Computer model predicts COVID-19 recovery

An international collaboration that tracked 215 patients infected with COVID-19 for up to a year post-infection has developed a model that predicts individual recovery chances.

The researchers measured a range of immune system cells and metabolic parameters in blood plasma to estimate the probabilities of unvaccinated patients making a full recovery from the disease. They discovered distinct ‘systemic recovery’ profiles, with specific progression and recovery of the inflammatory, immune cell, metabolic and clinical responses to COVID-19.

The predictive model applied to each individual is available here for other researchers.

The finding, published in Nature, Immunology, is a result of a three-year partnership between the University of Cambridge, The Cambridge Institute of Therapeutic Immunology and Infectious Disease-National Institute of Health Research (CITIID-NIHR) BioResource COVID-19 Collaboration and Australian National Phenome Centre (ANPC).

Professor Jeremy Nicholson, Pro-Vice Chancellor for Health Sciences and Director of the ANPC, said, “We have taken advantage of very well collected and curated patient blood samples from Cambridge University’s Addenbrooke’s Hospital in the UK, and combined our state-of-the-art technology at the ANPC to produce a uniquely detailed individual mapping of the disease progression and outcomes.”

“The research highlights multiple new features of the disease, which is highly variable in terms of severity and persistence.

“Importantly, the interplay between the immune system and the blood metabolic profiles gives deep insights into how individuals vary and how that variation impacts on long-term outcomes, including Long COVID or even death.

“What is worrying, is that the long-term disease trajectory appears to be set very early during the course of infection. That means for unvaccinated patients (all the patients in the study were unvaccinated) urgent action would be needed to mitigate the long-term negative consequences of the disease.”

The ANPC has already begun improving on this model to make it suitable for deployment in the clinic.

“Our next task is to operationalise this model by reducing it down to the most critical parameters and ones that can be measured rapidly ideally using only one analytical tool, and that is what we are working on now,” Nicholson said.

Explaining the model further, Professor Julien Wist, Deputy Director of the ANPC, said, “The model emphasises the importance of the different types of immune cell numbers and specific amino acid and lipoproteins metabolism pathways that are disturbed as a result of the inflammatory effects of COVID-19.”

“We have published the data and the mathematical model developed with Cambridge — currently for research use only — and it is available online to anyone who wants to examine how the different parameters interact.”

Professor Peter Klinken, Chief Scientist of Western Australia, said that through international collaborations such as these with Cambridge University, the ANPC and Western Australian science continues to have an impact on the world stage in critical areas of unmet medical need.

This latest research is being augmented and validated by thousands of blood plasma samples collected at Harvard University, which is also working closely with the ANPC to predict Long COVID outcomes.

The Australian funding for this research was provided by the WA Government, Spinnaker Health Research Foundation, the McCusker Charitable Foundation and the Medical Research Future Fund, as well as UK Medical Research Council and National Institute of Health Research. The ANPC is part of Murdoch University’s Health Futures Institute.

Image credit: iStockphoto.com/recep-bg