Decreased plasma gelsolin in the COVID-19-related acute respiratory distress syndrome

Introduction

Corona Virus Disease 2019 (COVID-19) mainly involves the respiratory system and has clinical features that range from asymptomatic to severe pneumonia resulting in acute respiratory distress syndrome (ARDS) [1]. ARDS is characterized histologically by diffuse alveolar damage with increased vascular permeability and reduced compliance, affecting gas exchange and leading to intractable hypoxemia. ARDS develops most commonly in the setting of pneumonia (bacterial and viral; fungal is less common), nonpulmonary sepsis, major trauma, and aspiration of gastric and/or oral and oesophageal contents [1, 2]. The majority of patients who die from COVID-19 die from pulmonary failure due to severe ARDS. Previous studies have shown that more than 30 % of hospitalised COVID-19 patients might develop ARDS, with a mortality rate of 15–52 % [3]. Although COVID-19 has been largely brought under control through vaccination, the search for therapeutic pathways for active patients remains important, especially because of the high mortality rate of severe ARDS and the high probability of sequelae even if the patient recovers [4, 5]. Moreover, new findings may provide additional information about other respiratory diseases with similar characteristics and the identification of molecular hallmarks associated with severe disease that could serve as prognostic biomarkers is relevant to guide therapeutic decisions. Over the last 50 years, numerous basic and clinical studies have been conducted on ARDS, but the morbidity and mortality rates of ARDS remain high and there is a lack of specific drugs [2]. Although there are still some differences between two form of ARDS, the pathophysiology of traditional ARDS and COVID-19-dependent ARDS share many of the same pathophysiological aspects (reduced lung parenchymal compliance, alveolar flooding, vasculopathy, and gas exchange impairment) and most patients with severe COVID-19 pneumonia fulfill the clinical criteria for ARDS [6, 7].

In addition, many discharged COVID-19-associated ARDS patients present with abnormal pulmonary architecture, reduced diffusion capacity, and functions characterized by a fibroproliferative response [8]. The fibrotic stage, which is mostly irreversible and considered the last stage of ARDS, has a relatively high incidence in COVID-19-mediated ARDS patients [9]. Nevertheless, it remains unclear why most patients with ARDS are able to recover from the inflammatory process while a subset undergoes an excessive fibrotic process. However, the fibrotic process, in which many factors such as lung epithelial and endothelial cell damage, immune system-mediated activation of profibrotic factors, and mechanical ventilation are involved, is in parallel to the severity of ARDS [10, 11]. This indicates that the fibrotic process can be estimated, albeit at a limited level, using parameters related to the severity of the ARDS.

Plasma gelsolin (pGSN), is the primary actin-sequestering protein in extracellular compartments, involved in the pathogenesis of several diseases and viral infections [12]. Decreased pGSN concentration is described in critically ill patients and those requiring rapid medical attention [13]. Moreover, it has been demonstrated that decreased pGSN levels are significantly correlated with sequential organ failure assessment scores in intensive care unit (ICU) patients [14].

Transforming growth factor-beta1 (TGF-β1) is a critical cytokine that has a key role during respiratory viral infections by mediating both suppression of immune responses and exracellular matrix remodeling which lead to lung fibrosis [11, 15]. The pivotal function of TGF-β1 in the immune system is to ensure the resolution of inflammatory responses through the regulation of immune cells such as lymphocytes, natural killer cells, dendritic cells, and macrophages [16].

In the same way as TGF-β1, lysophosphatidic acid (LPA), is another important profibrotic molecule. The immune modulatory properties of LPA have been shown especially in cancer progression. Circulating LPA is a bioactive lipid and is generated by the enzyme autotoxin (ATX) from membrane lysophospholipids or circulating lysophosphatidylcholine in plasma. In pulmonary fibrosis, LPA production rises and LPA binding to its receptors activates the pro-inflammatory signals from epithelial cells, the TGF-β signaling, and the stimulation of fibroblast accumulation [15].

The involvement of pGSN in inflammatory and fibrotic processes in association with both TGF-β1 and LPA has been demonstrated [17, 18]. Although the relationship between pGSN levels and prognosis of COVID-19 has been investigated in previous studies, no consensus has been reached 19], [20], [21. Moreover, there is no study that has evaluated pGSN levels together with the profibrotic and anti-inflammatory factors LPA and TGF-β1 in these patients. Therefore, the aim of this study was to evaluate the potential roles of pGSN, TGF-β1 and LPA as profibrotic and immune modulatory markers in severe/critical COVID-19 patients developing ARDS.

