Adjuvant Trastuzumab Without Chemotherapy For Treating Early HER2-Positive Breast Cancer In Older Patients: A Cohort Study Accompanying With The RESPECT Trial


 Purpose To gauge the effects of treatment practices on prognosis for all older patients with HER2-positive breast cancer, particularly to determine whether adjuvant trastuzumab alone can offer benefit over no adjuvant therapy. This report accompanies the RESPECT study, a randomized-controlled trial (RCT) comparing trastuzumab monotherapy with trastuzumab-plus-chemotherapy for early HER2-positive breast cancer.Patients and methods Patients who declined the RCT were treated based on the physician’s discretion. We studied the (1) trastuzumab-plus-chemotherapy group, (2) trastuzumab-monotherapy group, and (3) non-trastuzumab group (no therapy or anticancer therapy without trastuzumab). The primary endpoint was disease-free survival (DFS), which was compared using the propensity-score method.Results We enrolled 398 eligible patients, aged over 70 years, with HER2-positive invasive breast cancer, of whom 275 (69%) were in the RCT, and 123 (31%) were in this cohort group. The median age was 74.5 years. Among cohort group treatment categories were as follows: (1) trastuzumab-plus-chemotherapy group (n = 36, 30%), (2) trastuzumab-monotherapy group (n = 52, 43%), and (3) non-trastuzumab group (n = 32, 27%). A total of 73% of patients received trastuzumab-containing regimens, with or without chemotherapy. The 3-year DFS was 92.3% in the trastuzumab-plus-chemotherapy group, 89.2% in the trastuzumab-monotherapy group, and 82.5% in the non-trastuzumab group. DFS in the non-trastuzumab group was lower than in the trastuzumab-plus-chemotherapy and trastuzumab-monotherapy groups (propensity-adjusted HR: 3.29; 95% CI: 1.15–9.39; P = 0.026). The relapse-free survival in the non-trastuzumab group was lower than in the trastuzumab-plus-chemotherapy and trastuzumab monotherapy groups (propensity-adjusted HR = 7.80; 95% CI: 2.32–26.2, P < 0.0001). Chemotherapy with trastuzumab or trastuzumab monotherapy did not affect health-related quality of life (HRQoL) at 36 months.Conclusions Trastuzumab-treated patients had better prognoses than patients not treated with trastuzumab without deterioration of HRQoL. Thus, trastuzumab monotherapy can be considered for patients who reject chemotherapy.Trial registration number The protocol was registered on the website of the University Hospital Medical Information Network (UMIN), Japan (protocol ID: UMIN 000028476).


Introduction
This study accompanied the RESPECT study [1], which is a randomized controlled trial (RCT) designed to compare the value of trastuzumab monotherapy with the standard combination treatment with the value of chemotherapy in patients over 70 years, with human epidermal growth factor receptor type 2 (HER2)positive invasive breast cancer, who received curative surgery. This study aimed to determine the overall prognosis of older patients with HER2-positive breast cancer, including patients who did not agree to participate in the RCT despite meeting the eligibility criteria, and to compare the e cacies of trastuzumab monotherapy and trastuzumab-plus-chemotherapy. Particularly, we aimed to determine whether adjuvant trastuzumab alone can offer bene t over no adjuvant therapy. Before starting this RCT, we questioned whether acquiring consent to participate in this RCT might be di cult in older patients, because of the possibility of emphasizing treatment in accordance with the patient's wishes, considering the potential adverse events (AEs) of chemotherapy. It is currently unknown whether adjuvant trastuzumab therapy alone can offer a bene t over no adjuvant therapy. Although we sought to directly compare trastuzumab monotherapy with no treatment in older patients, we were concerned that such a study would not be feasible because some patients might refuse to participate in an arm without trastuzumab, despite having HER2-positive disease. In addition, only healthy patients can participate in the RCT. Thus, we designed a non-interventional cohort study in addition to the RCT to gauge the effects of treatment practices on prognosis for all older patients with HER2-positive breast cancer.

