Beyond conventional care: The therapeutic potential of hemoperfusion in severe COVID-19

Abstract

BACKGROUND

Hemoperfusion (HP) is an extracorporeal blood purification modality utilized to remove small- to medium-sized molecules, such as toxins and cytokines, that are difficult to remove by conventional hemodialysis. In clinical practice, HP has been successfully used as a salvage therapy for drug overdose and occasionally in patients with liver failure and sepsis.

AIM

To summarize the clinical outcomes of a series of patients with severe coronavirus disease 2019 (COVID-19) who received HP.

METHODS

Here, we summarize the clinical outcomes of a series of 18 patients with severe COVID-19 who received HP in our institution during the COVID-19 pandemic. A review of the literature was also performed.

RESULTS

HP was well-tolerated, and after an average of three sessions, respiratory and cardiovascular parameters as well as blood inflammatory markers improved in most patients. Ten patients were discharged alive. Our literature search identified a total of 20 studies (873 patients) in which HP was used for COVID-19. Nine studies reported improvements in respiratory parameters, and 13 studies (438 patients in total) reported better survival rates in patients undergoing HP.

CONCLUSION

HP was well-tolerated in patients with severe COVID-19, and most studies reported improved clinical parameters, including better survival rates, when HP was used in patients with severe COVID-19. Further research, especially prospective studies, is needed to evaluate the utility of HP as an early and supportive therapy for critically ill patients due to infectious diseases, such as those with COVID-19 or severe sepsis.

Key Words: Hemoperfusion; Cytokine storm; Severe COVID-19; Blood; Therapy; Survival

Core Tip: Hemoperfusion (HP) is an extracorporeal blood purification therapy that is increasingly being utilized in the intensive care unit. We show that HP improved respiratory and cardiovascular parameters and various inflammatory markers in a series of coronavirus disease 2019 patients in critical condition, in agreement with various reports from the literature. However, the absence of data from randomized controlled trials and a lack of consensus guidelines remain important issues for the utilization of HP in clinical practice.

INTRODUCTION

Hemoperfusion (HP), also referred to as hemadsorption, is an extracorporeal blood purification modality that consists of the passage of anticoagulated whole blood through a column device that contains adsorbent particles for the removal of small- to medium-sized molecules (ranging from 100 to 40000 Daltons) such as toxins and cytokines that are difficult to remove by conventional hemodialysis[1]. Adsorbent particles can be made of activated charcoal or resins, such as polystyrene and polymyxin B. Charcoal has a high affinity for water-soluble molecules. In contrast, resins have a greater affinity for lipid-soluble molecules, and polymyxin B is typically used for the removal of bacterial endotoxins in the setting of sepsis.

HP has been clinically available since the 1970s, and despite its relatively low cost and potential usefulness in a broad spectrum of clinical conditions, its major clinical application has been largely the removal of drugs or poisons from the blood in cases of acute poisoning or drug overdose. Less frequently, it has been used as a supportive treatment for patients with severe liver dysfunction before and after liver transplantation, and occasionally, in conjunction with other extracorporeal techniques. HP is utilized for removing waste products from the blood in patients with kidney failure[1,2]. In addition, HP and other extracorporeal therapies, such as therapeutic plasma exchange, have been successfully utilized in patients with sepsis, where they have been associated with better clinical outcomes and an improved survival rate[3,4]. HP also showed beneficial effects in patients with severe influenza A, ameliorating the harmful effects of the cytokine storm[5]. Therefore, based on the crucial role of the dysregulated inflammatory response in the pathogenesis and clinical course of coronavirus disease 2019 (COVID-19)[6,7], along with the lack of specific therapeutic options, the use of extracorporeal therapies was utilized in patients with severe COVID-19, and several case reports or small case series[8,9] reported from around the world have been reported in the literature with variable results. Here, we retrospectively analyzed the clinical outcomes of a series of patients with severe COVID-19 treated with HP in our institution. We also reviewed the literature of articles published up to May 31, 2023, utilizing defined search criteria for publications reporting the use of HP in critically ill COVID-19 patients.

