Vai al contenuto principale

Dott.ssa Sara Ferrini

Phd thesis

Cerebrospinal fluid as a tool to improve diagnosis and treatment in bovine neurology

Achieving an etiological diagnosis in bovine CNS infection is challenging because clinical signs and hematological changes are nonspecific. CSF collection can be easily and safely performed in the field. CSF analysis in the diagnosis of infection (INF) usually shows a moderate to marked increase in total nucleated cell count (TNCC) and microprotein concentration which are not specific for CNS INF, however. Additionally the timely analysis of CSF is essential, yet it is subject to operator proficiency and requires specialized facilities. Furthermore, bacterial identification in CNS INF relies on CSF culture, which lacks sensitivity.
To address these limitations, pioneering approaches might hold potential. Machine learning (ML) is revolutionizing disease diagnosis. Next-generation sequencing (NGS) could bypass the need for CSF culture. CSF lactate is recognized in human as CNS INF biomarker. Lastly, automated CSF analysis could be an alternative to the labor-intensive manual analysis.
Based on these premises, we developed 4 projects (P1, P2, P3, P4) exploiting these pioneering approaches applied on bovine CSF.

Objectives and methods:
With P1 we compared ML models to predict the likelihood of CNS INF in cattle. We created a web app based on the ML model for INF diagnosis. Different ML methods were compared for their ability to predict whether INF was present based on demographics, clinical findings and CSF analysis.
For P2 we exploited 16S rRNA sequencing to compare the CSF microbial composition in cattle with CNS INF and in cattle with other CNS disorders.
With P3 we established a Reference Interval (RI) for CSF lactate and assess its potential as a biomarker for detecting acute CNS INF. Statistical analysis was performed to disclose an association between CSF lactate levels and interpretation of CSF in sick animals.
P4 assessed the Idexx ProCyte Dx® hematology analyzer's capability in analyzing CSF in cattle. We established a RI for ProCyte Dx TNCC and investigated the instrument's performance in delivering TNCC and differential cell count compared to the gold standard laboratory analysis.

P1: 98 cattle with CNS INF and 86 with other CNS disorders were included. All methods had high prediction accuracy (≥80%). The accuracy of the logistic regression model was significantly higher (0.84; AUC 0.91) than the other models and was selected for implementation in a web application.
P2: 4 cattle with CNS INF and 6 with other CNS disorders were enrolled. Bacterial genetic material was identified in 6 and 2 groups were formed: an INF (n = 3) and a non INF group (n = 3). There were no detectable differences in the CSF microbial composition of the samples from the two groups. Pseudomonas (45%) was the most frequently expressed bacterial genus.
P3: 27 healthy and 86 sick cattle with either CNS INF (n=47) or other CNS disorders (n=39) were included. The RI for CSF lactate was 1.1-2.4 mmol/L. Based on a cut-off of 3.15 mmol/L, CSF lactate had diagnostic sensitivity and specificity for neutrophilic pleocytosis of 93% and 80%, respectively (AUC 0.92).
P4: CSF was collected from 113 cattle, 76 with normal TNCC and 37 with increased TNCC. The RI for Procyte Dx TNCC was 0-20 cells/µL. Procyte Dx TNCC correlated well with the laboratory TNCC (rho = 0.89; p < 0.01). Procyte Dx differential cell count was provided only for 4 cattle. Due to the small sample size, no population-wide inferences were performed.

Conclusion and clinical importance: 
Our findings support the use of ML algorithms as tools for veterinarians to improve diagnosis.
The results of 16S rRNA sequencing showed the presence of a microbial community in the CSF in cattle with neurological disorders but further studies are needed.
We determined the RI for CSF lactate in cattle and demonstrated its association with neutrophilic pleocytosis.
Procyte Dx worked well as an alternative for CSF cell counting but software modification are needed to assess differential cell count.
The findings of the projects improve CNS INF diagnosis and correct use of antimicrobial drugs.

1) Stokol T, Divers TJ, Arrigan JW, McDonough SP. (2009) Cerebrospinal fluid findings in cattle with central nervous system disorders: a retrospective study of 102 cases (1990-2008). Vet Clin Pathol. 38(1):103-112. doi:10.1111/j.1939- 165X.2008.00094.x

2) Fecteau G, Parent J, George LW. (2017) Neurologic Examination of the Ruminant. Vet Clin North Am Food Anim Pract. 33(1):1-8. doi:10.1016/j.cvfa.2016.09.001

3) Platt, S. R, and Olby, N. J. (2013). BSAVA manual of canine and feline neurology. Fourth edition. Quedgeley, Gloucester: British Small Animal Veterinary Association.

4) Ramachandran PS, Wilson MR. (2020) Metagenomics for neurological infections - expanding our imagination. Nat Rev Neurol. 16(10):547-556. doi:10.1038/s41582-020-0374-y.

5) Rajkomar A, Dean J, Kohane I. (2019) Machine Learning in Medicine. N Engl J Med. 380(14):1347-1358. doi: 10.1056/NEJMra1814259.

