A 14-year retrospective of HER2 + metastatic breast cancer treatment at one comprehensive cancer center: Impact of the trastuzumab/pertuzumab and trastuzumab-emtansine sequence versus trastuzumab on overall survival

Abstract

Introduction: Monoclonal antibodies represent a significant improvement in the treatment of the HER2 + metastatic cancer, which is associated with a worse prognosis. The objective of this study was to compare overall survival (OS) data in patients treated with trastuzumab + pertuzumab followed by trastuzumab-emtansine (T-DM1) in the second line of metastatic treatment (Arm A) versus patients treated with trastuzumab alone (Arm B). Progression-free survival (PFS) in first-line metastatic patients was also compared in both arms. Methods: This single-center retrospective study included patients from February 2008 to August 2022. OS and PFS of both arms were described and estimated using the Kaplan-Meier method. Data were extracted from electronic medical records and CHIMIO prescribing software. Results: The total duration of metastatic treatment of the 82 patients was significantly longer in the arm A (43.2 ± 28 months vs 33.6 ± 28.9 months), as was the median time to death (59 vs 52 months). The OS data showed a significant reduction in the risk of death in the arm A (Hazard Ratio = 0.59; 95% Confidence Interval [0.37–0.94]; p = 0.02). No difference was shown for PFS. Conclusion: The trastuzumab/pertuzumab/T-DM1 combination showed a significant improvement in OS. Our results are in line with the CLEOPATRA and EMILIA studies, but long-term responders in the arm A may have impacted our results. The absence of difference in term of PFS in first metastatic line may be explained by a selection bias, as patients in the arm A potentially have more aggressive forms.

Keywords

Breast neoplasms trastuzumab pertuzumab ado-trastuzumab emtansine

Introduction

Breast cancer is the leading cause of cancer in women, with an incidence doubling since the 1990s and it remains the leading cause of cancer death in women.¹ HER2 protein is found in 15–30% of invasive breast cancers and is associated with increased cell proliferation and motility, tumor invasiveness, progression of regional and distant metastasis, accelerated angiogenesis, and decreased apoptosis leading to a poor prognosis.^2,3 The first-line treatment of metastatic HER2 + breast cancer includes a combination of trastuzumab and pertuzumab ± taxane. The combination of trastuzumab-emtansine (TDM-1) can be used in second-line treatment (Table 1).⁴

Table 1.

Management of metastatic HER2 + breast cancer.

Line	Protocol
1^st line	12 months after last trastuzumab treatment or de novo stage IV: Trastuzumab 6 mg/kg (8 mg/kg LD) + Pertuzumab 420 mg (840 mg LD) + Paclitaxel 80 mg/m² or Docetaxel 75 mg/m² D1 D8 = D21 Maintenance: Trastuzumab 6 mg/kg + Pertuzumab 420 mg (+ Hormone therapy if HR+) Free interval less than 12 months: T-DM1 3,6 mg/kg D1 = D21
2^nd line	T-DM1 D1 = D21
3^rd line	T-DXd D1 = D21 Tucatinib/Trastuzumab/Capecitabine (progressive brain metastases after radiotherapy)
Other lines Several options	– Inclusion in a therapeutic trial – T-DXd – Tucatinib-trastuzumab-capecitabine – Trastuzumab + chemotherapy (vinorelbine, capecitabine, gemcitabine, VP16, platinum salts…) – Lapatinib + capecitabine – Trastuzumab + lapatinib (HR -) – Trastuzumab or lapatinib + hormone therapy – Resumption of anthracycline – Targeted therapy

D: Day; LD: Loading Dose; HR: Hormone Receptors; T-DM1: Trastuzumab-emtansine; T-DXd: Trastuzumab-deruxtecan.

