Secondary Logo

Journal Logo

Management of COVID-19 Patients

Treatment of Cytokine Storm in COVID-19 Patients With Immunomodulatory Therapy

Yessayan, Lenar*; Szamosfalvi, Balazs*; Napolitano, Lena; Singer, Benjamin; Kurabayashi, Katsuo§,¶; Song, Yujing§; Westover, Angela*; Humes, H. David*

Author Information
doi: 10.1097/MAT.0000000000001239
  • Free



COVID-19 develops into a multilobar viral pneumonitis and often progresses to respiratory failure and acute respiratory distress syndrome (ARDS) requiring mechanical ventilation. ARDS is the leading cause of death in COVID-19 patients.1 Accumulating evidence suggests that this progression arises from “cytokine storm” defined as excessive inflammation and uncontrolled release of proinflammatory cytokines leading to endothelial dysfunction and consequent organ failure.2,3 The excessive release of cytokines may be due to the viral illness itself, ventilator-associated lung injury and to extracorporeal membrane oxygenation (ECMO). The clinical approach to this excessive inflammatory process has been limited to lowering cytokine activity or levels with specific cytokine inhibitors or extracorporeal hemofiltration or sorbent-based cartridges with limited success.4 We present our experience with an immunomodulatory extracorporeal device (an investigational device not yet approved by the FDA) with a fundamentally different mechanism of action of regulating cytokine-producing white blood cells, called the selective cytopheretic device (SCD), to treat two COVID-19 patients with cytokine storm and severe ARDS requiring ECMO. The SCD has been tested in multiple large animal models of inflammation, including sepsis and ARDS, and has been evaluated in multiple FDA-approved clinical trials with an excellent safety profile and strong efficacy signals to enhance organ recovery and reduce mortality in intensive care unit (ICU) patients with multiorgan failure.5,6


The SCD is an extracorporeal membrane cartridge which continuously processes circulating neutrophils and monocytes to a less proinflammatory phenotype, thereby tempering the cytokine storm and subsequent tissue damage. The SCD is integrated into a continuous renal replacement therapy (CRRT) blood circuit to provide this immunomodulatory effect (Figure 1). The SCD sustains the same blood flow settings of the CRRT circuit. To attain this effect, the SCD requires a rigorously controlled low ionized calcium (iCa) concentration between 0.25 and 0.4 mM in the extracorporeal blood circuit achieved with regional citrate anticoagulation (RCA) and dialysate/replacement fluid with no calcium.7 In the low-iCa environment, the most activated neutrophils and monocytes in blood perfusing through the SCD bind to its membranes. This binding event and continuous exposure to low iCa promotes a change in circulating neutrophils and monocytes to a less proinflammatory state.8 As required in RCA protocols, replacement calcium is infused into the blood circuit exiting the SCD so that blood returning to the patient maintains a normal calcium level in the patient. A CRRT protocol with weight-based dialysate/replacement fluid rates and fixed citrate-to-blood flow ratio, as opposed to a titration approach, with personalized dosing approach for calcium supplementation based on precalculated effluent calcium losses is utilized to (1) achieve adequate circuit ionized calcium (iCa) (<0.4 mm/L) for any hematocrit level and hence plasma flow and (2) keep systemic citrate levels <2 mmol/L irrespective of body citrate metabolism and eliminate the risk of clinically significant hypocalcemia.7

Figure 1.
Figure 1.:
Selective cytopheretic inhibitory device use in continuous renal replacement therapy system. SCD, Selective cytopheretic inhibitory device.

