Septic shock is known to have a high mortality rate and carries the highest risk of death in intensive care (1). Albeit the increasing number of extracorporeal membrane oxygenation (ECMO) treatments, ECMO is not considered a standard treatment for adult patients with septic shock (2, 3) as opposed to the pediatric and neonatal populations (3–6). Even though ECMO has gained increased acceptance for treatment of adult severe respiratory failure, the controversy remains concerning its usefulness in septic shock (2, 7). During the last years, several publications question this standpoint (8–11) and indications emerge that ECMO may be performed with mortality rates comparable to or better than conventional treatment (12). Bréchot et al (8) reported a 70% survival rate and low post-ECMO mortality compared with a reference cohort. However, others have reported worse outcome especially in combination with extracorporeal cardiopulmonary resuscitation (ECPR) (10, 11).
This retrospective study was undertaken to describe a septic shock population with a combination of acute respiratory distress syndrome and cytotoxic cardiac failure or distributive shock submitted to ECMO treatment at a high-volume ECMO center.
MATERIALS AND METHODS
Our center is a high-volume ECMO unit treating approximately 90 respiratory and cardiac ECMO patients per year. The unit is an Extracorporeal Life Support Organization (Ann Arbor, MI) member since 1992 and Center of Excellence since 2007. Population served is 10 million (Sweden, all age groups). We also support Finland and Ireland concerning neonatal and pediatric respiratory ECMO. Hence, an ECMO transport organization is integrated with the department. Approximately 80% of the patients are referred from other hospitals, assessed and cannulated at the referring hospital, and then transferred to our unit after commencement of ECMO (13, 14).
Indications for ECMO may be either respiratory and/or circulatory. Septic shock should be considered for referral to an ECMO center experienced in treating septic shock (8, 15). In our department, respiratory support (i.e., venovenous ECMO) is offered if the Pao2-to-Fio2 ratio is less than 60–80 mm Hg (Fio2 1.0), that is, the resource is used as rescue treatment when conventional respiratory critical care fails. Venoarterial ECMO for cardiocirculatory support is considered after adequate fluid resuscitation together with at least one of the following criteria: persistent lactatemia greater than 5 mmol/L, mixed venous saturation (Svo2) less than 55%, or cardiac index less than 2 L/min/m2 (> 1 hr), rapidly deteriorating ventricular function, refractory arrhythmia, Vasoactive Inotropic Score (VIS) greater than 50 (> 1 hr), greater than 45 (> 8 hr), or greater than 40 if myocarditis. VIS (16, 17) was calculated as ([(epinephrine + norepinephrine) μg/kg/min] × 100 + [(dobutamine + dopamine) μg/kg/min] + [milrinone μg/kg/min] × 15 + [vasopressin IU/kg/min] × 10,000). Echocardiography was performed in all patients before cannulation.
The ECMO pumps used were Centrimag (St. Jude Medical Europe, Zaventem, Belgium) or Cardiohelp (Maquet Cardiopulmonary AG, Rastatt, Germany). All cannulations were peripheral, performed by a dedicated ECMO surgeon. The drainage configuration applied was adapted to minimize differential hypoxemia in venoarterial ECMO (18, 19). Bio-Medicus 19–21F/18 cm, 23F/25 cm, 17–29F/50 cm (Medtronic International Trading Sàrl, Tolochenaz, Switzerland), or 25F/38 cm Maquet HLS (Maquet Cardiopylmonary AG, Rastatt, Germany) was used in both venovenous and venoarterial ECMO and implanted via the right jugular vein. The oxygenated blood was returned via a femoral vessel with a 15F/17 cm, 19–21F/18 cm Bio-Medicus arterial cannula. All venoarterial ECMO patients were catheterized with a 6F or 8F cannula for perfusion of the cannulated leg. Avalon Elite 31F (Maquet) was used for dual-lumen cannulation. The oxygenators were Hilite 7000LT (Medos Medizintechnik AG, Stolberg, Germany) or Quadrox (Maquet).
For anticoagulation, a bolus dose of unfractionated heparin was administered at cannulation followed by continuous infusion. Activated partial thrombin time was targeted between 1.5 and 2 times of normal. ECMO blood flow was maintained between 3.5 and 5.0 L/min to reach a preoxygenator oxygen saturation during venoarterial ECMO of greater than 65% (venovenous > 70%). Echocardiography was used to monitor cardiac function. Inopressors and/or inodilators were adjusted to maintain a mean arterial blood pressure (MAP) greater than 65 mm Hg.
