Introduction: We report an Intervention/outcome study of 33 severe COVID-19 subjects who received Seraph 100 Microbind Affinity Blood Filter (Seraph 100) hemoperfusion therapy (15 survivors, 18 non-survivors) under emergency authorization from the FDA. Our objective was to determine if Seraph 100 hemoperfusion reduces SARS-CoV-2 RNA titers and/or markers of inflammation and/or epi/endothelial damage. Methods: Viral RNA and 78 protein analytes related to endothelial/epithelial damage and/or inflammation were quantified in systemic blood samples from 33 severe COVID-19 subjects collected upon intensive care unit (ICU) admission and then immediately before and after blood passed through the heparin-based Seraph 100 filter at two time points on the first day of hemoperfusion. Viral RNA titers were quantified using droplet digital PCR. Protein analytes were quantified using multiplex/multi-analyte panels on MesoScale Discovery and ProteinSimple Ella platforms. Results: A total of 15/33 subjects had detectable viral RNA in baseline samples (shortly after ICU admission). These initial viremia levels were low, and they did not change uniformly post-perfusion. Five of 55 protein analytes that were upregulated 1.4–120X at ICU admission relative to healthy controls showed significant decreases across the filter during the indicated time points on the first day of hemoperfusion: IP-10/CXCL10, fms-like tyrosine kinase 1, MIG/CXCL9, hepatocyte growth factor (HGF), and receptor for advanced glycosylation end products (RAGE). Paired t tests identified 25 additional analytes that showed significant decreases (p < 0.05) only without Bonferroni correction. Conclusion: Initial freely circulating SARS-CoV-2 RNA levels of ICU-admitted subjects were low or undetectable. The Seraph 100 filter did not significantly reduce viral RNA titers in their plasma. However, multiple circulating proteins with roles in inflammation, endothelial/epithelial damage, and/or angiogenesis decreased significantly across the filter. Larger prospective trials will be required to determine if such transient reductions translate into improved patient outcomes. However, this study did not demonstrate a direct reduction of free SARS-CoV-2 viral RNA by the Seraph 100.

Multi-analyte methods were applied to evaluate performance of the Seraph 100, an extracorporeal hemoperfusion therapy given emergency use authorization to treat severe COVID-19. Goals: The objective of this study was to determine if Seraph 100 hemoperfusion reduces SARS-CoV-2 RNA levels and/or markers of inflammation and epithelial/endothelial damage by measuring viral RNA and 78 protein analytes in blood from severe COVID-19 subjects immediately before and after hemoperfusion. Results: Multiple inflammatory protein mediators, but not viral RNA, were reduced in blood following Seraph 100 treatment. Conclusion: Potential benefits of Seraph 100 hemoperfusion may be mediated by reductions in endogenous circulating blood proteins, but not to a direct reduction of free SARS-CoV-2 viral RNA.

Infection by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) results in a wide range of symptoms and coronavirus disease of 2019 (COVID-19) severity. Patients who develop acute respiratory distress syndrome (ARDS) have higher risks of intensive care unit (ICU) admission and mortality [1]. The development of “pathogen-agnostic” therapies to mitigate or reverse such severe responses is a compelling goal to improve patient outcomes, especially following exposure to new pathogens for which no effective vaccines or therapies are yet available. The Seraph 100 Microbind Affinity Filter (Seraph 100, ExThera Corp., Martinez, CA, USA) is an extracorporeal hemoperfusion device designed to remove systemic viral, bacterial, and fungal pathogens and inflammatory mediators whose overproduction contributes to ARDS and organ failure [2, 3]. The filter cartridge contains heparin-coated beads that mimic the heparan sulfate-rich glycocalyx on vascular endothelial cells, which protects the vasculature but is also an initial attachment surface for multiple blood-borne pathogens. Two prospective, randomized clinical trials are evaluating safety and outcomes for subjects with sepsis/septic shock treated with the Seraph 100: NCT05011656 (USA) and NCT04547257 (Europe). An interventional study (NCT02914132) demonstrated the safety of the Seraph 100 for treatment of 15 septic patients on renal replacement therapy [4], and a recent pharmacokinetics study demonstrated that the Seraph 100 did not remove antibiotics [5].

In April 2020, the FDA granted emergency use authorization of the Seraph 100 for patients admitted to an ICU with confirmed or imminent ARDS or respiratory failure (EUA200165). SARS-CoV-2 binds to the angiotensin-converting enzyme 2 receptor, which is expressed on lung epithelia and extrapulmonary sites, including the vascular endothelium [6‒10]. SARS-CoV-2 has also been shown to bind heparin, cell-surface heparan sulfate glycans, and synthetic heparin analogs [11, 12]. Therefore, we hypothesized that Seraph 100 hemoperfusion could reduce viral titers and inflammatory mediators [13‒17]. Early reports of SARS-CoV-2 RNA in blood of COVID-19 patients produced varying estimates of viral RNA levels [18‒20]. However, a significant correlation of RNA titers with disease severity was seen in some studies [21‒23], including a 2021 meta-analysis [24]. Interestingly, Jacobs et al. [23] found higher RNA copy numbers and virion particles in pellets after plasma centrifugation, which co-localized with CD41 (GPIIb/IIIa), indicating that some viral RNA may be encapsulated in extracellular vesicles or attached to platelets, thereby avoiding rapid degradation.

