Real world evidence of adjuvant trastuzumab in HER2 positive early breast cancer

Adjuvant trastuzumab in HER2+ breast cancer reduces recurrence and mortality, and has been the standard treatment since 2006. The objective was to analyze health outcomes in the real world. Observational, retrospective study of patients with HER2+ breast cancer, stages I–III, treated with adjuvant trastuzumab in the past 15 years in only one center and for the first time in Spain. Survival was analyzed according to the number of cycles and cardiotoxicity. Two hundred and seventy-five HER2positive patients (18.60%) out of 1479 received adjuvant (73%) or neoadjuvant/adjuvant (26%) trastuzumab, concomitantly (90%) or sequentially (10%) with chemotherapy. The probability of overall and disease-free survival (OS and DFS) at 5 years was 0.93 (95% CI 0.89–0.96), and 0.88 (95% CI 0.83–0.92). The number of cases with a significant and asymptomatic decrease in ventricular ejection fraction and heart failure were 54 (19.64%) and 12 (4.36%), respectively. Sixty-eight patients (24.70%) received 16 or fewer cycles, especially those older than 65 (OR 0.371, 95% CI 0.152–0.903; p = 0.029) and with cardiotoxicity (OR 15.02, 95% CI 7.437–30.335; p < 0.001). The risk of cardiotoxicity was associated with having received radiotherapy (OR 0.0362, 95% CI 0.139–0.938; p = 0.037). Arterial hypertension (HR 0.361, 95% CI 0.151–0.863, p = 0.022), neoadjuvant treatment (HR 0.314, 95% CI 0.132–0.750, p = 0.009) and cardiotoxicity (HR 2.755, 95% CI 1.235–6.143, p = 0.013) maintained significant association with OS. Only neoadjuvant treatment maintained a significant association with DFS (HR 0.437, 95% CI 0.213–0.899, p = 0.024). The effectiveness of neoadjuvant and adjuvant trastuzumab can be considered comparable to those of clinical trials. In the real world, factors such as age, hypertension, radiotherapy, neoadjuvant treatment, and cardiotoxicity should be taken into consideration to optimize outcomes.


Association between variables and cardiotoxicity.
Of all the variables none was significantly associated with cardiotoxicity in the bivariate analysis, and only the way of administering chemotherapy and trastuzumab (concomitant or sequential) and radiotherapy treatment, approached statistical significance in the multivariate analysis (Table 3). For the cardiotoxicity as a dependent variable only radiotherapy treatment showed statistical significance in the multivariate model (OR 0.0362, 95% CI 0.139-0.938; p = 0.037).

Discussion
The results presented confirm the effectiveness of trastuzumab in the real world, with OS and DFS that can be considered comparable to those of clinical trials [6][7][8] . In the Early Breast Cancer Trialists'' Collaborative group (EBCTCG) meta-analysis 17 , with a median follow-up of 10.7 years, 26.60% of patients had recurrence of HER2 breast cancer and 19.70% died. For our patients, with a shorter follow-up, 6 years, 11.34% had recurrence of breast cancer and 9.44% died. The real-world effectiveness of trastuzumab is also demonstrated when it was administered with neoadjuvant intent, with pathologic complete response in 48.61% of cases, better or similar to that obtained in trials involving chemotherapy and trastuzumab [18][19][20] .
The way of measuring cardiac toxicity, which was different for each study, means that the results are not entirely comparable. Some 18.6% of participants in the treatment arm with doxorubicin and cyclophosphamide followed by docetaxel plus trastuzumab, in the BCIRG 006 trial (follow-up 65 months) 8 , 14% in the chemotherapy plus trastuzumab treatment arm of the NSABP-31 trial (follow-up 24 months) 21 , and 7.08% in the chemotherapy arm followed by 1 year of trastuzumab in the HERA trial (follow-up 12 months) 6,22 , had significant decreases in LVEF. The percentage of grade 3 and 4 heart failure was 4.5% in a meta-analysis of five trastuzumab trials 23 . Our cardiotoxicity results were an asymptomatic decline in LVEF in 19.64% and heart failure in 4.36%. Cardiotoxicity was the main reason for discontinuing trastuzumab, with 24.70% receiving 16 or fewer cycles.
