This vignette provides all reproducible codes for our article:

Oral microbial signatures of head and neck cancer patients with diverse longitudinal oral mucositis severity patterns

Jiadong Mao, Saritha Kodikara, Erin Marie D San Valentin, Kim-Anh Do, Cielito C. Reyes-Gibby, Kim-Anh L^e Cao

Abstract:

Background: Oral mucositis (OM) is a painful and debilitating side effect of cancer treatment in patients with head and neck cancer (HNC). Alterations in the oral microbiome have been implicated in OM development, making microbial signatures potential targets for therapeutic interventions. This study aimed to characterize longitudinal patterns of OM severity among HNC patients, determine clinically relevant patient clusters based on symptom trajectories, and identify microbial signatures specific to these clusters.

Results: We analyzed longitudinal data from 140 HNC patients, employing non-negative sparse principal component analysis (NSPCA) to summarize seven OM-related symptom ratings into a calibrated OM score. Functional data analysis and hierarchical clustering identified three distinct patient groups with differing trajectories of OM progression. One patient group exhibited a notably rapid increase in OM severity post-treatment initiation, while the other groups displayed more gradual progression. Demographic analyses revealed significant differences in age and weight distributions between clusters, with older, lighter patients more common in the group experiencing gradual OM progression. Partial least squares knockoff (PLSKO) analysis identified group-specific microbial signatures: notably, Prevotella spp. positively associated with calibrated OM score across all groups, while Alloprevotella (Alloprevotella0302) was significantly enriched only in patients experiencing rapid OM progression. Conversely, genera associated with oral health, including Haemophilus, Rothia, and Actinomyces, were negatively correlated with calibrated OM score.

Conclusions: Distinct longitudinal patterns of OM score among HNC patients are associated with specific oral microbial signatures and demographic characteristics. The identification of these group-specific microbiome profiles underscores the potential for microbiome-targeted interventions to manage OM severity, highlighting Prevotella as a central microbial marker across patient clusters.