Patients
The trial protocol is described within the full text of this article. We recruited patients, aged 70-80 years old, with HER2-positive invasive breast cancer who underwent curative surgery. The patient-inclusion criteria were as follows: patients with invasive breast cancer histologically diagnosed as HER2-positive breast cancer, who underwent curative surgery for stage I (pathological tumor size > 0.5 cm), IIA, IIB, or IIIA disease. HER2-positivity was de ned by the ASCO/CAP guidelines [2], which lay down the following criteria: immunohistochemical staining of 3+ (uniform, intense membrane staining of > 30% of invasive tumor cells) and a uorescence-in situ hybridization (FISH) result of more than six HER2 gene copies per nucleus or a FISH ratio (HER2 gene signal: chromosome 17 signal) of more than 2.2. Other key eligibility criteria were as follows: a baseline left ventricular-ejection fraction of ≥55% (measured by echocardiography) within 4 weeks of registration, an Eastern Cooperative Oncology Group performance status (PS) score of 0 or 1, and su cient organ function that meets the prescribed criteria in laboratory tests performed within four weeks of registration. The key exclusion criteria were as follows: the presence of active multiple primary cancer (synchronous multiple primary cancer and invasive cancer of other organs); ≥4 histological axillary lymph node metastases; no histological evaluation of axillary lymph nodes; a histologically con rmed positive margin found during breast-conservation surgery; a history of drug-related allergy that could hinder planned treatment; any history of or complication following cardiac disorders; a history of congestive heart failure; cardiac infarction; complications requiring treatment, such as ischemic cardiac disorder, arrhythmia, valvular-heart disease, poorly-controlled hypertension, and poorly controlled diabetes; di culty in regularly attending a medical institution due to a deterioration in the ability to perform the activities of daily living; di culty participating in the trial because of a psychiatric disorder or psychiatric symptoms; or ineligibility for enrollment according to the decision of an investigator based on contacts and locations.

Trial design and oversight
In general, the physicians involved in this study explained and encouraged patients to participate in the RCT. Patients who did not agree to participate in the RCT (despite meeting the eligibility criteria and being given an explanation of the RCT by the investigator) were included in the cohort study with written informed consent. Patients who consented to participate in the RCT were randomly assigned at a 1:1 ratio to the trastuzumab monotherapy group or trastuzumab-plus-chemotherapy group [1]. In this cohort study, treatment was chosen based on the discretion of the treating physician and the patients' wishes without intervention, and patients were prospectively surveyed based on routine medical records. The cohort group was divided into 1) the trastuzumab-plus-chemotherapy group, 2) the trastuzumab-monotherapy group, and 3) a group that received no therapy at all or received any anticancer therapy without trastuzumab (the non-trastuzumab group). The purpose of this prospective cohort study was to assess the overall effect of adjuvant therapy on HER2-positive primary breast cancer in older patients (≥ 70 years) and to investigate the e cacy and safety of trastuzumab-pluschemotherapy, trastuzumab monotherapy, and non-trastuzumab treatment.
This study was reviewed and approved by independent ethics committees and institutional review boards. The study conformed with the Declaration of Helsinki and the "Ethical Guidelines for Clinical Research" guidelines of the Ministry of Health, Labor and Welfare. Written informed consent was obtained from all patients.

End points
The primary endpoint of the cohort study was disease-free survival (DFS), and the secondary endpoints were overall survival (OS), relapse-free survival (RFS), AEs, and health-related quality of life (HRQoL). DFS was de ned by the occurrence of any of the following: a diagnosis of recurrence after breastconservation therapy, local (ipsilateral chest wall) recurrence, regional lymph node recurrence or distant organ metastasis; a diagnosis of metachronous breast cancer or secondary cancer (not including cutaneous basal cell carcinoma, squamous cell cancer, or endometrial intraepithelial carcinoma); and all deaths (regardless of cause). RFS was de ned by the occurrence of any of the following: any local recurrence (including local recurrence after breast-conservation therapy), regional lymph node recurrence or distant organ metastasis (not including metachronous breast cancer or secondary cancer), and death (regardless of cause). HRQoL was assessed using the Functional Assessment of Cancer Therapy-general (FACT-G) scale [3].

Assessment
In the cohort group, patients were surveyed using routine medical records without interventions, such as prospective treatment and testing. Tumor grades and types of AEs were determined according to Common Terminology Criteria for Adverse Events v3.0. The survival data and AEs were reviewed every year, beginning from the time of rst enrollment to the end of the study. HRQoL of the study population was assessed at registration, and after 36 months.