A retrospective cohort study of 106 severe COVID-19 subjects, half of whom received Seraph 100 therapy, was recently completed. Seraph 100-treated subjects had more vasopressor-free days and were less likely to die. While these differences persisted after adjustment for potentially confounding variables, a post hoc analysis referencing an external control group did not suggest a mortality benefit [17]. The purpose of the present study was to (1) determine if the Seraph 100 reduces SARS-CoV-2 RNA titers in ICU-admitted COVID-19 subjects and (2) perform targeted proteomics on their plasma/serum samples, focusing on markers of inflammation or epithelial/endothelial damage, to better define the potential effects of hemoperfusion on clinically relevant circulating blood components.

This study was conducted using serial blood samples collected from 33 severe COVID-19 patients enrolled shortly after ICU admission. All subjects met the Emergency Use Authorization criteria for treatment with the Seraph 100. RNA and protein analytes were measured at baseline to post-filter at t = 4 h after filter application, pre- and post-filter at t = 1 h after filter application, pre- and post-filter at t = 4 h after filter application, and pre-filter at t = 1 h versus post-filter at t = 4 h after filter application (shown in Fig. 1). SARS-CoV-2 RNA levels in plasma were quantified by droplet digital (dd)PCR and nucleocapsid protein was measured by ELISA. Additional protein analytes were measured as described below (detailed methods are in online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000542995).

Fig. 1.

Schematic showing the Seraph 100 filter placement, the simultaneous withdrawal of blood samples from lines entering and exiting the filter at two time points, and the collection of venous blood samples on days 0, 4, 7, and 28 based on patient availability.

Fig. 1.

Schematic showing the Seraph 100 filter placement, the simultaneous withdrawal of blood samples from lines entering and exiting the filter at two time points, and the collection of venous blood samples on days 0, 4, 7, and 28 based on patient availability.

Close modal

MSD Platform Panels and Assays

Fifty-six protein biomarkers related to inflammation or endothelial/epithelial damage were measured in serum samples using MesoScale Discovery (MSD, Rockville, MD) panels according to the manufacturer’s instructions, including three panels developed in-house to assess additional inflammatory and endothelial/epithelial damage biomarkers. Healthy control serum levels were determined using samples from 25 local donors. Isoelectric points for the analytes were obtained from published databases [25, 26].

ProteinSimple Ella Platform Panels and Assays

Thirty-seven protein markers related to inflammation or endothelial/epithelial damage were measured in plasma samples using the ProteinSimple Ella immunoassay (Bio-Techne, Minneapolis, MN, USA) according to the manufacturer’s instructions. Concentrations were also measured in 20 healthy control plasmas (Precision Biospecimen Solutions, Inc.; Bethesda, MD, USA). Ten analytes were assayed on both the MSD and Ella platforms: angiopoietin-1, EGF, IFN-gamma, IL-10, IL-1RA, IL-5, IL-6, IL-8, TNF-alpha, and VEGF-A.

Demographic and Clinical Data

The duration of Seraph 100 treatments was typically 4–5 h. Demographic data are summarized in Table 1. No significant differences were noted between survivors (N = 15) and non-survivors (N = 18) in terms of age, sex, race and ethnicity, body mass index, APACHE II score, Charlson Comorbidity Index, mechanical ventilation, kidney replacement therapy (KRT), or extracorporeal membrane oxygenation (ECMO). Seraph treatments were performed alone for 11 subjects, in conjunction with KRT for 7 subjects, with ECMO for 13 subjects, and with both KRT and ECMO for 2 subjects.

Table 1.

