Real‐world clinical outcomes of the combination of anti‐PD‐1 antibody, trastuzumab, and chemotherapy for HER2‐positive gastric/gastroesophageal junction cancer

Abstract Background Previous clinical trials indicated the addition of anti‐PD‐1 antibody remarkably improved the efficacy of trastuzumab and chemotherapy in patients with HER2‐positive gastric/gastroesophageal junction (GEJ) cancer. However, no real‐world experiences have been reported yet. Methods We retrospectively analyzed 1212 patients with gastric/GEJ cancer treated at Nanjing Drum Tower Hospital between 2019 and 2022. Among 138 patients with HER2‐positive gastric/GEJ cancer, 47 patients receiving at least two doses of the combination regimen with anti‐PD‐1 antibody, trastuzumab, and chemotherapy were recruited in the study population, and 38 out of 47 patients with measurable disease were included in the efficacy population. Progression‐free survival (PFS), objective response rate (ORR), disease control rate (DCR), and toxicity profiles were reported. Results In the study population, 37 (78.7%) received the study therapy as a first‐line treatment. In the efficacy population, the ORR and DCR were 76.3% and 94.7%, respectively. The overall median PFS was 9.1 months (95% confidence interval [CI] 6.3–11.9 months). For the first‐line treatment, the mPFS was 10 months, and 7 months for the second‐line. Among 14 patients who failed the study treatment, three (21.4%) developed brain metastasis as the first failure site. No significant association was found between PFS and the expression of PD‐L1. 22.2% of patients developed grade 3 treatment‐related adverse events (TRAEs). No treatment‐related grade ≥4 adverse events or deaths occurred. Conclusion This real‐world study validated the combination regimen's high efficacy and good tolerance in patients with HER2‐positive gastric/GEJ cancer. An increased incidence of brain metastasis was observed in patients who failed this regimen.


| INTRODUCTION
Gastric cancer (GC) is the fifth most commonly diagnosed cancer and the fourth leading cause of cancer-related death worldwide and ranks second for incidence and third for mortality in China. 1 Currently, the overall 5-year survival rate is only 35.1% in China. 2 Over 80% of GC patients are diagnosed at an advanced stage in China 3 and with a median overall survival (OS) of 10-14 months. 4 HER2 is overexpressed or amplified in about 20% of patients with GC. 5 In the Chinese population, the positive rate is 10.8%-12.5%. 6,7 HER2 overexpression/amplification is solidly associated with a more aggressive disease in patients with GC. 8 HER2 positivity defined as HER2 immunohistochemistry (IHC) scores of 3+ or 2+ with fluorescent in situ hybridization-positive (FISH+). The phase III ToGA trial established the addition of trastuzumab (a humanized monoclonal antibody that targets HER2) to chemotherapy in advanced HER2-positive GC based on the improved median OS (16.0 vs. 11.8 months). 9 With the development of immunotherapy in recent years, the addition of checkpoint inhibitors to HER2targeted therapy significantly improved outcomes in patients with advanced HER2-positive GC. Two single-arm, phase II studies (NCT02954536, NCT02901301) have demonstrated the clinical efficacy and safety for the combination of anti-PD1 antibody, trastuzumab, and chemotherapy in metastatic HER2-positive GC patients. [10][11][12] To further investigate the combination, Janjigian et al. conducted a randomized, double-blind, placebo-controlled phase III KEYNOTE-811 trial. In the protocol-specified first interim analysis, the objective response rate (ORR) was significantly improved in the pembrolizumab group compared with the placebo group (74.4% vs. 51.9%) (NCT03615326). 13 Though impressive results were achieved in metastatic HER2-positive GC patients treated by the combination of pembrolizumab with chemotherapy/trastuzumab, to our knowledge, no real-world experiences have been reported yet. To address this gap, we retrospectively analyzed the metastatic HER2-positive GC patients treated by the combination of checkpoint inhibitor, chemotherapy, and trastuzumab between 2019 and 2022 in our institution in a real-world setting.