Statistical analysis
The target sample size for the cohort group was set to 200-400 patients. The survey of the clinical trial group indicated the presence of approximately 200 HER2-positive patients (aged ≥ 70 years) each year, who could potentially participate in either the RCT or the cohort group. Assuming that 40-50% of patients would provide consent for participating in the RCT and that approximately 70% would provide informed consent for the cohort study, we expected that approximately 50-65 patients would be included in the observational studies each year. Thus, it was considered possible to include 200 patients in the RCT over 4 years. The DFS was set as the primary endpoint, as was done for the RCT of the RESPECT study [1].
The DFS and other endpoints among trastuzumab-plus-chemotherapy, trastuzumab-monotherapy, and non-trastuzumab groups were compared using propensity score-based covariate adjustments by the Cox regression model. The propensity score was estimated for each participant using a multinomial regression model, based on the age (70-75 versus 76-80 years), hormone receptor status (positive versus negative), pathologic nodal status (positive versus negative), and PS (0 or 1) in the model. Group comparisons of FACT-G total and subdomain scores were performed using an Analysis of Co-Variance (ANCOVA). This model used scores at 36 months as an outcome, and the adjusted covariates were scored at baseline in addition to the same factors in the survival analysis (i.e., age, hormone receptor status, pathologic nodal status, and PS). The estimated score difference at 36 months from trastuzumabplus-chemotherapy (as a reference) in the trastuzumab-monotherapy group and the non-trastuzumab group, 95% con dence interval (CI), and p-value were calculated.
All collected data were analyzed using SAS ® version 9.4 (SAS Institute, Inc). A p-value of < 0.05 was considered to re ect a statistically signi cant difference. The end points, assessments, and statistical analyses are described in detail in the protocol.

Patients
We enrolled 398 eligible patients, aged over 70 years, with HER2-positive invasive breast cancer, from 114 institutions, between October 2009 and October 2014, of whom 275 (69%) were in the RCT group and 123 (31%) were in the cohort group. The CONSORT diagram is presented in Figure 1. All 275 patients who were randomly assigned to the RCT group were registered for full-set analysis: trastuzumab monotherapy (n = 137), and trastuzumab-plus-chemotherapy (n = 138). In the cohort group, three patients (2.4%) were excluded because all e cacy data was missing, leaving 120 patients for a full-set analysis; the treatment categories were as follows: 1) the trastuzumab-plus-chemotherapy group (n = 36, 30%), 2) the trastuzumab-monotherapy group (n = 52, 43%), and 3) the non-trastuzumab group (n = 32, 27%). A total of 73% of patients received trastuzumab-containing regimens, with or without chemotherapy. The patients' characteristics of the full analysis set are shown in Table 1 (n = 395). P values were assessed by chi squared test. The mean ages of the patients in the RCT and cohort groups at entry were 73.9 and 74.6, respectively, and the numbers of patients in the categories (<75 or ≥75) were different (P = 0.03). The groups had no differences in terms of the cancer stage, estrogen receptor (ER) and/or progesterone receptor (PgR) positivity, surgical procedure, lymph node metastasis, and co-morbidities (hypertension, diabetes mellitus, osteoporosis, or hyperlipidemia). The characteristics of the patients in the cohort group are shown in Table 2, according to the treatment options (n = 120). P values were assessed by chi squared test. Among the three subgroups in the cohort group, ER and/or PgR positivity were higher in the non-trastuzumab group (81.3% versus 48.1% in the trastuzumab group, 47.2% in the trastuzumab-pluschemotherapy group; P = 0.005). No differences existed among the groups in terms of the age category, stage, surgical procedure, lymph node metastasis, or co-morbidities.

DFS, RFS, and OS
The data cut-off date was October 31, 2017. The median follow-up time was 3.2 years (range: 0.9-7.0 years) in the cohort group. The details of DFS events are listed in Appendix Table A1. The DFS at 3 years was 96.7% in the trastuzumab-plus-chemotherapy group, 89.2% in the trastuzumab-monotherapy group, and 82.5% in the non-trastuzumab group (Figure 2). In the non-trastuzumab group, 26 of 32 patients (81.3%) were ER-positive; hormone therapy was initiated for 25 patients (78.1%), and breast irradiation was performed for four patients (12.5%). The DFS of patients in the non-trastuzumab group was lower than that of patients in the trastuzumab-plus-chemotherapy and trastuzumab monotherapy groups (propensity-adjusted HR = 3.29; 95% CI: 1.15-9.39, P = 0.026). DFS of the nontrastuzumab group also showed a worse prognosis than that of the trastuzumab-monotherapy group (propensity-adjusted HR: 2.15; 95% CI: 1.20-3.93, P = 0.012). The RFS of patients in the non-trastuzumab group was lower than that of patients in the trastuzumab-plus-chemotherapy and trastuzumab monotherapy groups (propensity-adjusted HR = 7.80; 95% CI: 2.32-26.2, P < 0.0001) (Figure 3). The OS of patients in the non-trastuzumab group was marginally lower than that of patients in the trastuzumabplus-chemotherapy and trastuzumab monotherapy groups (propensity-adjusted HR = 3.44; 95% CI: 0.75-15.67, P = 0.11) ( Figure A1).