Emergency/expanded access use of the SCD was approved in two ICU patients with COVID-19-ARDS following FDA guidelines. Clinical eligibility criteria for emergency use included the presence of COVID-19 adult patient with acute kidney injury (AKI) requiring CRRT or ARDS, provision of medical care in an ICU, and the intention to deliver full supportive care for a minimum of 96 hours. Exclusion criteria were as follows: cardiovascular instability that precludes initiation of renal replacement therapy, irreversible brain damage based on available historical and clinical information, presence of any solid organ transplant with full-dose immunosuppressant regimen, patients with stem cell transplant in the previous 100 days or who have not engrafted, acute or chronic use of circulatory support device other than ECMO such as ventricular assist devices (LVADs, RVADs, BIVADs), metastatic malignancy which is actively being treated or may be treated with chemotherapy or radiation during the subsequent 3 month period after treatment, chronic immunosuppression defined as >20 mg prednisone daily alone or in combination of other immunosuppressant medications (i.e., cyclophosphamide, azathioprine, methotrexate, rituximab, mycophenolate, and cyclosporine), moribund or chronically debilitated state for whom full supportive care is not indicated, and any reason the investigator deems exclusionary. Because the SCD immunomodulates the inflammatory state, our decision was further guided by the presence of elevated IL-6 levels (>100 pg/ml). The results for the rapid cytokine assay were reported within 12 hours from measurement.

Michigan Institute for Clinical & Health Research (MICHR) provided the investigators with regulatory support. The use was approved by the University of Michigan Institutional Review Board, and each patient’s legal representative provided written informed consent. Both patients had severe ARDS requiring advanced mechanical ventilation and veno-venous ECMO support for refractory hypoxemia. A CRRT pump system and blood circuit containing a hemofilter and an SCD integrated in series was placed in parallel to the ECMO circuit. The SCD was used continuously and exchanged every 24 hours until ECMO therapy was concluded by the attending ICU physicians.

In the absence of data about the optimal duration for treatment with the SCD in patients with COVID-19, we opted to continue the SCD for the duration of ECMO (up to 20 days) given the latter procedure is considered to elicit an inflammatory response. Before and during SCD treatment, the following laboratory measurements were obtained using a rapid cytokine panel which included interleukin-6, -10, and -1b (IL-6, IL-10, IL-1b) and tumor necrosis factor; the inflammatory markers lactate dehydrogenase (LDH), ferritin, C-reactive protein (CRP), and procalcitonin; and D-dimer a marker of hypercoagulable state. Arterial blood gases were measured by standard clinical laboratory procedures. Separate IRB approval informed consent was obtained for use of a research cytokine test for clinical decision making and monitoring, and the cytokine assay, though rapid, was only available 3 days per week. The trends of laboratory measurements, ventilator, and ECMO settings were monitored.


Patient 1 was a 26-year-old male with history of obesity (body mass index, 38.9 kg/m2) and obstructive sleep apnea. He presented to a referring hospital with a week history of productive cough, fever, and acutely worsening dyspnea. He was intubated shortly after arrival to the emergency room. He developed ARDS and was treated with hydroxychloroquine and azithromycin × 5 days for COVID-19. His ICU course was complicated by Klebsiella pneumonia confirmed by positive sputum culture and pneumomediastinum likely secondary to barotrauma. He developed severe hypoxemia and required proning. He was transferred to our hospital a week later for ECMO evaluation. He required inhaled nitric oxide during flight transfer and was cannulated the same day for refractory hypoxemia despite chemical paralysis while on pressure control ventilation with bilevel positive airway pressure (37/18 cm H2O). Over the next 3 days, he developed high fevers and hypotension requiring vasopressors, and his PaO2 remained marginal on ECMO, paralytics, and ventilator support. He did not qualify to receive IL-6 inhibitors because he was on mechanical ventilation for more than 24 hours per our institutional policy at the time. Rapid cytokine profile showed elevated IL-6 levels (231 pg/ml). The SCD device was started on the same day and after 3 days of ECMO. Within the first 52 hours after SCD initiation, his oxygenation improved with reduced O2 requirements, and Il-6 levels and all inflammatory markers were substantively reduced (see Table 1). At 36 hours, some improvement of the bilateral patchy infiltrates on chest X-ray was reported. His hospital course was complicated by bilateral pneumothoraces, which required bilateral chest tubes and therapy for ventilator-associated Klebsiella pneumonia. His inflammatory markers including procalcitonin, LDH, CRP, and D-dimer trended down during SCD therapy. Of importance, his IL-6 levels remained in the normal range and his IL-6/IL-10 ratios remained near or below 1.0 throughout the course of his treatment with the SCD. His oxygenation continued to gradually improve, while his ECMO flow and sweep were gradually weaned. He was taken off ECMO 20 days after initiation and 17 days after SCD treatment. He continued to improve until he was transitioned to room air. Nasopharyngeal PCR for SARS-CoV-2 was negative 21 days after stopping the SCD. Table 1 shows the trends of his select clinical and laboratory parameters.