| Patient eligibility
This study was performed under the guidelines of the Ethics Committee of Nanjing Drum Tower Hospital, which is an academic institution affiliated with Nanjing University. Patients treated in our hospital will be informed of the potential research use of personal information. Only patients who agreed with and signed the consent were included in our study. The eligibility criteria included: pathologically confirmed gastric/gastroesophageal junction (GEJ) cancer; at an advanced stage; HER2 IHC score 3+ or 2+ with FISH+; treated by the combination of immune checkpoint inhibitor, trastuzumab, and chemotherapy between 2019 and 2022. The exclusion criteria included: HER2-negative; patients with HER2positive GC receiving adjuvant chemotherapy or neoadjuvant therapy.

| Treatment
Trastuzumab was intravenously given (i.v.) with a loading dose of 8 mg/kg on Day 1 and a subsequent maintenance dose of 6 mg/kg every 3 weeks. Patients intravenously received a flat dose of pembrolizumab 200 mg or sintilimab 200 mg every 3 weeks. The chemotherapy regimen included platinum plus fluoropyrimidine. The first-line chemotherapy was given as follows: intravenous oxaliplatin 130 mg/m 2 on Day 1 and oral S-1 40-60 mg twice daily, Days 1-14, every 3 weeks. If S-1 was not tolerated, oral capecitabine 2-3 g was given twice daily, Days 1-14, every 3 weeks. The second-line chemotherapy regimen included intravenous nab-paclitaxel 100-120 mg/m 2 on Days 1, 8 every 3 weeks, or intravenous docetaxel 75 mg/ m 2 on Day 1 every 3 weeks, or irinotecan 150 mg/m 2 on Day 1 every 2 weeks. Dose modifications were indicated if untolerated toxicities occurred. The treatment was given until patients' refusal, disease progression, or unacceptable toxicities.

| Molecular biomarkers
We retrospectively collected the information of molecular biomarkers, including HER2 status, mismatch repair (MMR) status, expressions of PD-L1, and Epstein-Barr virus (EBV), from the department of pathology. HER2 IHC scores 3+ or 2+ with FISH-positive were defined as HER2positive. MMR status was estimated by using an MMR IHC panel to detect expressions of the MMR proteins including MLH1, PMS2, MSH2, and MSH6. Expressions of the four proteins indicate an MMR-proficient tumor, and loss of expressions suggests an MMR-deficient (dMMR) tumor. 14,15 The PD-L1 expression was detected by VENTANA PD-L1 (SP142) assay. The combined positive score (CPS) was defined as the number of PD-L1 positive cells including tumor, lymphocytes, and macrophages divided by total tumor cells. The FISH assay was used to determine the EBV status.

| Response evaluation and follow-up
Treatment response was assessed every 6-8 weeks with magnetic resonance imaging (MRI) or computed tomography according to response evaluation criteria in solid tumors (RECIST version 1.1). 16 In this study, we investigated ORR, disease control rate (DCR), median progression-free survival (PFS), and toxicities. The ORR is the proportion of patients who have a complete response (CR) or partial response (PR) to therapy, and DCR is ORR plus the proportion of patients with stable disease. Adverse events were evaluated during follow-up and for 30 days after the last treatment. All treatment-related toxicities were assessed according to the Common Toxicity Criteria for Adverse Events version 4.03.

| Statistical analyses
We used the SPSS 20.0 version to analyze the data. Associations between patients' and treatment-related factors and PFS were estimated with the Kaplan-Meier analysis and a Cox proportional hazards model. p-values ≤0.05 were considered statistically significant.

| Recruitment process of our study
The recruitment process of our study is shown in Figure 1. In total, 1212 patients were treated for gastric/GEJ cancer in our department from January 2019 to January 2022.
There were 138 HER2-positive patients (138/1212, 11.4%). Among HER2-positive patients, 50 patients received a combination regimen with anti-PD-1 antibody, trastuzumab, and chemotherapy for advanced or metastatic disease. We excluded three patients who received only one dose of the study therapy and refused the subsequent treatment. Finally, 47 patients were recruited for analyses of treatment efficacy and adverse events.

| Patients' characteristics
As shown in Table 1, the median age of the study population was 66 years (range 28-85 years), and most patients (80.9%) were male. In 42.6% of the recruited patients, the primary tumor location was at the GEJ. 78.7% of the patients received the combination regimen with anti-PD-1 antibody, trastuzumab, and chemotherapy as a first-line treatment. Most patients (74.5%) had ≥2 metastatic sites. And the most common metastatic sites were lymph nodes (80.9%), liver (48.9%), bone (19.1%), and lung (17.0%). No patients were with dMMR status. One patient (2.1%) was EBV-positive. In HER2-positive patients, 76.6% of patients were with HER2 IHC 3+, and 23.4% of patients were HER2 IHC 2+ and FISH-positive. Twenty-five patients (53.2%) had tumors with CPS ≥1.

