ABSTRACT
Introduction: Evidence from clinical trials strongly supports the safety and efficacy of the different COVID-19 vaccines. Indeed, the risk to develop a severe form of the disease, possibly leading to death, it is highly decreased in fully vaccinated individuals. Nowadays, vaccines effects and their possible ability to stimulate an autoimmune reaction are still poorly understood. The aim of this study was to check the development and /or persistence of antinuclear antibodies (ANA) in healthcare workers (HCPs) after mRNA based anti-SARS CoV-2 vaccines. Method(s): In this study, 77 HCPs were considered (60 females and 17 males, age range 26-67 years, median age 48) without any history of COVID-19 infection. All the subjects were vaccinated with 2 doses of BioNtech/Pfizer BNT162b2 mRNA. Furthermore, half of them received a third dose of the same vaccine, whereas the other half of Moderna (Spikevax). Blood Samples were collected before the inoculation of the vaccine (T0), at 3 (T1) and 12 months (T2) after the first dose. Therefore, at T1 all the subjects received two doses of vaccine and at T2 three doses. ANA presence was evaluated using indirect immunofluorescence on Hep-2 cells (EUROIMMUN test kit) at dilutions: 1:80, 1:160, 1:320, 1:640. Fisher and Wilcoxon statistical tests were performed using GraphPad Prism 9 Software. Result(s): Among 77 subjects enrolled, at T0 25 were positive for ANA (23 maintained this positivity also at T1 and T2) and 52 were negative. At T1, 46/52 remained negative, whereas 6/52 became ANA positive (5 maintained this positivity also at T2). At T2, 30/46 were still negative, instead 16/46 became ANA positive. In addition, from T1 to T2, it has been observed a statistically significant increase of ANA presence. Conclusion(s): Our results suggest that mRNA based anti-SARS CoV-2 vaccines seem to induce the onset of de novo ANA in 22/77 (28,57%) of subjects and that the percentage of positivity seems to directly correlate to the number of vaccine expositions: 6/77 (7,79%) after 2 doses;and 16/77 (20,78%) after 3 doses.
ABSTRACT
In 2020, severe coronavirus 2 respiratory syndrome (SARS-Cov-2) has quickly risen, becoming a worldwide pandemic that is still ongoing nowadays. Differently from other viruses the COVID-19, responsible for SARS-Cov-2, demonstrated an unmatched capability of transmission that led towards an unprecedented challenge for the global health system. All health facilities, ranging from Hospitals to local health surveillance units, have been severely tested due to the high number of infected people. In this scenario, the use of methodologies that can improve and optimize, at any level, the management of infected patients is highly advisable. One of the goals of Artificial Intelligence in medicine is to develop advanced tools and methodologies to support patient care and to help physicians and medical work in the decision-making process. More specifically, Machine Learning (ML) methods have been successfully used to build predictive models starting from clinical patient data. In our paper, we study whether ML can be used to build prognostic models capable of predicting the potential disease outcome. In our study, we evaluate different unsupervised and supervised ML approaches using SARS-Cov-2 data collected from the "Azienda Ospedaliera SS Antonio e Biagio e Cesare Arrigo"Hospital in Alessandria area, Italy, from 24th February to 31st October 2020. Our preliminary goal is to develop a ML model able to promptly identify patients with a high risk of fatal outcome, to steer medical doctors and clinicians towards the best management strategies. © 2021 IEEE.