Introduction

Systemic treatments targeting the human epidermal growth factor receptor 2 (HER2) have significantly improved outcomes for patients with HER2-positive breast cancer and have altered its natural course [1]. These have also led to meaningful improvements in survival for advanced/metastatic disease [2].

HER2-directed agents including monoclonal antibodies such as trastuzumab and pertuzumab, drug-antibody conjugates including ado-trastuzumab emtansine (T-DM1) and tyrosine kinase inhibitors such as lapatinib and neratinib are now available for routine use and more agents are currently being investigated [3]. In the first-line setting along with chemotherapy, the overall response rate is 80% on dual anti-HER2 blockade and 69% on trastuzumab, with a median duration of response of 20.2 months and 12.5 months, respectively, and a complete remission rate of 5.5% and 4.2%, respectively [4, 5]. Response to anti-HER2 treatment is confirmed as frequently durable also in population-based studies [6, 7].

We reviewed and updated a database of patients on first-line anti-HER2 treatment at our Institution [8] and selected those in remission for more than 1 year to assess its long-term efficacy and cardiac safety and we made a preliminary assessment of patients electively stopping treatment following prolonged complete remission.

Methods

We retrospectively identified and reviewed the medical records of all patients with locally advanced or metastatic breast cancer who received first-line anti-HER2 treatment in the palliative setting from 01/10/2001 to 30/11/2016 at The Royal Marsden NHS Foundation Trust for at least 1 year.

Patients’ characteristics (including age and performance status), early and advanced stage tumour characteristics (including stage, grade, histology, receptor status) and treatment characteristics in the curative and palliative setting (including chemotherapy regimen, anti-HER2 regimen, number of cycles administered, endocrine treatment, dose reductions and delays, early discontinuation, use of growth factors, enrolment within clinical trials, treatment for central nervous system [CNS] metastatic disease, bone therapy, outcome of treatments in the curative setting) were extracted from our electronic medical records, along with data on clinical and radiological response. Based on international guidelines [9, 10], oestrogen receptor (ER) and progesterone receptor (PgR) were defined as negative in case of Allred score ≤ 2, whereas HER2 was defined as negative in case of score ≤ 1 on immunohistochemistry or single-probe average HER2 copy number < 4.0 signals/cell or dual-probe HER2/CEP17 ratio < 2.0 with an average HER2 copy number < 4.0 signals/cell on in situ hybridization.

According to the Food and Drug Administration definitions [11], progression-free survival (PFS) was calculated as the time from commencement of treatment until disease progression. Overall survival (OS) was calculated as time from commencement of treatment until death from any cause. Best responses were also recorded. Overall response rate (ORR) was calculated as the proportion of patients achieving partial response (PR) or complete response (CR) on systemic therapy; clinical benefit rate (CBR) was calculated as the proportion of patients achieving stable disease (SD), PR or CR. This analysis was approved as a service evaluation by The Royal Marsden Hospital/Institute of Cancer Research Committee for Clinical Research.

Descriptive analysis method was used to summarize the data using counts and percentages for categorical variables and for the continuous non-normal variables using median and range or interquartile range. Chi squared and Fishers exact test were used to compare outcomes in different patient groups. The Kaplan–Meier method was utilized for the calculation of OS from date of treatment to death or last follow-up date. The Kaplan–Meier method was also used for the calculation of PFS from treatment to disease recurrence or death; progression-free and lost to follow-up patients were censored at last follow-up date. Median time to event was reported with 95% confidence interval and patient groups were compared using log-rank test.

Results

Two hundred and eight patients with HER2-positive advanced or metastatic breast cancer treated with first-line anti-HER2 treatment for at least 1 year were identified from the hospital database.

Baseline characteristics are reported in Tables 1 and 2. The median age at start of treatment was 54 years. One hundred and forty-three patients (68.7%) had a metastatic disease biopsy and 80 patients (55.9%) had ER-positive cancer on metastatic specimen. Seventy-nine patients (38.0%) were diagnosed with de novo metastatic disease, whereas 129 patients (62.0%) had previous early breast cancer. Sixty-six patients (50.8%) with relapsed disease had received chemotherapy in the adjuvant setting and 34 (26.1%) in the neoadjuvant setting. Fifty patients (54.4%) had already received trastuzumab in the curative setting but only one patient had received pertuzumab.

