Dynamic Shifts in the Oral Microbiota Following Cancer Surgery: A 172-Sample Longitudinal Study of Surgical Site Infection Risk

This longitudinal study of 45 oral cancer patients reveals that surgical site infections are associated with younger age, advanced tumor stage, and persistent post-operative dysbiosis dominated by *Prevotella* and *Porphyromonas*, while early post-surgical abundance of *Aggregatibacter* serves as a significant predictive biomarker for infection risk.

The Gist

The Big Picture: A Garden After a Storm

Imagine the inside of your mouth as a bustling, diverse garden. It has many different types of plants (bacteria) living together in a balanced ecosystem. Usually, this garden is stable.

However, when a patient undergoes surgery to remove oral cancer, it's like a massive storm hits that garden. The surgeon cuts out a large section of the land, and the environment changes drastically. This study followed 45 patients to see what happens to their "mouth gardens" before, during, and after this storm, specifically looking at why some patients got an infection (a "weed invasion") while others healed cleanly.

The Cast of Characters

• The Patients: 45 people with oral cancer.
• The "Storm": The cancer surgery.
• The "Garden": The oral microbiome (the community of bacteria in the mouth).
• The "Weed Invasion": Surgical Site Infection (SSI), which happened in about 31% of the patients (14 out of 45).

What the Researchers Found

1. The Surprise: Younger Patients Were at Higher Risk

Usually, we think older people are more likely to get infections because their immune systems are weaker. But this study found the opposite: younger patients were more likely to get infected.

The Analogy: Think of it like a construction site. The younger patients didn't have weaker "security guards" (immune systems); they simply had bigger construction projects. They had larger, more advanced tumors that required much more extensive surgery. The bigger the surgery, the more the "garden" was disturbed, making it harder to recover.

2. The Storm Hits: The Garden Gets Wiped Out

Immediately after surgery, the bacterial population in everyone's mouth dropped significantly. The storm had cleared out most of the plants. This happened to both the people who got infected and those who didn't.

3. The Divergence: Two Different Paths

Here is where the two groups went in different directions:

• The "Clean Heal" Group (No Infection): These patients were like resilient gardeners. Within about 5 to 12 days, their garden started to regrow the same plants that were there before the storm. The original dominant plants (like Rothia and Veillonella) came back, and the garden returned to its normal, balanced state.
• The "Infected" Group: These patients' gardens did not return to normal. Instead, a chaotic new ecosystem took over. The original plants stayed low, but two specific types of "weeds" exploded in number: Prevotella and Porphyromonas. By the time the infection was diagnosed, these two weeds made up about 40% of the entire garden, whereas they were barely present in the healthy group.

4. The Early Warning Signal: The "Keystone" Bacteria

The most exciting finding was a specific bacteria called Aggregatibacter.

The Analogy: Imagine a small, quiet construction crew arriving at a site before the big demolition. In the patients who eventually got infected, this crew (Aggregatibacter) showed up in higher numbers immediately after surgery (within 5 days), even before the infection was officially diagnosed.

The study suggests that Aggregatibacter acts like a "founder" or a "pioneer." It doesn't necessarily cause the infection itself, but it changes the environment (like loosening the soil or changing the oxygen levels) to make it perfect for the nasty weeds (Prevotella and Porphyromonas) to take over later.

• The Stat: If a patient had more than 0.044% of this Aggregatibacter bacteria right after surgery, they were nearly 6 times more likely to develop an infection later.

Summary of the Story

1. Before Surgery: Everyone's mouth garden looked similar.
2. During Surgery: The garden is destroyed for everyone.
3. After Surgery:
   • Healthy Recovery: The original plants grow back quickly.
   • Infection: The garden gets taken over by specific weeds (Prevotella and Porphyromonas).
4. The Clue: A specific "pioneer" bacteria (Aggregatibacter) appears in high numbers right after surgery in the patients who will eventually get infected. It paves the way for the weeds to arrive later.

The Conclusion

The study concludes that surgery disrupts the mouth's bacteria for everyone. However, patients who get infections fail to restore their original bacterial balance. Instead, they develop a persistent imbalance dominated by specific harmful bacteria.

The researchers found that looking for the "pioneer" bacteria (Aggregatibacter) in the days immediately following surgery could act as an early warning system, telling doctors which patients are on the path to infection before the infection actually becomes visible.