Table 1. - Trends of Laboratory Parameters, Ventilator Settings, and Vasopressor Use for Patient 1
Variable Before Therapy 6 hr: Posttherapy 30 hr: Posttherapy 52 hr: Posttherapy
Date 21 April 22 April 22 April 23 April
Time (hrs: min) 23:00 6:04 23:59 23:25
PaO2 mm Hg 55 77 74 120
FiO2 (%) 100 70 60 60
PaO2/FiO2 55 110 123 200
PEEP (cm H2O) 12 12 12 12
Procalcitonin (ng/ml) 3.23 2.74 1.8 1.15
d-Dimer (mg/l) >35 11.95 12.26 9.49
LDH (IU/l) 942 806 753 628
Ferritin (ng/ml) 1,902 1,775 1,820 1,873
CRP (mg/dl) 36.1 28.9 21.6 12.2
IL-6 (pg/ml) 231 NA 5.65 3.32
IL-10 (pg/ml) 19.5 NA 8.263 6.68
IL-6/IL-10 11.8 NA 0.7 0.5
CXR Extensive bilateral mixed interstitial and alveolar opacities Interval decrease in diffuse patchy parenchymal opacities
Vasopressors Norepinephrine vasopressin No vasopressors
CRP, C-reactive protein; CXR, chest X-ray; LDH, lactate dehydrogenase; NA, not available; PEEP, positive end-expiratory pressure.

Patient 2 was a 52-year-old male with history of diabetes, hypertension, renal transplant with chronic allograft failure (chronic kidney disease stage V), and obesity (body mass index, 36 kg/m2) who presented to the emergency department with 1 week history of cough and 1 day history shortness of breath. He was diagnosed with COVID-19 pneumonia and AKI. He required intubation and mechanical ventilation, immunosuppression medications were reduced, and hemodialysis was initiated for progressive AKI. He did not qualify to receive IL-6 inhibitors because he had chronic kidney stage 5 per our institutional policy at the time. Over the subsequent 10 days, his respiratory failure progressed to severe ARDS and hypoxemia with PaO2/FiO2 ratio of 55 on advanced mechanical ventilation (assist control ventilation; tidal volume, 400 ml; positive end-expiratory pressure, 16) and paralytics. Due to hemodynamic instability from COVID-19 septic shock and persistent hypoxemia, he required initiation of CRRT and ECMO support and antibiotics were initiated due to a concern for bacterial superinfection. Rapid cytokine profile showed elevated IL-6 levels (598 pg/ml). The SCD was started in conjunction with CRRT. Within the first 50 hours, his oxygenation improved, and IL-6 levels and all his inflammatory markers were substantively reduced (see Table 2). He continued to improve his oxygenation and was converted to pressure support mode of ventilation on day 4 after the SCD initiation and underwent tracheostomy on day 7. His inflammatory markers including LDH, ferritin, procalcitonin, and CRP trended down during SCD therapy. His IL-6 levels remained suppressed 8 days after initiation (108.60 pg/ml), and more importantly, his IL-6/IL-10 ratios remained below 1.0 (0.7). His oxygenation continued to gradually improve, while his ECMO flow and sweep were gradually weaned. He was taken off ECMO and SCD after 16 days of therapy. Nasopharyngeal PCR for SARS-CoV-2 was positive 15 days after stopping the SCD. His oxygenation continued to improve, and he was eventually transitioned to room air. He was discharged home 21 days after discontinuing the ECMO and SCD treatment. Table 2 shows the trends of his select clinical and laboratory parameters.