Table 1 Baseline characteristics
Full size table
Table 2 First-line treatment characteristics
Full size table

Eastern Cooperative Oncology Group (ECOG) performance status was 0–1 in 202 patients (97.1%). Metastatic disease involved the bones in 105 patients (50.5%), the lungs in 57 (27.4%), the liver in 91 (43.7%) and the brain in 22 (10.6%) before commencing anti-HER2 therapy. One hundred and seventy-two patients (82.7%) received chemotherapy along with anti-HER2 therapy, which included pertuzumab in 69 patients (33.2%) and trastuzumab in 202 patients (97.1%). Chemotherapy agents included docetaxel in 77 patients (35.6%), paclitaxel in 59 patients (28.4%), capecitabine in 13 patients (6.2%) and vinorelbine in 10 patients (4.8%). Two patients received single-agent T-DM1 and three patients had lapatinib alone. In 23 patients (11.1%), trastuzumab was given with endocrine treatment alone and in 7 patients it was given as a single agent. Ninety-nine patients (47.6%) received maintenance endocrine therapy following induction with chemotherapy. Bisphosphonates were given in 102 patients (49.0%). While on first-line therapy, 23 patients (11.1%) underwent whole brain radiotherapy, ten (4.8%) brain surgery, 15 (7.2%) stereotactic radiotherapy and one patient received intrathecal chemotherapy.

As shown in Table 3, in the overall cohort the median PFS was 39.5 months (95% CI 31.0–44.0) (Fig. 1). 52.5% of patients remained in remission at 24 months of follow-up and 29.4% at 5 years. The median OS was 81.0 months (95% CI 58.0–90.0) (Fig. 2), with 59.2% of patients alive at 5 years. Responses are presented in Table 4. 44 patients (21.1%) achieved complete remission, 138 (66.3%) partial response and 26 (12.5%) stable disease as best response on systemic treatment. In the overall cohort, ORR was 87.5%. Disease progression occurred in 108 patients (51.9%) during follow-up.

Table 3 Survival analysis
Full size table
Fig. 1
figure 1

Kaplan–Meier curve for progression-free survival

Full size image
Fig. 2
figure 2

Kaplan–Meier curve for overall survival

Full size image
Table 4 Responses
Full size table

As shown in Table 5, cardiac toxicity was reported in 10 patients (4.8%). No deaths related to trastuzumab were observed. Out of 177 patients receiving concurrent chemotherapy, fifty-one patients (28.8%) required dose interruptions and 54 (30.5%) dose reductions. Forty-three patients (24.3%) received granulocyte-colony stimulating factor support.

Table 5 Safety
Full size table

Fifteen patients (7.2%) had to stop the anti-HER2 treatment because of toxicity, including cardiac toxicity in six cases, hypersensitivity in one case and owing to coincidental multiple spinal surgeries requiring prolonged hospital admissions in another centre for severe symptomatic spondylolisthesis in another case.

In seven patients, the treatment was electively discontinued in the context of prolonged complete remission after 17, 41, 59, 66, 78, 85 and 87 months and at the time of updated data collection in June 2019 these patients were all still alive and in continuing remission for 79, 90, 92, 100, 123, 130 and 171 months. These seven cases are summarized in Table 6. On the other hand, of 140 patients in remission for more than 17 months, 45 (32.1%) eventually experienced disease progression.

