INTRODUCTION

Infection with the respiratory SARS-CoV2 virus causes the coronavirus disease 2019 (COVID-19) which ranges from asymptomatic infection through mild and severe disease (pneumonia) to critical disease which has the form of an acute respiratory distress syndrome (ARDS) or sepsis and septic shock (1). The range of severity is reflected in the mortality rates which reach approx. 40% in critically ill patients (2). Although other coronaviruses can also infect humans, COVID- 19 is characterized by a unique pathogenesis, making it a distinct disease (3). Aside from respiratory support, the only widely accepted treatment for COVID-19 is dexamethasone and recently tocilizumab (1).

There is an urgent need to develop efficient therapies for COVID-19 and it has become apparent that, as in the case of bacterial sepsis, a personalized biomarker-guided approach is a prerequisite for successful biological therapies (4). Since early in the COVID-19 pandemic, we and others have proposed that unbalanced inflammasome signaling may be a key process in the pathogenesis of COVID-19 (5, 6). Briefly, inflammasomes are multiprotein complexes activated by various stimuli such as dsDNA (Absent in Melanoma [AIM2]), potassium efflux and reactive oxygen species (NLR Family Pyrin Domain Containing 3 [NLRP3]) (7). Upon activation, the receptor proteins oligomerize and recruit adaptor proteins (apoptosis-associated speck-like proteins [ASC]) and pro-caspase-1. Then, pro-caspase-1 undergoes autoproteolysis to its active form, which then can cleave pro-interleukin-1α, β, –18, and gasdermin D (GSDMD). N-terminal GSDMD can form pores in the cell membrane, which leads to the release of the cleaved cytokines as well as other molecules such as lactate dehydrogenase or alarmins and lytic cell death called pyroptosis (7). Indeed, it has been demonstrated that inflammasomes are activated in the lungs of COVID-19 patients (8, 9) and can be activated in circulating monocytes (10). However, the status of inflamma- some reactivity in the blood myeloid cells appears complex (11). Although some studies showed an increase in the concentrations of pro-inflammatory cytokines with an increasing severity of COVID-19 (12), there are still limited studies focusing on critically ill COVID-19 patients who are at the highest risk of death. Consequently, there are ongoing clinical trials with direct (Dapansutrile, Colchicine) and indirect (Disulfiram, a GSDMD inhibitor) inflammasome inhibitors. In addition, two studies on the IL-1R antagonist in COVID-19 were recently published with contradictory results (13, 14).

In this study, we aimed to evaluate the prognostic utility of selected plasma proteins related to inflammasome activation in critically ill COVID-19 patients. Considering the high mortality rate and lack of effective therapies we focused on critically ill COVID-19 patients upon admission to the intensive care unit (ICU). We hypothesized that accurate biomarkers could serve to guide immunomodulatory therapies that targetinflammasomes.

PATIENTS AND METHODS Patients

We conducted a prospective observational cohort study. We prospectively collected peripheral blood samples from 45 patients with critical COVID-19 on admission to the ICU of the Wroclaw Medical University between October 2020 and March 2021. Critical COVID-19 was defined according to the COVID-19 treatment guidelines (1). Briefly, the patients met criteriaforARDS, sepsis, or septic shock, and required the introduction of life-sustaining therapies (1). The SARS-CoV2 infection was confirmed by a real-time reverse-transcriptase polymerase chain reaction (RT-PCR) assay of nasal or pharyngeal swab probes. All patients were treated according to WHO guidelines (1). All patients were administered intravenous dexamethasone at a dose of 6 mg per day before ICU admission and sample collection and this therapy was continued in the ICU. The study protocol complies with the 1975 Declaration of Helsinki as revised in 1983. The Wroclaw Medical University Bioethics Committee approved this study (consent No. 394/2021). The Committee also approved the collection of plasma from traumatic brain injury (TBI) patients (consent No.KB-391/2015). Written informed consent was obtained from the patient or a legally authorized representative. In addition, blood samples collected from patients with TBI were used to compare markers of inflammasome activation between cases with SARS-CoV2 infection (the COVID-19 group) and cases without any signs of infection (the TBI group). Blood was collected from the TBI patients within 24 h of ICU admission. In 60% of the TBI patients %) multiple locations of hemorrhagic lesions were observed on the computed tomography scan of the head: subarachnoid (n = 3), subdural (n = 8), intracerebral (n = 5) hemorrhage were diagnosed. The clinical status of the patients was determined with the Acute Physiology and Chronic Health Evaluation (APACHE) II score and Sequential Organ Failure Assessment (SOFA) score on admission to the ICU. The APACHE II score is routinely used as a predictive tool for ICU patients and includes 12 physiological variables (the fraction of inspired oxygen, partial pressure of oxygen, body temperature, mean arterial pressure, blood pH, heart rate, respiratory rate, serum sodium, serum potassium, serum creatinine, hematocrit, white blood cell count, and the Glasgow Coma Scale) and two disease-related variables (a history of having severe organ failure or being immunocompromised and the type of ICU admission). The SOFA score is routinely used in the ICU for monitoring the severity of a patient's clinical condition based on the status of the following systems: respiratory (PaO2/FiO2 index), cardiovascular (mean arterial pressure and the dose of vasopressors), hepatic (bilirubin level), coagulation (platelets level), renal (creatinine level/urine output), and neurological (Glasgow coma scale).

