Medical prevention of congenital pneumonia in preterm newborns: a microbiome-associated approach

Objective. To scientifically substantiate a microbiome-associated approach to medical prevention of congenital pneumonia in premature newborns based on determination of the oropharyngeal microbiome, and identification of signs of chronic intrauterine hypoxia. Materials and methods . 75 premature newborns with congenital pneumonia against the background of chronic intrauterine hypoxia (the main group) and 79 infants without congenital pneumonia with infectious diseases specific for the perinatal period (the comparison group) were examined. A new generation sequencing was performed with the MiSeq (Illumina) apparatus. Statistical data processing was carried out in the R programming environment (version4.3.1), RStudio program (2023.09.1+494). The significance level was accepted equal to 0.05. Results. The median gestational age in the main group was 28.00 [26.00; 30.00] weeks, and in the comparison group it was 33.00 [31.00; 35.00] weeks, with p-val <0.001. During the sequencing process of the oropharyngeal biomaterial, microbiome-associated biomarkers of congenital pneumonia were identified in premature newborns at the genus level: Brucella (≥5.8%, Se=0.88 and Sp=0.57); Achromobacter (≥ 3.1%, Se=0.667 and Sp=0.658); Ralstonia (≥0.3%, Se=0.653 and Sp=0.709); Stenotrophomonas (≥ 9.0%, Se=0.64 and Sp=0.671); Klebsiella (≥ 0.02%, Se = 0.693 and Sp=0.595); Pseudomonas (≥1.5%, Se=0.6 and Sp=0.684). Obtaining one or more microbiome-associated biomarkers in the form of the above-mentioned bacteria in a nasopharyngeal sample is a basis for definition of pneumonia probability in premature newborns within the method for medical prevention of pneumonia in premature newborns. Conclusion . With regard to identified microbiome-associated biomarkers (Brucella ≥5.8%, Achromobacter ≥3.1%, Ralstonia ≥ 0.3%, Stenotrophomonas ≥9.0%, Klebsiella ≥0.02%, Pseudomonas ≥1.5%) and the level of hypoxia-inducible factor (HIF-1-alpha ≥0.017 ng/ml), a medical software has been developed on the ground of artificial neural networks allowing to determine likelihood of having congenital pneumonia in premature newborns. Based on the identified biomarkers, the “Method of Medical Prevention of Pneumonia in Premature Newborns” has been developed and implemented in practical healthcare, approved by the Ministry of Health of the Republic of Belarus as of 26.05.2025, with the registration number 005-0225, and in the form of an instruction for use.