Usability of rectal swabs for microbiome sampling in a cohort study of hematological and oncological patients

Large-scale clinical studies investigating associations between intestinal microbiota signatures and human diseases usually rely on stool samples. However, the timing of repeated stool sample collection cannot be predefined in longitudinal settings. Rectal swabs, being straightforward to obtain, have the potential to overcome this drawback. Therefore, we assessed the usability of rectal swabs for microbiome sampling in a cohort of hematological and oncological patients.

We used a pipeline for intestinal microbiota analysis from deep rectal swabs which was established and validated with test samples and negative controls. Consecutively, a cohort of patients from hematology and oncology wards was established and weekly deep rectal swabs taken during their admissions and re-admissions.

Validation of our newly developed pipeline for intestinal microbiota analysis from rectal swabs revealed consistent and reproducible results. Over a period of nine months, 418 rectal swabs were collected longitudinally from 41 patients. Adherence to the intended sampling protocol was 97%. After DNA extraction, sequencing, read pre-processing and filtering of chimeric sequences, 405 of 418 samples (96.9%) were eligible for further analyses. Follow-up samples and those taken under current antibiotic exposure showed a significant decrease in alpha diversity as compared to baseline samples. Microbial domination occurred most frequently by Enterococcaceae (99 samples, 24.4%) on family level and Enterococcus (90 samples, 22.2%) on genus level. Furthermore, we noticed a high abundance of potential skin commensals in 99 samples (24.4%).

Deep rectal swabs were shown to be reliable for microbiome sampling and analysis, with practical advantages related to high sampling adherence, easy timing, transport and storage. The relatively high abundance of putative skin commensals in this patient cohort may be of potential interest and should be further investigated. Generally, previous findings on alpha diversity dynamics obtained from stool samples were confirmed.