Trastuzumab and pertuzumab are humanized IgG1 monoclonal antibodies. Trastuzumab targets subdomain IV of the extracellular domain of HER2, while pertuzumab targets subdomain II. Pertuzumab is always combined with trastuzumab to enhance anti-HER2 + tumor activity.⁵ The combination inhibits the proliferation of tumor cells overexpressing HER2 and both drugs are potent mediators of antibody-dependent cellular cytotoxicity (ADCC) directed against these cells. Trastuzumab-Pertuzumab combination is recommended for the treatment of both advanced and early-stage breast cancer in patients with HER2 + tumors. The CLEOPATRA trial, which compared trastuzumab and docetaxel with pertuzumab or placebo, showed an improvement in Progressive-Free Survival (PFS) of 6.1 months (Hazar Ratio (HR) for PFS = 0.62; 95% Confidence Interval (CI) [0.51–0.75]; p < 0.001) and Overall Survival (OS) with à 5% survival improvement in patients previously treated with trastuzumab and the maintenance of this benefit over time after a median follow-up of 99 months (HR = 0.64 (95% CI [0.47–0.88]; p = 0.005).⁶

T-DM1 is a conjugated monoclonal antibody containing trastuzumab covalently linked to DM, a cytotoxic microtubule inhibitor component derived from emtansine (maytansinoid) through a stable thioether linker MCC, to limit systemic release of DM1: emtansine refers to the MCC-DM1 complex. Trastuzumab targets tumor cells that overexpress HER2 and DM1 induces cell death by apoptosis. Results from in vitro cytotoxicity assays show that DM1 is 20 to 200 times more potent than taxanes and vinca alkaloids. It is indicated as a single agent and as second-line therapy in metastatic breast cancer with HER2.⁷ The KATHERINE study comparing trastuzumab with T-DM1 in early-stage patients with residual invasive disease showed a clinically and statistically significant improvement in PFS and 3-year Invasive Disease-Free Survival (IDFS) in patients who received T-DM1 versus trastuzumab (HR for invasive disease or death = 0.50; 95% CI [0.39–0.64]; P < 0.001).⁸ The percentage of patient having T-DM1 with an IDFS at 3 years was 88.3% and 77% for patients having trastuzumab. In the EMILIA study comparing T-DM1 with lapatinib plus capecitabine in advanced (unresectable or metastatic) HER2 + breast cancer patients also showed a significant improvement in OS (29.9 months (95% CI [26.3–34.1]) versus 25.9 months (95% CI [22.7–28.3]) for patients receiving T-DM1 versus lapatinib and capecitabine, respectively (HR for overall survival = 0.75; 95% CI [0.64–0.88]).⁹

The introduction of pertuzumab in 2013 and T-DM1 in 2014, respectively, has further improved even more the management of metastatic HER2 + metastatic breast cancer. Here we want to describe with different temporality the OS and PFS of patients with HER2 + metastatic breast cancer who received chemotherapy based on the 1^st metastatic line (L1) of trastuzumab, pertuzumab and paclitaxel and from 2^nd or more metastatic line of T-DM1 versus the OS of patients who received only trastuzumab in L1.

Method

Study design

This was a monocentric retrospective study at one comprehensive cancer center. Eligible patients were women older than 18 years with HER2 + metastatic breast cancer who received at least one anticancer regimen for metastatic stage. Arm A included patients who received trastuzumab + pertuzumab combination in L1 from January 2010 to June 2022 and T-DM1 in 2^nd metastatic line. Arm B included patients who received trastuzumab alone at least in L1 from February 2008 to June 2018. The patients’ hormonal status was homogeneous between the two groups. The data of the patients studied were obtained from the local database that feeds the national Epidémio-Stratégie Médico-Economique (ESME) database of UNICANCER yearly update. Patients in this database did not object to the collection of their data, which were anonymized before analysis. The database for this study was created according to the reference methodology MR004 of the Commission Nationale de l'Informatique et des Libertés (CNIL). As this study was non-interventional, retrospective, monocentric and implied only medical-file data collection of deceased patients, no informed consent or additional ethical committee review was required.

Data collection

Clinical and demographic data were obtained from the center database that feeds the UNICANCER's ESME database, verified and completed using patient electronic health records from the UNICANCER network iDocMed 2.7.3 license GLP v2 software and the Computerized Provider Order Entry (CPOE) CHIMIO^® v5.9, Computer Engineering, Paris, France. These data were collected on a standardized data collection spreadsheet (EXCEL^TM, Microsoft). Demographic and administrative data collected for all patients were: age, weight, height, age at diagnosis and age at death. Clinical and medical data collected were: risk factors, comorbidities, cancer history, menopausal status, date of diagnosis and tumor type, TNM stage, grade, hormonal status and HER2 status at diagnosis. The elements of metastatic character analyzed were date of diagnosis, metastatic site(s), different lines of treatment and number of lines, addition of hormonal therapy and radiotherapy. Anticancer history, place of trastuzumab in combinations or alone, and treatment duration were also collected.