Materials and methods

The study included all the patients admitted and hospitalized at the Kayseri City Hospital (Turkey) with SARS-CoV-2 pneumonia between September 20 and December 20, 2021, who agreed to participate. The study protocol was approved by the Institutional Ethics Committee (prot. no.2021/25), after obtaining scientific research approval for the study from the Ministry of Health General Directorate of Health Services. Written informed consent was collected from all participants prior to enrolment.

COVID-19 pneumonia was confirmed by chest computed tomography (CT) scan showing typical ground-glass abnormalities with a positive SARS-CoV-2 RT-PCR performed on a nasopharyngeal swab in all patients. A “Bio-Speedy^® SARS-CoV-2 + VOC202012/01 RT-qPCR” kit from Bioeksen R&D Technologies Istanbul-Turkey was utilized to conduct the RT-PCR procedure. The amplification was performed with a CFX96 Touch Real-Time PCR Detection System Instrument (Bio-Rad Laboratories, Inc, USA). vNAT^® tubes were also obtained from Bioeksen R&D Technologies Istanbul-Turkey.

The patients were divided into two groups according to whether ARDS developed or not, and the study was carried out with a total of three groups with a control group.

The control group is formed of volunteers who applied to the outpatient clinic with suspicion of COVID-19 but did not have SARS-CoV-2 RT-PCR (+) test results and CT findings related to COVID-19, whose age distribution was appropriate for the patient groups, and whose routine hematology and biochemistry tests were within the reference range were included.

The patient group who did not develop ARDS comprised patients who did not develop ARDS from the time of hospital admission until discharge. Blood samples were taken from these patients within 24–48 h before discharge. The patients in this group were those with mild/moderate COVID-19 pneumonia [22] requiring hospitalization in conventional wards.

The patient group who developed ARDS comprised patients who developed ARDS according to the Berlin Criteria at any time from the moment they were admitted to the hospital until discharge. Blood samples were taken from these patients within 24–48 h after the diagnosis of ARDS. The patients in this group were those with severe and critical COVID-19 pneumonia [22] requiring hospitalization in ICU. According to the current Berlin definition, ARDS is characterized by refractory hypoxemia, respiratory failure not explained by cardiac failure or fluid overload, and chest imaging showing bilateral opacities not fully explained by effusions, lobar/lung collapse, or nodules; presentation within 1 week of a known clinical insult or worsening respiratory symptoms; and oxygenation: (1) 200 mmHg <PaO₂/FIO₂ ≤300 mmHg with PEEP or CPAP ≥5 cmH₂O: mild; (2) 100 mmHg <PaO₂/FIO₂ ≤200 mmHg with PEEP ≥5 cmH₂O: moderate; (3) PaO₂/FIO₂ ≤100 mmHg with PEEP ≥5 cmH₂O: severe [23].

The study exclusion criteria for both patient groups were defined as age <18 years, the presence of any chronic inflammatory and autoimmune disease, malignancy, or advanced organ failure (e.g., liver failure; congestive heart failure; severe renal failure, glomerular filtration rate <30 mL/min; or chronic obstructive pulmonary disease). None of the patients were vaccinated before or during this period. Patients received antiviral therapies (lopinavir/ritonavir, ribavirin, and remdesivir) and non-antiviral adjunctive therapies (steroids, anticoagulation, and antibiotics) in addition to appropriate supportive care. Complete blood count was analyzed in whole blood samples (BD Vacutainer K2E 5.4 mg, 3 mL/Siemens Advia 2120i) and coagulation tests [Prothrombin time (PT); International normalised ratio (INR); Partial thromboplastin time (aPTT)] in plasma (BD Vacutainer 9NC 0.105 M, 4.5 mL/Sysmex CN 6000). Routine biochemistry analyses including blood urea nitrogen (BUN), creatinine, sodium, estimated glomerular filtration rate (eGFR) and C-reactive protein (CRP), etc. measurements were performed in serum samples of all patients (BD Vacutainer SST II Advance 5 mL/Cobas c 702; Roche Diagnostics).

