Events version 4.03, with causality to treatment recorded. The immune-mediated AEs were also noted and graded. Patients who discontinued treatment for reasons other than disease progression were followed up until progression.
Outcomes
The primary endpoints were safety in the dose-escalation phase and ORR in the dose-expansion phase. Secondary endpoints included duration of response (DoR), disease control rate [DCR, defined as the proportion of patients with complete or partial response or durable stable disease (≥ 6 weeks)], clinical benefit response rate (CBR, defined as the proportion of patients with complete or partial response or stable disease at week 24), PFS, overall survival (OS) and 12-month survival rate. Exploratory analysis included the correlation of response with the PD-L1 tumor proportion score (TPS), TMB, and cancer genomic findings.
PD-L1 expression analysis
PD-L1 protein expression was centrally assessed by IHC (PD-L1 IHC 22C3 pharmDx) using archived or fresh tumor tissues harvested prior to the treatment. The expression level was determined using TPS, defined as the percentage of viable tumor cells showing partial or complete membrane staining at any intensity. Specimens were considered PD-L1-positive if TPS ≥1%.
Cancer genomics and TMB analysis
The next-generation sequencing (NGS) was performed using the BGI Oseq pan-cancer panel (covers 636 genes and 1.95 Mb) on the MGISEQ-2000 platform using peripheral blood samples (i.e., liquid biopsy), and pretreatment tumor biopsy or archival tissue samples. Genomic alterations, including single base substitutions (single-nucleotide variants), short and long insertions/deletions, copy-number variants, and gene rearrangement and fusions, were all assessed. The TMB was determined by analyzing somatic mutations, including the coding base substitution and indels per megabase and was categorized as high or low with the median as cutoff (25).
Statistical analysis
Demographics, baseline characteristics, safety, and tolerability were descriptively summarized. Tumor response was analyzed in the full analysis set (FAS), defined as all participants who received at least one dose of the investigational product, as well as the efficacy-evaluable set, defined as subjects who received at least one dose of investigational product, and had at least one postbaseline radiographic evaluation for efficacy, as prespecified in study protocol and statistical analysis plan. The 95% confidence interval (CI) for ORR was estimated using the Clopper–Pearson method. Time-to-event endpoints were estimated by the Kaplan–Meier approach. Exploratory analyses of response in correlation to tumor PD-L1 status, TMB, and cancer genomic findings were also conducted. Subgroups were compared using a two-sided log-rank test for PFS and OS, and a χ2 test for ORR and DCR. Twelve-month survival rate was compared using a two-sided test based on log(−log(.)) transformation (26). Analyses were conducted using SAS statistical software version 9.4 (SAS Institute).
Results
Patient characteristics and disposition
Between March 21, 2017 and October 11, 2018, a total of 105 enrolled patients with advanced nonsquamous NSCLC without targetable driver mutations received camrelizumab and 250 mg daily apatinib (Fig. 1). All patients were chemotherapy pretreated but immunotherapy naïve. Because of the similarity in tumor histology and treatment efficacy in the 23 patients from the phase Ib study and the 82 patients from cohort 1 of the phase II study, their data were combined for analysis. Of these patients, the median age was 58 years (range, 35–70), 59.0% were current or former smokers, 62.9% were PD-L1-negative, and 2.9% had a PD-L1 TPS ≥50%. In addition, 21.9% of patients received ≥2 prior systemic regimens, and 25.7% had metastasis involving more than two organs (Table 1).