Statistical analysis

Quantitative variables were described as mean ± standard deviation [minimum and maximum] and qualitative variables were represented by their number and percentage. Demographic and clinical data were subjected to univariate analysis and results are presented according to p-value. Survival analysis was performed considering the time to the last news date, which was defined by the occurrence of death or the last chemotherapy treatment for the patients still on treatment. PFS and OS were estimated using the Kaplan-Meier method (Prism GraphPad 8, San Diego, USA) and were compared using the log-rank test.

Results

Patient characteristics and treatments

A total of 82 women with HER2 + breast cancer were included, 39 (48%) for arm A and 43 (52%) for arm B. The mean age of the patients was 53 (±11) years at initial diagnosis and 55 (±11) years at diagnosis and initiation of metastatic therapy. Baseline data characteristics are described in the Table 2 for all patients and detailed in Table 3 for each arm . Half of the patients had an initial metastatic cancer (n = 41; 50%) and more than a half had a grade 3 tumor (n = 45; 55%), were hormone-dependent (n = 48; 59%) and postmenopausal (n = 49; 60%). The tumors were invasive ductal carcinoma (IDC) in 90% of cases (n = 74).

Table 2.

Baseline data for all patients included in the retrospective study.

	Number of patients (%) or m ± SD [min – max]
Total patients	82
Trastuzumab + Pertuzumab / T-DM1 (A)	39 (48)
Trastuzumab (B)	43 (52)
Age (years)
At diagnosis	53 ± 11 [28–82]
At the time of metastatic diagnosis	55 ± 11 [31–83]
At the beginning of L1	55 ± 11 [32–83]
Upon death	58 ± 12 [34–85]
Body mass index (BMI, kg/m²)	27.9 ± 5.92 [16.3–41.9]
History of other cancer
Personal history	7 (9)
Family history of breast or ovarian cancer	29 (35)
Family history of other cancer	19 (23)
Comorbidities	23 (28)
Histological grade at diagnosis
1–2	36 (44)
3	45 (55)
NA	1 (1)
Positive HR status	48 (59)
Positive menopausal status	49 (60)
Mean time from discovery of metastasis to L1 (days)	53 ± 67
Tumor type at diagnosis
IDC	75 (91)
DCIS	5 (6)
ILC	2 (2)
TNM stage at diagnosis
T
T0	1 (1)
T1	9 (11)
T2	29 (35)
T3	19 (23)
T4	20 (24)
NA	4 (5)
N
N0	17 (21)
N1	43 (52)
N2	9 (11)
N3	9 (11)
NA	4 (5)
M
M0	41 (50)
M1 – immediately metastatic	41 (50)

DCIS: Ductal Carcinoma in situ; HR: Hormone Receptors; IDC: Invasive Ductal Carcinoma; ILC: Invasive Lobular Carcinoma; L1: First Line Metastatic Treatment; NA: Not Applicable; TNM: Tumor, Node, Metastasis; m ± SD: mean ± Standard Deviation.

Table 3.

Comparison of baseline patient characteristics.

	Arm A (Number of patients (%) or m ± SD [min – max])	Arm B (Number of patients (%) or m ± SD [min – max])	p-value
Total	39	43
Age (years)
At diagnosis (mean ± SD [min - max])	51 ± 9 [30–68]	55 ± 13 [28–82]	0.119
At the time of metastatic diagnosis (mean ± SD [min - max])	53 ± 9 [34–71]	57 ± 12 [31–83]	0.118
At the beginning of L1 (mean ± SD [min - max])	53 ± 10 [34–72]	57 ± 12 [32–83]	0.115
Body mass index (kg/m²)	27 ± 5.8 [16,3–41,5]	28.7 ± 5.9 [19.3–41.9]	0.220
History of cancer
Personal history	4 (10)	3 (7)	0.70
Family history of breast or ovarian cancer	20 (51)	9 (21)	0.01
Family history of other cancer	10 (26)	9 (21)	0.79
Comorbidities	9 (23)	14 (33)	0.34
Histological grade at diagnosis			0.90
1–2	16 (41)	20 (47)
3	22 (56)	23 (53)
NA	1 (3)	0 (0)
Positive HR status	25 (64)	23 (53)	0.33
Menopausal status	22 (56)	27 (63)	0.56
Mean time from discovery of metastasis to L1 (days)	40 ± 27 [7–123]	65 ± 88 [0–409]	0.09
Tumor type at diagnosis			0.16
IDC	36 (92)	39 (91)
DCIS	1 (3)	4 (9)
ILC	2 (5)	0 (0)
TNM stage at diagnosis
T			0.35
T0	1 (3)	0 (0)
T1	6 (15)	3 (7)
T2	14 (36)	15 (35)
T3	6 (15)	13 (30)
T4	9 (23)	11 (26)
NA	3 (8)	1 (2)
N			0.76
N0	9 (23)	8 (19)
N1	20 (51)	23 (53)
N2	4 (10)	5 (12)
N3	3 (8)	6 (14)
NA	3 (8)	1 (2)
M			0.047
M0	15 (38)	26 (60)
M1	24 (62)	17 (40)

