PDF
Citation
Print
Introduction
Gastric cancer is the fifth most common cancer and the fourth leading cause of cancer mortality worldwide [1]. Systemic chemotherapy is the primary treatment for most patients with metastatic or recurrent advanced gastric cancer (AGC) [2,3]. Human epidermal growth factor receptor 2 (HER2) is part of a family of receptors associated with tumor cell proliferation, apoptosis, adhesion, migration, and differentiation [4]. Trastuzumab, an anti-HER2–targeted agent, has immensely improved the survival outcomes. Moreover, a combination of trastuzumab and a doublet of fluoropyrimidine and platinum has become the standard of care in patients with metastatic or unresectable HER2-positive AGC [2,5]. However, most patients eventually develop resistance to trastuzumab-based chemotherapy, and further therapeutic strategies are required [6,7].
Because the HER family members synergistically interact with the same or other HER subfamily members, HER2 activation results in increased epidermal growth factor receptor (EGFR or HER1) signaling, suggesting that the dual inhibition of HER2 and EGFR could overcome resistance to trastuzumab [8]. Therefore, pan-HER inhibitors are considered more potent than those inhibiting only one or two HER subfamily receptors. Several pan-HER inhibitors, such as lapatinib [9,10], dacomitinib [11], and afatinib [12] have been investigated for their antitumor activity in AGC patients; however, their efficacy has not been promising.
Varlitinib is a reversible small molecule pan-HER inhibitor targeting EGFR, HER2, and HER4 [13]. In addition, varlitinib effectively inhibits heterodimers with HER3, which lacks a kinase domain [14]. Preclinically, varlitinib inhibited multiple proliferation and anti-apoptosis pathways, including AKT, phosphoinositide 3-kinase, and survivin pathways, in a gastric cancer xenograft model [15,16]. Furthermore, early clinical studies suggested varlitinib as a potent inhibitor of signal transduction in EGFR and HER2 co-expressing AGC, with promising antitumor activity in combination with paclitaxel with/without carboplatin for several solid tumors [17-19]. Based on the previous data, we explored the clinical efficacy of a combination of varlitinib and paclitaxel in patients with EGFR/HER2 co-expressing AGC as a second-line treatment.
We determined the maximum tolerated dose (MTD) and recommended phase II dose (RP2D) of varlitinib when combined with paclitaxel (phase Ib part), and assessed the antitumor efficacy and safety of this combination in previously treated patients with EGFR/HER2 co-expressing AGC (phase II part).
Materials and Methods
1. Patients
This prospective phase Ib/II study was conducted at six centers in the Republic of Korea. Patients were included if patients aged ≥ 19 years; histopathologically confirmed unresectable, recurrent, or metastatic gastric adenocarcinoma (including adenocarcinoma of the esophagogastric junction); both EGFR and HER2 immunohistochemistry (IHC) expression ≥ 1+, or EGFR amplification/mutation on next-generation sequencing (NGS) and HER2 IHC ≥ 1+; the presence of one or more evaluable or measurable lesions by Response Evaluation Criteria in Solid Tumors (RECIST, ver. 1.1); after failure of the first-line chemotherapy, including fluoropyrimidine and/or platinum-based chemotherapy, irrespective of trastuzumab exposure; and adequate bone marrow and hepatic functions with good performance (Eastern Cooperative Oncology Group Performance Status [ECOG PS], 0-1). The key exclusion criteria were a history of other malignancies, interstitial lung disease, uncontrolled cardiac dysfunction, previous exposure to taxanes, and symptomatic central nervous system metastasis.