. | Total (n = 105) . |
---|---|
Age | |
Median (range), y | 58 (35–70) |
Gender, n (%) | |
Male | 79 (75.2) |
Female | 26 (24.8) |
ECOG performance status, n (%) | |
0 | 13 (12.4) |
1 | 92 (87.6) |
Disease stage, n (%) | |
IIIB | 5 (4.8) |
IV | 100 (95.2) |
Tumor histology, n (%) | |
Adenocarcinoma | 99 (94.3) |
Others | 6 (5.7) |
Smoking status, n (%) | |
Current or former smoker | 62 (59.0) |
Never smoked | 43 (41.0) |
No. of prior systemic therapy, n (%) | |
1 | 82 (78.1) |
≥2 | 23 (21.9) |
No. of organs with metastasis, n (%) | |
>2 | 27 (25.7) |
≤2 | 78 (74.3) |
PD-L1 tumor proportion score, n (%) | |
<1% | 66 (62.9) |
≥1% | 25 (23.8) |
≥50% | 3 (2.9) |
Not available | 14 (13.3) |
. | Total (n = 105) . |
---|---|
Age | |
Median (range), y | 58 (35–70) |
Gender, n (%) | |
Male | 79 (75.2) |
Female | 26 (24.8) |
ECOG performance status, n (%) | |
0 | 13 (12.4) |
1 | 92 (87.6) |
Disease stage, n (%) | |
IIIB | 5 (4.8) |
IV | 100 (95.2) |
Tumor histology, n (%) | |
Adenocarcinoma | 99 (94.3) |
Others | 6 (5.7) |
Smoking status, n (%) | |
Current or former smoker | 62 (59.0) |
Never smoked | 43 (41.0) |
No. of prior systemic therapy, n (%) | |
1 | 82 (78.1) |
≥2 | 23 (21.9) |
No. of organs with metastasis, n (%) | |
>2 | 27 (25.7) |
≤2 | 78 (74.3) |
PD-L1 tumor proportion score, n (%) | |
<1% | 66 (62.9) |
≥1% | 25 (23.8) |
≥50% | 3 (2.9) |
Not available | 14 (13.3) |
At the data cutoff of April 10, 2020, the median follow-up was 15.2 months (range, 1.2–33.6). Among all 105 patients, 15 (14.3%) were still on treatment at the time of analysis, including one who was receiving treatment beyond disease progression. The most common reasons for discontinuing treatments were disease progression (radiographic or clinical, n = 65), AEs (n = 11), and patients' withdrawal of consent (n = 7; Fig. 1).
Efficacy
Of the 105 patients in the FAS, 1 (1.0%) patient had confirmed complete response, 28 (26.7%) had confirmed partial response, and 48 (45.7%) had stable disease (including 2 with unconfirmed partial response; Table 2). In the efficacy-evaluable population (n = 94; Supplementary Table S1), 74.5% (70/94) of patients archived shrinkage of the target lesions (Table 2; Fig. 2A) and the confirmed ORR was 30.9% (29/94; 95% CI, 21.7–41.2). Median time to response was 1.9 months (range, 1.7–8.4) and median DoR was 12.6 months [95% CI, 6.5–not reached (NR)] with treatment still ongoing in 10 of 29 responders (34.5%) at the time of the data cutoff (Fig. 2B). Furthermore, 77 patients achieved disease control, with a DCR of 81.9% (95% CI, 72.6–89.1), and 49 had CBR, with a CBR rate of 52.1% (95% CI, 41.6–62.5; Supplementary Fig. S3). An ad hoc analysis including 4 patients with clinical progression or death from cancer as patients with progressive disease showed similar ORR, DCR, and CBR rate (Supplementary Table S2).