DCIS: Ductal Carcinoma in situ; HR: Hormone Receptors; IDC: Invasive Ductal Carcinoma; ILC: Invasive Lobular Carcinoma; L1: First Line Metastatic Treatment; NA: Not Applicable; T-DM1: Trastuzumab-emtansine; TNM: Tumor, Node, Metastasis; m ± SD: mean ± Standard Deviation.

In univariate analysis, there were no statistical differences in baseline characteristics between the two treatment arms for age, body-mass index (BMI), comorbidities, histologic grade, hormonal and menopausal status at diagnosis, tumor type and T and N stage at diagnosis with a p-value always > 0.05.

There was a significant difference in metastatic status at diagnosis, with 24 patients in the arm A and 17 patients in the arm B having metastatic tumors at baseline, respectively, and 20 patients in the arm A and 9 patients in the arm B having a family history of breast or ovarian cancer, respectively (p = 0.01). There was no difference in the mean time from detection of metastatic disease to initiation of L1. Data on TNM stage were missing for 4 patients and data on histologic grade at diagnosis were missing for one patient.

Overall survival

At initial diagnosis

More than half patients in both arms, mostly in the arm B, received concurrent radiotherapy with a significant difference in the distribution (p = 0.02). RH + patients received hormonal therapy with anti-estrogens or anti-aromatase, with no difference between the two arms (p = 0.48). Some patients received anthracyclines in metastatic disease with a significant difference between the two arms (p = 0.03), with patients in the arm B receiving more. There was also a significant difference in the use of taxanes, which was greater in the arm A (p = 0.01) (Table 4).

Table 4.

Other treatments received by patients.

	No. of patients A (%)	No. of patients B (%)	p-value
Radiotherapy	25 (64)	37 (86)	0.02
Hormone therapy	23 (59)	22 (51)	0.48
Anti-oestrogens	9 (23)	9 (21)	0.81
Anti-aromatase	19 (49)	16 (37)	0.29
Anthracyclines	5 (13)	14 (33)	0.03
Taxanes	38 (97)	33 (77)	0.01

The total duration of metastatic treatment was significantly different between the arm A: 43.2 ± 28 months and the arm B: 33.6 ± 28.9 months, p = 0.04. Patients in the arm A received a mean of 4.6 ± 1.9 lines of metastatic treatment, while patients in the arm B received a mean of 3.2 ± 1.8 lines of metastatic treatment, with a significant difference (p = 0.0007) (Table 5).

Table 5.

Comparison of metastatic treatments in the two arms.

	Arm A (m ± SD [min – max])	Arm B (m ± SD [min – max])	p-value
Total duration of metastatic treatment (month)	43.2 ± 28 [5–147]	33.6 ± 28.9 [0–128]	0.04
No. of lines of metastatic treatment	4.6 ± 1.9 [2–9]	3.2 ± 1.8 [1–7]	0.0007
Median survival time to initial diagnosis (month)	59	52
Average survival time to initial diagnosis (month)	76	58

m ± SD: mean ± Standard Deviation.

