Abstract
Background
Real‐world safety and effectiveness data for trastuzumab plus chemotherapy treatment of patients with HER2‐positive metastatic gastric cancer (mGC) in China are lacking.
Patients and Methods
EVIDENCE was a prospective, multicenter, noninterventional registry study evaluating the safety and effectiveness of trastuzumab in five cohorts of Chinese patients with gastric cancer, stratified by HER2 status and trastuzumab treatment. Effectiveness was analyzed for cohorts I (HER2‐positive, trastuzumab treated), II (HER2‐positive, trastuzumab untreated), and IV (HER2‐negative, trastuzumab untreated); trastuzumab‐related adverse events (AEs) were analyzed for cohort I.
Results
Cohorts I, II, and IV included 174, 113, and 422 patients, respectively. Most patients received first‐line chemotherapy (87.6%). Median overall survival (OS1) for first‐line treatment was 22.3, 17.2, and 17.4 months in cohorts I, II, and IV, respectively. After excluding patients who had surgery, respective median OS1 was 19.9, 15.3, and 12.9 months. Respective first‐line progression‐free survival (PFS1) was 8.2, 6.9, and 6.2 months; and respective first‐line response rates (RR) were 51.7%, 18.4%, and 32.8%. Cohort I was significantly favored over cohort II for propensity score–matched first‐line median OS1 (hazard ratio [HR], 0.61), PFS1 (HR, 0.64), and RR (odds ratio, 4.93). Trastuzumab‐related AEs, grade 3–5 AEs, serious AEs, and AEs with a fatal outcome occurred in 23.6%, 3.4%, 2.3%, and 0.6% of cohort I patients, respectively.
Conclusion
Safety profiles were consistent with those known for trastuzumab and chemotherapy; trastuzumab treatment improved outcomes. Our study provides real‐world data supporting first‐line trastuzumab plus chemotherapy in Chinese patients with HER2‐positive mGC.
Implications for Practice
This prospective, noninterventional registry study aimed to provide safety and effectiveness data for the use of trastuzumab in combination with chemotherapy in Chinese patients with HER2‐positive metastatic gastric cancer (mGC) from the real‐world clinical setting. Trastuzumab plus first‐line chemotherapy was shown to be safe and to improve outcomes when compared with patients treated with chemotherapy alone. Trastuzumab was effective within a range of treatment regimens; subgroup analysis showed that trastuzumab paired most effectively with the XELOX regimen. This study provides real‐world clinical safety and effectiveness data supporting the use of trastuzumab in the treatment of Chinese patients with HER2‐positive mGC.
Keywords: Gastric cancer, Epidermal growth factor receptor, Trastuzumab, China, Registries
Short abstract
This article reports the results of the EVIDENCE study, which evaluated the effectiveness and safety of trastuzumab in patients with HER2‐positive metastatic gastric cancer in a real‐world setting in China.
Introduction
Gastric cancer (GC) is the third leading cause of cancer‐related deaths globally [1] and the third most common cause of cancer‐related deaths in China [2]. Currently, curative treatment is limited to surgical resection combined with perioperative or adjuvant chemotherapy; in patients with advanced unresectable or metastatic GC (mGC), chemotherapy is the major therapeutic option [3, 4, 5]. Although chemotherapy can improve patient survival, there are high rates of relapse and resistance that necessitate other treatment options [6, 7]. Targeted therapies have recently begun to play important roles in treatment of advanced GC in China, and immunotherapies such as immune checkpoint inhibitors are being investigated [8, 9].
Human epidermal growth factor receptor 2 (HER2), which is overexpressed or amplified in 6%–36% of patients with GC [10], is an important target of novel mGC treatment strategies [11]. The humanized monoclonal antibody, trastuzumab (Herceptin; Roche, Shanghai Roche Pharmaceuticals Ltd., Shanghai, China), binds HER2 and inhibits the proliferation and survival of HER2‐dependent tumors. In the phase III ToGA trial of patients with advanced GC or gastrointestinal cancer, trastuzumab improved progression‐free survival (PFS) and overall survival (OS); this was the first study to report a median OS of >1 year in patients with GC treated with trastuzumab plus chemotherapy [12, 13]. This was also the first phase III study of this patient population, of whom 51% were Asian, to use a molecular marker (HER2 tumor overexpression) to select patients for treatment with a targeted agent [12]. Trastuzumab was approved in China in 2012 for the treatment of HER2‐overexpressing mGC and has since become a standard treatment in combination with cisplatin and capecitabine or 5‐fluorouracil (5‐FU) [14].
Although the safety profile of trastuzumab has been well documented and evaluated in several randomized studies [15, 16, 17, 18], questions remain regarding its safety and efficacy among the general patient population in a real‐world clinical setting in China. Registry studies collecting observational data from a real‐world setting are important for contributing to informed treatment decisions, and such studies would greatly benefit both clinicians and patients with cancer in China considering treatment regimens that include trastuzumab [19, 20, 21]. The aim of this EVIDENCE study was to gather data and evaluate the effectiveness and safety of trastuzumab in patients with HER2‐positive mGC in China in a real‐world, multicenter setting and to evaluate treatment patterns and clinical outcomes in this patient population.
Subjects, Materials, and Methods
Patients
Patients with histologically confirmed recurrent or metastatic disease documented to involve at least one organ or tissue, or with inoperable locally advanced disease, and who have accessible medical records and a known HER2 status were eligible to participate in this study. The key exclusion criterion was participation in a blinded study after enrollment. All patients provided written informed consent prior to participation.
Study Design and Treatments
EVIDENCE was a prospective, multicenter, noninterventional, registry study conducted in China. All eligible enrolled patients were assigned to one of five parallel study cohorts according to HER2 status, disease stage, and treatment with trastuzumab. Tumors were evaluated for HER2 status by immunohistochemistry (IHC) and in situ hybridization (ISH): HER2 positivity was defined as either IHC 3+ or IHC 2+ and ISH+ (≥2.0 HER2 gene copy number to chromosome enumeration probe 17 ratio) in either primary or metastatic tumors, and HER2 negativity as IHC 2+ without ISH information.
The cohorts were as follows: cohort I, 200 patients with HER2‐positive mGC who were treated with trastuzumab; cohort II, 200 patients with HER2‐positive mGC who were not treated with trastuzumab; cohort III, 400 patients with HER2‐positive non‐mGC; cohort IV, 400 patients with HER2‐negative mGC; and cohort V, 400 patients with HER2‐negative non‐mGC. Recruitment was initiated in April 2013 and continued until the target patient number was reached or until the last target patient was enrolled on December 31, 2016, whichever occurred first. Data from patients in cohorts I, II, and IV are reviewed in the present analysis. Patients in cohorts I, II, and IV were followed up until death, withdrawal, 1 year after the date of enrollment of the last patient in their specified cohort, or December 31, 2017, whichever occurred first.
Treatment, procedures, and scheduled visits were determined at the investigator's discretion and not specified in the study protocol. Data collection intervals were aligned with the expected schedule of routine care. Key data collected at baseline included demographic information, risk factors, Eastern Cooperative Oncology Group (ECOG) performance status (PS), GC history and prior treatment, cardiovascular risk factors, laboratory tests, record of hepatitis, selected concomitant medications, left ventricular ejection fraction (LVEF), and HER2 testing. Key data collected at follow‐up included ECOG PS, selected concomitant medications, LVEF, disease response status, and safety data.
Approval was obtained from the institutional review boards of the participating centers and the study was conducted in accordance with the Declaration of Helsinki, Pharmaceutical Affairs Law, Good Clinical Practice, and associated Chinese regulations. This study was registered at ClinicalTrials.gov (NCT01839500).
Effectiveness Outcomes
The major effectiveness outcomes of interest were OS, PFS, and response rate (RR). OS was assessed as OS1, defined as the time from the date of diagnosis of first recurrence/metastasis to death from any cause (cohorts I, II, and IV); first‐line OS2, defined as the time from the start date of trastuzumab to death from any cause (cohort I only); and second‐/third‐line OS2, defined as the time from 1 day after the date of first/second investigator‐assessed disease progression to death from any cause (cohort I only).