PGSN, TGF-β1, and LPA levels were measured using a commercial ELISA kit [Cloude-Clone Corp, USA, Cat No: SEA372Hu (Version:12.0); SEA124Hu (12th Edition); CEK623Ge (Version:13.0) respectively], according to the manufacturer’s instructions. The inter-assay and intra-assay coefficient of variability (%CV) values were <12 and <10 %, respectively for three of kits. The pGSN, TGF-β1 and LPA kits’ detection ranges were 31.2–2.000 pg/mL; 15.6–1.000 pg/mL, 123.5–10.000 ng/mL and the limit of quantication (LoQ) values were 15.2 pg/mL; 5.7 pg/mL; 50.5 ng/mL, respectively.

The blood samples obtained during the study were centrifuged at 3,500 rpm at +4 °C for 15 min and the plasma samples were stored in aliquots at −80 °C until the study day.

Plasma samples were activated before the study for TGF-β1. For this purpose, 50 µL of sample was mixed with 10 µL of HCl (1 M) and incubated for 10 min at room temperature, after which it was neutralized with a solution containing 10 µL of 1.2 N NaOH and 0.5 M HEPES. Standard dilution solution (80 µL) was added to each sample and made ready for the study. Finally, all results were multiplied with three in accordance with the TGF-β1 kit manual. To examine pGSN, the samples were diluted 1/50,000 in phosphate buffered saline based on our own laboratory experience.

Results

There was no difference between the groups in terms of gender (p=0.416). When the patient and control groups were compared in terms of age, there was no difference between the control group and ARDS (−) patients. It was observed that the ARDS (+) patient group was older than both the control and ARDS (−) patient groups (p=0.002 and p=0.001, respectively) (Table 1).

The clinical characteristics of the patients are presented in Table 2.

In the examination of the biochemical and hematological parameters, the COVID-19 patients were determined to have increased CRP, neutrophil, and PT, and decreased albumin and calcium compared to the control group. With the progression of the disease to ARDS in COVID-19 patients, there was seen to be increased BUN, lactate dehydrogenase (LDH), CRP, neutrophil, D-Dimer, fibrinogen, PT, INR, and aPTT levels. Albumin, hemoglobin, hematocrit, calcium and lymphocyte levels were decreased in the ARDS (+) group compared to the ARDS (−) patient group (Table 3).

The pGSN levels increased in COVID-19 patients, but in the comparisons between the two patient groups a significant decrease was determined in the ARDS (+) patients. TGF-β1 and LPA levels were decreased in COVID-19 patients compared to the control group, with a greater decrease determined in the ARDS (+) patients compared to the ARDS (−) group (Table 4).

On univariate analysis, the following variables were found to be significant predictors of ARDS: age, eGFR, CRP, neuthrophil, lymphocyte, Fibrinogen, PT, pGSN, TGF-β1, and LPA. A multivariate analysis of the same data demonstrated that CRP and pGSN remained as independent predictors of the development of ARDS (Table 5).

By receiver operating characteristic analysis, pGSN, TGF-β1, and LPA predicted development of ARDS in COVID-19 patients; the area under the curve (AUC) of 0.977 (95 % CI 0.900 to 0.999); 0.741 (95 % CI 0.612 to 0.846), and 0.723 (95 % CI 0.592 to 0.831), respectively (Figure 1). Also, when AUCs were compared, there was no difference between TGF-β1 and LPA (p=0.853), while pGSN was found to be significantly higher (p<0.0004 and p< 0.0009 respectively).

Figure 1:

Receiver operating curve analysis of pGSN, TGF-β1, and LPA.

The cut-off values for pGSN, TGF-β1, and LPA for predicting development of ARDS in COVID-19 patients, 4,573 ng/mL (90 % sensitivity, 99 % specificity), 1,358 pg/mL (66.7 % sensitivity, 76.6 % specificity), and 568 ng/mL (99 % sensitivity, 56.7 % specificity), respectively.

Discussion

ARDS, which begins with injury through activation of alveolar macrophages by microbial or cell injury products, is difficult to control in the proliferative phase and especially in the fibrotic phase, and therefore, early treatment is critical for prognosis in managing and improving symptoms [10].