Safety
Patients who registered for the cohort group (n = 120) were included in the safety analysis. The incidences of grade 3 or 4 non-hematological AEs were higher in the RCT group than in the cohort group (18.5% versus 10.8%, P = 0.05). All serious AEs resolved. Common AEs are listed in Table A2.
HRQoL Mean scores and 95% CI for the FACT-G total and sub-domain at each survey point are presented in Appendix Table A3 and Figure 4. ANCOVA showed that there were no signi cant differences in FACT-G total score after 36 months between the trastuzumab-plus-chemotherapy and trastuzumab monotherapy groups, and the trastuzumab-plus-chemotherapy and non-trastuzumab groups (Appendix Table A4). The only difference between the groups was that the social and family well-being (SFWB) score at 36 months between the trastuzumab-plus-chemotherapy group and the trastuzumab monotherapy group (estimated value = -5.5, P = 0.036), and SFWB score of trastuzumab-plus-chemotherapy group at 36 months was better than that of trastuzumab monotherapy group.

Discussion
The RESPECT study is the rst randomized adjuvant trial comparing trastuzumab monotherapy with trastuzumab-plus-chemotherapy for patients with HER2-positive breast cancer [1]. The RESPECT study was accompanied by a cohort study for patients who refused to participate in the RCT. By adding the cohort group to the RCT, we could evaluate the overall e cacy of adjuvant therapy for HER2-positive breast cancer in 398 patients over 70 years of age in detail, which enabled us to determine the prognoses of patients who did not receive trastuzumab prospectively, despite meeting the criteria for the RCT. We found that the DFS of the non-trastuzumab group was signi cantly lower than that of the trastuzumabplus-chemotherapy and the trastuzumab-monotherapy groups. Trastuzumab with chemotherapy has been approved as a standard adjuvant therapy for HER2-positive primary breast cancer based on previous studies that compared chemotherapy with trastuzumab-plus-chemotherapy [4][5][6][7]. Since 2005, no data have been generated through clinical trials regarding (i) trastuzumab without chemotherapy, and (ii) no trastuzumab treatment, because all patients receive trastuzumab.
Here, for the rst time, we added an implication to this issue as prospective data with propensityadjustment analysis in a non-interventional cohort study group. In a previous study, trastuzumab-pluspertuzumab was tested in a neoadjuvant setting as a treatment regimen without chemotherapy [8], but chemotherapy was administered after surgery, and the study lacked a no-treatment arm without anti-HER2 therapy. Recent retrospective data obtained using the National Cancer Database revealed no signi cant difference in survival by administering HER2 to patients who did not receive chemotherapy [9].
Data from a large observational study suggested that trastuzumab-plus-chemotherapy should remain the preferred option for all patients indicated for adjuvant treatment, and that a low proportion of patients need an alternative treatment approach, either because of contraindications or the patient's preference. Trastuzumab monotherapy might be a reasonable option [10], but to our knowledge, no prospective data exist to suggest that prospective adjuvant trastuzumab alone can offer a bene t over no adjuvant therapy.
The RESPECT study was designed to investigate whether older patients with HER2-positive breast cancer should undergo chemotherapy; however, before starting the trial, we were concerned that the consent rate for participation might be low. We expected it to be 40-50%, although almost 70% of the patients provided informed consent to participate in the RCT. The RCT employs a standard method for determining a standard therapy, but it remains controversial as to whether patients would bene t from participating in the RCT. Some ndings showed improved outcomes in trial participants as compared to those in non-participants, including those with breast cancer [11][12][13] whereas other ndings revealed no improvement in outcomes [14,15]. A systematic review indicated that insu cient data were available to conclude positive trial outcomes [16]. Notably, the rate of ER-positivity was higher in the cohort group, suggesting that ER expression in patients with HER2-positive breast cancer may have in uenced their decision to participate in the study or to choose hormonal therapy as a treatment option. In the nontrastuzumab group, ER-positivity was 81.3%, and the patients only received hormonal therapy, although it was associated with a worse prognosis compared to the chemotherapy-plus-trastuzumab, and trastuzumab groups. Older patients are at an increased risk for severe chemotherapy-induced toxicity [17][18][19]. Regarding the safety of trastuzumab in older patients, the results of a large observational study indicated that the risk of cardiac function toxicity was 5.7% [20] and that it was associated with age [20,21], although it remained manageable [20], and the risks associated with trastuzumab were outweighed by the bene ts [20,22]. In terms of the balance between bene t and harm, we recommend trastuzumab monotherapy if the patients do not receive chemotherapy, based on our current ndings. Besides the incidence of AEs, HRQoL is also important, because chemotherapy causes signi cant deterioration of HRQoL in older patients [23,24]. We observed a clinically signi cant HRQoL rate of deterioration between 2 months and 1 year into the RCT, which recovered after 3 years [1,25]. As a result of the QoL evaluation in this cohort study, chemotherapy with trastuzumab or trastuzumab monotherapy as postoperative adjuvant therapy did not affect global QoL at 36 months. We also observed the impact of chemotherapy on cognitive functioning in the RCT [26], the information would be important to share decision making between clinicians and patients.
Comprehensive geriatric-assessment (CGA) screening tools for older patients can be useful for predicting severe AE [27], and it is important to intensify supportive care and develop modi ed treatment regimens in vulnerable patients who may subsequently experience greater toxicity [28]. For the RCT group, we developed a CGA screening tool at baseline and at each point to analyze the relationship between AEs and DFS.
This study has a few limitations. Although 275 patients in the RCT group and 120 patients in the cohort group were treated and assessed, no de nitive conclusions regarding trastuzumab without chemotherapy (compared with trastuzumab-plus-chemotherapy) can be made, because there were fewer events than expected because > 80% of the patients enrolled had stage I or stage IIA breast cancer. Even in the nontrastuzumab arm of the cohort group, the 3-year DFS was over 80%. Although more patients were needed for a higher number of events, it was di cult to complete enrollment, because of the low number of older HER2-positive patients and disease heterogeneity [29]. We could have extended the follow-up period to detect more events, but it was assumed that non-breast cancer deaths as well as recurrences would accumulate, because eight years passed after the rst patient was enrolled. However, a longer follow-up period is needed to shed light on patient prognosis.
In conclusion, we found that patients who received any trastuzumab-containing regimen, even trastuzumab monotherapy, had a better prognosis than those who were not treated with trastuzumab, without deterioration of HRQOL. Although trastuzumab-plus-chemotherapy remains a standard of care, trastuzumab monotherapy could be considered for selected older patients, even if the patients are not willing to receive chemotherapy.