Table 6 Summary of the seven cases who discontinued anti-HER2 therapy electively including patient and disease characteristic at metastatic presentation and systemic treatments given
Full size table

Discussion

Treatments targeting HER2 have become important agents in the management of metastatic HER2- positive breast cancer and have altered significantly the natural course of this disease. [1] Patients may have dramatic responses to anti-HER2 treatment and achieve prolonged remissions. Our analysis adds to the findings previously reported by Yeo et al. [8] and contributes to the limited evidence on outcomes in long-term responders to anti-HER2 treatments. Our findings are consistent with a previously published series of long-term responders enrolled in the HER-OS registry which documented a progression-free rate of 47.1% at 5 years and a median time to progression of 4.5 years in a population of 268 patients free from disease progression for more than 2 years [7]. Our cohort had a similar age and performance status and a similar proportion of ER-positive cases. We included a higher proportion of patients with liver or CNS involvement which may explain the different efficacy outcomes in our series. Nevertheless, our findings support an active and potentially more effective treatment of CNS disease in HER2-positive breast cancer patients, especially in patients achieving a prolonged response to systemic therapy.

The median PFS in our analysis is almost twice as long as the median PFS documented in the experimental arm of the CLEOPATRA study. [4, 5] However, only one-third of our patients received pertuzumab, which became available in the United Kingdom in 2012 and publicly funded in 2013. Moreover, a higher proportion of patients in our series had received trastuzumab in the curative setting. Therefore, these findings suggest that patients responding to trastuzumab for more than 1 year have a good chance of long-term outcome lasting several years, even if they have previously received it as adjuvant treatment.

Our study raises the question of the optimal duration of anti-HER2 therapy in long-term responders with advanced/metastatic disease, given that almost one-third of our patients remained in remission at 5 years of follow-up. In early breast cancer, three randomized trials have explored different durations of trastuzumab in the adjuvant setting [12,13,14] and two out of three found that duration of 6 months failed to show non-inferiority compared with a year. No similar trials have been reported in advanced disease, and yet this issue is relevant regarding the debate on the increasing healthcare costs and the cost-effectiveness of anti-HER2 treatments [15]. Seven patients in our series had their treatment electively discontinued based on prolonged complete remission, and none of these have so far relapsed, suggesting that cures may sometimes be possible even in patients with metastatic disease. These prolonged remissions may be related either to a clearance of the HER2-positive clones or to immune surveillance. A randomized trial of discontinuation of anti-HER2 therapy following prolonged complete remission would be challenging to conduct owing to the small proportion of potentially eligible patients and potential concerns amongst both patients and health care professionals about those randomized to stop treatment. Nevertheless, at the very least a prospective national or international data base on patients electively stopping is strongly warranted.

Our analysis indicates also that long-term anti-HER2 treatment remains safe and well tolerated. Continuation of HER2-directed treatment can increase the risk of cardiotoxicity and regular monitoring of cardiac function with echocardiogram or multigated acquisition scan is mandatory [16]. The incidence of cardiotoxicity varies according to patient-related factors, such as previous chemotherapy, comorbidities and age [17, 18]. A 2012 meta-analysis of trials investigating trastuzumab documented and increased risk for severe heart failure of 2.5 versus 0.4% [relative risk (RR) 5.11, 90% CI 3.00–8.72] and reduction in left ventricular ejection fraction (RR 1.83, 90% CI 1.36–2.47) [19]. The subsequent APT study of adjuvant weekly paclitaxel chemotherapy plus trastuzumab in node-negative, HER2-positive tumours measuring up to 3 cm documented grade 3 left ventricular systolic dysfunction only in 0.5% of patients and asymptomatic LVEF decline in 3% [20]. Our analysis did not demonstrate any increased cardiac toxicity in patients receiving anti-HER2 therapy for more than 1 year, and the treatment was discontinued due to cardiac adverse events in only six patients, which suggests that most cases are reversible following dose interruptions and adequate medical therapy.

New agents with documented efficacy are currently being investigated in the management of advanced HER2-positive breast cancer [21]. Such compounds include Fc-engineered antibodies that can delay further disease progression in the context of heavily pre-treated disease [22] and are likely be tested in the near future in treatment-naïve patients. The roles of trastuzumab and pertuzumab are now very well established in the first-line setting; identifying predictors of long-term response to these would be important in selecting which patients might benefit from entry into future clinical trials assessing the long-term benefit of these newer agents in addition.