MATERIALS AND METHODS

This case series included a total of 18 adult patients in critical condition associated with COVID-19 who were consecutively admitted to the Baptist Hospital of Nicaragua (BHN) between August 2021 and November 2021. Epidemiological, clinical, and laboratory data were sourced and extracted from the hospital information system. Informed consent was obtained from all patients or their next of kin to publish their cases. Institutional Review Board approval was not required for this observational, retrospective study of routinely transmitted patient information, and written informed consent was waived owing to the rapid emergence of this infectious disease. The criteria for HP therapy are indicated in Supplementary Table 1. Non-eligible patients for HP were those with more than 15 days since starting COVID-19, patients with thrombocytopenia (< 60000), or those with an obvious bleeding tendency and active bleeding.

Laboratory data

Blood cell count and differential were measured at the clinical laboratory of BHN with an ABX Pentra XL 80 automated cell counter system (Horiba Ltd., Kyoto, Japan). Additional blood examinations included liver function tests, ferritin, magnesium, C-reactive protein (CRP), lactate dehydrogenase (LDH), D-dimer, and procalcitonin.

HP procedure

HP was carried out using the HA380 disposable HP cartridge (Jafron Biomedical Co., Ltd., Beijing, China), which is filled with neutral macroporous resin, mainly adsorbing molecules from 10 kDa to 60 kDa. In addition to supportive care and antiviral therapy, the standard treatment protocol for patients with acute respiratory failure included the use of noninvasive ventilation (helmet) as the first choice, and when these interventions were insufficient and patients were awake and cooperative, prone positioning with noninvasive ventilation/HFNC was used. However, when clinical conditions worsened, patients were intubated, and mechanical ventilation with a lung protective strategy (tidal volume of 4-6 mL/predicted body weight and adjustment of it based on driving pressure of 14 cm H₂O and a plateau pressure of 30 cm H₂O) was performed.

Literature search and review

A literature review was performed using standardized search terms, MeSH (medical subject headings) in MEDLINE and PubMed, and Emtree in Embase, organized in a hierarchal structure. We included in the analysis only peer-reviewed original articles that were published in the English language. The search terms were "hemoperfusion COVID-19", "hemoperfusion coronavirus", "hemoperfusion SARS-COV-2", and "hemoperfusion in COVID". Articles reporting data from more than 10 patients in which all the HP strategies adopted to reduce cytokines and pro-inflammatory molecules in the context of COVID-19 were included in the review, including artificial liver support, polymyxin B HP, CytoSorb (Cytosorbents Corporation, Princeton, NJ, United States), oXiris (Baxter, Deerfield, IL United States), plasmapheresis, and HP 440/380/280/230. Case reports or studies including less than 10 patients and articles reporting the use of other blood purification strategies not aimed at reducing cytokine levels, such as hemodialysis and continuous renal replacement therapy, were excluded from the review. We initially searched relevant publications between November 2022 and December 2022. Considering the rapid increase in scientific literature on the COVID-19 pandemic, we performed a second search of the literature between March 2023 and May 31, 2023, and a third updated search for articles published up to July 2024. Following these criteria, our search yielded 1357 articles, among which 998 were rapidly excluded due to their lack of relevance. The 359 articles that passed this first screening were retrieved and further evaluated for relevance and eligibility, and 339 of them were excluded because some of them were case reports, some studies were review articles, or they reported the use of HP in other medical conditions. We finally identified 20 articles that fulfilled our eligibility criteria and were therefore subjected to a full review. The outline of the literature search is shown in Figure 1.

Figure 1 Outline of the study search, screening strategy, and review of the literature.

Statistical analysis

Categorical variables were described as the total number and percentages, and continuous variables were described as either the means with SD or the median interquartile range. Data analysis was performed using Stata version 16.1 (Stata Corp., College Station, TX, United States), and statistical significance was set at a P value of ≤ 0.05.