In our study, only having received radiotherapy was associated with an increased risk of cardiotoxicity, without detecting, as in other studies 15 , any association with age, arterial hypertension or obesity. The risk of cardiotoxicity associated with radiation received by the heart, and administration of potentially cardiotoxic trastuzumab is unclear. It is important to highlight that in an analysis of 1503 patients treated with adjuvant radiation in the randomized NCCTG N9831 trial, no increased cardiotoxicity was observed, even in patients treated with concomitant trastuzumab and radiotherapy 22 . www.nature.com/scientificreports/ In an attempt to identify the variables associated with a lower number of trastuzumab cycles-being over 65 years of age, and suffering from cardiotoxicity showed significant odds ratios in the multivariable analysis, so we can confirm the association between not completing the 12 months of trastuzumab and age, and cardiotoxicity. Of the women studied, 17.8% were older than 65 years. In real-world studies, these older patients specifically are not well represented in clinical trials, and patients who are poorly represented experience worse outcomes than those who are well represented 23 . www.nature.com/scientificreports/ One of the consequences of trastuzumab cardiotoxicity is the obligation to discontinue treatment, causing patients to be unable to receive all 17-18 cycles in the 12 months of treatment. In our real world, although patients received a median of 17 cycles, the 68 (24.70%) who received 16 or fewer cycles did not have a worse OS than those who received more than 16 cycles. These results are at odds with the systematic review of trials on deescalation in the duration of adjuvant trastuzumab therapy 14 , where a shorter duration of adjuvant trastuzumab is associated with significantly worse DFS and OS. Other studies and meta-analyses 12,24 , however, have shown non-inferiority of 6 months compared to 12 months. These results should be interpreted with caution, as they are based primarily on the Persephone trial 12 , an older trial that employed chemotherapy-trastuzumab regimens considered nonstandard (taxane-free anthracyclines in 90% and sequential trastuzumab in 53% of cases), in a relatively low-risk population (59% of node-negative patients). www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ In the case of adjuvant treatment of early breast cancer with trastuzumab, short-and long-term cardiovascular side effects become crucially important in health outcomes 25 and this is what we observed also in our real-world study, as patients who did not develop cardiotoxicity, and who do not have hypertension, had better OS than those who did have heart failure or significant decline in LVEF and hypertension. Other authors 26 have already established the consensus that patients who do not complete adjuvant trastuzumab because of cardiotoxicity have a significantly increased risk of relapse and death from breast cancer. These findings also support 1 year of adjuvant trastuzumab 26 .
The use of trastuzumab with chemotherapy in the neoadjuvant setting results in a higher rate of complete disease responses, but the effects on DFS and OS are unclear compared to adjuvant treatment. One study supports early initiation of neoadjuvant trastuzumab, especially in patients with hormone-resistant and HER2-positive tumors 27 , so we cannot give a plausible explanation for our findings of worse OS and DFS in patients treated in the neoadjuvant setting.
This study was a retrospective observational evaluation of data intended to assess the effectiveness of adjuvant trastuzumab in the real world and is therefore subject to limitations. There may be unmeasured confounders unequally distributed among the groups. There may also be uncaptured patient comorbidities. Another limitation is the need for longer follow-up, as patients with hormone-sensitive breast cancer, even if they are HER2+, may have late recurrences.
There are no similar publications in Spain with detailed data on the effectiveness and cardiac toxicity of adjuvant trastuzumab in the real world and with an extended follow-up of 15 years. For this reason, the results presented here constitute valuable information with which other centers can make comparisons.
In conclusion, our results confirm the significance of the use of trastuzumab in the adjuvant treatment of early-stage breast cancer in the real world, with survival comparable to those of clinical trials. In the real world, factors such as age, hypertension, radiotherapy, neoadjuvant treatment and cardiotoxicity should be taken into consideration to optimize outcomes.

Methods
This is an observational, descriptive, retrospective, single-center, study on a medicinal product for human use. It was carried out at the Medical Oncology Department of a university hospital. All cases treated since the official approval of trastuzumab in 2006 were selected. Patients were identified from the Medical Oncology department's database, and data related to patients and treatment received were retrospectively collected from Medical Oncology's medical records. All experimental protocols were approved by the Cádiz Research Ethics Committee.
Patients with breast cancer undergoing surgical treatment with a pathology result showing HER2-positive infiltrating carcinoma were included. Those suitable for inclusion in the study had to have been treated with trastuzumab, either alone or with chemotherapy and/or hormone therapy, between 1 January 2006 and 31 December 2020, and followed up until 1 September 2021.