Cytokine analysis

Blood samples were collected via routinely inserted arterial cannula. 2.7 mL of peripheral blood was collected into tubes containing 0.109 M sodium citrate (BD Vacutainer, BD, NJ). Plasma was collected after centrifugation at 2,000 × g for 10 min and stored at –70°C. IL-1α, IL-1β, IL-18, IL-1RA, and galectin-1 concentrations were analyzed by a Milliplex Human Multiplex Assay (Merck, Germany) using the Magpix System. ACS protein was analyzed using the commercially available ELISA kit (Cusabio Technology, Tex). Human GSDMD was analyzed using a commercial ELISA kit (Abclonal, Mass). Other analytes were measured at the certified hospital clinical laboratory.

Caspase-1 immunoprecipitation

Equal volumes of plasma (200 μL) from the control group (TBI) and the COVID-19 patients were treated with 50 μM Biotin-FAD-FMK (SantaCruz Biotechnology, Calif) with gentle rotation at 4°C overnight. Then, 10 μL of Streptavidin Sepharose Bead Conjugate (Cell Signalling Technology, Mass) was added to the plasma samples and incubated for 5 h with gentle rotation at 4°C. The biotin-streptavidin bead complexes were centrifuged for 5 min at 2,300 × g and washed five times using 1 × Ripa Lysis buffer (Millipore) supplemented with a Protease Inhibitor Cocktail (Millipore) and 1 mM PMSF. Pellet samples were suspended in 4 × Laemmli buffer and boiled for 3 min at 94°C. For the detection of the active forms of caspase-1, the pulled down immunocomplexes were separated in 15% SDS—PAGE gels followed by wet electrotransfer onto PVDF membranes. Non-specific bindings were blocked in TBS-T with 5% non-fat milk for 1 h at room temperature. Membranes were incubated with anti-Caspase-1 (cleaved 297; 1:1,000 dilution; GenTex) overnight at 4°C. Rabbit polyclonal antibody (1: 10,000 dilution; Vector Laboratories, Inc, CA) was used to detect the primary antibody. Signals were visualized by a WesternBright ECL HRP substrate (Advansta) using the UVITEC gel documentation system. Densitometric analysis was done by GelAnalyzer 19.1. All samples were analyzed simultaneously.

Statistical analysis

Normality of the data was tested using the Shapiro-Wilk test. Continuous variables are expressed as median values and interquartile ranges, whereas categorical variables are presented as frequencies with percentages. Categorical variables were compared using Fisher exact test. Continuous variables were compared using the Mann—Whitney test. Receiver operating characteristic (ROC) curves were constructed to calculate the area under the curve (AUC) of the investigated parameters. The cut-off values were indicated using You- den's index method. Correlations between variables were tested using Spearman test. Logistic regression analysis was performed to test the relationship between biomarkers and the risk of death or the development of secondary infections. Statistica 13.1 software was used for the calculations and GraphpadPrism 9.0 was used to create the graphs. Statistical significance was determined as P < 0.05.

RESULTS Clinical characteristics of the patient population

We enrolled 45 critically ill COVID-19 patients with an RT- PCR confirmed SARS-CoV2 infection. The clinical and demographic characteristics of the group are presented in Table 1. This group included 24 males (53%) and 21 females and the median age was 58 years (IQ: 43–68). On admission, the median APACHE II score of the patients was 14 (IQ: 1122) and the SOFA score was 9 (IQ:7–10). Hospital mortality reached 62% and was higher than the 28-day mortality, which was 53%. The patients who did not survive were significantly older (median 64 (IQ:52–70) vs. 50 (IQ: 33–60) yrs) and had a higher APACHE II score in comparison with the survivors (20 (IQ:14–25) vs. 12 (IQ:8–13)). The most common comorbidities were hypertension and cancer; however, their occurrence was not related with survival. Extra corporeal membrane oxygenation therapy was used in 29% of non-survivors and 18% of survivors. Twelve patients were admitted with coinfections with Acinetobacter baumannii and Candida albicans being the most common pathogens (3/12 each), followed by the methicillin-resistant Staphylococcus aureus (MRSA, 2/12). Twenty-one patients (47%) developed secondary infections with A baumannii and C albicans as the most frequent pathogens (6/21 each) followed by MRSA and Stenotrophomonas maltophilia (4/21 each).