. | All patients . | Efficacy-evaluable . |
---|---|---|
. | (n = 105) . | patients (n = 94) . |
Median follow-up (range), months | 15.2 (1.2–33.6) | — |
Best overall response, n (%) | ||
Complete response (CR; confirmed) | 1 (1.0) | 1 (1.1) |
Partial response (PR; confirmed) | 28 (26.7) | 28 (29.8) |
Stable disease (SD) | 48 (45.7) | 48 (51.1) |
Progressive disease | 16 (15.2) | 16 (17.0) |
Not evaluable | 12 (11.4) | 1 (1.1) |
Objective response rate, % (95% CI) | 27.6% (19.3–37.2) | 30.9% (21.7–41.2) |
Disease control rate, % (95% CI) | 73.3% (63.8–81.5) | 81.9% (72.6–89.1) |
Clinical benefit response rate | 49 (46.7%) | 49 (52.1%) |
(CR/PR/SD ≥24 weeks), n (%) (95% CI) | (36.9–56.7) | (41.6–62.5) |
. | All patients . | Efficacy-evaluable . |
---|---|---|
. | (n = 105) . | patients (n = 94) . |
Median follow-up (range), months | 15.2 (1.2–33.6) | — |
Best overall response, n (%) | ||
Complete response (CR; confirmed) | 1 (1.0) | 1 (1.1) |
Partial response (PR; confirmed) | 28 (26.7) | 28 (29.8) |
Stable disease (SD) | 48 (45.7) | 48 (51.1) |
Progressive disease | 16 (15.2) | 16 (17.0) |
Not evaluable | 12 (11.4) | 1 (1.1) |
Objective response rate, % (95% CI) | 27.6% (19.3–37.2) | 30.9% (21.7–41.2) |
Disease control rate, % (95% CI) | 73.3% (63.8–81.5) | 81.9% (72.6–89.1) |
Clinical benefit response rate | 49 (46.7%) | 49 (52.1%) |
(CR/PR/SD ≥24 weeks), n (%) (95% CI) | (36.9–56.7) | (41.6–62.5) |
Efficacy by PD-L1 and TMB status
Of the 105 patients analyzed, 91 (86.7%) had tumor PD-L1 expression data. There was no significant difference in ORR between patients with PD-L1–positive (n = 25) and PD-L1–negative tumors (n = 66; 36.0% vs. 22.7%, P = 0.20; Supplementary Table S3; Fig. 3A). Similarity was observed for DCR (72.0% vs. 71.2%; Supplementary Table S3), median DoR (NR vs. 13.4 months), median PFS (6.8 months vs. 5.1 months, P = 0.29), and median OS (NR vs. 11.4 months, P = 0.10; Fig. 3C and F).
TMB data were available in 46 patients, with a median TMB of 8.21 mutations/Mbp. Patients with high TMB (n = 24) did not show improvement in ORR compared with those with low TMB (n = 22; 29.2% vs. 36.4%, P = 0.60; Fig. 3A). Similarly, no significant difference was observed in PFS or OS (Fig. 3D and G).
Efficacy by STK11/KEAP1 mutation status
In the 46 patients with tissue NGS results, 14 participants harbored mutations in STK11 (n = 4) or KEAP1 (n = 7) alone, or both (n = 3; Supplementary Fig. S4). Compared with double wild-type STK11/KEAP1 participants, the ORR was 42.9% versus 28.1% (P = 0.33; Fig. 3A); DCR was 92.9% versus 65.6% (P = 0.053; Fig. 3B); median PFS was 9.4 versus 5.3 months (P = 0.64) and median OS was 17.9 months versus NR (P = 0.69; Fig. 3E and H). In addition, difference was observed in the 12-month survival rate (85.1% vs. 53.1%, P = 0.010). To further validate this finding, we also included 48 cases who had NGS results only from the peripheral blood; among them, eight cases harbored mutations of either STK11 or KEAP1 or both. We observed a similar trend toward a better ORR (37.5% vs. 27.5%, P = 0.68), DCR (87.5% vs. 65.0%, P = 0.41), PFS (8.3 vs. 5.4 months, P = 0.18), and OS (19.3 vs. 10.7 months, P = 0.49) for patients with STK11 and/or KEAP1 mutation (Supplementary Fig. S5).