PFS was assessed as first‐line PFS1, defined as the time from the start of treatment to investigator‐assessed disease progression or death from any cause, whichever occurred first (cohorts I, II, and IV); second‐/third‐line PFS1, defined as the time from one day after the date of first/second investigator‐assessed disease progression to the next investigator‐assessed disease progression or death from any cause, whichever occurred first (cohorts I, II, and IV); first‐line PFS2, defined as the time from the start of trastuzumab to investigator‐assessed disease progression or death from any cause, whichever occurred first (cohort I only); and second‐/third‐line PFS2, defined as the time from 1 day after the date of first/second investigator‐assessed disease progression to the next investigator‐assessed disease progression or death from any cause, whichever occurred first (cohort I only).
RR was defined as the proportion of patients achieving complete response or partial response based on their best investigator‐assessed overall response.
Safety Outcomes
Safety outcomes were examined in cohort I only and included nonserious adverse events (AEs) other than AEs of special interest (AESIs), all AESIs, and all serious AEs (SAEs). AEs were coded according to the Medical Dictionary for Regulatory Activities version 21.0.
Statistical Analysis
As this was a registry study, no formal calculation was performed for the sample size. It was estimated that approximately 1,600 patients with GC would be registered at approximately 85 hospitals in 7 regions of China within 4 years. This sample size was sufficient to estimate the confidence interval (CI) with the half‐width within 0.049 times the SD for normally distributed variables. The full analysis set (FAS) was the main analysis population and comprised all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations. Patients who participated in a blinded study after being enrolled in the EVIDENCE study were considered withdrawn. All eligible patients were included in the study and were assigned to each of the cohorts according to their HER2 status, disease state, and whether they were treated with trastuzumab; the decision to treat with trastuzumab was made by the treating physician. Subgroup analyses by treatment line, de novo (i.e., if the date of the first metastasis was 31 days or more after the date of primary GC diagnosis) or recurrent (i.e., if the first metastasis occurred before the date of primary GC diagnosis or within the first 31 days after diagnosis) disease, history of surgery, and first‐line treatment regimen were performed for each reported cohort. Missing or loss to follow‐up data were not included in the efficacy or safety analyses except for partial missing dates in which the month or day was missing. If an endpoint start date was missing, in most cases, December was imputed for the missing month or the last day of the month was imputed for the missing day. If an endpoint end date was missing, in most cases, January was imputed for the missing month or the first day of the month was imputed for the missing day. In an effort to minimize lost to follow‐up bias, additional information from public sources was sought to obtain date of death for patients who were lost to follow‐up, except in cases of withdrawn consent.
Categorical data were summarized by rate and 95% CI using the Clopper‐Pearson Exact method. Continuous data were summarized using mean, SD, coefficient of variation, geometric mean, median, minimum, and maximum. Safety and effectiveness outcomes were evaluated using univariate analysis, stratified analysis by relevant baseline risk factors, and multivariable analysis to adjust for potential confounding factors. Using the Kaplan‐Meier method, OS1 and PFS1 of each line were summarized by cohort and by subgroups for cohorts I, II, and IV; and OS2 and PFS2 of each line were summarized by cohort and by subgroups for cohort I. Propensity score, defined as the conditional probability of being treated given the covariates, was used to reduce bias between cohorts I and II by balancing the covariates within the treatment groups. All statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC). The accepted level of significance was 5% and all tests were two‐sided.
Results
Patients
Between April 2013 and June 2018, 1,556 patients were enrolled (and provided informed consent) at 85 hospitals in China. Of these, a total of 772 patients were enrolled in cohorts I (196), II (122), and IV (454); and 709 patients were included in the FAS (cohort I, 174; cohort II, 113; and cohort IV, 422; supplemental online Fig. 1). Nine (5.2%), 7 (6.2%), and 15 (3.6%) patients from cohorts I, II and IV, respectively, withdrew from the study. Reasons for withdrawal included patients enrolling in a blinded interventional study, patient's request to withdraw, physician's unwillingness to follow up, and other. Death was reported for 78 (44.8%), 48 (42.5%), and 184 (43.6%) patients in cohorts I, II and IV, respectively (supplemental online Fig. 1). The mean ± SD duration of follow‐up was as follows: cohort I, 422.5 ± 320.5 days; cohort II, 287.5 ± 284.1 days; and cohort IV, 277.5 ± 243.3 days.
Patients’ demographic and clinical/disease characteristics were representative of the intended population (Table 1). Patients were similar in age and predominantly male. All three cohorts comprised a similar distribution for ECOG PS, Lauren classification, TNM status, and primary GC diagnosis. Patients in cohorts I and II were all HER2 positive, whereas patients in cohort IV were HER2 negative. Most patients in cohorts I, II, and IV received first‐line chemotherapy (168 [96.6%], 94 [83.2%], and 359 [85.1%], respectively; Table 2).
Table 1.
Patient baseline demographics and clinical characteristics
| Characteristics | Cohort I (N = 174) | Cohort II (N = 113) | Cohort IV (N = 422) |
|---|---|---|---|
| Age, mean ± SD, yr | 59.7 ± 12.3 | 60.4 ± 11.3 | 56.3 ± 12.4 |
| Sex, male | 132 (75.9) | 95 (84.1) | 285 (67.5) |
| ECOG performance | |||
| Grade 0–1 | 118 (67.8) | 63 (55.8) | 279 (66.1) |
| Grade 2–5 | 24 (13.8) | 15 (13.3) | 29 (6.9) |
| Unknown | 32 (18.4) | 35 (31.0) | 114 (27.0) |
| Primary GC diagnosis | |||
| Antrum | 43 (24.7) | 25 (22.1) | 124 (29.4) |
| Body | 32 (18.4) | 24 (21.2) | 89 (21.1) |
| Gastroesophageal junction | 51 (29.3) | 25 (22.1) | 60 (14.2) |
| T status | |||
| Tx | 24 (20.5) | 13 (14.6) | 43 (15.9) |
| T0 | 0 | 0 | 1 (0.4) |
| Tis | 1 (0.9) | 0 | 0 |
| T1 | 3 (2.6) | 1 (1.1) | 3 (1.1) |
| T1a | 2 (1.7) | 0 | 0 |
| T1b | 1 (0.9) | 1 (1.1) | 2 (0.7) |
| Unknown | 0 | 0 | 1 (0.4) |
| T2 | 3 (2.6) | 3 (3.4) | 13 (4.8) |
| T3 | 21 (17.9) | 12 (13.5) | 35 (13.0) |
| T4 | 65 (55.6) | 60 (67.4) | 175 (64.8) |
| T4a | 33 (28.2) | 31 (34.8) | 82 (30.4) |
| T4b | 17 (14.5) | 16 (18.0) | 54 (20.0) |
| Unknown | 15 (12.8) | 13 (14.6) | 39 (14.4) |
| N status | |||
| N0 | 8 (6.8) | 10 (11.2) | 19 (7.0) |
| N1 | 6 (5.1) | 10 (11.2) | 33 (12.2) |
| N2 | 23 (19.7) | 13 (14.6) | 51 (18.9) |
| N3 | 42 (35.9) | 35 (39.3) | 114 (42.2) |
| N3a | 5 (4.3) | 8 (9.0) | 19 (7.0) |
| N3b | 9 (7.7) | 7 (7.9) | 9 (3.3) |
| Unknown | 18 (15.4) | 6 (6.7) | 42 (15.6) |
| M status | |||
| M0 | 36 (23.2) | 31 (30.1) | 73 (23.3) |
| M1 | 116 (74.8) | 66 (64.1) | 227 (72.5) |
| Unknown | 3 (1.9) | 6 (5.8) | 13 (4.2) |
| Lauren classification | |||
| Intestinal | 35 (20.1) | 15 (13.3) | 26 (6.2) |
| Diffuse | 15 (8.6) | 6 (5.3) | 50 (11.8) |
| Mixed | 11 (6.3) | 2 (1.8) | 19 (4.5) |
| Unknown | 113 (64.9) | 90 (79.6) | 327 (77.5) |
| HER2 status | |||
| Positive | 174 (100) | 113 (100) | 0 |
| IHC 3+ | 139 (83.2) | 100 (88.5) | 0 |
| ISH positive | 41 (23.6) | 12 (10.6) | 3 (0.7) |
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations. Data are number of patients (%) unless otherwise stated.
Abbreviations: ECOG, Eastern Cooperative Oncology Group; GC, gastric cancer; IHC, immunohistochemistry; ISH, in situ hybridization; M, metastasis; N, number of lymph nodes; T, size of the primary tumor.