Regarding OS, the median time to death was 59 months (95% CI [49–84]) in the arm A and 52 months (95% CI [43–64]) in the arm B, respectively, corresponding to a HR of 0.59 (95% CI [0.37–0.94]), a significant 41% reduction in the risk of death in the arm A (p = 0.02). Figure 1 shows OS for the two arms. The 1-year overall survival rate (SR) was 97% in the arm A and 93% in the arm B (p < 0.05), the 2-year SR was 93% in the arm A and 82% in the arm B, the 6-year SR was 40% in the arm A and 27% in the arm B, and the 10-year SR was 24% in the arm A and 5% in the arm B, respectively.

Progression-free survival

Median PFS during L1 was not significantly different between the two arms with a median duration of 11 (95% CI [6–16]) months in the arm A and 10 (95% CI [6–13]) months in the arm B, with a p = 0.28 by log-rank test. The mean L1 duration was 18.1 ± 21.5 months in the arm A and 14.4 ± 18 months in the arm B, which was not significantly different (p = 0.39). The HR for progression was 0.79 (95% CI [0.51–1.22]). Figure 2 shows the evolution of PFS during L1 as a function of time in months according to the Kaplan-Meier model.

Discussion

The present study in one cancer comprehensive center is a description of breast cancer innovation implementation in real-life practice. The results strongly suggest that the use of T-DM1 has led to an improvement in the survival of patients compared to the therapies that have been used to date.

The two groups of patients were comparable, but the statistically significant difference in early metastatic disease was an important difference that may bias our results: patients in the arm A had more early metastatic tumors than those in the arm B. HER2 + status was reported as positive in the data collection regardless of its score (1, 2, and 3 = +).

The combination of pertuzumab and trastuzumab in first-line metastatic breast cancer with second-line T-DM1 prolongs overall survival in patients with HER2-positive metastatic breast cancer. Only one patient in the trastuzumab arm was still alive at the end of data collection and has been treated for metastatic breast cancer for 12 years, whereas 12 patients in the trastuzumab + pertuzumab + T-DM1 arm were still alive and have been treated for metastatic breast cancer for 5 ± 3 [2–12] years. This introduces a bias in the analysis of the survival data and reduces the statistical power because the data remain incomplete. In fact, some patients in the arm A are long-term responders: 5 patients have been on metastatic treatment for more than 7 years and 7 for less than 4 years, which may have an impact on the OS or PFS results. However, the results remain consistent: treatment with pertuzumab plus trastuzumab/T-DM1 resulted in a significant reduction in the risk of death from both initial diagnosis and metastatic diagnosis compared with trastuzumab alone. There was also a 7-month increase in median survival. Additionally, patients in the arm A were treated for a median time of 43.2 months versus 33.6 months in the arm B and received a median of 1.4 additional lines of metastatic therapy.

In the CLEOPATRA trial comparing trastuzumab alone with trastuzumab + pertuzumab, the median time to death was higher in the trastuzumab + pertuzumab arm. In our trial, the median time to death in the arm receiving both molecules plus T-DM1 was 59 months with an HR of 0.59. Based on the CLEOPATRA study, it is then possible to demonstrate the benefit of adding T-DM1 to pertuzumab in the treatment of HER2 + metastatic breast cancer.⁶ This therapeutic benefit was also observed in the EMILIA trial comparing T-DM1 with lapatinib plus capecitabine.⁹ The difference in OS between both arms may also be influenced by several biases: patients in the arm A statistically received more taxanes and fewer anthracyclines during their metastatic treatment. In addition, patients in the arm B received more radiotherapy, which may be explained by the fact that more patients in this arm were treated in the adjuvant stage and received adjuvant radiotherapy, whereas patients in the arm A were treated more from the metastatic stage. We also note that the overall survival data are similar at 70 months, which may be explained by the small number of patients enrolled in the study, with less variability in the data than in the CLEOPATRA or EMILIA studies.

Our study did not show a significant difference in PFS between the two arms in first-line metastatic therapy comparing trastuzumab + pertuzumab versus trastuzumab alone with an HR of 0.79 (95% CI [0.51–1.22]). This may be explained by selection bias: indeed, we selected patients who received first-line trastuzumab and pertuzumab and then metastatic second-line T-DM1, which necessarily implies progression and excludes long responders. Thus, our selection includes more patients with a potentially more aggressive form and who will relapse earlier. Furthermore, our study does not take into account the history before metastatic progression: the CLEOPATRA trial showed that the use of adjuvant trastuzumab improved the reduction of the risk of death.⁶ Since our group of interest had more patients with early metastatic cancer, this benefit of adjuvant trastuzumab could not be used and patients had more advanced disease.