RESULTS

Case series

A total of 18 patients with severe COVID-19 underwent HP, which was performed after the diagnosis of cytokine storm within the first 48 hours of hospital admission. Most patients (77.7%) in this series were men and relatively young (mean age 53 ± 15.80), and the most common comorbidity was hypertension (33.3%) followed by type 2 diabetes mellitus (27.7%). An incremental dosage of anticoagulation was used both to maintain circuit patency and to control thrombophilia. Unfractionated heparin was started at 10 IU/kg/h, but in some patients, a higher dosage of up to 20 IU/kg/h was required to ensure circuit patency. The demographic baseline characteristics of patients and clinical parameters before and after HP are shown in Table 1.

Table 1 Key demographic characteristics of the patients included in this case series.

Variable	Data
Sex	Male: 77 (n = 14)
	Female: 23 (n = 4)
Body mass index, mean ± SD	30.6 ± 7.85
Age, mean ± SD	53 ± 15.80
Comorbidities
Hypertension	33.33 (n = 6)
Heart disease	0
Diabetes mellitus	27.77 (n = 5)
CKD	22.22 (n = 4)
Asthma	0
Smoking	5.55 (n = 1)
Days from onset to severe pneumonia diagnosis, mean ± SD	8.94 ± 3.90
APACHE II score, median (range)	14.5 (29-6)
Initial ARDS respiratory support
Low flow oxygen cannula	5.55 (n = 1)
Oxygen mask with bag	44..44 (n = 8)
Invasive mechanical ventilator	50 (n = 9)

APACHE II: Acute Physiology and Chronic Health Evaluation II; ARDS: Acute respiratory distress syndrome; CKD: Chronic kidney disease.

The median (IRQ) length of the inpatient stay was 8 ± 4 days. Among the patients included in this study, eight were intubated. Nine patients received four or more sessions of HP, three patients received three sessions, and six patients received one or two sessions. Overall, improvements in respiratory cardiac parameters were evident following the application of HP, with oxygen saturation improving in 10 patients and reaching 95% in nine of them. Similarly, CRP levels and total leukocyte counts decreased in the majority of patients; however, ferritin levels slightly increased after receiving HP (Table 2 and Supplementary Table 2). Finally, 10 patients survived, and eight died. We believe that in patients who underwent HP at the early stages of severe COVID-19, clinical deterioration was prevented, which in turn may have obviated the need for intubation. Finally, 10 patients (55.5%) survived and eight (44.5%) died. Of note, among the expired patients, six had acute kidney injury (AKIN) stage 3 and underwent HP with continuous renal replacement therapy, and the other two patients did not develop AKIN during their intensive care unit (ICU) stay.

Table 2 Vital signs, lung ventilation parameters, and key laboratory markers before and after hemoperfusion.

Variable	Before HP	After HP	Difference	P value
SBP, mmHg	102.77 ± 25.48	119.66 ± 30.93	+ 16.9	0.01
DBP, mmHg	57.11 ± 17.94	71.77 ± 16.16	+ 14.6	0.11
RR, breath/min	23 ± 3.89	21.94 ± 3.07	- 1.06	0.19
HR, pulse/min	92.55 ± 17.15	87.05 ± 24.73	- 5.5	0.14
Temperature, °C	36.26 ± 0.53	36.32 ± 0.82	- 0.06	0.94
PEEP	12.94 ± 1.77	12.22 ± 1.20	- 0.72	0.12
SpO2, %	89.11 ± 7.87	91.94 ± 5.16	+ 2.83	0.04
CRP, mg/dL	158.56 ± 73.98	85.68 ± 98.02	- 72.88	0.01
Ferritin, ng/mL	2199.44 ± 2577.07	2294.46 ± 2548.26	+ 95	0.09
Leukocytes/per mm3	14.85 ± 4.92	9.81 ± 3.66	- 5.04	0.05
Lymphocytes/per mm3	9.99 ± 5.61	8.01 ± 5.75	- 1.98	0.07
Neutrophils/per mm3	84.94 ± 5.71	86.85 ± 7.07	+ 1.91	0.13

CRP: C-reactive protein; DBP: Diastolic blood pressure; HP: Hemoperfusion; HR: Heart rate; PEEP: Positive end-expiratory pressure; RR: Respiratory rate; SBP: Systolic blood pressure.

