Description and comparison of PIMS-TS innate cell signature and immunophenotype with a cohort of healthy children, severe viral and bacterial infections and Kawasaki Disease

A new clinical syndrome associated to SARS-CoV-2 has been described in children. It has been named as Pediatric Inflammatory Multisystem Syndrome Temporally Associated with SARS-CoV-2 (PIMS-TS). This new disease is a main cause of hospital and pediatric intensive care unit (PICU). In this work we describe the innate cell signature and immunophenotype of children admitted to PICU because of PIMS-TS. Also, we compare it with healthy controls and children admitted to PICU because bacterial infection, viral infection and Kawasaki disease. We made a prospective-retrospective observational study in a tertiary pediatric hospital. Children admitted to PICU because of PIMS-TS from March 2020 to September 2020 were consecutively included. They were compare with previous cohorts from our center. A total of 247 children were included: 183 healthy controls, 25 viral infections, 20 bacterial infections, 6 Kawasaki disease and 13 PIMS-TS. PIMT-TS showed the lowest percentage of lymphocytes and monocytes with higher relative numbers of CD4+ (p =0,000). At the same time, we describe a differential expression of CD64, CD11a and CD11b. Monocytes and neutrophils in PIMS-TS showed higher levels of CD64 expression compared to all groups (p = 0,000). Also, proteins involved in leukocyte tissue migration, like CD11a and CD11b were highly expressed compare to other severe viral or bacterial infections (p = 0,000). In PIMS-TS this increased CD11a expression could be a sign of the activation and trafficking of these leukocytes. These findings are congruent with an inflammatory process and the trend of these cells to leave the bloodstream. In conclusion, we compare for the first time the innate cellular response of children with PIMS-TS with other severe forms of viral or bacterial infection and Kawasaki disease. Our findings define a differential cell innate signature. These data should be further studied and may facilitate the diagnosis and management of these patients.


Description and comparison of PIMS-TS innate cell signature and immunophenotype with a cohort of healthy children, severe viral and bacterial infections and Kawasaki Disease.
A new type of affectation temporarily linked to the new coronavirus SARS-CoV-2 in childhood has been described. Europe, Great Britain and the United States have been the regions with the higher number of cases [1][2][3]. This new clinical syndrome has been named as Pediatric Inflammatory Multisystem Syndrome Temporally Associated with SARS-CoV-2 (PIMS-TS). It has clinical and analytical similarities to Kawasaki disease and suppose a main cause of hospital and pediatric intensive care unit (PICU) admission [4].
A majority of PIMS-TS cases are not related to active infections. An immune basis has been proposed for its establishment. In this point, there has been described the presence of immunity dysregulation or the release of autoantibodies. Related to that, immunoglobulin and corticoids seem to have a preeminent role. The clinical response is usually quick and with improvement in a short period of time [2,5].
Related to the probable role of leukocyte dysregulation the study of the cellular response in PIMS-TS could be of interest. Also, its comparison with healthy children and other causes of PICU admissions may help to describe and know the basis of their inflammation. Nowadays there are scarce data about this in children [6].
In this work we describe the innate cell signature and immunophenotype of children admitted to PICU because of PIMS-TS. Also, we compare it with healthy controls and children admitted to PICU because bacterial infection, viral infection and Kawasaki disease.

Material and methods
Prospective-retrospective observational study conducted in a tertiary pediatric hospital after Ethics Committee for clinical research approval. Done prospectively in children admitted to PICU because of PIMS-TS from March 2020 to September 2020 were consecutively included. They were compare with previous cohorts from our center (see above). In all cases one peripheral blood sample was extracted after parents or legal guardians consent at PICU admission. A previously established intravenous line was always used. The volume obtained was 0.5 ml per sample and collected in sterile EDTA tube. The sample handling was based on item 59 of the Spanish law on Biomedical Research. Study was designed to not influence the treatment of the participating patients.

Sample processing and analysis by flow cytometry
Samples were collected in sterile EDTA at room temperature or refrigerated at 4°C, used for CD45+

Cohorts of analysis
Four cohorts of analysis were described:

Results
A total of 247 children were included: 183 healthy controls, 25 viral infections, 20 bacterial infections, 6 Kawasaki disease and 13 PIMS-TS. In the PIMS-TS group 5 children were RT-PCR positive, 7 were Ig G positive and one RT-PCR and IgG positive at diagnosis. There were sex differences only in KD group. There were observed differences in age (Table 1). A negative correlation for lymphocytes with age was observed (r= -0,240, p = 0,001). The CD64 expression in monocytes and neutrophils was not correlated with age or influenced by sex. The percentage of leukocytes and MFI of each type of leukocytes are described in table 1 and Figure 1. The significative differences between PIMS-TS and other groups are showed in Figure 1.