Cohort characteristics

Full cohort (N = 33)Survivors (N = 15)Non-survivors (N = 18)p value
Age, mean (SD), years 44.6 (13.6) 45.1 (13.2) 44.3 (14.3) 0.87 
Sex, N (%)    0.29 
 Female 12 (36.4) 4 (26.7) 8 (44.4)  
 Male 21 (63.6) 11 (73.3) 10 (55.6)  
Race and ethnicity, N (%)    0.36 
 Non-Hispanic white 13 (39.4) 8 (53.3) 5 (27.8)  
 Non-Hispanic black 2 (6.1) 0 (0.0) 2 (11.1)  
 Non-Hispanic Asian 1 (3.0) 1 (6.7) 0 (0.0)  
 Hispanic 14 (42.4) 5 (33.3) 9 (50.0)  
 Unknown 3 (9.1) 1 (6.7) 2 (11.1)  
Body mass index, median (IQR) 35.5 (29.4–43.1) 32.0 (28.9–43.9) 36.3 (31.6–43.1) 0.69 
APACHE II, mean (IQR) 13 (9–22) 13 (8–22) 12.5 (9–22) 0.82 
Charlson Comorbidity Index, median (IQR) 1 (0–1) 1 (0–1) 1 (0–2) 0.74 
Required ECMO, N (%) 16 (48.5) 5 (33.3) 11 (61.1) 0.11 
Required mechanical ventilation, N (%) 30 (90.9) 12 (80.0) 18 (100) 0.08 
Required KRT, N (%) 18 (54.6) 8 (53.3) 10 (55.6) 0.9 
Treated with corticosteroids, N (%) 33 (100) 15 (100) 18 (100) 
Treated with remdesivir, N (%) 26 (78.8) 15 (100) 14 (77.8) 
Vasopressor-free daysa, median (IQR) 15 (8–28) 24 (12–28) 8 (5–12) <0.01 
Mechanical ventilation-free days, median (IQR) 6 (2–22) 18 (6–26) 3.5 (1–8) <0.01 
ICU-free daysa, median (IQR) 0 (0–7) 12 (0–20) 0 (0–0) <0.01 
Days until treatmentb, median (IQR) 3.6 (1.7–5.7) 2.6 (1.5–5.7) 4.4 (3.5–7.3) 0.09 
Length of stayc, median (IQR) 28 (9–50) 
Days until deathd, median (IQR) 36 (16–80) 
Full cohort (N = 33)Survivors (N = 15)Non-survivors (N = 18)p value
Age, mean (SD), years 44.6 (13.6) 45.1 (13.2) 44.3 (14.3) 0.87 
Sex, N (%)    0.29 
 Female 12 (36.4) 4 (26.7) 8 (44.4)  
 Male 21 (63.6) 11 (73.3) 10 (55.6)  
Race and ethnicity, N (%)    0.36 
 Non-Hispanic white 13 (39.4) 8 (53.3) 5 (27.8)  
 Non-Hispanic black 2 (6.1) 0 (0.0) 2 (11.1)  
 Non-Hispanic Asian 1 (3.0) 1 (6.7) 0 (0.0)  
 Hispanic 14 (42.4) 5 (33.3) 9 (50.0)  
 Unknown 3 (9.1) 1 (6.7) 2 (11.1)  
Body mass index, median (IQR) 35.5 (29.4–43.1) 32.0 (28.9–43.9) 36.3 (31.6–43.1) 0.69 
APACHE II, mean (IQR) 13 (9–22) 13 (8–22) 12.5 (9–22) 0.82 
Charlson Comorbidity Index, median (IQR) 1 (0–1) 1 (0–1) 1 (0–2) 0.74 
Required ECMO, N (%) 16 (48.5) 5 (33.3) 11 (61.1) 0.11 
Required mechanical ventilation, N (%) 30 (90.9) 12 (80.0) 18 (100) 0.08 
Required KRT, N (%) 18 (54.6) 8 (53.3) 10 (55.6) 0.9 
Treated with corticosteroids, N (%) 33 (100) 15 (100) 18 (100) 
Treated with remdesivir, N (%) 26 (78.8) 15 (100) 14 (77.8) 
Vasopressor-free daysa, median (IQR) 15 (8–28) 24 (12–28) 8 (5–12) <0.01 
Mechanical ventilation-free days, median (IQR) 6 (2–22) 18 (6–26) 3.5 (1–8) <0.01 
ICU-free daysa, median (IQR) 0 (0–7) 12 (0–20) 0 (0–0) <0.01 
Days until treatmentb, median (IQR) 3.6 (1.7–5.7) 2.6 (1.5–5.7) 4.4 (3.5–7.3) 0.09 
Length of stayc, median (IQR) 28 (9–50) 
Days until deathd, median (IQR) 36 (16–80) 

IQR, interquartile range.

aNumber of days the subject was alive and not on vasopressors, mechanical ventilation, or in ICU in the first 28 days after ICU admission.

bNumber of days from ICU admission until first Seraph 100 treatment.

cNumber of days from ICU admission until discharge among survivors.

dNumber of days from ICU admission until death among non-survivors.

SARS-CoV-2 RNA Titers and Nucleocapsid Protein Levels

Plasma samples were analyzed by ddRT-PCR, quantifying two distinct viral RNAs (N1 and N2). Fifteen subjects had detectable viral RNA at baseline, ten had no detectable viral RNA in their samples, and 8 had varying levels over the course of their hospital stay (shown in Fig. 2, online suppl. Table S1). Thus, the filter did not consistently affect viral RNA levels. Only one subject had detectable nucleocapsid protein on days 4 and day 7 after Seraph 100 placement, respectively (not shown). No SARS-CoV-2 RNA was detected for this subject at any time point.

Fig. 2.

Seraph 100 filtration did not remove SARS-CoV-2 RNA from plasma. Viral amplicons N1 and N2 quantified by ddPCR from plasma samples of survivors (a) and those that died (b), obtained pre- and post-filtration at t = 1 h and 4 h, and in samples obtained at days 1, 4, 7, and 28. Box plots show medians with interquartile ranges of cDNA amplicon levels in samples from subjects that had measurable cDNA at the indicated time points. None of the comparisons between group means or medians showed statistically significant differences. Lines connect measurements for the same subject before and after filter application and show either decreased (solid line) or increased (dashed line) viral RNA levels across the filter. Dots at the bottom of the plot correspond to no viral RNA. The heat maps summarize the SARS-CoV-2 N1 and N2 amplicon levels measured at all time points for all 33 subjects.

Fig. 2.

Seraph 100 filtration did not remove SARS-CoV-2 RNA from plasma. Viral amplicons N1 and N2 quantified by ddPCR from plasma samples of survivors (a) and those that died (b), obtained pre- and post-filtration at t = 1 h and 4 h, and in samples obtained at days 1, 4, 7, and 28. Box plots show medians with interquartile ranges of cDNA amplicon levels in samples from subjects that had measurable cDNA at the indicated time points. None of the comparisons between group means or medians showed statistically significant differences. Lines connect measurements for the same subject before and after filter application and show either decreased (solid line) or increased (dashed line) viral RNA levels across the filter. Dots at the bottom of the plot correspond to no viral RNA. The heat maps summarize the SARS-CoV-2 N1 and N2 amplicon levels measured at all time points for all 33 subjects.

Close modal

Plasma and Serum Protein Levels

Of 78 analytes measured in plasma and serum (shown in online suppl. Fig. S1), 55 circulated at 1.4–120× higher levels in COVID-19 baseline (pre-Seraph 100 treatment) samples compared to healthy controls (shown in Fig. 3; online suppl. Table S2). Of 14 analytes measured on both platforms, only EGF showed discrepant results. Therefore, EGF was removed from subsequent analyses. IL-1alpha, IL-2 and IL-4 were below the MSD instrument detection level in healthy controls and therefore could not be compared to levels in COVID-19 subjects. Of the remaining comparisons of healthy controls to severe COVID-19 baseline samples, 61 met the Bonferroni-corrected thresholds for significance: 51 upregulated and 6 downregulated. Levels of E-selectin, periostin/OSF-2, fibrinogen D-dimer, galectin-1 and IL-6 were >50× higher in COVID-19 than in healthy controls. An additional 15 analytes were >10× higher, 11 were >4× and <10× higher, and 24 were 1.4–4× higher. PIGF, eotaxin, G-CSF, bFGF, endothelin-1 and IL-12/IL-23p40 were significantly lower in COVID-19 baseline samples than in healthy controls (shown in online suppl. Fig. S2).