Table 1 - Clinical and demographical characteristics of critically ill COVID-19 patients Characteristic Survivors (n = 17) Non-survivors (n = 28) P Age 50 (33–60) 64 (52–70) 0.0030 Sex (male), n (%) 8 (47) 16 (57) 0.1248 Hospitalization days 28 (13–45) 17 (7–28) 0.0349 ICU days 21 (6–35) 12 (4–18) 0.2508 28-day mortality 0 24 0.1336 SOFA score 7 (6–9) 9 (8–12) 0.0024 APACHE II score 12 (8–13) 20 (14–25) 0.0001 Comorbidities:  Co-infections n (%) 3 (18) 9 (33) 0.4295  Cancer, n (%) 0 (0) 5 (18) 0.0773  Hypertension, n (%) 8 (47) 14 (50) 0.8482  Diabetes, n (%) 2 (12) 5 (18) 0.6931  Asthma, n (%) 1 (6) 3 (10) 1.0000  COPD, n (%) 0 (0) 1 (4) 1.0000  Obesity, n (%) 3 (18) 3 (10) 0.6581  Coronary artery disease, n (%) 0 (0) 4 (15) 0.2812  Chronic renal disease, n (%) 0 (0) 2 (7) 0.7029  Pregnancy, n (%) 3 (18) 0 (0) 0.0160  Secondary infections 9 (53) 12 (42) 0.7522 Treatment:  CRRT n (%) 0 (0) 6 (21) 0.1100  ECMO, n (%) 3 (18) 8 (29) 0.6390

CCRT indicates continuous renal-replacement therapy; COPD, chronic obstructive pulmonary disease; ECMO, extra-corporeal membrane oxygenation; ICU, intensive care unit.

Markers of inflammasome activation are not outcome related in critically ill COVID-19 patients

First, we compared the levels of IL-1α, IL-1 β, IL-18, IL- 1RA, galectin-1, ASC, and gasdermin D in the study group with the age and sex matched group of TBI patients. Interestingly, only galectin-1 was significantly different in critically ill COVID- 19 patients in comparison with the TBI patients (17,101.84 (IQ: 10,836.62–25,551.30) pg/mL vs. 30,764.20 (IQ: 23,150.6735,759.48) pg/mL; P = 0.007; Fig. 1). Then, we compared the concentration of the inflammasome related markers between COVID-19 survivors and non-survivors. None of the analyzed proteins (IL-1α, IL-1 β, IL-18, IL-1RA, galectin-1, ASC, and GSDMD) differed between these groups (Fig. 2). Moreover, LDH and ferritin, which can serve as surrogate markers of cell death and IL-1 signaling, respectively, were not increased in patients with an unfavorable outcome (Table 2). Other routinely monitored markers of inflammation, such as white blood count, IL-6 or CRP, were also similar in these two groups. Of the analyzed parameters, only the international normalized ratio (INR) was significantly higher in the non-surviving COVID- 19 patients. However, logistic regression analysis did not reveal a relationship between INR, any of the other analyzed parameters or mortality (P > 0.05). Also, there were no differences in the concentrations of the analyzed biomarkers of the inflammasome response between patients who were admitted with co-infections and those who were not. These results indicate that inflamma- some-related proteins are less elevated in critically ill COVID-19 patients at ICU admission than in TBI patients and are not different between survivors and non-survivors.

Fig. 1:

Comparison of the inflammasome-related markers in traumatic brain injury (TBI) and critical COVID-19 patients. A, Interleukin-1α. B. Interleukin-1β. C, Interleukin-18. D, Interleukin-1 receptor antagonist. E, Apoptosis-associated speck-like protein containing a CARD. F Gasdermin D. G, Galectin-1. Concentrations of plasma proteins were compared between TBI patients (n = 10) and critically ill COVID-19 patients (n = 45) with the Mann-Whitney test. ∗∗P < 0.05.

