BACKGROUND: Homozygous familial hypercholesterolaemia is a genetic disorder characterised by substantially raised LDL cholesterol, reduced LDL receptor function, xanthomas, and cardiovascular disease before age 20 years. Conventional therapy is with statins, ezetimibe, and apheresis. We aimed to assess the long-term safety and efficacy of the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor evolocumab in a subset of patients with homozygous familial hypercholesterolaemia enrolled in an open-label, non-randomised phase 3 trial. METHODS: In this interim subset analysis of the TAUSSIG study, which was undertaken at 35 sites in 17 countries, we included patients aged 12 years or older with homozygous familial hypercholesterolaemia who were on stable LDL cholesterol-lowering therapy for at least 4 weeks; all patients received evolocumab 420 mg subcutaneously monthly, or if on apheresis every 2 weeks. Dosing could be increased to every 2 weeks after 12 weeks in patients not on apheresis. The primary outcome of the TAUSSIG study was treatment-emergent adverse events; secondary outcomes were the effects of evolocumab on LDL cholesterol and other lipids. We analysed patients on an intention-to-treat basis, and all statistical comparisons were done post hoc in this interim analysis. The TAUSSIG study is registered with ClinicalTrials.gov, number NCT01624142, and is ongoing. FINDINGS: 106 patients were included in this analysis, 34 receiving apheresis at study entry and 14 younger than 18 years. The first patient was enrolled on June 28, 2012, and the cutoff date for the analysis was Aug 13, 2015; mean follow-up was 1·7 years (SD 0·63). After 12 weeks, mean LDL cholesterol decreased from baseline by 20·6% (SD 24·4; mean absolute decrease 1·50 mmol/L [SD 1·92]); these reductions were maintained at week 48. 47 of 72 patients not on apheresis at study entry increased evolocumab dosing to every 2 weeks, with an additional mean reduction in LDL cholesterol of 8·3% (SD 13·0; mean absolute decrease 0·77 mmol/L [SD 1·38]; p=0·0001). In a post-hoc analysis, mean reductions in LDL cholesterol in patients on apheresis were significant at week 12 (p=0·0012) and week 48 (p=0·0032), and did not differ from reductions achieved in patients not on apheresis (p=0·38 at week 12 and p=0·09 at week 48). We noted a small reduction (median -7·7% [IQR -21·6 to 6·8]) in lipoprotein(a) at week 12 (p=0·0015), with some additional reduction at week 48 (-11·9% [-28·0 to 0·0]; p<0·0001). HDL cholesterol was increased by a mean of 7·6% (SD 18·1) at week 12 (p<0·0001) and 7·6% (SD 20·6) at week 48 (p=0·0007). Evolocumab was well tolerated; 82 (77%) patients reported treatment-emergent adverse events, which were mostly minor. The most common were nasopharyngitis (14 patients [13%]), influenza (13 [12%]), headache (11 [10%]), and upper respiratory tract infection (11 [10%]). Serious adverse events occurred in 18 (17%) patients, with the most common being cardiovascular events (four patients [4%]). There were no deaths and four positively adjudicated cardiovascular events, one (3%) among patients on apheresis and three (4%) among patients who did not receive apheresis. INTERPRETATION: Our interim results suggest that evolocumab is an effective additional option to reduce LDL cholesterol in patients with homozygous familial hypercholesterolaemia, with or without apheresis. FUNDING: Amgen.