Fig. 3.

Ratios of plasma and serum analytes in plasma-serum from severe COVID-19 subjects to healthy controls. a–d Fifty-five of the 78 MSD and Ella platforms were significantly higher in the plasma/serum from severe COVID-19 subjects at ICU admission (baseline samples) compared to plasma (ELLA platform) or serum (MSD platform) healthy control samples measured in-house on the same instruments. For analytes measured by both MSD and ELLA, the averaged ratio is shown; in all cases, these ratios were highly similar.

Fig. 3.

Ratios of plasma and serum analytes in plasma-serum from severe COVID-19 subjects to healthy controls. a–d Fifty-five of the 78 MSD and Ella platforms were significantly higher in the plasma/serum from severe COVID-19 subjects at ICU admission (baseline samples) compared to plasma (ELLA platform) or serum (MSD platform) healthy control samples measured in-house on the same instruments. For analytes measured by both MSD and ELLA, the averaged ratio is shown; in all cases, these ratios were highly similar.

Close modal

Effects of Seraph 100 Hemoperfusion on Dysregulated Plasma/Serum Proteins

None of the 5 analytes that were upregulated >50× in severe COVID-19 compared to healthy controls showed a significant change in concentration after Seraph 100 hemoperfusion on day 1. Five of the analytes upregulated 1.4–50X in severe COVID-19 showed significant decreases, after Bonferroni correction, over one or more time intervals during the first 4 h of hemoperfusion (shown in Fig. 4; online suppl. Fig. S3). An additional 25 analytes showed decreases that were significant (p < 0.05) after pairwise comparisons without Bonferroni correction (shown in online suppl. Fig. S3–S5; Tables S3–S5).

Fig. 4.

Levels of several protein analytes changed during Seraph 100 hemoperfusion. Analytes in severe COVID-19 plasma (ELLA assays) or serum (MSD assays) that were upregulated between 1.4–50X in baseline samples from severe COVID-19 subjects and that showed significant changes at one or more intervals between baseline (ICU admission) and after 4 h of hemoperfusion with the Seraph 100 on day 1. Solid horizontal lines indicate changes over the indicated interval that met the Bonferroni-corrected criteria of p < 0.00022 (MSD platform) or p < 0.00034 (ELLA platform). Dotted horizontal lines indicate changes over the indicated interval with p < 0.05 after pairwise t tests comparing all measurements pre- and post-Seraph 100 perfusion. IP-10, Flt-1, MIG, and HGF showed the largest significant changes across the filter. Changes of additional analytes are summarized in online supplementary Figures S3–S5.

Fig. 4.

Levels of several protein analytes changed during Seraph 100 hemoperfusion. Analytes in severe COVID-19 plasma (ELLA assays) or serum (MSD assays) that were upregulated between 1.4–50X in baseline samples from severe COVID-19 subjects and that showed significant changes at one or more intervals between baseline (ICU admission) and after 4 h of hemoperfusion with the Seraph 100 on day 1. Solid horizontal lines indicate changes over the indicated interval that met the Bonferroni-corrected criteria of p < 0.00022 (MSD platform) or p < 0.00034 (ELLA platform). Dotted horizontal lines indicate changes over the indicated interval with p < 0.05 after pairwise t tests comparing all measurements pre- and post-Seraph 100 perfusion. IP-10, Flt-1, MIG, and HGF showed the largest significant changes across the filter. Changes of additional analytes are summarized in online supplementary Figures S3–S5.

Close modal

The intervention in this study was the use of the Seraph 100 Microbind Affinity Blood Filter device to perfuse blood of severe COVID-19 subjects. Similar to previous studies of severe COVID-19 [21‒23], most of the subjects (23/33) had detectable viral RNA in their blood during at least one point of the study. Quantification of viral RNA in plasma immediately pre- and post-filter showed that Seraph 100 hemoperfusion did not consistently reduce viral RNA copies. It is possible that other heparin-binding blood components out-competed the virus or viral particles, particularly if the level of viremia was low. We used ddRT-PCR in this study to quantify viral RNA copies in the blood; thus, it was not possible to discriminate between live virus, viral particles, or free viral RNA in the blood; free viral RNA may be less likely to be captured by the filter and is a potential explanation for why RNA copies were not reduced by filtration.

Another limitation of the study was that the protocol specified collection of platelet-poor plasma. A subsequent publication [23] showed SARS-CoV-2 RNA and viral particles were present in pellets after plasma centrifugation, indicating encapsulation of viral RNA in platelets and/or extracellular vesicles. Thus, the levels of virus in blood were likely higher, in some cases, than those measured in our plasma samples. An independent study reported the Seraph 100 reduces plasma levels of the SARS-CoV-2 nucleocapsid protein [13], but nucleocapsid protein was undetectable in all but one of our subjects.