Fig. 2:

Inflammasome-related markers in the plasma of critically ill COVID-19 patients at low and high risk of death. The high risk of death was stratified by a threshold of more than nine points in the SOFA score. A, Interleukin-1α. B, Interleukin-1 β. C, Interleukin-18. D, Interleukin-1 receptor antagonist. E, Apoptosis-associated speck-like protein containing a CARD. F Gasdermin D. G, Galectin-1. H, Lactate dehydrogenase. I, Ferritin. Concentrations of plasma proteins were compared between patients at high risk (n = 15) and patients at low risk of death (n = 30) with the Mann-Whitney test. ∗P < 0.05.

Table 2 - Laboratory findings in critically ill COVID-19 patients upon ICU admission Characteristic Survivors (n = 17) Non-survivors (n = 28) P White blood count [x106/L] 12.9 (10.8–15.5) 15.7 (11.6–18.9) 0.3736 Platelets [× 109 /L] 287 (216–311) 241 (178–332) 0.2323 APTT [s] 33.1 (30.6–38.8) 36.0 (30.6–42.9) 0.3800 INR 1.10 (1.05–1.17) 1.24 (1.15–1.37) 0.0095 D-dimers [ug/mL] 1.80 (1.20–7.00) 6.45 (1.90–23.70) 0.0918 C-reactive protein [mg/L] 109.0 (67.0–149.0) 110.0 (67.5–195.0) 0.8149 Procalcitonin [ng/mL] 0.30 (0.11–0.43) 0.51 (0.14–1.95) 0.1528 Interleukin-6 [pg/mL] 75.85 (10.40–217.00) 88.80 (40.05–1,001.50) 0.1939 Glucose [mg/dL] 136.0 (107.0–159.0) 172.5 (127.0–214.0) 0.1340 Ferritin [ng/mL] 698.5 (292.5–1,563.5) 1,156.0 (568.0–2,002.0) 0.2089 Lactate dehydrogenase [U/L] 618.0 (480.0–709.0) 583.0 (427.0–930.0) 0.9906 Galectin-1 [pg/mL] 16,424 (12,957–22,040) 21,760 (10,759–29,643) 0.4329 IL-1α [pg/mL] 4.80 (4.80–11.97) 5.46 (4.80–8.68) 0.9905 IL-1β [pg/mL] 7.39 (2.69–11.46) 8.08 (2.69–13.23) 0.9252 IL-1RA [pg/mL] 43.11 (17.87–68.17) 23.26 (10.71–87.75) 0.7519 IL-18 [pg/mL] 57.78 (41.11–94.15) 66.58 (24.14–96.61) 0.9160 ASC [pg/mL] 0.00 (0.00–0.00) 76.70 (18.70–134.700) 1.0000 GSDMD [pg/mL] 0.02 (0.02–0.27) 0.010 (0.06–0.12) 0.5509

APTT indicates activated partial thromboplastin time; ASC, apoptosis-associated speck-like containing a CARD; GSDMD, gasdermin D.

Galectin-1 and IL-1RA can stratify the severity of critically ill COVID-19 on admission

As the SOFA score is one of the most widely used tools to stratify critically ill patients, including COVID-19, we analyzed its predictive capacity in our group of patients. Indeed, the ROC analysis for in-hospital mortality was characterized by an AUC of 0.824 (95% CI: 0.703–0.944), which is of moderate accuracy (15). As a predictor of death, a SOFA score of 9 points was characterized by sensitivity of 75.0% and specificity of 64.7%. Then, we applied this cut-off point to divide the patients into two groups of high- and low risk of death. The high-risk patients had a significantly higher level of galectin-1 in comparison to individuals at low-risk (25,551.30 (IQ: 15,628.8541,048.02) pg/mL vs. 16,302.72 (IQ: 8,785.86–22,040.25) pg/ mL; P = 0.014, respectively). Interestingly, the high-risk group had a significantly lower level of IL-1RA in comparison with the low-risk patients (14.50 (IQ: 9.28–47.01) pg/mL vs. 39.35 (IQ: 17.87–88.01) pg/mL; P = 0.047). The high-risk group of patients was also characterized by higher white blood count (17.00 (IQ: 13.40–20.30) × 106/mL vs. 12.75 (IQ: 10.2016.80) × 106/mL; P = 0.034), procalcitonin (2.10 (IQ: 0.404.00) ng/mL vs. 0.25 (IQ: 0.11–0.60) ng/L; P = 0.001 and IL-6 (700.00 (IQ: 680.70–1890.00) pg/mL vs. 66.60 (IQ: 16.00107.00) pg/mL; P = 0.001). The frequency of co-infections was not significantly different between these groups; however, chronic comorbidities such as coronary artery disease, heart failure, and renal disease were significantly more common in the high-risk patients. Interestingly, patients who developed secondary infections had a lower level of galectin-1 than those who did not (14,908.51 (IQ: 8,394.48–21,911.48) pg/mL vs. 23,839.62 (IQ: 15,628.85–29,846.66) pg/mL; P = 0.039; however, logistic regression analysis did not confirm this relationship (OR = 1.00, P = 0.042). There were no differences among the analyzed inflammasome-related biomarkers and any standard inflammatory parameters (IL-6, PCT, WBC) between patients who developed secondary infections during the ICU stay and those who did not. Altogether, these results suggest that the threshold of nine SOFA score points indicates low- and high-risk critically ill COVID-19 patients who differed in the inflammasome and inflammatory profile.