2. Study design
This study comprised two phases: phase Ib and phase II. Phase Ib was designed to determine the safety, tolerability, MTD, and RP2D of varlitinib administered daily in combination with weekly paclitaxel (80 mg/m2; days 1, 8, and 15) during the first cycle, based on the traditional 3+3 design, with a starting dose (dose level 1) and a reduced dose (dose level –1). Dose level 1 was varlitinib 300 mg twice daily with weekly paclitaxel; dose level –1 was varlitinib 300 mg bid for 4 days per week (4 days on, 3 days off, intermittent dose). Administration of study drugs at dose level 1 was planned for three patients until the first dose-limiting toxicity (DLT) was observed for 4 weeks. If DLT was not observed, phase II was conducted at dose level 1. If one or more DLTs were observed in three patients, the cohort was expanded to six. MTD was defined as the highest dose at which fewer than two of the six patients experienced DLT. If two or more DLTs were observed in six patients at dose level 1, the cohort was de-escalated to a lower level (dose level –1). Phase II study evaluated the safety and tolerability of varlitinib and paclitaxel combination therapy in RP2D and investigated its antitumor activity. Oral varlitinib was administered based on the RP2D, and paclitaxel was administered weekly at 80 mg/m2 via intravenous infusion for 3 weeks every 4 weeks. It was discontinued in the event of disease progression, toxicity above an acceptable level, patient refusal, withdrawal of consent, or death.
3. Safety analyses and efficacy
Adverse events (AEs) were evaluated using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE, ver. 4.03). DLTs were defined as grade ≥ 3 non-hematologic toxicities (except for alopecia); grade ≥ 3 diarrhea, nausea, and vomiting, despite maximal dosage of anti-diarrhea and/or anti-emetic medication (as applicable); and grade 4 neutropenia sustained for ≥ 7 days, grade 3-4 neutropenia with fever or infection, grade 4 thrombocytopenia and grade 3 thrombocytopenia sustained for ≥ 7 days, or with accompanying bleeding, or requiring transfusion for 28 days. All patients who received at least one dose of varlitinib were included in the safety analysis. Tumor response was evaluated by conducting chest and abdomen/pelvis computed tomography or magnetic resonance imaging every 6 weeks from the start of the study to the 24th week, after which the frequency was reduced to once every 12 weeks until disease progression using the RECIST ver. 1.1, according to the investigator’s assessment.
4. Circulating tumor DNA assay for monitoring response to therapy
The circulating tumor DNA (ctDNA) was extracted from the patient’s plasma (10 mL) at two time points (before treatment and at the first response evaluation or end of the treatment). The DNA NGS library and solution-based target enrichment were performed at IMBdx, Inc. (Seoul, Korea), using AlphaLiquid100 as previously described [20]. The targeted gene panel of 118 cancer-related genes was used to detect the four classes of variations (single-nucleotide variants, indels, fusions, and copy number variations) in 118 cancer-related genes. The captured DNA libraries were sequenced using the Illumina NovaSeq 6000 platform (Illumina, San Diego, CA) in the 2×150 bp paired-end mode.
5. Statistical analysis
The primary objective of phase Ib was to determine the MTD of varlitinib in combination with paclitaxel. The primary objective of the phase II trial was to study progression-free survival (PFS) with varlitinib combined with paclitaxel. PFS was defined as the period from enrollment to the day on which (1) radiological or clinical progression was evident, (2) subsequent treatment was indicated, or (3) the patient died. Secondary objectives were to assess the safety, tolerability, objective response rate (ORR), disease control rate (DCR), and overall survival (OS). For phase II, the sample size was calculated according to statistical assumptions and checked using the null hypothesis (PFS of 2.9 months in the single paclitaxel group) and alternative hypothesis (PFS of 4.4 months in the paclitaxel and ramucirumab combination group) based on the RAINBOW study [21]. Therefore, the sample size was determined to be 50 subjects with a two-sided significance level of 10% at a power of 70%. Considering a dropout rate of 10%, the total number of subjects in phases Ib/II was 55.