Table 2. - Trends of Laboratory Parameters, Ventilator Settings, and Vasopressor Use for Patient 2
Variable Before Therapy 5 hr: Posttherapy 26 hr: Posttherapy 38 hr: Posttherapy 50 hr: Posttherapy
Date April 21–30 May 2 May 2 May 3 May 3
Time 02:22 23:55 11:40 23:59
PaO2 (mm Hg) 55 100 77 81 77
FiO2 (%) 95 50 44 40 40
PO2/FiO2 58 200 175 202 192
PEEP (cm H2O) 12–16 14 14 14 14
Procalcitonin (ng/ml) 12–17 1.58 5.67 5.85 4.85
d-Dimer (mg/L) 4.0–10.0 4.16 5.27 NA 7.63
LDH (IU/L) 499–828 497 332 366 304
Ferritin (ng/ml) 5,559–9,247 5,237 5,279 5,441 3,873
CRP (mg/dl) 28–33 29 25 NA 20
IL-6 (pg/ml) 598 NA NA NA 116
IL-10 (pg/ml) 33 NA NA NA 185
IL-6/IL-10 18 NA NA NA 0.62
CXR Diffuse bilateral parenchymal airspace abnormalities Interval improvement of pulmonary edema
Vasopressors Norepinephrine vasopressin Vasopressors discontinued
CRP, C-reactive protein; CXR, chest X-ray; LDH, lactate dehydrogenase; NA, not available; PEEP, positive end-expiratory pressure.


Since accumulating evidence suggests that patients with COVID-19 viral pneumonia progress to severe ARDS due to excessive lung inflammation and cytokine storm, we treated under emergency/expanded use of the SCD two COVID-19 ARDS patients requiring ECMO support. SCD therapy was chosen in these cases because of prior preclinical and clinical research studies suggesting potential benefit. In this regard, SCD therapy improved multiorgan dysfunction and survival time in porcine models of septic shock and ARDS.5,6 Multiple FDA-approved SCD clinical protocols have demonstrated an excellent safety profile and clinical efficacy in adult and pediatric ICU patients with multiorgan failure and AKI requiring CRRT.9,10

Treatment with the SCD in both COVID-19 severe ARDS patients resulted in significant reductions in all measured inflammatory markers including procalcitonin, D-dimer, LDH, Ferritin, CRP, IL-6, and IL-6/IL-10 ratio within hours of SCD initiation. The SCD effect to consistently lower the IL-6-to-IL-10 ratio is especially noteworthy. A lower ratio of these two cytokines has been a reliable marker to discern the severity of pneumonia, with ratios near 1 associated with less clinical disease.11 These improvements correlated in time with improvement in the patients’ PaO2 with a reduction in FiO2. By 24 hours, this improvement was evident in chest X-ray findings and discontinuation of vasopressors to treat hypotension. The inflammatory markers in general had gradual but continued reductions over the next several days. This treatment effect of the SCD to reduce levels of a full array of inflammatory biomarkers coincident with the noticeable improvement in clinical parameters suggests an important role that the “cytokine storm” plays in the progression of lung injury and deterioration of respiratory function resulting in poor outcomes in COVID-19 patients with ARDS.

Given the hypothesized immunomodulatory mechanism of SCD, patients were selected for high levels of IL-6. There seems to be a strong bias to believe that many critically ill COVID patients have cytokine storm, and this may not be true. Six additional COVID-19 patients with ARDS were screened (four receiving ECMO) with IL-6 levels less than 100 pg/ml averaging 29 ± 32 pg/ml (mean ± SD). We observed similar heterogeneity of cytokine levels in 20 additional critically ill COVID-19 patients with respiratory failure requiring mechanical ventilation. In 14 patients who did not receive tocilizumab, IL-6 was 167 ± 190 pg/ml. As expected, four patients whose IL-6/IL-6 receptor interaction was disrupted by tocilizumab displayed higher IL-6 levels of 2,917 ± 190 pg/ml.12 IL-10 levels were not different between these two groups (46 ± 44 and 50 ± 63 pg/ml, respectively). The effect of SCD on lowering IL-6 and raising IL-10 differed markedly from the effect of tocilizumab.