All patients were treated with β-lactam antibiotics and vancomycin by continuous infusion. For monitoring of infection, cultures are performed twice weekly and when needed. Plasmapheresis is used on occasion as an adjunct with IV immunoglobulins. For ventilation, rest settings are applied where positive end-expiratory and driving pressures are reduced (20).
Retrospective data were obtained from our local database (PasIva; Otimo Data AB, Kalmar, Sweden) and from patient charts. All adult patients admitted to our unit between January 2012 and December 2017 with septic shock were eligible for inclusion. Inclusion criteria were as follows: greater than 18 years old; septic shock (International Classification of Diseases, 10th Edition [ICD-10]: R65.1, R57.2) at acceptance for ECMO; fulfillment of the “Sepsis-3” definition (21); presence of cardiocirculatory failure requiring a support equivalent to a VIS greater than 50 (internal guidelines accessible on www.ecmo.se) to reach a target MAP greater than 65 mm Hg; and signs of organ hypoperfusion at the time of ECMO initiation. Patients not treated at our unit during the whole ECMO run and patients with ongoing cardiopulmonary resuscitation (CPR) at the time of ECMO initiation were excluded. However, patients who had received CPR and achieved return of spontaneous circulation prior to ECMO initiation were included. Data were coded and cannot be traced back to any certain individual. The study was approved by the Stockholm Regional Ethical Board.
Data were collected from the time of the first phone-call until discharge from the ECMO department or death. The following data were recorded: age, sex, body mass index, comorbidities, time for onset of sepsis, isolated pathogen/s, and focus of infection. At decision for ECMO, arterial blood gas, circulation, and ventilation parameters including echocardiography were recorded. At initiation of ECMO, Simplified Acute Physiology Score (SAPS-3) and Sequential Organ Failure Assessment Score (SOFA) were calculated. Doses of vasoactive agents, ECMO mode, and CPR during the last 24 hours were noted. The number of failing organs was evaluated as defined by SAPS-3. At arrival at our ECMO unit, cardiac function was assessed by echocardiography and the following laboratory samples recorded: WBC count, p-C-reactive protein, and p-procalcitonin. Antibiotic coverage was reevaluated, and patients were treated with at least one antibiotic active against the identified or suspected (empirical) pathogen. All adverse events during treatment were noted.
Primary outcome variables were survival to discharge from ECMO unit, hospital survival, and survival at follow-up of at least 6 months. Secondary outcomes were days on ECMO and ECMO-associated complications.
Statistical and Data Analysis
Data are presented as median (interquartile range [IQR]) unless the mean ± sd is given. For comparisons, Mann-Whitney U test for nonparametric data was performed. For categorical data, Fisher exact test was used. The standardized mortality ratio was calculated from the observed hospital mortality rate and the estimated mortality rate obtained from the median SAPS-3 score. Subsequently, the number needed to treat could be calculated. p value less than 0.05 was considered significant.
One hundred forty-eight septic shock patients according to ICD-10 coding were identified. Thirty-seven were included in the study (Fig. 1). Patient characteristics are presented in Supplemental Digital Content 1 (http://links.lww.com/CCM/E601). Plasma procalcitonin was 86 μg/L (IQR, 49–236), p-C-reactive protein 249 mg/L (194–293), and WBC count 8.5 × 109/L (2.3–19.8). Positive blood cultures were obtained in 73% of the cases. In 54%, the primary site of infection was the lung (Table 1). Primary cultures showed Gram-negative (G–) infections in 12 patients and Gram-positive (G+) in 20 patients (Table 2). In one patient, both G+ and G– were found. The G– group had a higher requirement for epinephrine (p = 0.018) before ECMO, but the total amount of vasoactive support was similar. Otherwise, no differences were found (Table 2).
The median time from first contact to start of ECMO was 5 hours 45 minutes. Ten patients (37%) were commenced on venovenous and 27 on venoarterial ECMO. For further ECMO data, see Supplemental Digital Content 2 (http://links.lww.com/CCM/E602).