Literature review

We identified 20 publications (873 total patients) that met the criteria for inclusion in this review (Figure 1). Among them, seven articles described retrospective cohorts (n = 320 patients), three were prospective studies, and 10 were case series. Most reported studies utilized non-selective HP with the help of the CytoSorb^® cartridge, which has been described as an extracorporeal therapy with anti-inflammatory properties[1]. Regarding the HP strategy utilized, intermittent therapy was used in 11 studies (409 patients in total)[10-21], while in eight studies (total patients = 324)[22-28] HP was carried out using continuous therapy.

Among the selected studies, seven were conducted in Iran (= 342 patients in total)[10,12-14,22,28,29], three studies were from China[17,23,27], two from Italy[15,25] and Thailand[11,21], and the remaining seven were one each from Turkey[30], Saudi Arabia[24], Japan[16], North Macedonia[26], Germany[18], Colombia[19], and the United States of America[20]. The majority of included studies had small sample sizes, with an average of 37 patients participating. Table 3 summarizes the main findings of the studies included in our review of the literature.

Table 3 Review of the literature.

Ref.	Number of subjects, age, M/F ratio	Study design/HP strategy	Effects on inflammatory markers	Effects on respiratory markers	Main clinical outcomes
Barriga-Moreno et al[19]	n = 116, Controls = 84; HP = 32; 57 year (range: 47-71 year); Males: 65%	Matched control retrospective Jafron© HA330 cartridge	Serum creatinine (1.4 mg/dL vs 0.5 mg/dL); Serum ferritin (2868 vs 1675)	NR	Mortality rate: 61% in controls (receiving standard care) vs 31% in HP group
Chiewroongroj et al[21]	n = 272	Matched control retrospective	HP patients had less DIC (13% vs 33%; P = 0.046) and less sepsis (38% vs 64%; P = 0.02)	HP patients had reduced mechanical ventilation duration (15 days vs 35 days; P < 0.001) and less pulmonary complications (20% vs 42%; P = 0.04)	Mortality rates did not differ between the groups (33% vs 38%, P = 0.83)
	Controls = 227; HP = 45; 57 year (range: 47-71 year); Males: 65%	Jafron© HA330 cartridge			HP group had a significantly shorter ICU stay (22 days vs 32 days; P = 0.017)
Hayanga et al[20]	n = 100 (63% male); 44 ± 11 year	Multicenter combination of VV-ECMO and CytoSorb hemoadsorption. Two post hoc groups: Early HP: ≤ 87 hours and late HP > 87 hours		Early HP had shorter median duration of mechanical ventilation (7 days)[2-26] than late HP (17 days)[7-37], P = 0.02	Survival rates were 86% at 30 days and 74% at 90 days. Earlier HP was associated with shorter need for organ support and ICU stay
Uysal et al[30]	55 (34, 61.8%) Mean age 585 ± 12.5 year. Mean (SD), year = 59.6. Male sex, n (%) = 82 (64.1)	Case series. Resin-directed Jafron© (HA330) hemadsorption cartridges	Fibrinogen, LDH, CRP, and platelets decreased after HP	NR	In total, 9 patients (16.4%) survived. Ferritin levels correlated with survival status
Alavi Darazam et al[22]	128 (55 HP vs 73 controls). Mean age, (SD), year = 59.6. Male sex, n (%) = 82 (64.1)	Matched control retrospective study. Jafron© (HA330) and CytoSorb® 300 cartridges	NR	Median SpO2 statistically higher in HP group than in the controls and median PaCO2 was lower	Mortality rate in the HP group significantly lower than controls (67.3% vs 89%; P = 0.002). The median length of ICU stay was 8 days in HP group vs 12 days in controls (P < 0.001)
Mikaeili et al[10]	68 (35 HP vs 33 controls)	Prospective study	NR	Dyspnea decreased significantly in HP group	Significant mortality rate reduction in HP group compared with controls (37.1% vs 63.6%; P = 0.02)
	Mean age: Control 5748 ± 15.63, 18/15; HP group 56.62 ± 15.60, 23/12; sex (M/F): HP group: 32/12; Control: 18/15	HA330 D HP for 4 hours, in 3 consecutive days. Early started		Saturation of oxygen (SpO2) and P/F ratio significantly increased after HP unlike the control group	After 2 weeks, imagen lung opacities improved in 55% of the patients in HP vs 15.6% in the controls
Surasit et al[11]	29 (15 HP vs 14 controls). HP group: 12/15, 54.5 ± 14.4; Control group: 7/7, 64.3 ± 10.2	Prospective study. HA-330 3 HP sessions. Early started	Except for Hb levels (higher in HP group), blood and inflammatory parameters (CRP, ferritin, LDH, CBC) were not statistically different	NR	Significantly lower 28-day mortality rate in HP group compared to control group. Improvement of CXR RALE score and decreased SOFA score in HP group compared to control group