Results
1. Patients’ characteristics
From January 2020 to December 2021, 31 patients were enrolled from six centers in the Republic of Korea. Although 22 patients were enrolled during the phase II study, this study was terminated early owing to the supply constraints of varlitinib following the sponsor’s decision to stop further development. Baseline characteristics of the study population are summarized in Table 1. The median age was 64 years (range, 34 to 83 years), and 18 patients (67%) were males. The first-line treatment regimens consisted of the following: fluoropyrimidine and/or platinum-based chemotherapy (17 patients), fluoropyrimidine/platinum with trastuzumab (6 patients), and fluoropyrimidine/platinum with pembrolizumab (4 patients). There were no statistical differences in best response and PFS according to the first-line treatment regimens (S1 Table).
2. Determination of RP2D
Although nine patients were enrolled at the dose 1 level of varlitinib 300 mg twice daily in the phase Ib study, three patients were excluded because of protocol violations (withdrawal, emergency surgery, and non-adherence to the study drug). Among the six patients, one of the first three experienced DLT of grade 4 aspartate aminotransferase/alanine aminotransferase elevation and hyperbilirubinemia in the first cycle. Three additional patients were enrolled, and there were no DLT at a dose level of 1. Therefore, RP2D was decided to be varlitinib 300 mg twice daily combined with weekly paclitaxel (80 mg/m2 on days 1, 8, and 15) every 4 weeks.
3. Efficacy and survival outcomes
Among the 27 patients, including 22 patients treated with RP2D and five patients from the phase Ib study (one patient who experienced DLT grade 4 was not evaluated) for efficacy evaluation, the median follow-up duration was 15.7 months (95% confidence interval [CI], 11.3 to 20.1). The median PFS and OS were 3.3 months (95% CI, 1.7 to 4.9) and 7.9 months (95% CI, 5.0 to 10.8), respectively (Fig. 1A). Among the 16 patients with measurable disease who were administered RP2D, tumor shrinkage was observed in 11 patients (69%); therefore, confirmed objective response and disease control were observed in five patients (ORR, 31%; 95% CI, 9 to 54) and 14 patients (DCR, 88%; 95% CI, 71 to 100), respectively (Fig. 2).
Based on EGFR and HER2 expression, the patients were divided into three subgroups as follows: (1) low expression (EGFR 1+ and HER2 1+), (2) strong EGFR expression (EGFR 3+ and any HER2; EGFR 2+ and HER2 2+), and (3) strong HER2 expression (HER2 3+ and any EGFR; HER2 2+ and EGFR 1+) (S2 Table). Patients with strong HER2 expression (n=8) had higher ORR and longer OS than those in the other groups (Fig. 1B). In contrast, patients with strong EGFR (n=3) had poor outcomes for ORR, PFS, and OS. The low-expression group (n=16) comprised the largest number of patients and showed similar efficacy to the strong HER2 group (Table 2, Fig. 2C and D). Therefore, we could not establish a relationship between EGFR and HER2 expression and treatment efficacy.
4. Toxicity and safety
The safety analysis set consisted of six patients in phase Ib, as described in Table 3. Grade 3 or higher neutropenia was observed in three patients (50.0%). One of the first three patients experiencing grade 4 showed increased aspartate aminotransferase (AST)/alanine transaminase (ALT)/bilirubin at dose level 1 and recovered after 3 weeks. Among the 27 patients, the most common AEs were neutropenia (51.8%; 22.2% of grade 3/4), AST/ALT elevation (22.2%; 11.1% of grade 3/4), diarrhea (27.3%), and nausea (18.5%). No treatment-related deaths or unexpected AEs resulting from treatment cessation were observed in patients with RP2D.
