Scientists explore pulmonary aspergillosis associated with COVID-19

In a recent study published on the bioRxiv* preprint server, researchers computationally investigated the causes of coronavirus disease 2019 (COVID-19) associated with pulmonary aspergillosis (CAPA).

Study: Pulmonary aspergillosis associated with COVID in immunocompetent patients. Image credit: Nhemz/Shutterstock

Aspergillus fumigatus, an opportunistic fungal pathogen, infects the lungs of immunocompromised individuals, including patients with chronic granulomatous disease and those undergoing organ transplantation or chemotherapy. Recently, patients with severe COVID-19 have been affected by CAPA. It has been reported that 10% of intensive care patients with COVID-19 develop invasive aspergillosis.

One study observed that conventional risk factors for aspergillosis were absent among CAPA patients, who were predominantly male, and 88% of them had comorbidities. The fact that CAPA is related to comorbidities and not to conventional risk factors for aspergillosis justifies newer hypotheses explaining the susceptibility of these hosts to CAPA.

About the study

In the present study, the researchers hypothesized and tested that in severely affected patients with COVID-19, the destruction of the lung epithelium facilitates colonization by opportunistic pathogens and that the depletion of the immune system hinders the ability to kill the fungus . They used a computational model to test the hypothesis and identify the parameters/conditions that explain CAPA.

A virtual patient population was created by sampling the parameter space of the model, with a (virtual) patient represented by a selection of parameters. The team used a previously published model, which simulated the innate response to A. fumigatus in immunocompetent subjects. The same model was modified to simulate A. fumigatus infection in a host with concurrent viral pneumonia.

In this model, neutrophils would kill hyphae by direct contact, and in an alternative scenario, granules secreted by neutrophils diffuse and kill the fungus. Neutrophils become active upon interaction with the fungus and secrete granules, the contents (molecules) of which would kill the hyphae at a certain rate and disintegrate with a half-life.

The study focused on five parameters: 1) intrinsic growth rate of hyphae, probability of 2) neutrophils and 3) monocytes to kill hyphae, the number of 4) monocytes and 5) neutrophils. The simulation analysis started with these small changes. A regular grid of 10 x 10 x 10 voxels was initially used with 640 type II pneumocytes and 20 germinated conidia. The number of neutrophils and monocytes was variable (from 360 to 960).

In general, the number of leukocytes was constant throughout the simulation. Each simulation was run for 24 hours or 720 iterations. The researchers performed Latin hypercube sampling (LHS) on the five parameters and generated 24,000 parameter sets (representing the equivalent number of dual-infected virtual hosts). Another 36,000 virtual hosts were created using the alternative neutrophil mechanism (granule secretion).

They measured the observed growth rate, defined as the collective effect of the intrinsic growth rate and the counter-effects of the immune response. The intrinsic growth rate described the time required to create new fungal cells, while the observed growth rate was the number of cells generated minus cells killed by leukocytes. The partial rank correlation coefficient of the five parameters was calculated; partial correlations were squared and 100 bootstraps were performed.


The observed growth rate distribution showed a prominent negative peak, indicating resolution of fungal infection, and a smaller peak with a positive tailed observed growth rate. Therefore, the patients were grouped into two groups, one where the patients cleared the infection and the other where the infection did not resolve.

The second cluster was subdivided into three groups representing progressively worsening infection. Clusters 2B and 2C represented CAPA patients. Cluster 1A had a negative observed growth rate and represented patients who cleared the infection but could develop CAPA. Groups 2B and 2C indicated that higher intrinsic growth and lower immune response could explain the occurrence of CAPA in immunocompetent subjects.

Additionally, classification trees were constructed to identify essential parameters for CAPA. This revealed that the dominant parameters were the intrinsic growth rate and the leukocyte and monocyte death rates. The virtual patient population was represented in three-dimensional space using the three parameters. The plot showed that clusters 1 and 2A intercalated and cluster 2C segregated into a zone of high intrinsic growth rate.

The authors then analyzed variations in the CAPA probability by varying the two most influential parameters (intrinsic growth rate and neutrophil death rate). The plot illustrated that the probability of CAPA increased with a higher intrinsic growth rate or a lower neutrophil death rate. The correlation (partial ranking) between the observed growth rate and the neutrophil death rate was negative, while the correlation between the observed growth rate and other parameters was insignificant.


Simulation analysis showed that 31% of the virtual population of COVID-19 patients exposed to the fungus developed CAPA. This was qualitatively consistent with empirical figures that only 10% of patients with COVID-19 in intensive care developed CAPA. This discrepancy could be explained by the fact that not all hospitalized patients with COVID-19 are exposed to the fungus.

However, in the alternative scenario (secretion of granules), only 17% of patients developed CAPA. The team observed that neutrophils efficiently regulated infection when the intrinsic growth rate was moderate/low. In conclusion, the authors proposed that CAPA was probably due to increased lung permissiveness due to better nutrient supply and reduced neutrophil effectiveness (to kill the fungus).

*Important news

bioRxiv publishes preliminary scientific reports that are not peer-reviewed and therefore should not be considered conclusive, guide clinical practice/health-related behavior, or be treated as established information.

Leave a Comment

Your email address will not be published. Required fields are marked *