Female and male patients with HER2-positive infiltrating breast carcinoma (immunohistochemical score 3+ and 2+ with gene amplification), confirmed by histology, at disease stage I, II and III were included. Patients could have received adjuvant or neoadjuvant systemic treatment with hormone therapy, chemotherapy, or neither. Patients who received systemic neoadjuvant therapy were treated with chemotherapy plus trastuzumab, or plus trastuzumab and pertuzumab prior to surgery. These patients did not receive adjuvant chemotherapy but they did receive adjuvant trastuzumab with a goal of 12 months of treatment. Cases of carcinoma in situ, without infiltrating component and patients with metastases were excluded.
Age, sex, functional capacity (measured by the ECOG scale) 28 , menopausal status, comorbidity (measured by the Charlson scale) 29 , body mass index and arterial hypertension were the patient-related independent variables.
The independent variables related to the breast tumor were: tumor size at definitive surgery, disease stage (American Joint Committee on Cance r 8th edition), histologic grade, histologic type, extensive peritumoral vascular invasion, estrogen and progesterone receptor status (considered positive if immunohistochemical expression was equal to or greater than 1%), HER2 (positive if immunohistochemical score was 3+ or 2+ with gene amplification by in situ hybridization), Ki67 proliferative index, number of axillary nodes with metastases, number of total axillary nodes analyzed.
The independent variables related to treatment were: type of surgery, radiotherapy, type of adjuvant hormonal treatment, type of adjuvant chemotherapy, route of administration of trastuzumab and form of administration of chemotherapy-trastuzumab (concomitant or sequential). Weekly trastuzumab was administered intravenously at a starting dose of 4 mg/kg of body weight and maintenance dose of 2 mg/kg for 12 months. Trastuzumab was administered three times a week either intravenously at a starting dose of 8 mg/kg and maintenance dose of 6 mg/kg, or subcutaneously at a dose of 600 mg, both also for 12 months.
The independent variables related to safety were cardiotoxicity (subclassified into symptomatic heart failure and asymptomatic LVEF decline), trastuzumab discontinuation (patients were classified into those who had completed the 17-18 three-weekly cycles of trastuzumab treatment and those who have not completed treatment with ≤ 16 cycles), and the reason trastuzumab was discontinued. In case of weekly trastuzumab administration, 3 weeks was considered to be 1 cycle. A baseline LVEF of at least 50% was required to initiate trastuzumab treatment and the test was repeated every 3 months during the year of treatment. Trastuzumab was discontinued if symptomatic heart failure or an asymptomatic decrease in LVEF of 15 points, or 10 points if this put the level below 50%, occurred. In cases of asymptomatic LVEF decline, angiotensin-converting enzyme inhibitors and beta-blockers were prescribed at low doses and repeated on a month-by-month basis. If levels recovered, trastuzumab was reintroduced.
The efficacy dependent variables analyzed were overall survival and disease-free survival 30 . www.nature.com/scientificreports/ A descriptive analysis of the variables (absolute and relative frequency, mean, median and standard deviation) was made. For the comparison of qualitative variables, the chi-square test with Fisher's correction was applied and, for quantitative variables, the Student's t-test.
The influence of independent variables on the number of trastuzumab cycles administered was measured using binary logistic regression, using odds ratio (OR) for risk estimation with a corresponding CI of 95%.
Logistic regression analysis with an independent variable was performed to calculate the odds ratio (OR) of each variable in relation to trastuzumab cardiotoxicity.
The Kaplan Meier method and the Log-Rank test for the comparison of survival curves were used to calculate survival. Cox regression with an independent variable was used to calculate the hazard ratio (HR).
Landmark analysis was carried out to monitor immortal time bias in the survival analysis among subgroups with ≤ 16 cycles and 17-18 cycles of trastuzumab. The landmark time is established as the duration of time that it would take to complete 16 cycles. Any patients experiencing event (DFS or OS) prior to this landmark time were dropped from the analytical sample. The groups (≤ 16 cycles versus 17-18 cycles) were defined, using only patients in the group without events prior to the landmark time. The survival analysis was conducted as usual, with t0 set equal to the landmark time.
The SPSS program, version 21, was used for statistical analysis of the data. In the statistical analysis, p < 0.05 was considered to indicate statistical significance.
Ethical considerations. The work described was carried out in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments on human subjects. All experimental protocols were approved by the Cádiz Research Ethics Committee (code 152.21), which authorized the waiver of informed consent.