The immunomodulatory effect of the SCD in the excessive inflammatory state and cytokine storm differs from other current approaches to treat this dysregulated immunologic state. Sorbent-based blood-purifying technologies which adsorb cytokines have not shown selectivity and durability of lowering blood cytokine levels nor efficacy to improve clinical outcomes.4 A systematic review of the literature has also concluded that the use of high-volume hemofiltration to convectively remove serum cytokines does not confer any measureable clinical benefit in critically ill patients.13 In contrast, the membranes of the SCD in a low ionized calcium environment promoted with RCA selectively bind the most activated circulating neutrophils and monocytes. These binding events result in resetting the circulating neutrophils and monocytes to a less proinflammatory state.6,8 The SCD modulates the cytokine storm not by reducing the blood levels of the soluble mediators of inflammation but by diminishing the proinflammatory activity of the central cellular elements of the innate immunologic system, the neutrophils and monocytes.6,8 Several clinical studies have demonstrated improved clinical outcomes in ICU patients with SCD therapy.6,9,10 The precise mechanism of action of the SCD to immunomodulate neutrophils and monocytes in an excessive inflammatory state is being further evaluated in ongoing preclinical and clinical research studies.

The results observed with SCD therapy on these two critically ill COVID-19 patients with severe ARDS and septic shock is encouraging but requires further evaluation in additional COVID-19 patients. Accordingly, a multicenter clinical trial is underway with an FDA-approved Investigational Device Exemption to evaluate the potential of SCD therapy to effectively treat COVID-19 ICU patients.


1. Richardson S, Hirsch JS, Narasimhan M, et al.: Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA. 2020.323:2052–2059.
2. Guan W-j, Ni Z-y, Hu Y, et al.: Clinical characteristics of coronavirus disease 2019 in China. N Eng J Med. 2020.382:1708–1720.
3. Tisoncik JR, Korth MJ, Simmons CP, Farrar J, Martin TR, Katze MG: Into the eye of the cytokine storm. Microbiol Mol Biol Rev. 2012.76:16–32.
4. Honore PM, Hoste E, Molnár Z, et al.: Cytokine removal in human septic shock: Where are we and where are we going? Ann Intensive Care. 2019.9:56
5. Ding F, Song JH, Jung JY, et al.: A biomimetic membrane device that modulates the excessive inflammatory response to sepsis. PLoS One. 2011.6:e18584
6. Pino CJ, Westover AJ, Johnston KA, Buffington DA, Humes HD: Regenerative medicine and immunomodulatory therapy: insights from the kidney, heart, Brain, and Lung. Kidney Int Rep. 2018.3:771–783.
7. Szamosfalvi B, Frinak S, Yee J: Automated regional citrate anticoagulation: technological barriers and possible solutions. Blood Purif. 2010.29:204–209.
8. Szamosfalvi B, Westover A, Buffington D, Yevzlin A, Humes HD: Immunomodulatory device promotes a shift of circulating monocytes to a less inflammatory phenotype in chronic hemodialysis patients. ASAIO J. 2016.62:623–630.
9. Tumlin JA, Chawla L, Tolwani AJ, et al.: The effect of the selective cytopheretic device on acute kidney injury outcomes in the intensive care unit: a multicenter pilot study. Semin Dial. 2013.26:616–623.
10. Tumlin JA, Galphin CM, Tolwani AJ, et al.; SCD Investigator Group: A multi-center, randomized, controlled, pivotal study to assess the safety and efficacy of a selective cytopheretic device in patients with acute kidney injury. PLoS One. 2015.10:e0132482
11. de Brito RdCCM, Lucena-Silva N, Torres LC, Luna CF, Correia JdB, da Silva GAP: The balance between the serum levels of IL-6 and IL-10 cytokines discriminates mild and severe acute pneumonia. BMC Pulm Med. 2016.16:170
12. Nishimoto N, Terao K, Mima T, Nakahara H, Takagi N, Kakehi T: Mechanisms and pathologic significances in increase in serum interleukin-6 (IL-6) and soluble IL-6 receptor after administration of an anti-IL-6 receptor antibody, tocilizumab, in patients with rheumatoid arthritis and Castleman disease. Blood. 2008.112:3959–3964.
13. Lehner GF, Wiedermann CJ, Joannidis M: High-volume hemofiltration in critically ill patients: a systematic review and meta-analysis. Minerva Anestesiol. 2014.80:595–609.

ARDS; COVID-19; COVID treatment; immunomodulation; extracorporeal therapy

Copyright © 2020 by the ASAIO