Darban et al[29]	40 patients, 57.5 ± 15.9; year 24/16	Retrospective cross-sectional study to determine the complications of HP patients with COVID-19 hospitalized in ICU	NR	NR	Arrhythmia, bleeding, thrombocytopenia, and coagulation disorders were the most common short-term complications mostly occurring on the second and third days after HP. Mortality occurred in 20 (50%) patients
Peng et al[23]	10 patients, median age 677 year (range = 50-85). Sex (male/female): 8/2	Case series Cytosorb HP. Median: 3 HPs (range = 1-6)	The level of IL-6 significantly decreased after HP. Lactate levels improved after HP	Significant improvement was found in PaO2/FiO2 [118 (81-220) mmHg vs 163 (41-340) mmHg, P = 0.04]	Albumin mildly decreased after HP. No significant changes were found in WBC
Abbasi et al[12]	37 patients, 55 ± 14.1; year 12/25	Retrospective cross-sectional study. Patients divided into three groups: (1) HP without (MV); (2) HP before MV; and (3) HP after MV	CRP and ferritin significantly improved after HP in the three groups	HR, RR, and PaO2/FIO2, significantly improved after HP in all groups	No statistically significant difference between the three groups in terms of length of hospital stay and ICU stay
Soleimani et al[13]	48 patients in total. HP cases: 24; Controls: 24	Retrospective observational study. HP: HA330 and HA280 filters for four hours	CRP levels decrease in HP patients compared to controls	RR and HR decreased after HP. SpO2 Levels significantly increased after HP. in the and a significant (P = 0.009)	HP improved respiratory distress in patients with severe COVID-19 but has no effect on mortality
Alharthy et al[24]	50 patients, 50 ± 9; Sex: Male (n, %): 39 (78%)	Case series	Nonsurvivors had higher levels of inflammatory biomarkers (CRP, ferritin, and IL-6), and more unresolved shock, ARDS	Pulmonary emboli more common in non survivors	Posttherapy values of IL-6 predicted in-hospital mortality for critically ill COVID-19 patients. No side effects of therapy were recorded
		All received CRRT and CytoSorb. Compared outcomes and biomarkers between survivors (39) and non survivors (11)
Villa et al[25]	37 patients, 31/6; 59 ± 9 year	Prospective non randomized. Blood purification with AN69ST (oXiris) hemofilter	Levels of IL-6 lowered in the first 24 hours of HP. IL-6 levels correlated with organ function	NR	Early HP treatment yielded the best outcomes. A slight decrease in observed vs predicted mortality rates observed. No complications reported
Ugurov et al[26]	15 patients, 13/2; 60 ± 13 year	Case series; Blood purification using the AN69ST (oXiris) hemofilter	CRP (mg/L) pre 109 (73) Decreasing levels of IL-6, IL-8 and TNF-α	Blood purification was associated with decreasing levels of IL-6, IL-8, TNF-α, and CRP	Median intensive care unit length of stay was 9.3 days. 2 out of 15 patients died
Dai et al[27]	Treatment group: 40/10, 60 ± 13 year; control group: 35/16, 60 ± 15 year	Case-control multicenter, prospective; Plasma exchange + HP (50 treated + 51 controls)	IL-6 Level decreased in the treatment group and increased in the control group	No statistical differences in the duration of invasive assisted ventilation between the two groups	The 28-day mortality rates were 16% (8/50) in the treatment group and 50.98% (26/51) in the control group
Hashemian et al[14]	14 patients, 9/6; 58 ± 12 year	Case-series Plasmapheresis Cytosorb	Inflammatory mediators (CRP, TNF-, IL-6, and Ferritin) were significantly reduced after plasmapheresis within a week	Plasmapheresis was associated with significant and rapid improvements in oxygenation status	Nine out of fifteen patients on NIPPV survived while the six patients undergoing IMV died
De Rosa et al[15]	12 patients with COVID-19 + septic shock 9/3; 60 ± 10 year	Case series Polymyxin-B HP (PMX-HP)	SOFA score progressively improved over the next 120 hours following PMX-HP and it was	NR	COVID-19 patients with endotoxic shock, PMX-HP was associated with organ function recovery, hemodynamic improvement, and contemporary EAA level reduction. No PMX-HP-related complications were observed
Asgharpour et al[28]	11 patients, 57 ± 18 year	Cases series; Tree sessions of extracorporeal resin-directed hemoadsorption	Serum IL-6 and CRP improved after intervention	Mean SpO2 increased after the three HP sessions	Six out of ten patients improved after the intervention
Katagiri et al[16]	15 patients, 66 (47) year	Case series Toramixyn	Except for one non-survivor, IL-8, IL-10, and IL-17 Levels remained almost unchanged or trended downward	NR	25% (NR)
Guo et al[17]	17 patients, 62 ± 14 year	Case series	32 cytokines (including IL-6 and TNF-α) out of 34 dosed were significantly decreased after each ALS course	NR	100%