Correlations of inflammasome-related markers with other inflammatory markers in critically ill COVID-19 patients

Additionally, we tested whether the concentrations of inflammasome-related proteins were related with other measures. Spearman's correlation test revealed a weak but significant correlation between IL-1α and IL-1β (r = 0.40) and between IL-18 and IL-1RA (r = 0.60) and galectin-1 (r = 0.27). Gasdermin D correlated with the ASC protein (r = 0.44) and ferritin correlated with INR (r = 0.39) and LDH (r = 0.45). There was also an inverse correlation between IL-1α and platelets and IL-1 β and the platelet count (r = –0.37 and r = –0.36, respectively). No significant correlations were found between the inflammasome-related markers and CRP, PCT, or IL-6. Although the correlation coefficients between the inflammasome-related markers are fair to moderate they suggest unidirectional regulation.

Cell free caspase-1 is not ubiquitously present in the circulation of critical COVID-19 patients

As secretion of IL-1 and pyroptosis can result from the activation of cellular pathways other than those of inflammasomes (16), we decided to confirm the activation of inflammasomes in COVID-19 patients by the development of the inhibitor-based immunoprecipitation of active caspase-1 followed by the Western Blot detection of specific fragments. Interestingly, the dominant form of caspase-1 present in the plasma was the intermediate form p35 found in 12 of 22 COVID-19 patients (Fig. 3A), while it was undetected in the TBI patients. In 6 of 22 COVID-19 patients the mature p20 form was also detected (Fig. 3A). The presence of active caspase-1 was not related to outcome nor the severity of the disease. There were also no differences in caspase-1 between patients with coinfections. However, patients at high-risk of death, as defined by a SOFA score >9, had a significantly higher amount of p35 in the densitometric analysis (Fig. 3B) Of interest, patients with detectable p35 had a higher fibrinogen level (7.30 g/L (IQ: 5.50–8.70) vs. 5.45 g/L (IQ: 4.10–6.30); P = 0.020) and lower d- dimers (1.35 μg/mL (IQ: 1.20–4.35) vs. 10.25 pg/mL (IQ: 6.3020.90); P = 0.019).

Fig. 3:

Active caspase-1 is unambiguously present in the plasma of critical COVID-19 patients. A, Representative immunoblots labeled with anti-caspase-1 of the pull-downed FAD-FMK bonded caspases from the plasma of TBI patients (3TBI, 5 TBI) and COVID-19 patients (numbers 822). A mature p20 band is present in approx. Half of the patients with the intermediate p35 form. B, Patients at high risk of death (SOFA>9) had more circulating p35 as revealed by densitometric analysis. Data were compared with the Mann-Whitney test. ∗P < 0.05.

DISCUSSION

Our study is the first to our knowledge that has comprehensively investigated plasma biomarkers related to inflammasome regulation in a well-defined group of critically ill COVID-19 patients on ICU admission. We found that the marked activation of inflammasomes defined as the presence of active capsase-1 in the patient plasma was not ubiquitous and was not related with outcome. Also, other quantitative inflamma- some-related proteins and cytokines were not related with outcome and only galectin-1 and IL-1RA differed in patients with a low- or high risk of death.