Table 2.
Treatment patterns
| Treatment pattern | All cohorts (N = 709) | Cohort I (N = 174) | Cohort II (N = 113) | Cohort IV (N = 422) |
|---|---|---|---|---|
| 1st line chemotherapya | 621 (87.6) | 168 (96.6) | 94 (83.2) | 359 (85.1) |
| 2nd line chemotherapya | 215 (30.3) | 74 (42.5) | 28 (24.8) | 113 (26.8) |
| 3rd line chemotherapya | 73 (10.3) | 31 (17.8) | 9 (8.0) | 33 (7.8) |
| Any surgery since 1st recurrence/metastatic disease diagnoseda | 203 (28.6) | 31 (17.8) | 32 (28.3) | 140 (33.2) |
| Any radiation since 1st recurrence/metastatic disease diagnoseda | 64 (9.0) | 21 (12.1) | 7 (6.2) | 36 (8.5) |
| Any gastric cancer‐related surgery before metastasis/recurrencea | 206 (29.1) | 54 (31.0) | 37 (32.7) | 115 (27.3) |
| Any adjuvant chemotherapy before metastasis/recurrencea | 102 (14.4) | 35 (20.1) | 13 (11.5) | 54 (12.8) |
| Any adjuvant radiation before metastasis/recurrencea | 6 (0.8) | 3 (1.7) | 0 | 3 (0.7) |
| Time from diagnosis of gastric cancer to metastasis/recurrence, mean ± SD, yr | 0.3 ± 0.9 | 0.5 ± 1.0 | 0.3 ± 0.7 | 0.3 ± 0.8 |
| 1st line chemotherapy drugsb , c | ||||
| n | 621 | 168 | 94 | 356 |
| XELOX | 72 (11.6) | 30 (17.9) | 12 (12.8) | 30 (8.4) |
| 5‐FU + C/XEL + C | 23 (3.7) | 19 (11.3) | 1 (1.1) | 3 (0.8) |
| 5‐FU/S‐1/XEL/tegafur/UFT | 46 (7.4) | 17 (10.1) | 9 (9.6) | 20 (5.6) |
| Epirubicin/doxorubicin/liposome adriamycin/pirarubicin + 5‐FU/S‐1/XEL + C/OX | 13 (2.1) | 2 (1.2) | 0 | 11 (3.1) |
| Paclitaxel/docetaxel + 5‐FU/ S‐1/XEL + C/OX | 33 (5.3) | 8 (4.8) | 6 (6.4) | 19 (5.3) |
| 5‐FU/S‐1/XEL/tegafur/UFT + C/OX/nedaplatin | 242 (39.0) | 75 (44.6) | 36 (38.3) | 131 (36.5) |
| Other | 287 (46.2) | 66 (39.3) | 43 (45.7) | 178 (49.6) |
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations. Data are number of patients (%) unless otherwise stated.
Percentages were calculated based on N.
Percentages were calculated based on n.
Only counted first chemotherapy in the first‐line chemotherapy.
Abbreviations: 5‐FU, 5‐fluorouracil; C, cisplatin; OX, oxaliplatin; UFT, tegafur/uracil; XEL, capecitabine; XELOX, capecitabine + oxaliplatin.
Effectiveness
The median OS1 was longest in cohort I at 22.3 (95% CI, 19.5–25.7) months, followed by cohort IV and cohort II at 17.4 (95% CI, 14.8–20.7) and 17.2 (95% CI, 14.1–28.2) months, respectively (supplemental online Fig. 2). The proportion of patients who had a history of surgery was 39.7%, 57.5%, and 58.5% in cohorts I, II, and IV, respectively. After exclusion of these patients, median OS1 for the respective cohorts was 19.9, 15.3, and 12.9 months.
Median OS2 of first‐, second‐, and third‐line treatment of cohort I was 22.1 (95% CI, 17.8–29.2), 12.0 (95% CI, 9.0–17.5), and 10.1 (95% CI, 7.5–14.9) months, respectively (Table 3). Median OS2 of first‐line treatment for cohort I patients analyzed by subgroup (first‐line treatment regimen) was 34.6 (95% CI, 34.6–not reached) months for patients receiving trastuzumab with capecitabine + oxaliplatin (XELOX), 27.4 (95% CI, 25.6–29.2) months for patients receiving trastuzumab + docetaxel/paclitaxel + oxaliplatin (OX)/cisplatin (C)/carboplatin/nedaplatin +5‐FU/S‐1/capecitabine (XEL), 34.6 (95% CI, 23.7–not reached) months for patients receiving trastuzumab + OX/C/carboplatin/nedaplatin +5‐FU/S‐1/XEL, 17.8 (95% CI, 7.8–not reached) months for patients receiving trastuzumab +5‐FU/S‐1/XEL, and 17.8 (95% CI, 13.8–22.1) months for patients receiving trastuzumab + other chemotherapy (Table 3).
Table 3.
Effectiveness outcomes: overall survival
| Endpoint | n | n (%) of censoringa , b | Median estimate (95% CI), months |
|---|---|---|---|
| OS1 | |||
| Cohort I (N = 174) | 169 | 101 (59.8) | 22.3 (19.5–25.7) |
| Cohort II (N = 113) | 109 | 69 (63.3) | 17.2 (14.1–28.2) |
| Cohort IV (N = 422) | 402 | 243 (60.4) | 17.4 (14.8–20.7) |
| Cohort I/II/IV (N = 709) | 680 | 413 (60.7) | 19.1 (17.0–20.9) |
| OS1 | |||
| Cohort I | |||
| Recurrent | 111 | 69 (62.2) | 24.1 (19.5–38.7) |
| De novo | 58 | 32 (55.2) | 22.0 (16.5–25.7) |
| Cohort II | |||
| Recurrent | 78 | 49 (62.8) | 17.5 (14.9–28.2) |
| De novo | 31 | 20 (64.5) | 14.1 (8.4–not reached) |
| Cohort IV | |||
| Recurrent | 292 | 174 (59.6) | 17.0 (14.5–20.7) |
| De novo | 110 | 69 (62.7) | 17.9 (12.6–not reached) |
| OS2 of first‐line treatmentc | |||
| Cohort I (N = 174) | 162 | 100 (61.7) | 22.1 (17.8–29.2) |
| OS2 of second‐line treatmentc | |||
| Cohort I (N = 174) | 57 | 29 (50.9) | 12.0 (9.0–17.5) |
| Without first‐line (N = 11) | 9 | 2 (22.2) | 15.9 (3.8–17.7) |
| With first‐line (N = 163) | 48 | 27 (56.3) | 11.6 (9.0–22.0) |
| OS2 of third‐line treatmentc | |||
| Cohort I (N = 174) | 24 | 14 (58.3) | 10.1 (7.5–14.9) |
| Without second‐line (N = 117) | 4 | 4 (100.0) | not reached (not reached–not reached) |
| With second‐line (N = 57) | 20 | 10 (50.0) | 8.9 (5.8–14.2) |
| OS2 of first‐line treatmentc | |||
| 1st line trastuzumab + XELOX | 35 | 28 (80.0) | 34.6 (34.6–not reached) |
| 1st line trastuzumab + 5‐FU + C or trastuzumab + XEL + C | 19 | 16 (84.2) | not reached (not reached–not reached) |
| 1st line trastuzumab + docetaxel/paclitaxel + OX/C/carboplatin/nedaplatin + 5‐FU/S‐1/XEL | 5 | 3 (60.0) | 27.4 (25.6–29.2) |
| 1st line trastuzumab + OX/C/carboplatin/nedaplatin + 5‐FU/S‐1/XEL | 73 | 54 (74.0) | 34.6 (23.7–not reached) |
| 1st line trastuzumab + 5‐FU/S‐1/XEL | 13 | 5 (38.5) | 17.8 (7.8–not reached) |
| 1st line trastuzumab + other chemotherapy | 66 | 37 (56.1) | 17.8 (13.8–22.1) |
| 1st line trastuzumab + targeted therapy | 0 | 0 | not reached (not reached–not reached) |
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations.
Patients were censored on the last known date if he/she was alive when he/she came off the study.
Percentage of censoring was based on n.
Cohort I only.