The results also show a significant difference in the time from metastatic diagnosis to the start of L1: the first line of metastatic treatment was started statistically later in the arm B. This may be explained by differences in treatment prior to 2013, and it could be observed in the study that adjuvant therapy could be continued for a shorter or longer period after detection of metastatic disease. However, the lack of data on prior treatment does not allow us to conclude whether this bias influences the outcome.

The incidence and severity of adverse events were not assessed in this study. The CLEOPATRA study showed that pertuzumab may cause an increase in grade 3 diarrhea and febrile neutropenia and an increase incidence of other adverse events of grade 1–2, but without an increase in cardiac toxicity.⁸ The EMILIA study demonstrated a reduction in adverse events with T-DM1 compared with lapatinib + capecitabine, highlighting a very good safety profile.

Conclusion

To conclude, our study shows an improvement in overall and metastatic survival with the combination of trastuzumab + pertuzumab in the first line of metastatic treatment and T-DM1 from the second line of metastatic treatment compared to trastuzumab alone. Our results are supported by the CLEOPATRA and EMILIA trials in which patients received either the combination of trastuzumab + pertuzumab or T-DM1 alone in the treatment of HER2 + metastatic breast cancer, but not the combination of these three molecules. If the PFS was not statistically significantly different in the two arms, it seems essential to reproduce this study in a larger group of patients by conducting a multicenter study, which would allow to avoid the bias of the practices of each center. Given that the tolerability profiles of the new drugs improve the quality of life of patients and prolong survival, it seems important to continue in this direction and to develop other antibody-based molecules to improve the quality of life of patients whose disease has become chronic.

Footnotes

Author contributors

VL: Conceptualization, Data curation, Formal analysis, Methodology, Investigation, Writing – original draft, Writing – review & editing

JC: Formal analysis, Writing – original draft, Writing – review & editing

BF: Formal analysis, Writing – original draft, Writing – review & editing

FS: Writing – review & editing

CJ: Formal analysis, Methodology, Supervision, Writing – review & editing

DP: Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Investigation, Writing – original draft, Writing – review & editing

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

V Lepage

References

Santé Publique France. Cancer du sein [Internet], https://www.santepubliquefrance.fr/maladies-et-traumatismes/cancers/cancer-du-sein (accessed 27 December 2021).

Ross

Slodkowska

Symmans

, et al. The HER-2 receptor and breast cancer: ten years of targeted anti-HER-2 therapy and personalized medicine. Oncologist 2009 Apr; 14: 320–368.

Iqbal

. Human epidermal growth factor receptor 2 (HER2) in cancers: overexpression and therapeutic implications. Mol Biol Int 2014; 2014: 852748.

Institut Curie, Gustave Roussy, APHP. Cancers et pathologies du sein : attitudes diagnostiques et thérapeutiques, protocoles de traitement 2021-2022 [Internet]. https://www.gustaveroussy.fr/sites/default/files/referentiel-senorif-2021-2022.pdf (accessed 4 April 2022).

Agence Européenne du Médicament. HERCEPTIN – Résumé des Caractéristiques du Produit [Internet], https://www.ema.europa.eu/en/documents/product-information/herceptin-epar-product-information_fr.pdf (accessed 24 June 2022).

Baselga

Cortés

Kim

, et al. Pertuzumab plus trastuzumab plus docetaxel for metastatic breast cancer. N Engl J Med 2012 Jan; 366: 109–119.

Agence Européenne du Médicament. KADCYLA – Résumé des Caractéristiques du Produit [Internet], https://www.ema.europa.eu/en/documents/product-information/kadcyla-epar-product-information_fr.pdf (accessed 24 June 2022).

von Minckwitz

Huang

Mano

, et al. Trastuzumab emtansine for residual invasive HER2-positive breast cancer. N Engl J Med 2019 Feb; 380: 617–628.

Diéras

Miles

Verma

, et al. Trastuzumab emtansine versus capecitabine plus lapatinib in patients with previously treated HER2-positive advanced breast cancer (EMILIA): a descriptive analysis of final overall survival results from a randomised, open-label, phase 3 trial. Lancet Oncol 2017 Jun; 18: 732–742.