Published online Apr 27, 2022. doi: 10.4254/wjh.v14.i4.719
Peer-review started: May 18, 2021
First decision: July 8, 2021
Revised: July 20, 2021
Accepted: March 25, 2022
Article in press: March 25, 2022
Published online: April 27, 2022
Processing time: 338 Days and 20.4 Hours
Classification of the pattern of periportal fibrosis (PPF) is essential in the prognostic evaluation of patients with Schistosomiasis mansoni.
There is a need for novel minimally invasive methods and new biomarkers for the diagnosis Schistosomiasis mansoni.
To develop metabolic models, based on 1H-nuclear magnetic resonance spectra, that allow the classification of the pattern of PPF and its associated metabolites in patients with Schistosomiasis mansoni.
Metabonomics models (MMs) were built to differentiate requirements with mild PPF and significant PPF. An analysis of the performance of MMs was performed for the prediction of PPF, using ultrasonography as a reference standard and the description of the main metabolites present in each PPF group and their relationship with serum markers.
The partial least squares-discriminant analysis (PLS-DA) and orthogonal projections to latent structures discriminant analysis (OPLS-DA) formalisms discriminated spectral regions between the groups as follows: carbohydrates and valine, more concentrated in those of the group with mild FPP; N-Acetylglycosamines, Alanine, Glycolaldehyde more concentrated in the samples of the group with significant PPF. OPLS-DA showed accuracy, sensitivity, and specificity, were equal to 92.7%, 90.3%, and 100% for the diagnosis of significant PPF.
The constructed MMs were able to discriminate between mild and significant FPP in patients with schistosomiasis with good accuracy.
This technique will be able to detect even low-intensity infections, overcoming the limitations of current diagnostic techniques, with the use of a single serum sample. These models can be inserted in the propaedeutic arsenal in clinical practice for the measurement of PPF in remote areas.