“LUX-Breast 3” was added to the title of the Clinical Trial Protocol (CTP) to indicate that this trial is part of the LUX-Breast programme. The study title and inclusion criterion 6 were corrected to allow recruitment of patients with CNS progression during or after both trastuzumab and lapatinib. “Targeted treatment other than trastuzumab” was deleted from exclusion criterion 1 to allow recruitment of patients who had received other anti-HER2 monoclonal antibodies in addition to trastuzumab. A statement on vinorelbine dose adjustment was added to allow patients who benefited from vinorelbine therapy to continue treatment on a regular dosing schedule. A description of the clinical evaluation and reporting of Drug-induced liver injury (DILI) was added at the request of the Regulatory Authority. The reporting requirements for worsening of underlying disease or other pre-existing conditions, and changes in vital signs, Electrocardiogram (ECG), physical examination and laboratory tests results were modified at the request of the Regulatory Authority. Information on the DMC was added to ensure blinding of the trial to the study team. The number of patient’s characteristics to be collected at baseline was reduced because it was non-pertinent data. The NSS worksheet had been omitted from the original protocol and was added in an appendix. Other changes were made to clarify the study procedures or to update the information in the CTP.

It was originally planned that the DMC meeting would be held when 20 patients had been treated for at least 12 weeks in each treatment arm (or progressed before Week 12). However, there was a risk that recruitment into the trial would be complete before the benefit-risk analysis was performed and so the analysis was performed on the first 60 patients treated, regardless of treatment duration. A caution was added if afatinib was to be combined with potent P-glycoprotein modulators as exposure to afatinib could be decreased. However, it ceased to be a requirement to discontinue a patient from the trial if they required concurrent therapy. Information was added on the potential adverse effects of keratitis and ulcerative keratitis, which have been added to the class labelling for Epidermal growth factor receptor (EGFR) inhibitors. It was clarified that afatinib could be paused for up to 7 days but dose reductions of afatinib were not to be performed in the event of AEs or SAEs that were notm treatment-related. The threshold for withholding vinorelbine infusion on the basis of platelet counts was increased from <75 to <100 x 109/L. The dose of vinorelbine could be reduced to 20 mg/m2 in patients with severe hepatic impairment. A patient diary was added for optional use during the trial to collect information on the time of study treatment, food intake, and occurrence of acute side effects. It ceased to be necessary to repeat the tumour assessment at the Screening Visit if valid results were available from routine clinical practice within 28 days prior to start of treatment. Minor errors were corrected and inconsistencies clarified.