Abbreviations: 5‐FU, 5‐fluorouracil; C, cisplatin; CI, confidence interval; First‐line OS2, time from start date of trastuzumab to death from any cause (applies only to cohort I); OS, overall survival; OS1, time from the date of diagnosis of first recurrence/metastasis to death from any cause; OX, oxaliplatin; Second/third‐line OS2, time from 1 day after the date of first/second investigator‐assessed disease progression to death from any cause (applies to cohort I only); XEL, capecitabine; XELOX, capecitabine + oxaliplatin.
Median PFS1 of first‐line therapy was longest in cohort I at 8.2 (95% CI, 6.8–10.1) months, followed by cohort II and cohort IV at 6.9 (95% CI, 4.4–8.6) and 6.2 (95% CI, 5.6–6.8) months, respectively. The median PFS1 of second‐line treatment was longest in cohort II (4.8 months vs. 3.6 months in cohort I and 2.7 months in cohort IV), as was the median PFS1 of third‐line treatment (5.4 months in cohort II vs. 3.2 months in cohort I and 3.1 months in cohort IV). Median PFS2 of first‐, second‐, and third‐line treatments for patients in cohort I were 7.8 (95% CI, 6.6–9.2), 3.8 (95% CI, 3.1–5.8), and 3.9 (95% CI, 2.8–8.5) months, respectively. Patients with de novo carcinoma had a longer median PFS2 than those with recurrent carcinoma in all treatment lines (Table 4).
Table 4.
Effectiveness outcomes: progression‐free survival
| Endpoint | n | n (%) of censoringa , b | Median estimate (95% CI), months |
|---|---|---|---|
| PFS1 of first‐line treatmentc | |||
| Cohort I (N = 174) | 170 | 48 (28.2) | 8.2 (6.8–10.1) |
| Cohort II (N = 113) | 94 | 38 (40.4) | 6.9 (4.4–8.6) |
| Cohort IV (N = 422) | 359 | 127 (35.4) | 6.2 (5.6–6.8) |
| Cohort I/II/IV (N = 709) | 623 | 213 (34.2) | 6.7 (6.3–7.3) |
| PFS1 of second‐line treatmentc | |||
| Cohort I (N = 174) | 72 | 11 (15.3) | 3.6 (2.9–5.7) |
| Cohort II (N = 113) | 21 | 3 (14.3) | 4.8 (1.9–6.5) |
| Cohort IV (N = 422) | 83 | 15 (18.1) | 2.7 (2.4–3.6) |
| Cohort I/II/IV (N = 709) | 176 | 29 (16.5) | 3.4 (2.7–3.8) |
| PFS1 of third‐line treatmentc | |||
| Cohort I (N = 174) | 31 | 2 (6.5) | 3.2 (2.0–4.5) |
| Cohort II (N = 113) | 7 | 1 (14.3) | 5.4 (0.6–7.1) |
| Cohort IV (N = 422) | 27 | 2 (7.4) | 3.1 (2.3–5.9) |
| Cohort I/II/IV (N = 709) | 65 | 5 (7.7) | 3.1 (2.4–4.0) |
| PFS1 of first‐line treatmentc | |||
| Cohort I | |||
| Recurrent (N = 111) | 111 | 29 (26.1) | 8.1 (6.6–10.0) |
| De novo (N = 58) | 56 | 17 (30.4) | 9.8 (5.9–11.2) |
| Cohort II | |||
| Recurrent (N = 78) | 64 | 26 (40.6) | 6.9 (4.4–8.6) |
| De novo (N = 31) | 26 | 10 (38.5) | 6.1 (2.6–9.0) |
| Cohort IV | |||
| Recurrent (N = 292) | 254 | 87 (34.3) | 6.2 (5.4–6.7) |
| De novo (N = 110) | 93 | 36 (38.7) | 6.1 (4.6–7.5) |
| PFS1 of second‐line treatmentc | |||
| Cohort I | |||
| Recurrent (N = 111) | 47 | 8 (17.0) | 3.6 (2.3–5.7) |
| De novo (N = 58) | 23 | 3 (13.0) | 3.7 (2.8–7.9) |
| Cohort II | |||
| Recurrent (N = 78) | 17 | 2 (11.8) | 4.8 (1.9–6.9) |
| De novo (N = 31) | 4 | 1 (25.0) | 2.4 (0.6–6.0) |
| Cohort IV | |||
| Recurrent (N = 292) | 58 | 11 (19.0) | 2.7 (2.2–4.2) |
| De novo (N = 110) | 22 | 4 (18.2) | 3.3 (1.7–4.3) |
| PFS1 of third‐line treatmentc | |||
| Cohort I | |||
| Recurrent (N = 111) | 22 | 2 (9.1) | 2.8 (1.6–4.2) |
| De novo (N = 58) | 9 | 0 | 3.2 (1.1–8.9) |
| Cohort II | |||
| Recurrent (N = 78) | 7 | 1 (14.3) | 5.4 (0.6–7.1) |
| De novo (N = 31) | 0 | 0 | not reached |
| Cohort IV | |||
| Recurrent (N = 292) | 21 | 1 (4.8) | 3.0 (1.7–4.0) |
| De novo (N = 110) | 4 | 1 (25.0) | 6.6 (3.4–7.1) |
| PFS2 of first‐line treatmentd , e | |||
| Cohort I (N = 174) | 162 | 48 (29.6) | 7.8 (6.6–9.2) |
| PFS2 of second‐line treatmentd , e | |||
| Cohort I (N = 174) | 51 | 5 (9.8) | 3.8 (3.1–5.8) |
| With first‐line (N = 163) | 42 | 5 (11.9) | 4.2 (2.9–6.2) |
| Without first‐line (N = 11) | 9 | 0 | 3.8 (0.6–8.6) |
| PFS2 of third‐line treatmentd | |||
| Cohort I (N = 174) | 19 | 2 (10.5) | 3.9 (2.8–8.5) |
| With second‐line (N = 57) | 16 | 1 (6.3) | 3.7 (2.7–8.5) |
| With third‐line (N = 117) | 3 | 1 (33.3) | 3.9 (3.2–not reached) |
| PFS2 of first‐line treatmentd , e | |||
| Cohort I | |||
| Recurrent (N = 111) | 105 | 29 (27.6) | 7.5 (6.4–9.1) |
| De novo (N = 58) | 54 | 17 (31.5) | 8.3 (4.9–10.5) |
| PFS2 of second‐line treatmentd , e | |||
| Cohort I | |||
| Recurrent (N = 111) | 33 | 4 (12.1) | 3.7 (2.9–5.8) |
| With first‐line (N = 105) | 29 | 4 (13.8) | 3.7 (2.9–6.2) |
| Without first‐line (N = 6) | 4 | 0 | 3.7 (0.6–5.7) |
| De novo (N = 58) | 16 | 1 (6.3) | 5.8 (2.9–8.6) |
| With first‐line (N = 54) | 12 | 1 (8.3) | 4.2 (1.5–8.3) |
| Without first‐line (N = 4) | 4 | 0 | 8.1 (2.9–18.3) |
| PFS2 of third‐line treatmentd | |||
| Cohort I | |||
| Recurrent (N = 111) | 15 | 2 (13.3) | 3.7 (2.2–8.5) |
| With second‐line (N = 37) | 12 | 1 (8.3) | 2.8 (1.6–8.5) |
| Without second‐line (N = 74) | 3 | 1 (33.3) | 3.9 (3.2–not reached) |
| De novo (N = 58) | 4 | 0 | 6.7 (2.7–11.5) |
| With second‐line (N = 18) | 4 | 0 | 6.7 (2.7–11.5) |
| Without second‐line (N = 40) | 0 | 0 | not reached |
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations.
Patients were censored on the last tumor assessment date even if the event did not occur.
Percentage of censoring was based on n.
Recurrent/de novo data were missing for 3 patients who had first‐line PFS1 and 2 patients who had second‐line PFS1 because the date of diagnosis for first recurrence/metastasis was missing in cohort I; recurrent/de novo data were missing for 4 patients who had first‐line PFS1 because the diagnosis for first recurrence/metastasis was missing in cohort II; and recurrent/de novo data were missing for 12 patients who had first‐line PFS1, 3 patients who had second‐line PFS1, and 2 patients who had third‐line PFS1 because the date of diagnosis for first recurrence/metastasis was missing in cohort IV.
Cohort I only.
Recurrent/de novo data were missing for 3 patients who had first‐line PFS2 and 2 patients who had second‐line PFS2 because the date of diagnosis for first recurrence/metastasis was missing.