Declarations Acknowledgments
We acknowledge Toru Watanabe and Yasuo Ohashi for valuable advice on the research plan. We thank the patients who participated in this trial and their families and caregivers, all investigators involved in this study, and all members of the independent data-monitoring committee for their contributions to the study. Data availability The datasets analyzed during the current study are available from the corresponding author on reasonable request.
Code availability The codes used to analyze the datasets during the current study are available from the corresponding author on reasonable request.
Ethical approval The trial protocol was approved by the institutional review boards of all participating institutions.
Informed consent All patients provided written informed consent. Figure 1 CONSORT diagram for the RESPECT study. Patients who provided informed consent to participate in the randomized controlled trial (RCT) were assigned, at a 1:1 ratio, to the trastuzumab monotherapy and trastuzumab-plus-chemotherapy groups. Patients who met the eligibility criteria but did not agree to participate in the RCT were included in the cohort group with written informed consent. In the RCT group, treatment was selected for each patient based on the discretion of the treating physician and the patient's wishes, without intervention.

Figure 2
Kaplan-Meier estimates of disease-free survival (DFS). The DFS at 3 years was 96.7% in the trastuzumab-plus-chemotherapy group, 89.2% in the trastuzumab-monotherapy group, and 82.5% in the non-trastuzumab group. The DFS period of the non-trastuzumab group was lower than that of the trastuzumab-plus-chemotherapy and the trastuzumab monotherapy groups (propensity-adjusted HR: 3.29; 95% CI: 1.15-9.39; P = 0.026). The DFS in the non-trastuzumab group also showed a worse prognosis compared with the trastuzumab monotherapy group (propensity-adjusted HR: 2.15; 95% CI: 1.20-3.93; P = 0.012). Tick marks indicate censored data. Kaplan-Meier estimates of relapse-free survival (RFS). The RFS of patients in the non-trastuzumab group was lower than that of patients in the trastuzumab-plus-chemotherapy and trastuzumab monotherapy groups (propensity-adjusted HR = 7.80; 95% CI: 2.32-26.2, P < 0.0001). Tick marks indicate censored data.

Figure 4
Means and 95%CI of FACT-G scores at each survey point. Mean value and 95% con dential interval (95%CI) of A) Functional Assessment of Cancer Therapy-general (FACT-G) total, B) physical well-being (PWB), C) social and family well-being (SFWB), D) emotional well-being (EWB) and E) functional wellbeing (FWB) scores at baseline, and after 36 months in each group.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. AppendixTable.docx