A total of 78 markers of inflammation and epithelial/endothelial damage were chosen as possible systemic biomarkers in severe COVID-19. Many of the MSD and ELLA platform analytes measured have high isoelectric points (positive charge) and have been shown to bind to heparin in vitro (shown in online suppl. Table S6). The fact that most of the analytes showed little or no change in concentration after passing through the Seraph 100 suggests that their interactions with the negatively-charged filter were transient and/or competed by other blood components, notably including the glycocalyx lining the vasculature. Nevertheless, among the 55 measured analytes that circulated in severe COVID-19 subjects at 1.4–120X healthy control levels, five decreased significantly over 4 h of hemoperfusion. This suggested either their direct removal by the filter or a decrease that happened in tandem with hemoperfusion, for example, due to endogenous metabolism. Regarding potential safety issues, an earlier pharmacokinetic/pharmacodynamic analysis established that Seraph 100 hemofiltration did not affect the levels of antibiotics administered to several of the 33 subjects [5].

The primary comparisons of interest in this study were (1) levels of viral RNA and analytes related to inflammation and epithelial/endothelial damage in severe COVID-19 (measured shortly after ICU admission) compared to healthy controls and (2) comparisons to determine which of these were reduced or eliminated by Seraph 100 hemoperfusion. Samples drawn simultaneously from lines going into and out of the Seraph 100 filter at 1 h and 4 h following filter application indicated the effect of a single pass through the filter. In several cases, the levels decreased at both of these time points (notably IP-10, fms-like tyrosine kinase 1 [Flt-1], and MIG), suggesting a consistent effect of the filter on these analytes. Others showed a decrease over the first 4 h of hemoperfusion, indicating either a cumulative effect of the filter or correlative changes in the subjects’ clinical status. Among these, IP-10, Flt-1, and hepatocyte growth factor (HGF) met the Bonferroni thresholds for significant decreases over the first 4 h of hemoperfusion, and 10 additional analytes showed decreases with p < 0.05 over the same period (IL-10, MMP-9, angiopoietin-1, VEGF-A, LCN2, IL-16, bFGF, IL-17A, PIGF, and eotaxin). The ELLA platform results are broadly in agreement with a previous study using this same panel, where significant increases in markers of inflammation were observed in COVID-19 patients relative to healthy controls [27].

The most highly upregulated markers, E-selectin-1, periostin/OSF-2, fibrinogen D-dimer, galectin-1 and IL-6, have all been noted as correlates of COVID-19 severity [28‒32], as have most of the other markers measured in this study. The five upregulated analytes that met the Bonferroni-corrected threshold for significant decreases during Seraph 100 hemoperfusion were IP-10, Flt-1, MIG (CXCL9), HGF, and receptor for advanced glycosylation end products (RAGE), while eotaxin-2 levels were depleted in the severe COVID-19 subjects and decreased further across the filter.

Regarding the biological roles of analytes that changed across the filter, sustained upregulation of IP-10, together with IL-6 and IL-10, has been shown to have a greater prognostic power for COVID-19 severity than other commonly measured inflammatory markers including total lymphocytes, D-dimer, CRP, ferritin, and albumin [33]. The Flt-1/PIGF ratio, a clinical indicator of endothelial damage correlated with preeclampsia risk in pregnant women, is also highly upregulated in COVID-19 [34]; The Flt-1/PIGF ratio in the present study decreased an average of >6X (from 288 to 45.8) in severe COVID-19 subjects in samples collected after 1 h and 4 h of Seraph 100 hemoperfusion. Interestingly, it has recently been demonstrated that the SARS-CoV-2 receptor-binding domain binds to both soluble Flt-1 and Flt-1 expressed on A549 cells, and anti-receptor-binding domain antibodies were shown to cross-react with sFlt-1 [35], while the SARS-CoV-2 N-protein binds to RAGE and activates pro-inflammatory signaling in vitro [36]. RAGE has also been implicated in COVID-19 pathology/severity [37, 38]. An analysis of 444 COVID-19 subjects identified both elevated MIG levels and decreased eotaxin-1 levels in those admitted to the ICU [39]. Elevated HGF has also been associated with COVID-19 severity [39‒41], consistent with tissue damage in the lung and other organs. In summary, all five upregulated blood analytes that were seen to decrease significantly after Seraph 100 hemoperfusion (IP-10, Flt-1, MIG, HGF, and RAGE), as well as other analytes that were decreased with p < 0.05 but that did not meet the stringent Bonferroni-corrected threshold for significance, have been associated with COVID-19 severity. Such reductions across the filter may provide clinical benefit through a temporary reduction in inflammatory mediators by allowing endogenous tissue repair to proceed, thereby promoting recovery. As a cautionary note, however, beneficial substances in blood could also be reduced during hemoperfusion. For example, eotaxin levels in the severe COVID-19 baseline samples were ∼1% of healthy control levels, and after 4 h of Seraph 100 treatment, they decreased further to 0.4% of healthy control levels. This difference is unlikely to have clinical significance, as both levels reflect a profound depletion. We cannot rule out reductions of additional blood components, however, that could potentially impede, rather than promote, healing. This study had several limitations: (1) small cohort; (2) no control group consisting of severe COVID-19 subjects not treated with the Seraph 100; (3) viral RNA levels in plasma cannot account for viral RNA in circulating live virus or viral particles, free viral RNA, or viral RNA sequestered by extracellular vesicles, platelets, or other cells and tissues.