| Treatment efficacy and clinical outcomes in the efficacy population
In the study population, the treatment responses of six patients have not been assessed until the last follow-up. Two F I G U R E 1 Flowchart of the recruitment process. FISH, fluorescence in situ hybridization; GC, gastric cancer; GEJ, gastroesophageal junction; HER2, human epidermal growth factor receptor 2; IHC, immunohistochemistry; PD-1, programmed cell death 1.
patients were given the anti-PD-1 antibody after the treatment failure of trastuzumab and were not included in the efficacy population. But the addition of immunotherapy did not control the disease progression in these two patients. One patient had a non-measurable lesion (ascites) by RECIST 1.1 criteria and was excluded in the efficacy population either. Eventually, 38 patients with measurable disease according to RECIST 1.1 criteria were included in the efficacy population.
In the efficacy population, there were 10 patients receiving the study therapy as a second-line treatment ( Table 2). These 10 patients developed disease progression during or within 6 months after the completion of adjuvant chemotherapy. Therefore, the study therapy was considered a second-line treatment for these 10 patients. For all patients in the efficacy population, six patients (15.8%) achieved CR, and 23 patients (60.5%) achieved PR. The overall ORR was 76.3% (29/38), and the overall DCR was 94.7% (36/38). Best treatment responses are shown in Figure 2. The ORR for the first-line and second-line treatments was 82.1% and 60%, respectively. Until the latest follow-up, 14 patients failed the study treatment. Eight patients were in the first-line treatment group and six patients were in the second-line treatment group. The most common first failure sites were the lymph nodes (5/14,  35.7%) and the liver (3/14, 21.4%). Interestingly, we found three patients developed brain metastasis as the first failure site (21.4%, 3/14) ( Table 2).
In the efficacy population, the overall mPFS was 9.1 months (95% confidence interval [CI]; 6.3-11.9 months) ( Figure 3A). For the first-line treatment, the mPFS was 10 and 7 months for the second line. The difference was not statistically significant (hazard ratio [HR] 2.126, 95% CI 0.693-6.524, p = 0.187) ( Figure 3B). Next, we analyzed the associations between patients' and treatment-related factors and PFS. As indicated in Figure 4A

| Treatment-related adverse events
In general, the study therapy was well-tolerated. There were neither grade 4 treatment-related adverse events (TRAEs) nor treatment-related deaths in our study. As shown in Table 3, the most common TRAE was grade 1 hypoalbuminemia (57.8%). 22.2% of patients developed grade 3 TRAEs. The most frequent grade 3 TRAEs were hypokalemia (6.7%) and thrombocytopenia (4.4%).