Safety
The median duration of drug exposure was 5.5 months (range, 0.5–25). A total of 104 (99.0%) patients experienced at least one treatment-related AE (trAE), and 73 (69.5%) patients experienced at least one trAE of grade 3 or higher (Table 3; Supplementary Table S4). The most common trAEs of grade 3 or higher were hypertension [19 (18.1%)], palmar–plantar erythrodysesthesia syndrome [14 (13.3%)], increased gamma-glutamyltransferase [10 (9.5%)], proteinuria [8 (7.6%)], and abnormal hepatic function [7 (6.7%); Supplementary Table S5]. Reactive cutaneous capillary endothelial proliferation (RCCEP), a common and self-resolving trAE of camrelizumab, was observed in 21.9% of patients, with no grade 3 or higher events reported. All RCCEP occurred on skin, with pathology showing capillary endothelial hyperplasia and capillary hyperplasia in dermis. Twenty-three patients (21.9%) experienced dose reduction of apatinib due to palmar–plantar erythrodysesthesia syndrome, proteinuria, or abnormal hepatic function. Eight patients (7.6%) discontinued treatment because of trAEs (two cases of pneumonitis, one case each of abnormal hepatic function, epistaxis, esophageal fistula, malnutrition, intestinal obstruction, and cerebral infarction). Two deaths (1.9%) were considered related to study treatment by the investigator: one from hemoptysis and one from refractory malignant ascites within 30 days after treatment discontinuation. Immune-related AEs (irAE) of grade 3 or higher were reported in 18 patients (17.1%), with the most frequent being immune-mediated pneumonitis [2 (1.9%)], increased gamma-glutamyltransferase [2 (1.9%)], and abnormal hepatic function [2 (1.9%)]. No unexpected AEs were observed.
. | All patients, n (%) . | |
---|---|---|
Adverse events . | Any grade . | Grade 3–5 . |
Any adverse event | 104 (99.0) | 73 (69.5) |
Proteinuria | 64 (61.0) | 8 (7.6) |
Hypertension | 60 (57.1) | 19 (18.1) |
Aspartate aminotransferase increased | 53 (50.5) | 1 (1.0) |
Palmar–plantar erythrodysesthesia syndrome | 46 (43.8) | 14 (13.3) |
Alanine aminotransferase increased | 42 (40.0) | 1 (1.0) |
Decreased appetite | 37 (35.2) | 0 |
Rash | 36 (34.3) | 1 (1.0) |
Asthenia | 35 (33.3) | 0 |
Blood bilirubin increased | 30 (28.6) | 1 (1.0) |
Anemia | 29 (27.6) | 0 |
Neutrophil count decreased | 28 (26.7) | 2 (1.9) |
White blood cell count decreased | 25 (23.8) | 1 (1.0) |
Diarrhea | 24 (22.9) | 1 (1.0) |
Reactive capillary endothelial proliferation | 23 (21.9) | 0 |
Platelet count decreased | 22 (21.0) | 1 (1.0) |
Mouth ulceration | 21 (20.0) | 1 (1.0) |
Bilirubin conjugated increased | 21 (20.0) | 0 |
Pyrexia | 20 (19.0) | 1 (1.0) |
Blood thyroid-stimulating hormone increased | 19 (18.1) | 0 |
Blood creatinine increased | 18 (17.1) | 0 |
Gamma-glutamyltransferase increased | 17 (16.2) | 10 (9.5) |
Dysphonia | 17 (16.2) | 0 |
Hepatic function abnormal | 15 (14.3) | 7 (6.7) |
Weight decreased | 14 (13.3) | 3 (2.9) |
Hypothyroidism | 14 (13.3) | 1 (1.0) |
Blood alkaline phosphatase increased | 13 (12.4) | 2 (1.9) |
Hematuria | 11 (10.5) | 0 |
. | All patients, n (%) . | |
---|---|---|
Adverse events . | Any grade . | Grade 3–5 . |
Any adverse event | 104 (99.0) | 73 (69.5) |
Proteinuria | 64 (61.0) | 8 (7.6) |
Hypertension | 60 (57.1) | 19 (18.1) |
Aspartate aminotransferase increased | 53 (50.5) | 1 (1.0) |
Palmar–plantar erythrodysesthesia syndrome | 46 (43.8) | 14 (13.3) |
Alanine aminotransferase increased | 42 (40.0) | 1 (1.0) |
Decreased appetite | 37 (35.2) | 0 |
Rash | 36 (34.3) | 1 (1.0) |
Asthenia | 35 (33.3) | 0 |
Blood bilirubin increased | 30 (28.6) | 1 (1.0) |
Anemia | 29 (27.6) | 0 |
Neutrophil count decreased | 28 (26.7) | 2 (1.9) |
White blood cell count decreased | 25 (23.8) | 1 (1.0) |
Diarrhea | 24 (22.9) | 1 (1.0) |