Abbreviations: CI, confidence interval; First‐line PFS1, time from start date of treatment to investigator‐assessed disease progression or death from any cause; First‐line PFS2, time from start date of trastuzumab to investigator‐assessed disease progression or death from any cause (applies to cohort I only); PFS, progression‐free survival; Second/third‐line PFS1, time from 1 day after first/second investigator‐assessed disease progression to the next investigator‐assessed disease progression or death from any cause; Second/third‐line PFS2, time from 1 day after the date of first/second investigator‐assessed disease progression to death from any cause (applies to cohort I only).
RR was highest in cohort I for all treatment lines. Respective first‐, second‐, and third‐line treatment RRs (%) were 51.7% (95% CI, 42.4–60.9), 13.0% (95% CI, 4.9–26.3), and 10.0% (95% CI, 2.8–23.7) for cohort I; 18.4% (95% CI, 7.7–34.3), 0%, and 0% for cohort II; and 32.8% (95% CI, 25.9–40.3), 8.3% (95% CI, 2.3–20.0), and 5.6% (95% CI, 0.7–18.7) for cohort IV (Table 5). The RR in cohort I and cohort IV was estimated to be greater in recurrent versus de novo cases in first‐line treatment (cohort I: 54.3% vs. 46.0%; cohort IV: 41.6% vs. 11.6%).
Table 5.
Effectiveness outcomes: response rate
| Endpoint | N | Number of responders | RRa estimate (95% CI), % |
|---|---|---|---|
| RR of first‐line treatmentb | |||
| Cohort I (N = 174) | 120 | 62 | 51.7 (42.4–60.9) |
| Cohort II (N = 113) | 38 | 7 | 18.4 (7.7–34.3) |
| Cohort IV (N = 422) | 174 | 57 | 32.8 (25.9–40.3) |
| Cohort I/II/IV (N = 709) | 332 | 126 | 38.0 (32.7–43.4) |
| RR of second‐line treatmentb | |||
| Cohort I (N = 174) | 46 | 6 | 13.0 (4.9–26.3) |
| Cohort II (N = 113) | 12 | 0 | 0.0 (0.0–26.5) |
| Cohort IV (N = 422) | 48 | 4 | 8.3 (2.3–20.0) |
| Cohort I/II/IV (N = 709) | 106 | 10 | 9.4 (4.6–16.7) |
| RR of third‐line treatmentb | |||
| Cohort I (N = 174) | 40 | 4 | 10.0 (2.8–23.7) |
| Cohort II (N = 113) | 9 | 0 | 0.0 (0.0–33.6) |
| Cohort IV (N = 422) | 36 | 2 | 5.6 (0.7–18.7) |
| Cohort I/II/IV (N = 709) | 85 | 6 | 7.1 (2.6–14.7) |
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations.
95% CIs of RR were calculated according to Clopper‐Pearson Exact method.
Recurrent/de novo data were missing for two patients who had first‐line RR, two patients who had second‐line RR, and one patient who had third‐line RR because the date of diagnosis for first recurrence/metastasis was missing in cohort I; recurrent/de novo data were missing for three patients who had first‐line RR because the diagnosis for first recurrence/metastasis was missing in cohort II; recurrent/de novo data were missing for six patients who had first‐line RR, two patients who had third‐line RR, three patients who had second‐line RR, and two patients who had third‐line RR because the date of diagnosis for first recurrence/metastasis was missing in cohort IV.
Abbreviations: CI, confidence interval; RR, response rate.
Propensity score matching was performed for the following covariates: IHC/ISH HER2 status, sex, age (at diagnosis of first recurrence/metastasis, ≤50 vs. >50 years), ECOG PS, recurrent/de novo disease, Lauren classification, first‐line therapy type, TNM status of TNM classification, and number of recurrence/metastasis. The patient numbers for each treatment group before and after propensity score matching are shown in Table 6. Propensity score matching for cohort I compared with cohort II significantly favored cohort I for OS1, PFS1, and RR: OS1 hazard ratio (HR), 0.61 (95% CI, 0.39–0.94; p = .0260); PFS1 (first‐line treatment) HR, 0.64 (95% CI, 0.45–0.91; p = .0142); and RR (first‐line treatment) odds ratio, 4.93 (95% CI, 1.93–12.61; p = .0009; supplemental online Fig. 3A, 3B).
Table 6.
Comparison of variables after propensity score matching
| Variable | Unmatched | Matched | ||||
|---|---|---|---|---|---|---|
| Cohort I (N = 174), n (%) | Cohort II (N = 113), n (%) | p value | Cohort I (N = 113), n (%) | Cohort II (N = 113), n (%) | p value | |
| Sex | ||||||
| Male | 132 (75.9) | 95 (84.1) | .0947 | 92 (81.4) | 95 (84.1) | .5974 |
| Female | 42 (24.1) | 18 (15.9) | 21 (18.6) | 18 (15.9) | ||
| Age, yra | ||||||
| >50 | 132 (75.9) | 88 (77.9) | .8283 | 85 (75.2) | 88 (77.9) | .8807 |
| ≤50 | 37 (21.3) | 21 (18.6) | 23 (20.4) | 21 (18.6) | ||
| Unknown | 5 (2.9) | 4 (3.5) | 5 (4.4) | 4 (3.5) | ||
| ECOG performance at baseline (group) | ||||||
| Grade 0–1 | 118 (67.8) | 63 (55.8) | .0441 | 78 (69.0) | 63 (55.8) | .0952 |
| Grade 2–5 | 24 (13.8) | 15 (13.3) | 13 (11.5) | 15 (13.3) | ||
| Unknown | 32 (18.4) | 35 (31.0) | 22 (19.5) | 35 (31.0) | ||
| Recurrent /De novo | ||||||
| Recurrent | 111 (63.8) | 78 (69.0) | .5625 | 74 (65.5) | 78 (69.0) | .8374 |
| De novo | 58 (33.3) | 31 (27.4) | 34 (30.1) | 31 (27.4) | ||
| Unknown | 5 (2.9) | 4 (3.5) | 5 (4.4) | 4 (3.5) | ||
| Pathological classification (Lauren classification) | ||||||
| Intestinal | 35 (20.1) | 15 (13.3) | .0441 | 22 (19.5) | 15 (13.3) | .4477 |
| Diffuse | 15 (8.6) | 6 (5.3) | 7 (6.2) | 6 (5.3) | ||
| Mixed | 11 (6.3) | 2 (1.8) | 4 (3.5) | 2 (1.8) | ||
| Unknown | 113 (64.9) | 90 (79.6) | 80 (70.8) | 90 (79.6) | ||
| 1st line therapy type | ||||||
| docetaxel/paclitaxel + OX/C/carboplatin/nedaplatin + 5‐FU/S1/XEL | 8 (4.6) | 6 (5.3) | .0422 | 6 (5.3) | 6 (5.3) | .4214 |
| OX/C/carboplatin/nedaplatin + 5‐FU/S1/XEL | 81 (46.6) | 36 (31.9) | 44 (38.9) | 36 (31.9) | ||
| 5‐FU/S1/XEL | 18 (10.3) | 9 (8.0) | 13 (11.5) | 9 (8.0) | ||
| Other therapy | 67 (38.5) | 62 (54.9) | 50 (44.2) | 62 (54.9) | ||
| HER2 status | ||||||
| IHC 2+ & ISH positive | 20 (11.5) | 10 (8.8) | .0613 | 12 (10.6) | 10 (8.8) | .9039 |
| IHC 3+ | 138 (79.3) | 100 (88.5) | 98 (86.7) | 100 (88.5) | ||
| Unknown | 16 (9.2) | 3 (2.7) | 3 (2.7) | 3 (2.7) | ||
| T status of TNM classification | ||||||
| Tx | 24 (13.8) | 13 (11.5) | .1850 | 18 (15.9) | 13 (11.5) | .7888 |
| Tis | 1 (0.6) | 0 | 0 | 0 | ||
| T1 | 3 (1.7) | 1 (0.9) | 3 (2.7) | 1 (0.9) | ||
| T2 | 3 (1.7) | 3 (2.7) | 3 (2.7) | 3 (2.7) | ||
| T3 | 21 (12.1) | 12 (10.6) | 13 (11.5) | 12 (10.6) | ||
| T4 | 65 (37.4) | 60 (53.1) | 52 (46.0) | 60 (53.1) | ||
| Unknown | 57 (32.8) | 24 (21.2) | 24 (21.2) | 24 (21.2) | ||
| N status of TNM classification | ||||||
| N0 | 8 (4.6) | 10 (8.8) | .0408 | 7 (6.2) | 10 (8.8) | .6847 |
| N1 | 6 (3.4) | 10 (8.8) | 6 (5.3) | 10 (8.8) | ||
| N2 | 23 (13.2) | 13 (11.5) | 14 (12.4) | 13 (11.5) | ||
| N3 | 42 (24.1) | 35 (31.0) | 33 (29.2) | 35 (31.0) | ||
| Unknown | 95 (54.6) | 45 (39.8) | 53 (46.9) | 45 (39.8) | ||
| M status of TNM classification | ||||||
| M0 | 36 (20.7) | 31 (27.4) | .3344 | 27 (23.9) | 31 (27.4) | .8299 |
| M1 | 116 (66.7) | 66 (58.4) | 69 (61.1) | 66 (58.4) | ||
| Unknown | 22 (12.6) | 16 (14.2) | 17 (15.0) | 16 (14.2) | ||
| Number of recurrence/metastasis of location | ||||||
| 0 | 7 (4.0) | 5 (4.4) | <.0001 | 4 (3.5) | 5 (4.4) | .0987 |
| 1 | 56 (32.2) | 69 (61.1) | 48 (42.5) | 69 (61.1) | ||
| 2 | 57 (32.8) | 26 (23.0) | 38 (33.6) | 26 (23.0) | ||
| 3 | 39 (22.4) | 9 (8.0) | 18 (15.9) | 9 (8.0) | ||
| >3 | 14 (8.0) | 3 (2.7) | 4 (3.5) | 3 (2.7) | ||
| Unknown | 1 (0.6) | 1 (0.9) | 1 (0.9) | 1 (0.9) | ||
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations.