ALS: Artificial-liver blood-purification system; CBC: Complete blood cell count; CRP: C-reactive protein; CRRT: Continuous renal replacement therapy; CXR RALE score: Severity score of lung infiltration on the chest X-ray; DIC: Disseminated intravascular coagulopathy; NR: Not reported; F: Female; HA: Hemadsorption; Hb: Hemoglobin; HP: Hemoperfusion; HR: Heart rate; ICU: Intensive care unit; IMV: Mechanical ventilation; LDH: Lactate dehydrogenase; M: Male; NIPPV: Non-invasive positive-pressure ventilation; RR: Respiratory rate; SOFA: Sequential organ failure assessment score; VV ECMO: Veno-venous extracorporeal membrane oxygenation; WBC: Whole blood cell counts.

In the largest study conducted so far, authors retrospectively analyzed data from 128 COVID-19 patients in Iran, with 55 patients receiving HP (using either Jafron© HA330 or CytoSorb^® 300 HP cartridges) and 73 patients categorized as a matched control group. There was a significantly lower mortality rate in the HP group compared to the matched control group (67.3% vs 89%; P = 0.002), and patients receiving HP also showed a higher SpO₂ than those in the matched group[22].

Four publications reporting prospective studies of HP in COVID-19 were found. The first one was an open-label study conducted in China between January and May 2020, and it reported that the mortality rate at day 28 was significantly lower in the treatment group (plasma exchange with HP) than in the control group (8/50 vs 26/51; P < 0.001). Also, whereas serum IL-6 levels increased in the control group, IL-6 levels significantly decreased in the treatment group[27]. Another prospective study was conducted in Italy and examined the effects of HP on 37 patients with severe COVID-19 hospitalized in intensive care units between February and April 2020. In addition to a reduction in serum IL-6 Levels, there were improvements in markers of systemic inflammation and multiorgan dysfunction, and a reduction in the expected ICU mortality rate, which were attributed to HP therapy, although no control group was included in the study[25]. Another prospective study was conducted between April and June 2020 in Iran and included 68 patients with severe critical symptoms of COVID-19 receiving either single standard therapy (33 controls) or a combination of standard treatment for COVID-19 combined with HP (35 patients) for 4 hours, in 3 consecutive days. Patients in the HP group had a significantly lower mortality rate compared with controls (37.1% vs 63.6%; P = 0.02)[10]. A small randomized controlled study investigated the effects of early initiation of HP via Cytosorb^® therapy in patients with severe COVID-19 pneumonia requiring venous extracorporeal membrane oxygenation (ECMO). Among the 34 eligible patients, 17 received HP Cytosorb^® therapy, and 17 did not. Strikingly, after 72 hours, IL-6 concentrations were higher in the HP group than in controls and had a worse survival rate, leading to the conclusion that cytokine adsorption therapy should not be used during the first days of ECMO support in COVID-19[18]. Except for the above-mentioned trial where HP failed to remove serum IL-6 and was associated with adverse clinical outcomes among COVID-19 patients undergoing ECMO[18], most studies reported improvements in inflammatory markers such as IL-6, fibrinogen, CRP, LDH, and procalcitonin. Also, nine studies reported an improvement in respiratory parameters, while one study did not observe any beneficial effects of HP on the clinical outcomes of COVID-19. Remarkably, a total of 13 studies (615 patients in total)[10,11,14,15,19-23,25-28] reported an improvement in survival among patients undergoing HP in comparison with those who did not receive HP. In summary, despite the large heterogeneity of patient populations and the relatively small sample sizes, it can be concluded that, in general, HP was safe in patients with severe COVID-19 and was associated with improvements in clinical outcomes.

