BACKGROUND & AIMS:Fecal microbiota transplantation (FMT) is a highly effective therapy for recurrent Clostridium difficile infection (CDI). However, transferring undefined living bacteria entails uncontrollable risks for infectious and metabolic or malignant diseases, particularly in immunocompromised patients. We investigated whether sterile fecal filtrates (containing bacterial debris, proteins, antimicrobial compounds, metabolic products, and oligonucleotides/DNA), rather than intact microorganisms, are effective in patients with CDI. METHODS:We performed a clinical case series to investigate the effects of fecal filtrate transfer (FFT) in 5 patients with symptomatic chronic-relapsing CDI at the Department of Internal Medicine I at the University Hospital Schleswig-Holstein (Kiel, Germany). Patients were followed up for at least 6 months and for up to 33 months. Stool was collected from 5 donors selected by the patients, and fully characterized according to FMT standards. Stool was sterile-filtered to remove small particles and bacteria; the filtrate was transferred to patients in a single administration via nasojejunal tube. Fecal samples were collected from patients before and at 1 week and 6 weeks after FFT. Microbiome, virome, and proteome profiles of donors and patients were compared. RESULTS:In all 5 patients, FFT restored normal stool habits and eliminated symptoms of CDI for a minimum period of 6 months. Proteome analyses of selected FFT filtrates showed no obvious protein candidates associated with therapeutic efficacy. 16S ribosomal RNA gene sequencing detected diverse bacterial DNA signatures in the filtrates. Analysis of virus-like particles from a filtrate found to reduce symptoms of CDI showed a complex signature of bacteriophages. Bacterial phylogeny and virome profile analyses of fecal samples from recipients indicated longitudinal changes in microbial and viral community structures after FFT. CONCLUSIONS:A preliminary investigation of 5 patients with CDI shows that transfer of sterile filtrates from donor stool (FFT), rather than fecal microbiota, can be sufficient to restore normal stool habits and eliminate symptoms. This finding indicates that bacterial components, metabolites, or bacteriophages mediate many of the effects of FMT, and that FFT might be an alternative approach, particularly for immunocompromised patients.