Technical Summary

Technical Summary: Dynamic Shifts in the Oral Microbiota Following Cancer Surgery

Problem Statement
Surgical site infection (SSI) remains the leading cause of perioperative morbidity following oral cancer surgery, yet the specific role of the oral microbiota in SSI pathogenesis is poorly defined. While clinical risk factors such as advanced tumor stage and procedural complexity are known, the longitudinal microbial dynamics that precede and characterize SSI development in oral squamous cell carcinoma (OSCC) patients have not been thoroughly elucidated. This study aims to investigate these microbiota dynamics to identify microbial signatures associated with SSI risk.

Methodology
The study employed a prospective longitudinal design involving 45 patients diagnosed with OSCC undergoing tumor resection at the A.C. Camargo Cancer Center.

• Cohort and Sampling: A total of 172 oral swab samples were collected from 45 patients across four time points: baseline (~29 days pre-surgery), immediately pre-surgery (hospital admission), early post-surgery (within 5 days), and late post-surgery (6–15 days).
• Microbiome Profiling: Bacterial composition was analyzed via 16S-rDNA V3-V4 region sequencing (Illumina MiSeq), yielding 172 successfully sequenced samples. Data processing included adapter removal, human read filtering, quality filtering (Phred ≥10), and denoising via Deblur. Taxonomic classification was performed at the genus level using the SILVA database.
• Quantitative Analysis: Bacterial load and human:bacterial DNA ratios were quantified using qRT-PCR targeting the 16S-rDNA V1 region and the human beta-actin gene in 170 samples.
• Clinical Assessment: SSI was defined within 30 days post-surgery using adapted CDC criteria (presence of purulent material, wound dehiscence with signs of infection, or physician diagnosis).
• Statistical Analysis: Alpha and beta diversities were assessed using phyloseq (R). Differences between SSI and non-SSI groups were evaluated using Mann-Whitney U tests, ANOSIM, and LEfSe. Logistic regression models were constructed to assess the predictive value of clinical variables and specific bacterial abundances.

Key Results

• SSI Incidence and Clinical Risk: 14 of 45 patients (31.1%) developed SSI. Contrary to some literature suggesting older age is a risk factor, this study found a significant association between younger age and SSI (median age 53.2 years in SSI group vs. 67.4 years in non-SSI group; p=0.011). This correlation was driven by the fact that younger patients presented with larger, more advanced tumors (T3/T4), necessitating more extensive surgical procedures.
• Microbial Load and Diversity: Surgery induced a marked reduction in bacterial load and diversity across all patients. However, at the late post-surgery time point (Collection IV), patients with SSI exhibited significantly higher alpha-diversity (Observed, Shannon, and Simpson indices) compared to non-infected patients (p<0.05). Beta-diversity also differed significantly between groups at this stage (weighted UniFrac, p=0.043).
• Microbial Succession and Composition:
  • Non-SSI Patients: Showed a trend toward restoring their preoperative microbial profile within 12 days, with dominant genera like Rothia and Veillonella reappearing.
  • SSI Patients: Exhibited persistent dysbiosis. At the time of infection (Collection IV), Prevotella and Porphyromonas dominated, accounting for ~40% of reads in infected patients versus 9.5% in non-infected patients.
  • Predictive Indicator: Crucially, Aggregatibacter abundance at the early post-surgery time point (Collection III, within 5 days) emerged as a significant predictor of subsequent SSI. An increase of 1% in Aggregatibacter abundance conferred a 1.10-fold increased odds of SSI (p=0.012). Frequencies exceeding 0.044% were associated with a 5.7-fold higher risk.
• Pathogen Specificity: Common culture-based pathogens like Pseudomonas and Staphylococcus were not significantly associated with infection at the genus level in this cohort, suggesting the dysbiosis is driven by specific anaerobic shifts rather than typical opportunistic aerobes.

Significance and Claims
The authors claim that their longitudinal analysis demonstrates that while OSCC surgery profoundly disrupts the oral microbiota, non-SSI patients are capable of restoring their preoperative microbial profile within 12 days. In contrast, SSI is characterized by persistent dysbiosis dominated by Prevotella and Porphyromonas.

The study posits a specific ecological succession model: Aggregatibacter acts as an early "founder" or "keystone" pathogen in the immediate postoperative period. Its presence (specifically >0.044% abundance) appears to alter the local microenvironment, "preparing the soil" for the subsequent outgrowth of obligate anaerobes (Prevotella and Porphyromonas) that characterize the established infection.

The paper concludes that early post-surgical Aggregatibacter abundance may serve as a novel risk indicator for SSI. This finding potentially enables the identification of high-risk patients for timely preventive interventions, such as adjusted prophylactic antibiotic regimens, to disrupt this microbial succession cascade. The authors acknowledge limitations, including the sample size which precluded multiple logistic regression, but emphasize the value of the large, longitudinally collected dataset in tracing these specific microbial profiles.