DISCUSSION

In this article, we have described the clinical outcomes of a series of patients with life-threatening COVID-19 who were treated with intermittent HP. Our results showed that HP was associated with improvements in various indicators of respiratory and cardiovascular function, including oxygen saturation, blood pressure, and respiratory rate. There was also a consistent decrease in CRP levels and total leukocyte counts. Importantly, even though all patients in this series were in critical condition and eight of them required invasive ventilation, a total of 10 patients survived, which is highly remarkable considering that they were treated in a low-income country, which is a factor that has been shown to directly conditionate provision of healthcare[31].

The beneficial effects of HP observed in this series are consistent with those reported in the literature, as revealed by our systematic search of publications testing the use of HP in COVID-19 patients. Using specific search criteria, we performed a literature search for articles reporting the use of HP in patients with severe COVID-19 and found 20 publications (839 total patients), which included mainly retrospective cohorts and case series (17 articles). Notably, about half of the patients (seven studies, 342 patients in total)[10,12-14,22,28,29], were treated in Iran, including the largest study reported so far (128 patients). Even though most studies published so far are small and there is considerable heterogeneity in terms of study design, populations, and study endpoints, it is possible to conclude that the vast majority of these studies reported positive results. HP was safe in COVID-19 patients, with no apparent complications associated with the procedure. Moreover, HP resulted in measurable clinical benefits, including an improvement in respiratory and cardiovascular parameters, amelioration of inflammatory responses, and better survival rates compared to controls. However, it must be noted that there were no prospective randomized control studies.

Data from studies in patients with sepsis indicate that the beneficial effects of HP are associated with the removal of circulating cytokines triggered by the infection, which results in significant mitigation of the host's inflammatory responses[1,32]. Thus, considering that the emergence of a "cytokine storm" plays a central role in the severity of COVID-19[6,7], the removal of inflammatory cytokines with HP appears to be the central mechanism by which this extracorporeal therapy contributes to improving the clinical outcomes in patients with severe COVID-19.

Direct HP using polymyxin B-immobilized fiber column (PMX-DHP) has been associated with better survival rates among patients with sepsis as shown by previous systematic review and meta-analyses of 17 trials[33,34]. However, the beneficial effects of PMX-DHP appear to be dependent on the patient’s baseline Sequential Organ Failure Assessment (SOFA) score as suggested by a more recent nationwide study conducted in Japan that included 44177 patients with sepsis treated between April 2018 to March 2020, where PMX-DHP significantly improved the survival of the patients in the SOFA score categories of 7-9 and 10-12 but did not improve survival rate in those with SOFA score categories of 0-6, 13-15, and 16-24[35]. Among the 18 studies included in our review, two case series[15,16] reported the use of PMX-DHP in patients with COVID-19, although it is unclear if a concomitant sepsis was confirmed in those cases, and hence, it is unknown whether this strategy may result in any additive beneficial effects compared to HP without polymyxin in patients with COVID-19.