CD64 and CD11a neutrophils expression utility as a PIMS-TS biomarker
To study the usefulness of CD64 surface expression as a tool to predict PIMT-TS we evaluated the receiver operating characteristics curve. As seen in Figure 2 the mCD64, nCD64 and nCD11a area under the curve (AUC) that were near to 1. For the mCD64 it was 0,994 (p=0,000; with a cut point of 29098 MFI and 94,1% specificity and 100% sensitivity). For the nCD64 it was 0,992 (p=0,000; with a cut point of 8753 MFI and 90% specificity and 100% sensitivity). The nCD11a AUC was 0,992 (p=0,000; with a cut point of 5203 MFI and 91,7% specificity and 94,1% sensitivity).

Discussion
This paper compares for the first time the innate cellular signature and immunophenotyping of severe PIMS-TS cases with a large cohort of healthy control, other severe infections diseases and KD. In peripheral blood samples, we observed differences in almost all leukocyte populations. The most visible of these differences affect lymphocytes and monocytes, which have the lowest values in PIMS-TS. At the same time, we describe a differential expression of CD64, CD11a and CD11b.
These leukocyte surface proteins were exceptionally high in PIMS-TS in monocytes and neutrophils.
As said, the cases included showed a different distribution of leukocyte populations. This distribution of leukocyte populations is strikingly different in lymphocytes (Table 2 and Figure 1).
PIMT-TS showed the lowest percentage of lymphocytes with higher relative numbers of CD4+. This lymphopenia has already been described in the adult population and severe critical children because of SARS-CoV-2 infections or PIMS-TS [2,7]. Compare to other severe viral infections, we observed that there was also a low percentage of monocytes and neutrophils (Figure 1). It is known that PIMS-TS is usually not associated with active SARS-CoV-2 infection. An immune dysfunction hypothesis has been proposed and explored in this children disease. Monocytes, neutrophils and lymphocytes are critical cells in viral first response. Their migration and accumulation in infected tissues added to the SARS-CoV-2 capacity to dysregulate this response may cause this low cell count in peripheral blood [6]. Our group has previously described an increase in leukocyte CD18/CD11a complex (LFA-1) expression in two short series of PIMS-TS [8,9]. It is congruent with a state of increased cellular predisposition to leave the bloodstream. We could not analyze the LFA-1 expression in this paper because it was not available in all groups.
Concerning immunophenotyping, we should highlight the findings observed about the expression of CD64, CD11a and CD11b. About CD64 expression, the PIMS-TS showed higher levels compared to all groups. The lowest CD64 value was observed in viral infections. The CD64 indirectly reflect cytokine expression and activation. As shown in Figure 2 the CD64 levels are higher in PIMS-TS than in severe bacterial infections and KD. This CD64 expression affects monocytes and neutrophils and may inform about the hyperinflammatory status in PIMS-TS [7,8].
Compared to KD we observed differences in monocytes but not in neutrophils. The study of CD64 and CD11a expression in these leukocytes could help in the differential diagnosis of PIMS-TS as is explained in Figure 2.
To study the proteins involved in leukocyte tissue migration, we examined the percentage of CD11a and CD11b positive cells. In our series, we observed that both proteins were higher in neutrophils and monocytes than in viral or bacterial infections ( Figure 1). Also, they were higher than KD but without differences. Related to CD11a, the MFI was also higher in neutrophils of PIMS-TS cases.
In adults, the inflammation in the basis of SARS-CoV-2 showed a predominant presence of macrophages and neutrophils in the affected territory. In PIMS-TS this increased CD11a expression could be a sign of the activation and trafficking of these leukocytes. These findings, added to the previously commented, are congruent with an inflammatory process and the trend of these cells to leave the bloodstream. This supports the inflammatory clinical findings but also confirm the utility of anti-inflammatory drugs as a cornerstone in the management of these children [10][11][12][13].
This work has limitations. We observed age differences between the groups analyzed ( Table 1).
The distribution of leukocyte populations is influenced by this. Despite this, we observed internal homogeneity between groups, which strengthens the observations. Also, the expression of CD64, CD11a and CD11b did not show correlations with age or differences by sex. Finally, the possible relationship between the observed data and the PIMS-TS clinical courses were not studied. This was not the aim of this work and should be considered in future works.
In conclusion, we compare for the first time the innate cellular response of children with PIMS-TS with other severe forms of viral or bacterial infection. Also, we describe differences with children with KD. Our findings define a differential cell innate signature with consistent data related to the possible presence of inflammation and immune dysregulation. These data should be further studied and may facilitate the diagnosis and management of these patients.