The primary goal of this study was to evaluate changes in 78 blood analytes immediately pre- and post-filter in severe COVID-19 subjects treated with the Seraph 100. Analyses of dynamic changes in these blood components over the course of their hospital stays, and identification of correlates with mortality, will be published separately. The definitive tests of “pathogen-agnostic” blood filtration devices will necessarily be the clinical outcomes of treated versus non-treated patients. Although this study did not demonstrate a reduction in SARS-CoV-2 RNA by the Seraph 100, several pro-inflammatory proteins were reduced significantly immediately post-hemoperfusion. Further identification of analytes reduced in blood following hemoperfusion, as in the present study, could help determine specific variables that are impacted by hemoperfusion and also identify potential effects on inflammatory mediators and other blood proteins relevant to patient outcomes. Along this line, we recently showed that Seraph 100 hemoperfusion does not affect levels of either serum or mucosal IgG and IgA antibodies in the same severe COVID-19 subjects enrolled in the present study [42].

We extend our sincere thanks to the clinical, scientific, and administrative staff at all sites for their respective, essential roles in conducting this study during the pandemic. We are very grateful to all of the subjects and their families/representatives for their participation. Special thanks go to Dr. Kevin Chung for early support, to Devi Gunasekera, MBBS, for sample processing and excellent record keeping at USUHS, and to Ms. Catherine Grill for overall study management. Please see the online supplementary material for a full list of PURIFY-OBS-1 investigators. The contents of this presentation are the sole responsibility of the author(s) and do not necessarily reflect the views, opinions, or policies of Uniformed Services University of the Health (USUHS), The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., the Department of Defense (DoD), or the Departments of the Army, Navy, or Air Force. Mention of trade names, commercial products, or organizations does not imply endorsement by the US Government. This work was prepared as part of the author(s)’ official duties. Title 17 U.S.C. § 105 provides that “Copyright protection under this title is not available for any work of the USA Government.” Title 17 U.S.C. §101 defines US Government work as work prepared by a military service member or employee of the US Government as part of that person’s official duties.

The Blood Purification with Seraph 100 Microbind Affinity Blood Filter for the Treatment of Severe COVID-19: An Observational Study (PURIFY-OBS-1) was approved by the Advarra Institutional Review Board (IRB), approval # Pro00047577, in accordance with all applicable Federal regulations governing human research protections (Clinical Trials.gov identifier NCT04606498). This study included severe COVID-19 subjects and both contemporaneous and historical control groups. The study included both retrospective and prospective subjects. For retrospective subjects, the IRB provided a waiver for informed consent. For prospective subjects, written informed consent to participate in the study has been obtained from all participants and all vulnerable participants’ legal guardian/next of kin, consistent with the Principles of Helsinki.

The authors have no conflicts of interest to declare.

This work was funded by Health Affairs (Award #HU00012020070), which had no role in the design, data collection, data analysis, or reporting of this study.

Michael Rouse, PhD; Victor A. Sugiharto, PhD; and Ian J. Stewart, MD: data curation, formal analysis, investigation, methodology, validation, writing – original draft, and writing – review and editing. Eric R. Gann, PhD; Joost Brandsma, PhD; Henry Robertson, PhD; Pavol Genzor, PhD; Seth A. Schobel, PhD; Josh G. Chenoweth, PhD; Mark P. Simons, PhD, MSPH: data curation, formal analysis, investigation, methodology, validation, and writing – review and editing. Chen, Hua-Wie and Pooja Vir, PhD: data curation, investigation, methodology, validation, writing – reviewed and approved the manuscript. Sarah A. Jenkins, MD, and Danielle V. Clark, PhD: supervision of staff and writing – reviewed and approved the manuscript. Jeffrey Della Volpe, MD; Stephen Chitty, MD; Ian M. Rivera, MD; Michael Lewis, MD; Caroline Park, MD; and Parikh Amay, MD: enrolled patients and writing – reviewed and approved the manuscript. Kathleen P. Pratt, PhD: conceptualization, data curation, formal analysis, investigation, methodology, validation, writing – original draft, and writing – review and editing.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.