5. Genomic alteration landscape in ctDNA analysis
A total of 25 samples from 15 patients were analyzed, and 26 mutations and seven gene amplifications were detected (Fig. 3). Among the 15 patients included in the baseline analysis, genetic alterations were detected in 11 patients (73.3%), with a median of five mutations per patient (range, 1 to 11). The most common baseline somatic alteration was TP53 (n=7, 63.6%), followed by ERBB2 amplification (n=4, 36.4%). Among the four patients with HER2 3+ expression and genetic alterations detected in ctDNA, three (75.0%) had ERBB2 amplification. However, there was no EGFR amplification in the two patients with strong EGFR expression, and genetic alterations were detected in the ctDNA. No differences were observed in the survival outcomes according to the detection of genetic alterations at baseline in terms of PFS (log-rank test, p=0.581) (Fig. 1C) and OS (p=0.391). However, six patients whose genetic alteration in the ctDNA were decreased or cleared after chemotherapy among 10 patients who were checked for follow-up ctDNA analysis had a trend toward better survival in terms of PFS (median, 5.5 months [95% CI, 1.5 to 2.6] vs. 1.4 months [95% CI, 1.2 to 1.5]; p=0.132) and OS (median, 7.9 months [95% CI, 2.6 to 13.2] vs. 5.8 months [1.3 to 10.4]; p=0.449) compared to that of the other nine patients.
Discussion
In this prospective, multicenter, open-label phase Ib/II study, a combination of 300 mg of varlitinib twice daily and paclitaxel displayed manageable toxicity and moderate efficacy in previously treated patients with EGFR/HER2 co-expressing AGC.
In a previous phase I study [18], the RP2D of varlitinib was determined at a dose of 300 mg twice daily (4 days on, 3 days off). The varlitinib study was performed mostly in patients with breast cancer treated with paclitaxel with or without trastuzumab. Although we selected 300 mg twice every day with weekly paclitaxel as RP2D, treatment-related AEs were slightly decreased, such as grade 3-4 neutropenia (20% vs. 40%), anemia (4% vs. 5%), and febrile neutropenia (7% vs. 11%), compared to the previous phase I study. Grade 3-4 non-hematologic toxicities, including diarrhea (0% vs. 5%), asthenia (0% vs. 11%), and mucositis (0% vs. 5%), decreased, except for hyperbilirubinemia/elevated AST (11% vs. 5%). However, these findings should be interpreted with caution not only due to differences in varlitinib dose but also to differences in patient groups and duration of treatment (median, 3.0 months; range, 0.1 to 17.1) in the phase Ib study [18].
The ORR, median PFS, and OS were 31%, 3.3, and 7.6 months, respectively. Compared to the paclitaxel alone group (ORR, 16%; PFS, 2.9 months; and OS, 7.4 months) in the RAINBOW trial [21], these results suggest that varlitinib as a pan-HER inhibitor, combined with chemotherapy, could be more potent for the second-line setting than chemotherapy alone. However, these results were based on a limited number of patients and additional studies are required to evaluate the efficacy of this combination therapy.
The ctDNA analysis performed in a subgroup of AGC patients whose tumors co-expressed EGFR and HER2 revealed 26 gene mutations and seven gene amplifications. An association was also observed between HER2 expression and ERBB2 amplification. Furthermore, a few patients whose ctDNA mutations decreased or were cleared after treatment reported a longer PFS. In addition, our findings suggested that gastric cancer has a high level of genetic diversity, and ctDNA dynamics using longitudinally obtained blood samples could predict response or resistance during treatment [22,23].
Because pretreatment biopsy was not mandatory, patients were enrolled based on EGFR and HER2 expression at initial diagnosis. Furthermore, it remains unclear whether IHC expression is an appropriate target biomarker for varlitinib. In addition, statistically planned patients could not be enrolled because of limited drug support.
In conclusion, a combination of varlitinib and paclitaxel displayed manageable toxicity and antitumor activity in patients with EGFR/HER2 co-expressing AGC who progressed after first-line chemotherapy. Further clinical studies are required to confirm the efficacy of dual inhibitors of EGFR/HER2 combined with paclitaxel in patients with EGFR/HER2-positive gastric cancer with a large sample size.
XML Download