Reactive capillary endothelial proliferation | 23 (21.9) | 0 |
Platelet count decreased | 22 (21.0) | 1 (1.0) |
Mouth ulceration | 21 (20.0) | 1 (1.0) |
Bilirubin conjugated increased | 21 (20.0) | 0 |
Pyrexia | 20 (19.0) | 1 (1.0) |
Blood thyroid-stimulating hormone increased | 19 (18.1) | 0 |
Blood creatinine increased | 18 (17.1) | 0 |
Gamma-glutamyltransferase increased | 17 (16.2) | 10 (9.5) |
Dysphonia | 17 (16.2) | 0 |
Hepatic function abnormal | 15 (14.3) | 7 (6.7) |
Weight decreased | 14 (13.3) | 3 (2.9) |
Hypothyroidism | 14 (13.3) | 1 (1.0) |
Blood alkaline phosphatase increased | 13 (12.4) | 2 (1.9) |
Hematuria | 11 (10.5) | 0 |
Discussion
In this study, we firstly reported the efficacy and safety of combining a VEGF-TKI and an ICI in chemotherapy-pretreated and immunotherapy-naïve patients with advanced/metastatic nonsquamous NSCLC. Our study demonstrated that the combination of apatinib and camrelizumab could provide a promising, durable clinical benefit, even in PD-L1–negative patients. In addition, favorable clinical outcomes were observed for patients with mutations in KEAP1 and STK11, which are reported to confer primary resistance to chemotherapy (27) and immunotherapy (28), respectively.
Several studies have demonstrated that low-dose angiogenesis inhibitors could improve the TIME and potentiate the antitumor effects of ICIs across various types of solid tumors, including lung cancer (15, 23, 29, 30). This phase Ib/II study investigated the efficacy and safety of apatinib 250 mg daily plus camrelizumab in a second- or later-line setting in patients with nonsquamous NSCLC. Despite that only 2.9% patients had a PD-L1 TPS ≥50%, 62.9% had negative PD-L1 expression, 21.9% had received ≥2 prior lines of chemotherapy, 25.7% had metastasis involving more than two organs, 30.9% achieved an objective response, and 81.9% exhibited disease control. The median PFS was 5.7 months and OS was 15.5 months, with a 12-month survival rate of 57.0%. Although cross-trial comparison can be challenging, the efficacy endpoints (DCR and 12-month survival rate) seem to be more favorable in our study than those using ICIs alone in chemotherapy-pretreated and immunotherapy-naïve patients with NSCLC (1–3, 4). Importantly, nearly half of the patients derived clinical benefit response in this study, warranting further investigation of this combination. Consistent with our findings, a higher DCR and prolonged survival were also observed with the combination of ramucirumab and pembrolizumab in the JVDF study which included 27 patients with previously treated NSCLC (15), or lenvatinib plus pembrolizumab in a phase Ib/II study which included 21 patients with metastatic NSCLC (16).
We also evaluated treatment efficacy by tumor PD-L1 expression and TMB status. Both PD-L1–positive and PD-L1–negative subgroups had better treatment response than reported for camrelizumab monotherapy in patients with advanced NSCLC (31). Our results demonstrated the improved efficacy of combining an antiangiogenesis agent with an ICI regardless of PD-L1 status, which was consistent with previous studies of axitinib in combination with pembrolizumab or avelumab in patients with metastatic renal cell carcinoma (12, 13). We also noted a trend of improved efficacy for PD-L1–positive tumor versus PD-L1–negative tumor (ORR: 36.0% vs. 22.7%), though statistical analysis was limited by the relative sample size. Similarly, efficacy of combination therapy was observed across patients with high and low TMB. Consistent with previous studies reporting that TMB was not capable of predicting the efficacy of ICI in combination with chemotherapy (32, 33), no association of TMB with treatment response was observed in the current study, though the small sample size should also be taken into consideration.