| DISCUSSION
Since the success of the phase III trial ToGA in 2010, the combination of trastuzumab and chemotherapy has been the standard therapy for advanced HER2-positive GC. 9 The previously reported ORR and mPFS were 41%-64% and 6.7-9.0 months, respectively. [17][18][19] Our study demonstrated the high ORR (76.3%), DCR, and manageable safety for the combination of anti-PD-1 antibody, trastuzumab, and chemotherapy in patients with HER2-positive gastric/GEJ cancer. To our knowledge, this is the first study to verify the role of the foresaid treatment regimen in gastric/GEJ cancer patients with HER2-positive in a real-world setting.
As described in the Section 3, 10 patients received the study therapy as a second-line treatment. They developed disease progression during or within 6 months after adjuvant chemotherapy. Currently, there is no valid evidence to support the addition of trastuzumab to adjuvant chemotherapy in patients with resectable HER2-positive GC after D2 resection. Therefore, trastuzumab is not applied to such patients in our institution. Though the study therapy was administered as a second-line treatment, these 10 patients were immunotherapy and trastuzumab-naïve before the study therapy. We observed a lower ORR and DCR in the second-line treatment group, indicating the rationale for the early use of immunotherapy and trastuzumab in patients with HER2-positive advanced GC. Data from a phase II study (T-ACT) suggested that continuous use of trastuzumab beyond progression failed to improve PFS in patients with HER2-positive GC. 20 Our data indirectly supported this point. As mentioned in Section 3, the addition of anti-PD-1 antibody and continuous use of trastuzumab did not control disease progression after treatment failure to trastuzumab in two patients that were excluded from the efficacy population. The high efficacy of the above study therapy is also biologically reasonable. Trastuzumab can increase the HER2 internalization and cross-presentation by dendritic cells to promote HER2-specific T-cell responses. 21,22 Trastuzumab can also upregulate expressions of PD-1/PD-L1 and modulate the expression of major histocompatibility complex class II (MHC II). 23 The anti-PD-1 antibody can combine with trastuzumab to prompt HER2-specific T-cell response and lead to the expansion of peripheral memory T cells. 24,25 However, a higher clinical efficacy is still challenged by some resistance mechanisms, such as heterogeneous expressions of HER2 26 and heterogeneous distribution of trastuzumab, 27 poor penetration of trastuzumab through blood-brain barriers, 28 and less immunological activity in metastatic lesions in GC patients, 29 and the immune privilege of central nervous system under normal conditions. 30 Therefore, we wondered whether a high incidence of brain metastasis could occur in patients under the study therapy.
Patients with HER2-positive GC tended to relapse in distant sites other than peritoneal or local recurrence. 31,32 Brain metastasis from GC is uncommon, and the incidence rate is 0.47%-0.7%. 33,34 In patients with breast cancer, there is a positive correlation between HER2 overexpression and a higher risk of brain metastasis. 35 Previous studies showed that brain metastases were higher and relapsed more frequently in HER2-positive GC patients. 36,37 The brain metastasis was diagnosed in three patients by brain MRI or cerebrospinal fluid examination after the appearance of nervous system disorders. We observed a relatively high incidence of brain metastasis (21.4%, 3/14) in patients who failed the study therapy. F I G U R E 2 Best treatment responses. Only patients who had RECISTmeasurable disease at baseline and at least one post-baseline evaluation were included (N = 36). CR, complete response; PD, progressive disease; PR, partial response; RECIST, response evaluation criteria in solid tumors; SD, stable disease.

F I G U R E 3
Kaplan-Meier analyses of progression-free survival (A) and according to treatment line (B). CI, confidence interval; HR, hazard ratio; mPFS, median progression-free survival.
Oncologists should carefully differentiate between brain metastasis and chemotherapy-induced refractory nausea and vomiting. Once nervous system disorders appear, intensified follow-up and brain MRI should be considered to ensure an early diagnosis of brain metastasis.
Though impressive responses were achieved in some patients, there were still some patients not sensitive to the study therapy or developing resistance after response, indicating the need to identify predictive biomarkers for the patients' outcomes. Both the Panthera trial and another phase II trial tried to find some biomarkers for the study therapy. These two clinical trials indicated that there was no obvious correlation between PFS and the PD-L1 expression. 10,38 And similar results were also found in our study. Given the high efficacy of the study therapy, we infer that the use of immunotherapy should be considered in HER2-positive GC patients regardless of the PD-L1 expression level.
In our institution, the HER2-positive rate was 11.4% in patients with GC, which seemed less than the data reported by the Phase III trial ToGA 9 but was similar to other Chinese patients based studies. 6,7 We thought the actual HER2-positive rate would be slightly higher than 11.4% because 72 patients in our study were with HER2 IHC 2+ but were absent from the FISH test. Most of them only received adjuvant chemotherapy in our institution and chose to skip the FISH test, which was not covered by insurance.
Our study had some limitations. First, although we retrospectively collected over 1000 gastric/GEJ cancer patients in our institution during the past 3 years, only about 50 patients were recruited for analyses, which was due to the lower HER2-positive rate in all gastric/GEJ cancer patients. Second, similar to other retrospective studies, it was hard to collect all the adverse events information due to the incomplete records during treatment. In addition, this was a single-arm study without a control cohort. NGS was not performed to validate the potential biomarkers. Potential mechanisms were not included in our study. We are currently investigating the effects of the study therapy on reprogramming the tumor immune microenvironment, and related data will be released shortly.
In summary, our study manifested the high effectiveness of the combination regimen with anti-PD-1 antibody, trastuzumab, and chemotherapy in patients with HER2positive GC in a real-world setting for the first time. In patients who failed this regimen, an increased incidence of brain metastasis was observed.