In this study we focused on critically ill COVID-19 patients on admission to the ICU, as this group is at high risk of death (reflected by the 28-day mortality of 53% and an even higher in-hospital mortality of 62% in our group). This time period for the single-time analysis was chosen because: the need for risk stratification biomarkers is greatest at ICU admission; at this time of clinical deterioration any potential biological therapies are likely to be considered. Although many groups have tried to identify new potential biomarkers, we selected a panel of proteins related with the different stages of inflammasome activity, and taking into consideration the structural component of inflammasomes, the ASC protein, and effector proteins cleaved by caspase-1 (IL-1α, IL-1β, IL-18) and gasdermin D, which forms pyroptotic pores in cell membranes. We measured the IL-1 antagonist IL-1RA, as it is upregulated in response to IL-1 signaling. Also, we measured galectin-1 and LDH, which are markers of inflammatory cell death (17) and ferritin, which is induced by inflammasome activity (18). Moreover, we have established a new method of analyzing active caspase-1 from human plasma, which is based on immunoprecipitation with biotinylated FAD-FMK, a pan-caspase inhibitor followed by Western Blot to specifically detect caspase-1 forms.

The concentrations of the measured cytokines were not different between critical COVID-19 and TBI patients. We choose TBI patients for comparison as they are a relatively homogenous group with sterile injury, and comparable age and sex demographics. ASC, IL-1, and capase-1 were already shown to be upregulated in the plasma of TBI patients (19). Although increased galectin-1 was previously demonstrated in COVID-19 patients (20), the only comparison was to a healthy control group. Here, we found higher galectin-1 in TBI patients than in COVID-19 patients, which suggests that the latter are not characterized by an extreme over-inflammatory response, as suggested by some authors (3, 21). Our results expand the findings of others showing similar levels of inflammatory cytokines in COVID-19 and sepsis or other critical conditions (22–24).

None of the analyzed protein was related with mortality in critically ill COVID-19 patients Even though some of the cytokines, like IL-18 or IL-1RA, had been previously linked with the severity of COVID-19 (9, 25, 26). It should be highlighted that our study is focused on a defined population of critical COVID-19 patients, in contrast to most studies which compare cytokine levels in early COVID-19 patients with varying degrees of severity. As the analyzed cytokines can be produced in an inflammasome-independent manner, we assessed the presence of active caspase-1. We detected the intermediate p35 form of caspase-1 in 54% cases and the mature p20 form in 27% of patients. These findings clearly indicate that the level of inflammasome activation is heterogenous in critical COVID-19, and complete maturation of caspase-1 is not common in these patients. The status of circulating caspase-1 was not related to outcome, similarly to the findings of D. Zamboni et al., who utilized capsase-1 ELISA (9). We were able to detect ASC and GSDMD proteins much less frequently, which also indicates the infrequent activation of inflammasomes in critical COVID-19. Our results are in accordance with recent findings that showed only a small percentage of NLRP3 and AIM2 inflammasome activation in the circulating monocytes of COVID-19 patients (10, 27). Moreover, our data points to the monocytes and macrophages as the major lung cell types showing the activation of inflammasomes in COVID-19 (9, 10) patients. Our findings do not preclude the possibility of the hyperactivation of inflammasomes earlier in the course of the disease or at the tissue level. Additionally, we compared inflammasome-related proteins in the patients divided by their SOFA score on admission with the most accurate threshold to predict mortality (low- and high-risk groups). Notably, galectin-1 and caspase-1 p35 was increased in the high-risk group, while IL-1RA was lower. Such a pattern may indeed suggest the loss of control of this inflammatory pathway in the high-risk group of patients. The high-risk patients also had higher levels of leukocyte count, IL-6, procalcitonin, and INR; however, none of these parameters was significantly related with outcome in the logistic regression model.

As inflammasomes reportedly play a role in the susceptibility to secondary infections (28, 29), we tested for a relation between inflammasome activation markers and co-infections or the development of secondary infections; however, there were no significant correlations between these variables. We observed some weak but significant correlations between IL- 1α and IL-1β, ASC and GSDMD, and ferritin and LDH, which suggests coordinated regulation of these mediators. Interestingly, we observed a link between inflammasome activity and the dysregulated coagulation system. Patients with an unfavorable prognosis had marks of coagulopathy as indicated by higher INR values. In addition, there was a correlation between ferritin and INR and an inverse correlation between IL-1α and IL-1β and platelets. Altogether, these observations suggest a mechanistic link between inflammasome activation and coagulopathy and possibly the features of secondary hemophago- cytic lymphohistiocytosis (3).

This study has several limitations related to the small patient group and a lack of mechanistic investigations. Noteworthy is that our study group is well defined and limited to ARDS and the sepsis forms of COVID-19. Although we analyzed several proteins related with the activation of inflammasomes, we did not measure other relevant mediators that can interact with this pathway (e.g., interferons). The present study would also have benefited from flow cytometry data of the inflammasome activation in circulating blood cells. Moreover, we analyzed only plasma proteins, and it should be noted that they may not fully reflect the tissue response. Although the distribution of comorbidities (except for pregnancy) was balanced between survivors and non-survivors, we cannot preclude the impact of underlying chronic inflammatory diseases, such as cardiovascular disease, on the activity of the inflammasome pathway. Finally, we were not able to explain the heterogenic pattern of the inflammasome activity in these patients.