6) Nazir M, Wani WA, Malik MA, Mir MR, Ashraf Y, Kawoosa K, Ali SW. (2017) Cerebrospinal fluid lactate: a differential biomarker for bacterial and viral meningitis in children. J Pediatr (Rio J). 94(1):88-92. doi: 10.1016/j.jped.2017.03.007.



Research activities

Articles on indexed reviews


  • Ferrini S, Rollo C, Bellino C, Borriello G, Cagnotti G, Corona C, Di Muro G, Giacobini M, Iulini B, D’Angelo A. A novel machine learning-based web application for field identification of infectious and inflammatory disorders of the central nervous system in cattle.Journal of Veterinary Internal Medicine 2023. 37 (2), 766-773 Doi: 10.1111/jvim.16664


  • Cagnotti G, Ferrini S, Di Muro G, Avilii E, Favole A, D'Angelo A. Duration of constant rate infusion with diazepam or propofol for canine cluster seizures and status epilepticus. Frontiers in Veterinary Science 2023. 10, 1247100.


  • Borriello G, Valentini F, Rampinelli M, Ferrini S, Cagnotti G, D'angelo A, Bellino C. Ocular Ultrasonography in Healthy Calves with Different Transducers. Animals 2023. 13(4), 742.


  • Borriello G, Cagnotti G, Avedano E, Bergagna S, Iannello P, Di Muro G, Ferrini S, D'Angelo A, Bellino C. Qualitative and quantitative monitoring of antibiotics on dairy cattle farms in relation to animal welfare indicators. Italian Journal of Animal Science 2023. 22:1, 760-768, DOI: 10.1080/1828051X.2023.2241878


  • Ferrini S, Grego E, Ala U, Cagnotti G, Valentini F, Di Muro G, Iulini B, Stella MC, Bellino C, D'Angelo A. Feasibility of 16S rRNA sequencing for cerebrospinal fluid microbiome analysis in cattle with neurological disorders: a pilot study. Veterinary Research Communication 2022. 47(2), 373–383.


  • Cagnotti G, Ferrini S, Di Muro G, Borriello G, Corona C, Manassero L, Avilii E, Bellino C, D'Angelo A. Constant rate infusion of diazepam or propofol for the management of canine cluster seizures or status epilepticus. Frontiers in Veterinary Science 2022. 9, 1005948.


  • Cagnotti G, Ferrini S, Ala U, Bellino C, Corona C, Dappiano E, Di Muro G, Iulini B, Pepe I, Roncone S, D'Angelo A. Analysis of Early Assessable Risk Factors for Poor Outcome in Dogs With Cluster Seizures and Status Epilepticus. Frontiers in Veterinary Science 2020. 7, 575551.


Abstract of international congresses


  • Ferrini S, Cagnotti G, Ala U, Bellino C, Biasibetti E, Borriello G, Corona C, Di Muro G, Iamone G, Iulini B, Pezzolato M, D’Angelo A. Diagnostic accuracy of cerebrospinal fluid lactate levels in cattle with neurological disorders. 34th ECVN-ESVN Symposium (Palma de Mallorca 23rd-24th September 2022) (Flash presentation)
  • Ferrini S, Rollo C, Bellino C, Borriello G, Cagnotti G, Corona C, Di Muro G, Giacobini M, Iulini B, D’Angelo A. Development of a Machine Learning based Web-app for in field identification of central nervous system infectious-inflammatory disorders in cattle. 34th ECVN-ESVN Symposium (Palma de Mallorca 23rd-24th September 2022) (Flash presentation)

Abstract of National congresses

  • Ferrini S, Cagnotti G, Ala U, Bellino C, Biasibetti E, Borriello G, Corona C, Di Muro G, Iamone G, Iulini B, Pezzolato M, D’Angelo A. Accuratezza diagnostica del lattato nel liquido cefalorachidiano in bovini affetti da patologie del sistema nervoso centrale. 54th Congresso Nazionale Società Italiana di Buiatria (6-7 ottobre 2022 Piacenza) (Oral presentation)
  • Ferrini S, Rollo C, Bellino C, Borriello G, Cagnotti G, Corona C, Di Muro G, Giacobini M, Iulini B, D’Angelo A. Sviluppo di una web-app per l’identificazione dei disturbi infettivo-infiammatori a carico del sistema nervoso centrale nel bovino. 54th Congresso Nazionale Società Italiana di Buiatria (6-7 ottobre 2022 Piacenza) (Oral presentation)
  • Ala U, Ferrini S, Bianchi P, Bertero A, Colitti B, Cagnotti G, Bertolotti L. Microbiome and biological interaction: can sterile districts really be considered as such? XX Congresso Nazionale A.I.B.G. (Roma, 23-24 settembre 2022) (Oral presentation)
  • Ferrini S. Machine Learning algorithms as diagnostic tools for infectious-inflammatory central nervous system disorders in cattle. Game of Research Event (16th December 2021 Turin) (Oral presentation).
Last update: 23/09/2023 15:07

Non cliccare qui!