1.
Meyer
NJ
,
Gattinoni
L
,
Calfee
CS
.
Acute respiratory distress syndrome
.
Lancet
.
2021
;
398
(
10300
):
622
37
.
2.
Seffer
MT
,
Cottam
D
,
Forni
LG
,
Kielstein
JT
.
Heparin 2.0: a new approach to the infection crisis
.
Blood Purif
.
2021
;
50
(
1
):
28
34
.
3.
Seffer
MT
,
Eden
G
,
Engelmann
S
,
Kielstein
JT
.
Elimination of Staphylococcus aureus from the bloodstream using a novel biomimetic sorbent haemoperfusion device
.
BMJ Case Rep
.
2020
;
13
(
8
):
e235262
.
4.
Eden
G
,
Schmidt
JJ
,
Buttner
S
,
Kumpers
P
,
Hafer
C
,
Rovas
A
, et al
.
Safety and efficacy of the Seraph® 100 Microbind® Affinity Blood Filter to remove bacteria from the blood stream: results of the first in human study
.
Crit Care
.
2022
;
26
(
1
):
181
.
5.
Selig
DJ
,
Reed
T
,
Chung
KK
,
Kress
AT
,
Stewart
IJ
,
DeLuca
JP
.
Hemoperfusion with Seraph 100 Microbind affinity blood filter unlikely to require increased antibiotic dosing: a simulations study using a pharmacokinetic/pharmacodynamic approach
.
Blood Purif
.
2023
;
52
(
1
):
25
31
.
6.
Hoffmann
M
,
Kleine-Weber
H
,
Schroeder
S
,
Kruger
N
,
Herrler
T
,
Erichsen
S
, et al
.
SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor
.
Cell
.
2020
;
181
(
2
):
271
80.e8
.
7.
Wiersinga
WJ
,
Rhodes
A
,
Cheng
AC
,
Peacock
SJ
,
Prescott
HC
.
Pathophysiology, transmission, diagnosis, and treatment of coronavirus disease 2019 (COVID-19): a review
.
JAMA
.
2020
;
324
(
8
):
782
93
.
8.
Varga
Z
,
Flammer
AJ
,
Steiger
P
,
Haberecker
M
,
Andermatt
R
,
Zinkernagel
AS
, et al
.
Endothelial cell infection and endotheliitis in COVID-19
.
Lancet
.
2020
;
395
(
10234
):
1417
8
.
9.
Teuwen
LA
,
Geldhof
V
,
Pasut
A
,
Carmeliet
P
.
COVID-19: the vasculature unleashed
.
Nat Rev Immunol
.
2020
;
20
(
7
):
389
91
.
10.
Iba
T
,
Connors
JM
,
Levy
JH
.
The coagulopathy, endotheliopathy, and vasculitis of COVID-19
.
Inflamm Res
.
2020
;
69
(
12
):
1181
9
.
11.
Clausen
TM
,
Sandoval
DR
,
Spliid
CB
,
Pihl
J
,
Perrett
HR
,
Painter
CD
, et al
.
SARS-CoV-2 infection depends on cellular heparan sulfate and ACE2
.
Cell
.
2020
;
183
(
4
):
1043
57.e15
.
12.
Parafioriti
M
,
Ni
M
,
Petitou
M
,
Mycroft-West
CJ
,
Rudd
TR
,
Gandhi
NS
, et al
.
Evidence for multiple binding modes in the initial contact between SARS-CoV-2 spike S1 protein and cell surface glycans
.
Chemistry
.
2023
;
29
(
1
):
e202202599
.
13.
Kielstein
JT
,
Borchina
DN
,
Fuhner
T
,
Hwang
S
,
Mattoon
D
,
Ball
AJ
.
Hemofiltration with the Seraph® 100 Microbind® Affinity filter decreases SARS-CoV-2 nucleocapsid protein in critically ill COVID-19 patients
.
Crit Care
.
2021
;
25
(
1
):
190
.
14.
Rifkin
BS
,
Stewart
IJ
.
Seraph-100 hemoperfusion in SARS-CoV-2-infected patients early in critical illness: a case series
.
Blood Purif
.
2022
;
51
(
4
):
317
20
.
15.
Kelly
MM
,
Wilkinson
JD
,
Rastegar
M
,
Lewis
MS
,
Betancourt
J
.
Two patients with severe COVID pneumonia treated with the seraph-100 Microbind affinity blood filter
.
J Intensive Care Med
.
2021
;
36
(
10
):
1228
32
.
16.
Schmidt
JJ
,
Borchina
DN
,
Van't Klooster
M
,
Bulhan-Soki
K
,
Okioma
R
,
Herbst
L
, et al
.
Interim analysis of the COSA (COVID-19 patients treated with the Seraph(R) 100 Microbind(R) Affinity filter) registry
.
Nephrol Dial Transplant
.
2022
;
37
(
4
):
673
80
.
17.
Chitty
SA
,
Mobbs
S
,
Rifkin
BS
,
Stogner
SW
,
Lewis
MS
,
Betancourt
J
, et al
.
A multicenter evaluation of the Seraph 100 Microbind affinity blood filter for the treatment of severe COVID-19
.
Crit Care Explor
.
2022
;
4
(
4
):
e0662
.
18.
Wolfel
R
,
Corman
VM
,
Guggemos
W
,
Seilmaier
M
,
Zange
S
,
Muller
MA
, et al
.
Virological assessment of hospitalized patients with COVID-2019
.
Nature
.
2020
;
581
(
7809
):
465
9
.
19.
Wang
W
,
Xu
Y
,
Gao
R
,
Lu
R
,
Han
K
,
Wu
G
, et al
.
Detection of SARS-CoV-2 in different types of clinical specimens
.
JAMA
.
2020
;
323
(
18
):
1843
4
.
20.
Fajnzylber
J
,
Regan
J
,
Coxen
K
,
Corry
H
,
Wong
C
,
Rosenthal
A
, et al
.
SARS-CoV-2 viral load is associated with increased disease severity and mortality
.
Nat Commun
.
2020
;
11
(
1
):
5493
.
21.
Xu
D
,
Zhou
F
,
Sun
W
,
Chen
L
,
Lan
L
,
Li
H
, et al
.
Relationship between serum severe acute respiratory syndrome coronavirus 2 nucleic acid and organ damage in coronavirus 2019 patients: a cohort study
.
Clin Infect Dis
.
2021
;
73
(
1
):
68
75
.
22.
Chen
X
,
Zhao
B
,
Qu
Y
,
Chen
Y
,
Xiong
J
,
Feng
Y
, et al
.
Detectable serum severe acute respiratory syndrome coronavirus 2 viral load (RNAemia) is closely correlated with drastically elevated interleukin 6 level in critically ill patients with coronavirus disease 2019