CONCLUSION

Our findings are of important clinical relevance, as inflam- masome-targeted therapies are widely tested in COVID-19, mostly without the use of biomarker guidance. Negative results were recently published of a collaborative trial on colchicine, which inhibits inflammasome formation (30, 31). Also, trials targeting IL-1 in severe COVID-19 with anakinra were reported to be negative (14). At the same time, another study that used su-PAR levels to guide anakinra treatment in early COVID-19 showed significant benefit (13). The results of these studies are not surprising in the light of our findings showing heterogeneity in the inflammasome activation in critically ill COVID-19 patients and the lack of a direct relation with mortality. Yet, it could be speculated that in a selected subpopulation of patients with activated caspase-1 or high galectin-1 level, these therapies could be beneficial. Importantly, the absence of a relation between the inflammasome-related markers and outcome does not preclude the involvement of this pathway in the pathogenesis of the disease. Nevertheless, further mechanistic studies on the role of inflammasomes in the pathogenesis of COVID-19 are needed.

REFERENCES 2. Armstrong RA, Kane AD, Kursumovic E, Oglesby FC. Cook TM: Mortality in patients admitted to intensive care with COVID-19: an updated systematic review and meta-analysis of observational studies. Anaesthesia 76 (4):537–548, 2021. 3. Osuchowski MF, Winkler MS, Skirecki T, Cajander S, Shankar-Hari M, Lachmann G, Monneret G, Venet F, Bauer M, Brunkhorst FM, et al. The COVID-19 puzzle: deciphering pathophysiology and phenotypes of a new disease entity. Lancet Respir Med 9 (6):622–642, 2021. 4. Cavaillon JM, Singer M, Skirecki T. Sepsis therapies: learning from 30years of failure of translational research to propose new leads. EMBO Mol Med 12 (4):e10128, 2020. 5. Osuchowski MF, Aletti F, Cavaillon JM, Flohe SB, Giamaerllos-Bourboulis EJ, Huber-Lang M, Relja B, Skirecki T, Szabo A, Maegele M. SARS-CoV-2/COVID-19: evolving reality, global response, knowledge gaps, and opportunities. Shock 54 (4):416–437, 2020. 6. Yap JKY, Moriyama M. Iwasaki A: Inflammasomes and pyroptosis as therapeutic targets for COVID-19. J Immunol 205 (2):307–312, 2020. 7. Xue Y, Enosi Tuipulotu D, Tan WH, Kay C, Man SM. Emerging activators and regulators of inflammasomes and pyroptosis. Trends Immunol 40 (11):1035–1052, 2019. 8. Toldo S, Bussani R, Nuzzi V, Bonaventura A, Mauro AG, Cannata A, Pillappa R, Sinagra G, Nana-Sinkam P, Sime P, et al. Inflammasome formation in the lungs of patients with fatal COVID-19. Inflamm Res 70 (1):7–10, 2021. 9. Rodrigues TS, de Sa KSG, Ishimoto AY, Becerra A, Oliveira S, Almeida L, Goncalves AV, Perucello DB, Andrade WA, Castro R, et al. Inflammasomes are activated in response to SARS-CoV-2 infection and are associated with COVID-19 severity in patients. J Exp Med 218 (3):e20201707, 2021. 10. Junqueira C, Crespo A, Ranjbar S, Ingber J, Parry B, Ravid S, de Lacerda LB, Lewandrowski M, Clark S, Ho F, et al. SARS-CoV-2 infects blood monocytes to activate NLRP3 and AIM2 inflammasomes, pyroptosis and cytokine release. medRxiv 2021; 10.1101/2021.03.06.21252796. 11. Ferreira AC, Soares VC, de Azevedo-Quintanilha IG, da Silva Gomes Dias S, Fintelman-Rodrigues N, Sacramento CQ, Mattos M, de Freitas CS, Temerozzo JR, Teixeira L, et al. SARS-CoV-2 engages inflammasome and pyroptosis in human primary monocytes. Cell Death Discov 7 (1):43, 2021. 