BACKGROUND: Despite the high incidence of metastatic breast cancer and its related mortality in the elderly population, our knowledge about optimal treatment for older patients with cancer is far from adequate. We aimed to evaluate the efficacy of dual anti-HER2 treatment with or without metronomic chemotherapy in older patients with HER2-positive metastatic breast cancer. METHODS: We did a multicentre, open-label, randomised, phase 2 trial in 30 centres from eight countries in Europe, in patients with histologically proven, HER2-positive metastatic breast cancer, without previous chemotherapy for metastatic disease, who were 70 years or older, or 60 years or older with confirmed functional restrictions defined by protocol, and had a life expectancy of more than 12 weeks and a performance status according to WHO scale of 0-3. Eligible patients were randomly assigned (1:1) by an online randomisation system based on the minimisation method to receive metronomic oral cyclophosphamide 50 mg per day plus trastuzumab and pertuzumab, or trastuzumab and pertuzumab alone. Trastuzumab was given intravenously with a loading dose of 8 mg/kg, followed by 6 mg/kg every 3 weeks. Pertuzumab was given intravenously with a loading dose of 840 mg, followed by 420 mg every 3 weeks. Patients were stratified by hormone receptor positivity, previous HER2 treatment, and baseline geriatric screening. The primary endpoint was investigator-assessed progression-free survival at 6 months as per Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. A difference of 10% or greater between the two groups was sought. Efficacy analyses were by intention to treat; safety was assessed in all patients who received at least one dose of study treatment. In case of progression, all patients were offered trastuzumab emtansine. This trial is registered with ClinicalTrials.gov, number NCT01597414, and is completed. FINDINGS: Between July 2, 2013, and May 10, 2016, 80 patients, of whom 56 (70%) had a potential frailty profile according to the geriatric screening G8 score (≤14), were randomly assigned to receive trastuzumab and pertuzumab (n=39) or trastuzumab and pertuzumab plus metronomic oral cyclophosphamide (n=41). Estimated progression-free survival at 6 months was 46·2% (95% CI 30·2-60·7) with trastuzumab and pertuzumab versus 73·4% (56·6-84·6) with trastuzumab and pertuzumab plus metronomic oral cyclophosphamide (hazard ratio [HR] 0·65 [95% CI 0·37-1·12], p=0·12). At a median follow-up of 20·7 months (IQR 12·5-30·4), the median progression-free survival was 5·6 months (95% CI 3·6-16·8) with trastuzumab and pertuzumab versus 12·7 months (6·7-24·8) with the addition of metronomic oral cyclophosphamide. The most frequent grade 3-4 adverse events were hypertension (in six [15%] of 39 patients in the trastuzumab and pertuzumab group vs five [12%] of 41 in the trastuzumab and pertuzumab plus metronomic oral cyclophosphamide group), diarrhoea (four [10%] vs five [12%]), dyspnoea (two [5%] vs four [10%]), fatigue (three [8%] vs two [5%]), pain (two [5%] vs two [5%]), and a thromboembolic event (0 [0%] vs four [10%]). Severe cardiac toxicities were occasionally observed in both groups. In the trastuzumab and pertuzumab group four patients died without progression, due to cardiac arrest during treatment (n=1), peritoneal infection (n=1), respiratory failure (n=1), and sudden death without a specified cause (n=1). In the trastuzumab and pertuzumab plus metronomic oral cyclophosphamide group, one patient died from heart failure. INTERPRETATION: Addition of metronomic oral cyclophosphamide to trastuzumab plus pertuzumab in older and frail patients with HER2-positive metastatic breast cancer increased median progression-free survival by 7 months compared with dual HER2 blockade alone, with an acceptable safety profile. Trastuzumab and pertuzumab plus metronomic oral cyclophosphamide, followed by trastuzumab emtansine after disease progression, might delay or supersede the need for taxane chemotherapy in this population. FUNDING: F Hoffmann-La Roche.