This study was designed to assess the involvement of circulating markers of inflammation and fibrosis in the progression of COVID-19 in patients who were admitted to the hospital with mild/moderate COVID-19 and who developed COVID-19-related ARDS. At the same time, the patient groups were also compared with non-COVID-19 volunteers. The results of the study showed that TGF-β1 and LPA levels were lower in patients than in the control subjects, and the lowest levels were observed in patients who developed ARDS. Furthermore, the higher pGSN levels in COVID-19 patients were found to decrease with the development of ARDS, and in multivariate analysis, pGSN was identified as an independent risk factor. Moreover the current study investigated the efficacy of pGSN as a marker for developing ARDS in COVID-19 patients. At a pGSN cutoff value of 4,573 ng/mL, it demonstrated high sensitivity, specificity, and predictive value in identifying ARDS progression.

Similar to the current study, Mesner et al. [24] also reported decreased plasma pGSN levels in critically ill COVID-19 patients, and Abers et al. [25] showed that the lowest pGSN levels were seen in patients who subsequently died of COVID-19. These findings are consistent with the knowledge that pGSN sequestrate the increased free actin and is removed from plasma [12]. The increase in plasma actin levels in COVID-19 patients is thought to be especially related to T cell degradation, which is one of the primary features of COVID-dependent pneumonia, which is also shown in this study with lymphopenia [20, 21, 24].

However, increased pGSN levels in COVID-19 patients compared to the control group is one of the remarkable findings of the current study. In contrast, a previous study reported decreased pGSN levels compared to healthy volunteers [19]. However, in the same study, it was shown that pGSN levels measured at the time of hospitalization were higher in patients who were hospitalized in the ICU than in patients followed up in the wards, but lower in patients who died than in patients who could be discharged. The most important reason for these conflicting results may be the kinetics of pGSN levels during the progression of the disease. Therefore, it can be thought that the possible pGSN change during the course of the disease differs with the sampling time. The emergence of COVID-19-associated ARDS can be explained by the high or low angiotensin-converting enzyme two protein, uncontrollable cytokine storm and activation of the immune response, and triggering of the Fas/FasL signaling pathway to promote apoptosis [2]. Especially the cytokine storm, which many of these factors have in common, leads to poor outcomes and appears within 7–10 days in COVID-19 patients [26]. This time interval coincides with the sampling time in the current study patients, so it can be said that the sampling time reflects the plasma gelsolin levels during the cytokine storm, since ARDS development is known to occur during the period when the cytokine storm cannot be prevented. In addition, the effect of pGSN on the control of different cell groups and pathways that play a role in cytokine storm [27] and its upregulation with inflammatory cytokines [19] supports the increased pGSN levels compared to the control subjects. Similarly, it should be noted that the sampling time in this study for the non-ARDS group coincided with the recovery phase of COVID-19, in other words, the point at which the anti-inflammatory/pro-inflammatory balance was achieved. Therefore, it can be thought that increased pGSN levels are a compensatory mechanism to protect the organism from the viral effects.

Consistent with previous studies [28, 29], laboratory markers such as CRP, albumin and LDH were also shown to be associated with the development of ARDS in this study. Furthermore, in multivariate analyses, CRP was discovered to be an independent risk factor for developing ARDS. In addition, significant increases in coagulation parameters such as D-Dimer, fibrinogen, and INR, which are considered to be one of the results of cytokine storm [30], were also seen in COVID-19 patients who developed ARDS [31, 32].

Another key observation in this study was the decrease in TGF-β1 and LPA which are known to be pro-fibrotic factors and important players in the immune response in COVID-19 patients, compared to the control group, and the lowest values were obtained ​​in the ARDS (+) group. According to the current study results, it can be said that the low concentrations of TGF-β1 and LPA contribute to the severity of the disease. However, TGF-β1 and LPA, were identified as significant variables in univariate analysis but were unable to be identified as independent risk factors in multivariate analyses.