.
Clin Infect Dis
.
2020
;
71
(
8
):
1937
42
.
23.
Jacobs
JL
,
Bain
W
,
Naqvi
A
,
Staines
B
,
Castanha
PMS
,
Yang
H
, et al
.
Severe acute respiratory syndrome coronavirus 2 viremia is associated with coronavirus disease 2019 severity and predicts clinical outcomes
.
Clin Infect Dis
.
2022
;
74
(
9
):
1525
33
.
24.
Tang
K
,
Wu
L
,
Luo
Y
,
Gong
B
.
Quantitative assessment of SARS-CoV-2 RNAemia and outcome in patients with coronavirus disease 2019
.
J Med Virol
.
2021
;
93
(
5
):
3165
75
.
25.
Kozlowski
LP
.
Proteome-pI: proteome isoelectric point database
.
Nucleic Acids Res
.
2017
;
45
(
D1
):
D1112
6
.
26.
Kozlowski
LP
.
Proteome-pI 2.0: proteome isoelectric point database update
.
Nucleic Acids Res
.
2022
;
50
(
D1
):
D1535
40
.
27.
Brandsma
J
,
Chenoweth
JG
,
Gregory
MK
,
Krishnan
S
,
Blair
PW
,
Striegel
DA
, et al
.
Assessing the use of a micro-sampling device for measuring blood protein levels in healthy subjects and COVID-19 patients
.
PLoS One
.
2022
;
17
(
8
):
e0272572
.
28.
Oliva
A
,
Rando
E
,
Al Ismail
D
,
De Angelis
M
,
Cancelli
F
,
Miele
MC
, et al
.
Role of serum E-selectin as a biomarker of infection severity in coronavirus disease 2019
.
J Clin Med
.
2021
;
10
(
17
):
4018
.
29.
Cabalak
M
,
Dogan
S
,
Bal
T
,
Dikmen
N
.
Serum periostin levels in COVID-19: is it useful as a new biomarker
.
Int J Clin Pract
.
2021
;
75
(
11
):
e14728
.
30.
Zhou
F
,
Yu
T
,
Du
R
,
Fan
G
,
Liu
Y
,
Liu
Z
, et al
.
Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study
.
Lancet
.
2020
;
395
(
10229
):
1054
62
.
31.
Markovic
SS
,
Gajovic
N
,
Jurisevic
M
,
Jovanovic
M
,
Jovicic
BP
,
Arsenijevic
N
, et al
.
Galectin-1 as the new player in staging and prognosis of COVID-19
.
Sci Rep
.
2022
;
12
(
1
):
1272
.
32.
Potere
N
,
Batticciotto
A
,
Vecchie
A
,
Porreca
E
,
Cappelli
A
,
Abbate
A
, et al
.
The role of IL-6 and IL-6 blockade in COVID-19
.
Expert Rev Clin Immunol
.
2021
;
17
(
6
):
601
18
.
33.
Laing
AG
,
Lorenc
A
,
Del Molino Del Barrio
I
,
Das
A
,
Fish
M
,
Monin
L
, et al
.
A dynamic COVID-19 immune signature includes associations with poor prognosis
.
Nat Med
.
2020
;
26
(
10
):
1623
35
.
34.
Giardini
V
,
Carrer
A
,
Casati
M
,
Contro
E
,
Vergani
P
,
Gambacorti-Passerini
C
.
Increased sFLT-1/PlGF ratio in COVID-19: a novel link to angiotensin II-mediated endothelial dysfunction
.
Am J Hematol
.
2020
;
95
(
8
):
E188
91
.
35.
Mutahar
AZI
,
Devaramani
M
,
Dayal
R
,
Saini
DK
,
Salimath
PV
,
Salimath
BP
.
Vascular endothelial growth factor receptor, fms-like tyrosine kinase-1 (Flt-1), as a novel binding partner for SARS-CoV-2 spike receptor-binding domain
.
Front Immunol
.
2022
;
13
:
906063
.
36.
Xia
J
,
Wang
J
,
Ying
L
,
Huang
R
,
Zhang
K
,
Zhang
R
, et al
.
RAGE is a receptor for SARS-CoV-2 N protein and mediates N protein-induced acute lung injury
.
Am J Respir Cell Mol Biol
.
2023
;
69
(
5
):
508
20
.
37.
Sellegounder
D
,
Zafari
P
,
Rajabinejad
M
,
Taghadosi
M
,
Kapahi
P
.
Advanced glycation end products (AGEs) and its receptor, RAGE, modulate age-dependent COVID-19 morbidity and mortality. A review and hypothesis
.
Int Immunopharmacol
.
2021
;
98
:
107806
.
38.
Chiappalupi
S
,
Salvadori
L
,
Donato
R
,
Riuzzi
F
,
Sorci
G
.
Hyperactivated RAGE in comorbidities as a risk factor for severe COVID-19-the role of RAGE-RAS crosstalk
.
Biomolecules
.
2021
;
11
(
6
):
876
.
39.
Abers
MS
,
Delmonte
OM
,
Ricotta
EE
,
Fintzi
J
,
Fink
DL
,
de Jesus
AAA
, et al
.
An immune-based biomarker signature is associated with mortality in COVID-19 patients
.
JCI Insight
.
2021
;
6
(
1
):
e144455
.
40.
Wolszczak-Biedrzycka
B
,
Dorf
J
,
Wojewodzka-Zelezniakowicz
M
,
Zendzian-Piotrowska
M
,
Dymicka-Piekarska
V
,
Matowicka-Karna
J
, et al
.
Changes in chemokine and growth factor levels may be useful biomarkers for monitoring disease severity in COVID-19 patients; a pilot study
.
Front Immunol
.
2023
;
14
:
1320362
.
41.
Paranga
TG
,
Pavel-Tanasa
M
,
Constantinescu
D
,
Plesca
CE
,
Petrovici
C
,
Miftode
IL
, et al
.
Comparison of C-reactive protein with distinct hyperinflammatory biomarkers in association with COVID-19 severity, mortality and SARS-CoV-2 variants
.
Front Immunol
.
2023
;
14
:
1213246
.
42.
Conner
TL
,
Vir
P
,
Laing
ED
,
Stewart
IJ
,
Mitre
E
,
Pratt
KP
.
The Seraph 100® Microbind affinity blood filter does not alter levels of circulating or mucosal antibodies in critical COVID-19 patients
.
Antibodies
.
2024
;
13
(
3
):
65
.