N represents the number of patients included in the analysis.
Age at diagnosis of first recurrence/metastasis, ≤50 versus >50 years.
Abbreviations: 5‐FU, 5‐fluorouracil; C, cisplatin; ECOG, Eastern Cooperative Oncology Group; IHC, immunohistochemistry; ISH, in situ hybridization; M, metastasis; N, number of lymph nodes; T, size of the primary tumor; OX, oxaliplatin; XEL, capecitabine; XELOX, capecitabine + oxaliplatin.
Based on propensity score matching, median OS1 (24.1 months vs. 17.2 months, respectively; p = .0248) and PFS1 (7.5 months vs. 6.9 months, respectively; p = .0136) were significantly longer, and the RR was significantly greater (53% vs. 18%, respectively) in cohort I than cohort II for first‐line treatment.
Safety
AEs are summarized in Table 7 and were most commonly reported in the category of blood and lymphatic system disorders. Neutropenia and anemia were the most common AEs reported in 67 (38.5%) and 61 (35.1%) patients, respectively. Trastuzumab‐related AEs occurred in 41 (23.6%) patients and included neutropenia (14 [8.0%]), thrombocytopenia (11 [6.3%]), gastrointestinal disorders (10 [5.7%]), and white blood cell count decreased (10 [5.7%]). Trastuzumab‐related grade 3–5 AEs, SAEs, and AEs with a fatal outcome occurred in six (3.4%), four (2.3%), and one (0.6%) patients, respectively. One trastuzumab‐related, nonserious AESI was reported. SAEs, regardless of causality, occurred in 18 patients, of which neutropenia was the most common, followed by pneumonia. SAEs associated with trastuzumab included cardiac failure, intestinal obstruction, drug hypersensitivity, and pulmonary fibrosis (reported by one patient each); the outcome of the pulmonary fibrosis was reported as fatal.
Table 7.
Adverse events by system organ class and preferred term
| Adverse eventa | Cohort I (N = 174) | |
|---|---|---|
| Patients, n (%)b | Incidence ratec | |
| Any AE | 130 (74.7) | 475.5 |
| Any nonserious AEs occurring in ≥5% | 121 (69.5) | 352.3 |
| Blood and lymphatic system disorders | 102 (58.6) | 200.2 |
| Neutropenia | 67 (38.5) | 84.0 |
| Anemia | 61 (35.1) | 63.6 |
| Thrombocytopenia | 49 (28.2) | 52.7 |
| Investigations | 41 (23.6) | 42.7 |
| White blood cell count decreased | 25 (14.4) | 26.8 |
| Blood alkaline phosphatase increased | 16 (9.2) | 8.4 |
| Aspartate aminotransferase increased | 9 (5.2) | 7.5 |
| Gastrointestinal disorders | 39 (22.4) | 48.7 |
| Vomiting | 25 (14.4) | 22.4 |
| Nausea | 15 (8.6) | 12.9 |
| Diarrhea | 14 (8.0) | 13.4 |
| Hepatobiliary disorders | 31 (17.8) | 29.3 |
| Hyperbilirubinemia | 24 (13.8) | 19.4 |
| Hepatic function abnormal | 12 (6.9) | 9.9 |
| Metabolism and nutrition disorders | 22 (12.6) | 16.4 |
| Hyperglycemia | 14 (8.0) | 9.4 |
| Decreased appetite | 9 (5.2) | 7.0 |
| General disorders and administration site conditions | 12 (6.9) | 7.5 |
| Asthenia | 12 (6.9) | 7.5 |
| Skin and subcutaneous tissue disorders | 11 (6.3) | 7.5 |
| Palmar‐plantar erythrodysesthesia syndrome | 11 (6.3) | 7.5 |
| Any nonserious AESI | 3 (1.7) | 4.0 |
| Hepatobiliary disorders | 3 (1.7) | 3.0 |
| Hyperbilirubinemia | 3 (1.7) | 1.5 |
| Hepatic function abnormal | 2 (1.1) | 1.5 |
| Investigations | 1 (0.6) | 1.0 |
| Alanine aminotransferase increased | 1 (0.6) | 0.5 |
| Aspartate aminotransferase increased | 1 (0.6) | 0.5 |
| Grade 3–5 AEs | 59 (33.9) | 61.6 |
| Any SAE | 18 (10.3) | 9.9 |
| AE with fatal outcome | 1 (0.6) | 0.5 |
| AEs leading to study drug discontinuation | 7 (4.0) | 5.0 |
| AEs leading to study drug interruption | 39 (22.4) | 36.8 |
| Any trastuzumab‐related AEs | 41 (23.6) | 53.7 |
| Trastuzumab‐related AEs occurring in ≥5% | ||
| Blood and lymphatic system disorders | 23 (13.2) | 21.4 |
| Neutropenia | 14 (8.0) | 9.4 |
| Thrombocytopenia | 11 (6.3) | 6.5 |
| Gastrointestinal disorders | 10 (5.7) | 11.4 |
| Investigations | 10 (5.7) | 7.0 |
| White blood cell count decreased | 10 (5.7) | 7.0 |
| Trastuzumab‐related grade 3–5 AEs | 6 (3.4) | 4.5 |
| Trastuzumab‐related SAEs | 4 (2.3) | 2.5 |
| Trastuzumab‐related AE with fatal outcome | 1 (0.6) | 0.5 |
| Trastuzumab‐related nonserious AESI | 1 (0.6) | 1.0 |
| ALT or AST >3 × ULN and total bilirubin >2 × ULN | 1 (0.6) | 1.0 |
The full analysis set included all enrolled patients who had not been withdrawn, regardless of whether they had any protocol deviations.
Medical Dictionary for Regulatory Activities Version 21.0, displayed by descending order of frequency of system organ class and preferred term.
Percentages were calculated based on N.
Incidence rate was defined as the number of events per 100 patient‐years of follow‐up.
Abbreviations: AE, adverse event; AESI, adverse event of special interest; ALT, alanine aminotransferase; AST, aspartate aminotransferase; SAE, serious adverse event; ULN, upper limit of normal.
No significant differences in ECOG PS scores or laboratory data were observed. Deaths due to progressive disease (PD), SAE, or other were as follows: cohort I, 57 (32.8%), one (0.6%), and 20 (11.5%), respectively; cohort II, 31 (27.4%), 0, and 17 (15.0%), respectively; and cohort IV, 126 (29.9%), 0, and 58 (13.7%), respectively.