Thrombocytopenia and even the emergence of bleeding complications have been a recurrent concern for clinicians utilizing extracorporeal therapies[36]; however, in our study, no such complications were observed, which is consistent with what was reported in our review, where no significant changes in platelet counts or bleeding complications were reported[10,11,14,15,19-23,25-28].

Another important issue to consider is the potential influence of the filtration devices or cartridges and the courses of HP (number and length of sessions) on the efficacy of HP in severe COVID-19. Data from in vitro studies showed that both CytoSorb and oXiris were equally effective at removing cytokines and other inflammatory mediators, being drastically more effective at doing so than Toraymyxin (Toray Industries Inc., Tokyo, Japan). On the other hand, both oXiris and Toraymyxin were significantly more effective than CytoSorb in removing endotoxin. Indeed, the ability of CytoSorb to remove endotoxins was poor[37]. This is important considering that filtration devices can reduce the levels of cytokines, inflammatory markers, and perhaps viral particles in the blood of infected patients, which is the core mechanism for the beneficial effects of HP in severe COVID-19. Interestingly, in the study by Alavi Darazam et al[22], a lower mortality rate was observed among patients in the Jafron^® 330 cartridge group compared to those in which the CytoSorb^® cartridge was used, although the differences were not statistically significant. These results could be clarified using randomized controlled clinical trials with homogeneous populations and larger sample sizes.

It is important to notice that in our study, not all patients benefited from HD blood purification, and the reason for the failure of this therapy in some patients is currently unknown. It is plausible that genetic factors, variability in host immune phenotypes, and the viral load, among others, may affect the efficacy of the procedure. Further studies are needed to identify the factors that determine the efficacy of HP in severe COVID-19 but each patient may likely require a tailored approach.

HP is increasingly being utilized in the ICU, and with the advances in this technology and the development of new biocompatible devices, new indications for this extracorporeal technique are being developed[32]. While the use of polymyxin-B-bound HP is now a well-established therapy in patients with sepsis, where it can reduce 28-day mortality among patients with sepsis and septic shock[38]. New conditions where HP is currently utilized include bacterial endocarditis, liver failure, and severe viral infections[39]. However, the absence of consistent data derived from randomized controlled trials and a lack of consensus clinical guidelines and defined criteria for its utilization remain important issues regarding the utilization of HP in clinical practice.

There are limitations associated with this study, including the small number of patients treated with HP, and also due to the lack of resources, we were unable to directly measure cytokines before and after the HP session, so we could not directly assess if the beneficial effects of HP, if any, were mediated via a reduction in the cytokine levels. Finally, although the number of patients enrolled in this study was relatively small, the results are encouraging, considering that HP is a technology that could be affordable even in low-income countries. Based on the broad experience of the use of HP in patients with sepsis and the encouraging data from a few studies reporting a beneficial effect of this extracorporeal therapy in patients with other viral infections, its safety and clinical utility could be tested in the setting of other severe viral infections or critically ill patients with acute inflammatory syndromes associated with other pathogen infections. However, currently, most experts recommend the application of HP on a personalized basis. Therefore, further studies are needed, particularly using randomized controlled trials and recruiting a larger number of patients to verify the results presented here. In addition, some aspects and parameters of HP in the context of severe viral infections, including COVID-19, need to be refined. For example, the optimal time for its implementation, the number of sessions required, and its combination with any additional supporting therapies must be properly determined. In summary, the available evidence suggests that HP is safe when utilized as an early and supportive therapy for critically ill patients with COVID-19 and may offer beneficial effects.

CONCLUSION

This case series and the available evidence suggest that HP is safe when utilized as an early and supportive therapy for critically ill patients with COVID-19 and may offer beneficial effects.