12. Herr C, Mang S, Mozafari B, Guenther K, Speer T, Seibert M, Srikakulam SK, Beisswenger C, Ritzmann F, Keller A, et al. Distinct patterns of blood cytokines beyond a cytokine storm predict mortality in COVID-19. J Inflamm Res 14:4651–4667, 2021. 13. Kyriazopoulou E, Poulakou G, Milionis H, Metallidis S, Adamis G, Tsiakos K, Fragkou A, Rapti A, Damoulari C, Fantoni M, et al. Early treatment of COVID- 19 with anakinra guided by soluble urokinase plasminogen receptor plasma levels: a double-blind, randomized controlled phase 3 trial. Nat Med 27:1752–1760, 2021. 14. The REMAP-CAP Investigators, Derde LP. Effectiveness of Tocilizumab, Sarilumab, and Anakinra for critically ill patients with COVID-19 the REMAP-CAP COVID-19 Immune Modulation Therapy Domain Randomized Clinical Trial. medRxiv 2021; 2021.2006.2018.21259133. 15. Fischer JE, Bachmann LM, Jaeschke R. A readers’ guide to the interpretation of diagnostic test properties: clinical example of sepsis. Intensive Care Med 29 (7):1043–1051, 2003. 16. Pyrillou K, Burzynski LC, Clarke MCH. Alternative pathways of IL-1 activation, and its role in health and disease. Front Immunol 11:613170, 2020. 17. Russo AJ, Vasudevan SO, Mendez-Huergo SP, Kumari P, Menoret A, Duduskar S, Wang C, Perez Saez JM, Fettis MM, Li C, et al. Intracellular immune sensing promotes inflammation via gasdermin D-driven release of a lectin alarmin. Nat Immunol 22 (2):154–165, 2021. 18. Kernan KF, Carcillo JA. Hyperferritinemia and inflammation. Int Immunol 29 (9):401–409, 2017. 19. Kerr N, Lee SW, Perez-Barcena J, Crespi C, Ibanez J, Bullock MR, Dietrich WD, Keane RW, de Rivero Vaccari JP. Inflammasome proteins as biomarkers of traumatic brain injury. PLoS One 13 (12):e0210128, 2018. 20. Kazancioglu S, Yilmaz FM, Bastug A, Ozbay BO, Aydos O, Yucel C, Bodur H, Yilmaz G. Assessment of Galectin-1, Galectin-3, and PGE2 levels in patients with COVID-19. Jpn J Infect Dis 74 (6):530–536, 2021. 21. Tang Y, Liu J, Zhang D, Xu Z, Ji J, Wen C. Cytokine storm in COVID-19: the current evidence and treatment strategies. Front Immunol 11:1708, 2020. 22. Wilson JG, Simpson LJ, Ferreira AM, Rustagi A, Roque J, Asuni A, Ranganath T, Grant PM, Subramanian A, Rosenberg-Hasson Y, et al. Cytokine profile in plasma of severe COVID-19 does not differ from ARDS and sepsis. JCI Insight 5 (17):e140289, 2020. 23. Kox M, Waalders NJB, Kooistra EJ, Gerretsen J, Pickkers P. Cytokine levels in critically ill patients with COVID-19 and other conditions. JAMA 324 (15):1565–1567, 2020. 24. Stolarski AE, Kim J, Zhang Q, Remick DG. Reply letter to the editor regarding: cytokine drizzle-the rationale for abandoning the term cytokine storm. Shock 56 (5):873–874, 2021. 25. Satis H, Ozger HS, Aysert Yildiz P, Hizel K, Gulbahar O, Erbas G, Aygencel G, Tunccan OG, Ozturk MA, Dizbay M, et al. Prognostic value of interleukin-18 and its association with other inflammatory markers and disease severity in COVID-19. Cytokine 137:155302, 2021. 26. Zhao Y, Qin L, Zhang P, Li K, Liang L, Sun J, Xu B, Dai Y, Li X, Zhang C, et al. Longitudinal COVID-19 profiling associates IL-1RA and IL-10 with disease severity and RANTES with mild disease. JCI Insight 5 (13):e139834, 2020. 27. Courjon J, Dufies O, Robert A, Bailly L, Torre C, Chirio D, Contenti J, Vitale S, Loubatier C, Doye A, et al. Heterogeneous NLRP3 inflammasome signature in circulating myeloid cells as a biomarker of COVID-19 severity. Blood Adv 5 (5):1523–1534, 2021. 28. Roth S, Cao J, Singh V, Tiedt S, Hundeshagen G, Li T, Boehme JD, Chauhan D, Zhu J, Ricci A, et al. Post-injury immunosuppression and secondary infections are caused by an AIM2 inflammasome-driven signaling cascade.