Discussion
To date, this is the largest real‐world data study of patients with mGC in China. The overall safety profiles observed in patients with HER2‐positive disease treated with trastuzumab were consistent with the profiles of trastuzumab + chemotherapy reported in previous clinical studies of patients with mGC [10, 12, 22]. We report that patients with HER2‐positive mGC treated with trastuzumab had improved median OS, PFS, and RR in a real‐world clinical setting compared with patients who did not receive trastuzumab. This study also showed that trastuzumab was effective within a range of treatment patterns, which is in‐line with previous reports [10, 23, 24, 25]. The trastuzumab + XELOX and trastuzumab + OX/C/carboplatin/nedaplatin +5‐FU/S‐1/XEL regimens were the most promising, which agrees with the current placement of XELOX as standard first‐line treatment for mGC. Several recent phase II studies supported the efficacy and safety of XELOX + trastuzumab in patients with HER2‐positive advanced GC [10, 22]. The current findings provide real‐world clinical evidence to further support the feasibility of trastuzumab + XELOX as a treatment option in this patient population. In propensity score matching of cohorts I and II, trastuzumab improved survival in patients with HER2‐positive mGC compared with chemotherapy alone.
Compared with the randomized controlled ToGA trial, which was conducted internationally, the prospective EVIDENCE study conducted in China reports similar results. In the ToGA trial, the OS was improved in patients treated with trastuzumab + chemotherapy (13.8 months) compared with patients treated with chemotherapy alone (11.1 months) [12]. In the 84 Chinese patients included in the primary analysis of the ToGA trial, the median OS was 12.6 versus 9.7 months in the trastuzumab plus chemotherapy versus chemotherapy alone groups (HR, 0.72; 95% CI, 0.40–1.29), which was consistent with the whole population results [26]. This supports the use of trastuzumab in Chinese patients with mGC. The present EVIDENCE study reports a longer relative median OS (22.3 months) compared with previous studies for patients with HER2‐positive GC treated with trastuzumab plus chemotherapy (range, 13.8–21.0 months) [10, 12, 23, 24]. PFS was also relatively longer for the present trial (8.2 months) compared with the ToGA trial (6.7 months) [12]. This is likely attributable to the differences between a study done using real‐world data and a randomized controlled study such as ToGA [21]. Patients enrolled in the EVIDENCE trial are more likely to have a wider range of baseline levels (e.g., the ECOG PS was unknown for the majority of patients in the EVIDENCE trial) representing a more heterogeneous patient population compared with randomized controlled studies. Additionally, the treatment patterns may vary more widely as treating physicians are not restricted by a study protocol as in randomized controlled trials [21]. In fact, patients in the ToGA trial were restricted to two chemotherapy regimens (XEL + C or 5‐FU + C) [12], whereas those in the EVIDENCE trial were treated with a variety of chemotherapy regimens as mentioned above.
The difference in first‐line treatment RRs between cohort II (18.4%) and cohort IV (32.8%) is potentially of interest and may warrant further exploration. Prior to publication of findings from the ToGA trial, analysis of HER2 status was generally unavailable in clinical trials of chemotherapy; as a result, there were limited available data to evaluate and compare responses to chemotherapy in patients with HER2‐positive and HER2‐negative mGC. The results from the EVIDENCE study suggest that patients with HER2‐positive disease do not respond as well to chemotherapy alone, compared with patients with HER2‐negative mGC. However, no propensity score matching was performed for these two cohorts; furthermore, the noninterventional nature of this study and the response assessment used may affect the accuracy of the results. As such, any interpretation of these data should be made with caution.
Trastuzumab is currently the only targeted therapy approved for first‐line treatment in standard‐of‐care therapy for patients with mGC. Other targeted therapies have thus far failed to show superiority to chemotherapy alone as first‐line therapies in patients with mGC [27]. Exciting new treatment possibilities are emerging regarding the use of trastuzumab in combination with immune checkpoint inhibitors. Preclinical studies have shown that trastuzumab upregulates programmed death–ligand one in a transgenic mouse model of breast cancer [28]. Promising preclinical data have led to the initiation of several early‐phase clinical trials testing immune checkpoint inhibitor drugs in combination with trastuzumab [29], including a phase II study (PANACEA, NCT02129556) investigating pembrolizumab (anti–programmed death‐one [PD‐1]) in combination with trastuzumab in patients with metastatic breast cancer who have progressed on trastuzumab. Nivolumab (anti‐PD1) was recently approved in Japan for the treatment of patients with pretreated advanced GC and is recommended as third‐line therapy [30]. Clinical trials investigating trastuzumab and immune checkpoint inhibitor combination treatment approaches in patients with mGC are warranted.
To date, evidence for continuation of trastuzumab in patients who have become refractory to trastuzumab‐containing first‐line treatment is somewhat inconclusive, as reports of survival benefit are mixed [31, 32, 33, 34]. Further studies to clarify the benefit of continued/multiline trastuzumab treatment are warranted.
This study had several limitations. The observational study data and the proportion of nonevaluable patients for the effectiveness analysis had some missing values at postbaseline visits, which may confound the data. Regarding bias from multiline chemotherapy, given the differences in proportions of patients receiving second‐ or later‐line chemotherapy, this bias is not likely adequately adjusted by propensity score analysis. This study was conducted in China, which likely limits the significance of current findings to Chinese patients. Previous studies have addressed the benefits of trastuzumab globally [12]. However, studies specifically relevant to the Chinese patient population with mGC in a real‐world clinical setting were lacking prior to this report. Within the Chinese population, the birthplace and place of residence for most patients was East China, and they were predominately of Han ethnicity. Thus, the representativeness and generalizability of this study within the Chinese population may be limited. Finally, AE data collection in this observational study, as with any observational study, is more limited compared with a typical randomized controlled study.
Conclusion
The results of this prospective EVIDENCE study provide valuable information on the outcomes and systemic treatment of patients with HER2‐positive mGC in a real‐world clinical setting, which remains a major therapeutic challenge, particularly in China. The OS data for trastuzumab treatment as first‐line therapy for patients with HER2‐positive mGC reported in the present study are quite promising; these data add valuable real‐world evidence to previous clinical study data and may play a central role in improving care and outcomes for people with HER2‐positive mGC by supporting the positioning of trastuzumab as standard first‐line therapy for this patient population.
Author Contributions
Conception/design: Shukui Qin, Jiafu Ji, Lin Shen
Provision of study material or patients: Shukui Qin, Jiafu Ji, Rui‐hua Xu, Wei Wang, Yong Tang, Feng Bi, Jin Li, Kang Wang, Jian‐ming Xu, Qingxia Fan, Wuyun Su, Lin Shen
Collection and/or assembly of data: Shukui Qin, Jiafu Ji, Rui‐hua Xu, Wei Wang, Yong Tang, Feng Bi, Jin Li, Kang Wang, Jian‐ming Xu, Qingxia Fan, Wuyun Su, Lin Shen
Data analysis and interpretation: Shukui Qin, Jiafu Ji, Rui‐hua Xu, Wei Wang, Yong Tang, Feng Bi, Jin Li, Kang Wang, Jian‐ming Xu, Qingxia Fan, Wuyun Su, Lin Shen
Manuscript writing: Shukui Qin, Jiafu Ji, Lin Shen
Final approval of manuscript: Shukui Qin, Jiafu Ji, Rui‐hua Xu, Wei Wang, Yong Tang, Feng Bi, Jin Li, Kang Wang, Jian‐ming Xu, Qingxia Fan, Wuyun Su, Lin Shen
Disclosures
The authors indicated no financial relationships.
Supporting information
See http://www.TheOncologist.com for supplemental material available online.
Appendix S1. Figures.
TABLE S1 STROBE Statement—Checklist of items that should be included in reports of cohort studies
Acknowledgments
We thank Sarah Bubeck, Ph.D., of Edanz Pharma for providing medical writing support, which was funded by Shanghai Roche Pharmaceuticals Ltd. This study was sponsored by Shanghai Roche Pharmaceuticals Ltd. Interim results of this study were previously published online for the 2017 Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, U.S., June 2–6, 2017, and presented at the 20th annual Meeting of Chinese Society of Clinical Oncology, Xiamen, Fujian, China, September 26–30, 2017. Updated results of this study were presented at the 2019 Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, U.S., May 31 to June 4, 2019.