Similar to the current study, it has been reported in previous studies that decreased serum TGF-β1 levels in COVID-19 patients and lower serum values of TGF-β1 were associated with unfavorable outcomes [33, 34]. Moreover, Montalvo et al. [35] showed significantly lower TGF-β1 expression in the nasopharyngeal swab samples of COVID-19 patients compared to the control group, and it was suggested that this is a mechanism to maintain homeostatic balance to protect the microenvironment of commensal agents in healthy individuals. Thus, it can be said that there is a decrease in the immunosuppressant effect secondary to the low TGF-β1 levels in COVID-19 patients, resulting in inadequate response to cytokine storm and thereby contributing to the development/severity of ARDS [16, 33].

It has been shown that TGF-β1 can participate in immune system regulation in different ways. Regulatory T (Treg) cells are known as immune system‐regulating cells by producing anti‐inflammatory cytokines, such as IL‐10, IL‐35, and TGF‐β1. In COVID‐19 patients, Treg cells have been shown to be significantly reduced in patients compared to a control group [36]. It has also been reported that TGF-β1 is a potent antagonist of interleukin (IL)-12-induced interferon-γ expression in natural killer (NK) cells and its expression coincides with the downregulation of NK cell responses during viral infection [16].

Conversely, it is confusing that an increase in TGF-β1 due to COVID-19 and its association with lung involvement and poor prognosis have been demonstrated [37, 38]. Witkovski et al. [37] suggested that “untimely” increased TGF-β1 levels in the first two weeks after the onset of the disease caused an inadequate response to control the virus as a result of inhibition of NK cells, and this resulted in an increase in disease severity. However, patient comorbidities and the treatment protocols may have an effect on the emergence of these contradictory findings.

LPA is a bioactive lipid, which is released into the circulation by hydrolysis of lysophosphotidyl derivatives that have been released into circulation via phospholipase A1 and A2, especially from membrane phospholipids, by ATX, which shows lysophospholipase D activity. LPA plays an important role in the control of different cellular activities with its differentiated receptor distribution (LPA1-6) [39]. It has been reported that COVID-19 increases ATX gene expression in respiratory epithelial and immune cells and the ATX/LPA axis contributes to cytokine storm and endothelial dysfunction [40]. However, the same study findings also showed that ATX levels in COVID-19 patients using dexamethasone were no different from those of the healthy control group. Sumida et al. also reported that steroids caused a dose-dependent decrease in ATX serum concentrations [41]. It should be noted here that steroid treatment was administered to 100 % of the current study ARDS group patients, and to approximately 45 % of the non-ARDS group. Thus it can be considered that the LPA results were affected by steroids.

In contrast, Shimura et al. [42] reported similar ATX levels in COVID-19 patients with and without steroid therapy with the same disease severity. These findings suggest that different pathways may be effective in reducing LPA, which would contribute to COVID-19 prognosis. Furthermore, in support of the current study findings, Shimura et al. [42] showed reduced ATX levels in male and female COVID-19 patients, separately, and these were more pronounced at different time intervals, particularly around 1 week after disease onset.

While LPA decreases the production of IL-10 and IL-2 in dendritic cells and naive CD4+ T cells, respectively, through LPA2 [43, 44], it plays a suppressive role in CD8+ T cells through LPA5 [45]. In addition, the production of IL-8, which acts as a neutrophil chemo-attractor in the airway epithelium in response to stimulation by allergens and viruses, is stimulated by LPA. Therefore, low levels may exacerbate disease severity by causing an increase in the virus load in patients [46].

Finally, It can be thought that the decreased serum LPA and TGF-β1 levels, especially in ARDS patients, contradict their profibrotic properties. Therefore, it is important to consider the progression of ARDS. Although TGF-β1 and LPA are referred to as profibrotic mediators [47], their inflammatory effects are thought to be more pronounced in this phase of ARDS, because, while the inflammatory phase of ARDS is more dominant in the first two weeks, the fibrotic process is the major long-term complication [48, 49].

Conclusions

Taken together, it can be said that pGSN, TGF-β1, and LPA concentrations can provide an idea about the poor prognosis of COVID-19. Especially when the role of these three parameters in the regulation of the immune system is considered, it is also possible to say that they play a role in suppressing the immune response, despite its established role in fibrosis.

There were a few limitations to this study, primarily the relatively small sample size and that only peripheral blood was evaluated. Samples taken from the lesion site could have provided more precise information, and samples taken at different times during the course of the disease rather than a single sample could have answered more questions. However, the workload of the hospital at that time prevented the study from being planned in that way.