The value of biofilm testing to guide antimicrobial stewardship in chronic respiratory diseases

Abstract

Introduction: Biofilm production is an important yet currently overlooked aspect of diagnostic microbiology that has implications for antimicrobial stewardship. In this study, we aimed to validate and identify additional applications of the BioFilm Ring Test® (BRT) for Pseudomonas aeruginosa (PA) isolates from patients with bronchiectasis (BE).

Materials and methods: Sputa were collected from BE patients who had at least one PA positive culture in the previous year. We processed the sputa to isolate both mucoid and non-mucoid PA, and determined their susceptibility pattern, mucA gene status, and presence of ciprofloxacin mutations in QRDR genes. The Biofilm production index (BPI) was obtained at 5 and 24 hours. Biofilms were imaged using Gram staining.

Results: We collected 69 PA isolates, including 33 mucoid and 36 non-mucoid. A BPI value below 14.75 at 5 hours predicted the mucoid PA phenotype with 64% sensitivity and 72% specificity.

Conclusion: Overall, our findings suggest that the fitness-cost associated with the mucoid phenotype or ciprofloxacin resistance is shown through a time-dependent BPI profile. The BRT has the potential to reveal biofilm features with clinical implications.

Keywords: Pseudomonas aeruginosa; antimicrobial agents; antimicrobial resistances; biofilm; biofilm diagnose.

Figure 1

Three different stages of PA mode of growth in sputum observed by light microscopy (100X). (A) Planktonic stage in which Gram-negative bacilli are seen free floating in planktonic mode of growth (circled an individual gram-negative bacilli). (B) Gram-negative bacilli grow aggregated forming immature biofilms. (circled an aggregate of gram-negative bacilli inside mucus) (C) Gram-negative bacilli can be found aggregated embedded in an optically distinguishable alginate extracellular matrix circled in the image.

Figure 2

The Biofilm index of mucoid and non-mucoid PA phenotypes and its time-dependent increase during incubation. (A) Boxplot showing BPI of the 69 PA isolates, by mucoid and non-mucoid PA, read at 5h as recommended by manufacturer. Median and interquartile ratio are represented by the box and whiskers show the maximum and minimum values. The non-mucoid vs mucoid PA phenotype is associated with an increased BPI (19,08 [10,63-20,00] vs 12,36 [5,55-18,74] p=0.006, respectively). (B) BPI at 5 vs 24h by phenotype. Median is represented by the central line whilst interquartile ratio is represented by the two lines at the extremes. A statistically significant increase of the BPI can be seen in mucoid strains when incubating at 24 h whilst a greater heterogenous non-statistically significant result is achieved in non-mucoid strains).

Figure 3

BPI in accordance to antibiotic resistance. The boxplot represents median and interquartile ratio, whiskers represent minimum and maximum values. Differences in the BPI in accordance to antibiotic resistance pattern where only statistically significant comparing ciprofloxacin resistant vs. susceptible PA (12,94 [6,58-19,37] vs. 19,38 [9,73-20,00], p=0.039, respectively).

Figure 4

BPI of mutant vs wild-type strains for mucA gene. This figure shows how mutant strains present a reduced BPI when compared to wild-type strains being this last population much more heterogeneous in BPI testing results.

Figure 5

ROC curves for sensitivity and specificity. (A) ROC curve to assess the best cut-off point of the BPI for PA phenotype determination. (B) ROC curve to assess the best cut-off point of the BPI for PA resistance to ciprofloxacin.

Figure 6

Biofilm dynamic during time and phenotype switch. The figure shows how the metabolic ratio of PA decreases whilst phenotype switches from a non-mucoid state to a mucoid phenotype and at the same time PA starts growing in biofilms instead of being found in planktonic state.