No part of this article may be reproduced, stored, or transmitted in any form or for any means without the prior permission in writing from the copyright holder. For information on purchasing reprints contact commercialreprints@wiley.com. For permission information contact permissions@wiley.com.
Disclosures of potential conflicts of interest may be found at the end of this article.
References
- 1.Bray F, Ferlay J, Soerjomataram I et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394–424. [DOI] [PubMed] [Google Scholar]
- 2.Chen W, Sun K, Zheng R et al. Cancer incidence and mortality in China, 2014. Chin J Cancer Res 2018;30:1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Takahashi T, Saikawa Y, Kitagawa Y. Gastric cancer: Current status of diagnosis and treatment. Cancers (Basel) 2013;5:48–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Orditura M, Galizia G, Sforza V et al. Treatment of gastric cancer. World J Gastroenterol 2014;20:1635–1649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Smyth EC, Verheij M, Allum W et al. Gastric cancer: EMSO Clinical Practice Guidelines for diagnosis, treatment and follow‐up. Ann Oncol 2016;27(suppl 5):v38–v49. [DOI] [PubMed] [Google Scholar]
- 6.Wagner AD, Grothe W, Haerting J et al. Chemotherapy in advanced gastric cancer: A systematic review and meta‐analysis based on aggregate data. J Clin Oncol 2006;24:2903–2909. [DOI] [PubMed] [Google Scholar]
- 7.GASTRIC (Global Advanced/Adjuvant Stomach Tumor Research International Collaboration) Group ; Oba K, Paoletti X et al. Role of chemotherapy for advanced/recurrent gastric cancer: An individual‐patient‐data meta‐analysis. Eur J Cancer 2013;49:1565–1577. [DOI] [PubMed] [Google Scholar]
- 8.Sun W, Yan L. Gastric cancer: Current and evolving treatment landscape. Chin J Cancer 2016;35:83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yan L.The journey of personalizing gastric cancer treatment. Chin J Cancer 2016;35:84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ryu MH, Yoo C, Kim JG et al. Multicenter phase II study of trastuzumab in combination with capecitabine and oxaliplatin for advanced gastric cancer. Eur J Cancer 2015;51:482–488. [DOI] [PubMed] [Google Scholar]
- 11.Gravalos C, Jimeno A. HER2 in gastric cancer: A new prognostic factor and a novel therapeutic target. Ann Oncol 2008;19:1523–1529. [DOI] [PubMed] [Google Scholar]
- 12.Bang YJ, Van Cutsem E, Feyereislova A et al. Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2‐positive advanced gastric or gastro‐oesophageal junction cancer (ToGA): A phase 3, open‐label, randomised controlled trial. Lancet 2010;376:687–697. [DOI] [PubMed] [Google Scholar]
- 13.Roukos DH. Targeting gastric cancer with trastuzumab: New clinical practice and innovative developments to over‐come resistance. Ann Surg Oncol 2010;17:14–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Gastric Cancer. Version 2.2018. Plymouth Meeting, PA: National Comprehensive Cancer Center, 2018. Available at http://www.nccn.org/professionals/physician_gls/pdf/gastric.pdf. Accessed March 6, 2019. [Google Scholar]
- 15.Romond EH, Perez EA, Bryant J et al. Trastuzumab plus adjuvant chemotherapy for operable HER2‐positive breast cancer. N Engl J Med 2005;353:1673–1684. [DOI] [PubMed] [Google Scholar]
- 16.Piccart‐Gebhart MJ, Procter M, Leyland‐Jones B et al. Trastuzumab after adjuvant chemotherapy in HER2‐positive breast cancer. N Engl J Med 2005;353:1659–1672. [DOI] [PubMed] [Google Scholar]
- 17.Slamon D, Eiermann W, Robert N et al. Adjuvant trastuzumab in HER2‐positive breast cancer. N Engl J Med 2011;365:1273–1283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Joensuu H, Kellokumpu‐Lehtinen PL, Bono P et al. Adjuvant docetaxel or vinorelbine with or without trastuzumab for breast cancer. N Engl J Med 2006;354:809–820. [DOI] [PubMed] [Google Scholar]
- 19.Dreyer NA, Garner S. Registries for robust evidence. JAMA 2009;302:790–791. [DOI] [PubMed] [Google Scholar]
- 20.Dreyer NA, Tunis SR, Berger M et al. Why observational studies should be among the tools used in comparative effectiveness research. Health Aff (Millwood) 2010;29:1818–1825. [DOI] [PubMed] [Google Scholar]
- 21.Kim HS, Lee S, Kim JH. Real‐world evidence versus randomized controlled trial: Clinical research based on electronic medical records. J Korean Med Sci 2018;33:e213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gong J, Liu T, Fan Q et al. Optimal regimen of trastuzumab in combination with oxaliplatin/capecitabine in first‐line treatment of HER2‐positive advanced gastric cancer (CGOG1001): A multicenter, phase II trial. BMC Cancer 2016;16:68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Meulendijks D, Beerepoot LV, Boot H et al. Trastuzumab and bevacizumab combined with docetaxel, oxaliplatin and capecitabine as first‐line treatment of advanced HER2‐positive gastric cancer: A multicenter phase II study. Invest New Drugs 2016;34:119–128. [DOI] [PubMed] [Google Scholar]
- 24.Roviello G, Petrioli R, Nardone V et al. Docetaxel, oxaliplatin, 5FU, and trastuzumab as first‐line therapy in patients with human epidermal receptor 2‐positive advanced gastric or gastroesophageal junction cancer: Preliminary results of a phase II study. Medicine (Baltimore) 2018;97:e10745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Gong J, Liu T, Fan Q et al. A multicenter, phase II study of trastuzumab plus capecitabine and oxaliplatin (XELOX) as first‐line chemotherapy for HER2‐positive advanced gastric cancer: Update results of efficacy and toxicity. J Clin Oncol 2014;32(suppl 3):102a. [Google Scholar]
- 26.Shen L, Xu JM, Feng FY et al. Trastuzumab in combination with chemotherapy versus chemotherapy alone for first‐line treatment of HER2‐positive advanced gastric or gastroesophageal junction cancer: A Phase III, multi‐center, randomized controlled trial, Chinese subreport [In Chinese]. Zhonghua Zhone Liu Za Zhi 2013;35:295–300. [DOI] [PubMed] [Google Scholar]
- 27.Apicella M, Corso S, Giordano S. Targeted therapies for gastric cancer: Failures and hopes from clinical trials. Oncotarget 2017;8:57654–57669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Chaganty BKR, Qiu S, Gest A et al. Trastuzumab upregulates PD‐L1 as a potential mechanism of trastuzumab resistance through engagement of immune effector cells and stimulation of IFNγ secretion. Cancer Lett 2018;430:47–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ayoub NM, Al‐Shami KM, Yaghan RJ.Immunotherapy for HER2‐positive breast cancer: Recent advances and combination therapeutic approaches. Breast Cancer (Dove Med Press) 2019;11:53–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Togasaki K, Sukawa Y, Kanai T et al. Clinical efficacy of immune checkpoint inhibitors in the treatment of unresectable advanced or recurrent gastric cancer: An evidence‐based review of therapies. Onco Targets Ther 2018;11:8239–8250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Narita Y, Kadowaki S, Masuishi T et al. Correlation between human epidermal growth factor receptor 2 expression level and efficacy of trastuzumab beyond progression in metastatic gastric cancer. Oncol Lett 2017;14:2545–2551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Li Q, Jiang H, Li H et al. Efficacy of trastuzumab beyond progression in HER2 positive advanced gastric cancer: A multicenter prospective observational cohort study. Oncotarget 2016;7:50656–50665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Palle J, Tougeron D, Pozet A et al. Trastuzumab beyond progression in patients with HER2‐positive advanced gastric adenocarcinoma: A multicenter AGEO study. Oncotarget 2017;8:101383–101393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Moser J, Solomon B, Nevala‐Plagemann CD et al. Survival of patients with metastatic HER2 positive gastroesophageal cancer treated with second‐line chemotherapy plus trastuzumab or ramucirumab after progression on frontline chemotherapy plus trastuzumab. J Clin Oncol 2019;37(suppl 4):69a. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
See http://www.TheOncologist.com for supplemental material available online.
Appendix S1. Figures.
TABLE S1 STROBE Statement—Checklist of items that should be included in reports of cohort studies