Longitudinal analysis at three oral sites links oral microbiota to clinical outcomes in allogeneic hematopoietic stem-cell transplant

Background Allogeneic hematopoietic stem-cell transplant (allo-HSCT) is a potentially curative therapy for several hematological disorders. Before stem-cell infusion, recipients undergo a conditioning regimen with chemo/radiotherapy and immunosuppressants, requiring the use of antibiotics to treat and prevent infections. This regimen promotes drastic alterations in the recipient’s microbiotas, including the oral microbiota, which have been associated with allo-HSCT complications and poor outcomes. However, long-term longitudinal studies on the oral microbiota of allo-HSCT recipients are scarce and disregard the existence of distinct microbiotas within the oral cavity. Here, we used 16S rRNA gene sequencing to characterize the microbiota dynamics (during and after allo-HSCT) of 31 allo-HSCT recipients at 3 oral sites (gingival crevicular fluid, oral mucosa, and supragingival biofilm).

Results Analysis of the oral microbiota dynamics during allo-HSCT revealed a significant decline in bacterial diversity and major shifts in microbiota composition in all oral sites, including blooms of potentially pathogenic genera. These blooms in some cases preceded respiratory infections caused by the blooming genera. We also noticed that differences in microbiota diversity and composition between oral sites were lost during allo-HSCT. Overall, oral microbiotas returned to their preconditioning state after engraftment. However, the ability to recover the initial bacterial composition varied between patients. After stratifying patients based on their ability to recover their preconditioning microbiota composition, we found that recovery of the oral mucosa microbiota composition was not associated with antibiotic usage but was associated with higher preconditioning diversity and earlier reconstitution of normal leukocyte counts. Most notably, oral mucosa microbiota composition recovery was an independent biomarker of better allo-HSCT outcomes.

Conclusion We observed clear patterns of microbiota dysbiosis in all three oral sites during allo-HSCT, however each oral site responded differently to the perturbations associated with allo-HSCT. Oral microbiota injury and recovery patterns were associated with allo-HSCT complications and outcomes. This study highlights the potential clinical impact of the oral microbiota in the allo-HSCT setting and the clinical value of tracking oral microbiota changes during allo-HSCT.