Tumor genomic alterations may affect TIME and subsequently influence treatment response. In the recent phase III MYSTIC study, STK11/KEAP1 mutations were found associated with inferior clinical outcomes in patients with NSCLC receiving ICIs. Interestingly, the presence of mutation in either STK11 or KEAP1 was correlated to a better clinical outcome in our study. The better efficacy observed might be due to the addition of apatinib because both STK11 and KEAP1 are involved in angiogenesis (34, 35). Because of their critical roles in regulating reactive oxygen species (36, 37), mutations in either STK11 or KEAP1 could disrupt the redox homeostasis and subsequently promotes tumor angiogenesis (38). Importantly, redox balance is critical for both cancer and T cells, and regulates T-cell immune response in the tumor microenvironment (39, 40).
No new safety signals were identified with the combination of apatinib and camrelizumab compared with monotherapy (27). The most frequent grade 3 or greater AEs were hypertension and increased gamma-glutamyltransferase, which are consistent with the classic effects of antiangiogenesis inhibitors and similar to those reported in renal cancers. We found no increase in the incidence of irAEs compared with that observed in the phase II trial of camrelizumab monotherapy in NSCLC, indicating that the addition of apatinib to camrelizumab did not increase irAEs. RCCEP was considered the most common AE associated with camrelizumab monotherapy, with incidences of 67%–97% reported previously (41–45). Our study showed that the incidence of RCCEP decreased to 21.9% when combined with apatinib.
We have to point out that this study was an open-label single-arm trial with no ICI monotherapy or chemotherapy control arm. Therefore, we cannot definitively claim that the combination of apatinib and camrelizumab is superior to either single-agent therapy. Our findings will need to be validated in larger-scale randomized trials preferably including non-Asian population as well. In addition, the proportion of PD-L1–positive patients was relatively low in this study, due to competitive patient recruitment to other trials that required positive PD-L1 expression. Nevertheless, encouraging efficacy was observed regardless of PD-L1 expression.
In summary, this phase Ib/II study has demonstrated that the combination of apatinib and caremlizumab is well tolerated, and has therapeutic value in pretreated patients with nonsquamous NSCLC irrespective of PD-L1 and TMB status. Furthermore, our preliminary biomarker analysis suggested that mutations in STK11 and KEAP1, which are speculated to confer resistance to (chemo-)immunotherapy as reported previously (27, 28), might dictate stronger benefit from such combination. All these potential values will be fully validated in our ongoing randomized phase III clinical trial using this combination in the first-line setting (NCT04203485).
Authors' Disclosures
Q. Wang reports other from Jiangsu Hengrui during the conduct of the study. X. Yang reports personal fees from Jiangsu Hengrui Medicine Co., Ltd during the conduct of the study and personal fees from Jiangsu Hengrui Medicine Co., Ltd outside the submitted work. No disclosures were reported by the other authors.
Authors' Contributions
C. Zhou: Conceptualization, supervision, funding acquisition, writing-review and editing. Y. Wang: Data curation, writing-original draft. J. Zhao: Data curation. G. Chen: Data curation. Z. Liu: Data curation. K. Gu: Data curation. M. Huang: Data curation. J. He: Data curation. J. Chen: Data curation. Z. Ma: Data curation. J. Feng: Data curation. J. Shi: Data curation. X. Yu: Data curation. Y. Cheng: Data curation. Y. Yao: Data curation. Y. Chen: Data curation. R. Guo: Data curation. X. Lin: Data curation. Z. Wang: Data curation. G. Gao: Data curation. Q. Wang: Conceptualization, formal analysis, supervision, writing-review and editing. W. Li: Formal analysis, writing-original draft. X. Yang: Formal analysis. L. Wu: Formal analysis, writing-original draft. J. Zhang: Conceptualization, supervision, writing-review and editing. S. Ren: Conceptualization, supervision, funding acquisition, writing-original draft.
Acknowledgments
The authors thank the patients and their families for their participation in this study, as well as the study teams at each of the study sites. This work was supported by Jiangsu Hengrui Medicine Co., Ltd, the Chinese National Natural Science Foundation Project (grant number 81772467; awarded to S. Ren), the Science and Technology Program of Shanghai (grant number 2017ZZ02012 and 19411950300; awarded to C. Zhou), and Shanghai Innovative Collaboration Project (grant number 2020CXJQ02; awarded to C. Zhou).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.