The Pao2-to-Fio2 ratio was 48.8 mm Hg (37.5–60 mm Hg) for venovenous and 60 mm Hg (50–74 mm Hg) for venoarterial (p < 0.03). Lactate was lower for the venovenous than for the venoarterial patients (3.65 mmol/L [3.3–7.0 mmol/L] vs 8.0 mmol/L [5.0–11 mmol/L]; p < 0.03). The ECMO flow was higher for venoarterial than for venovenous (4.6 L/min [4.3–5.1 L/min] vs 4.0 L/min [3.9–4.1 L/min]; p < 0.01), but the difference in effective ECMO blood flow was in reality higher because the recirculation fraction during venovenous-ECMO was not accounted for. There were no differences concerning MAP, heart rate, doses of norepinephrine and epinephrine, or VIS.
The subgroup of patients with left ventricular failure (LVF) at admission was supported with venoarterial rather than venovenous-ECMO (p < 0.05). The septic LVF patients had an ejection fraction of 25% (20–30) compared with 52.5% (40–60) for the non-LVF patients (p < 0.001). Hospital survival for the LVF patients was 90.0%, and long-term survival was 75.0%, in this group. The corresponding hospital survival for the non-LVF subjects was 64.7% (p = 0.044), and long-term survival was 47.1% (p = 0.081). In the whole sepsis cohort, survival from ECMO was 81.1%, hospital survival 78.4%, and at long-term follow-up (median, 46.1 mo), survival was 59.5% (Table 3). Five of the seven patients (71.4%) who experienced CPR before admission survived. Of the eight immunocompromised patients, one died during ECMO, the rest are still alive today (87.5%). Hospital survivors had a median SAPS-3 score of 85 (80.5–95) compared with 97 (87–100; p = 0.15) for the nonsurvivors. The deceased exhibited more organ failures than survivors, 6 (6–7) and 5 (4–6; p = 0.03), respectively. There was no difference in VIS and median SOFA score.
Commencement on venovenous-ECMO was associated with a higher risk for in-hospital death when compared with venoarterial (50% vs 11%; p = 0.011) and all-cause long-term mortality (70.0% vs 29.6%; p = 0.026), respectively. Long-term all-cause mortality was associated with more organ failures at admission (6 [5–6] vs 4.5 [4–6]; p = 0.036). However, there was no difference in the SAPS-3 score at admission, 85 (80–95) versus 87 (84–99), respectively, nor were there any differences concerning SOFA score, pH, lactate, Pao2/Fio2 doses of norepinephrine, epinephrine, or VIS.
Ventilator-associated pneumonia was reported in three of the eight registered complications (38%). Bleeding from cannulation site was seen in two patients, and two patients exhibited ischemia in the cannulated leg (venoarterial ECMO) which necessitated surgical care. All patients who expired during ECMO treatment were cases where ECMO treatment was withdrawn due to futility from intracranial events: major bleedings, ischemic events, or herniation (Supplemental Digital Content 2, http://links.lww.com/CCM/E602).
We investigated 37 patients with septic shock during 6 consecutive years. These subjects were identified from the ICD-10 codes, fulfillment of Sepsis-3 criteria, and the need for a significant amount of vasoactive support. The cohort included two subsets of patients: septic shock with LVF (septic cardiomyopathy) and patients with distributive shock. LVF showed a higher survival. The number of organ failures at admissions affected both in-hospital and long-term survival.
Why ECMO reduce mortality is generally related to the increase of oxygen delivery as well as offering circulatory support (22). Failing circulation in the septic shock patient is caused by cardiac depression but also vasoplegia and capillary leakage causing displacement of intravascular volume exacerbating hypovolemia. This considered, it suggests that venoarterial ECMO supports the failing heart but will have less or no direct impact on the other parts causing hypotension, that is, vasoplegia and capillary leakage. However, secondary beneficial effects from improved tissue oxygenation may play a role in stabilizing the circulation. Even though ECMO treatment per se does not restore microcirculation or cellular/mitochondrial oxygen uptake, it could increase the chances thereof by augmenting oxygen delivery and secondary improvement of tissue perfusion in general: milieu intérieur. Therefore, we postulate that in distributive septic shock, venoarterial ECMO supports the circulation by limiting the negative impact of generalized poor oxygenation.
Lactic acidosis may be caused by factors other than anaerobic metabolism. It has been pointed out that hyperlactatemia may be seen before there is an imbalance between oxygen delivery and oxygen demand (23, 24). In septic shock, hypotension may exist with no or mildly elevated lactate. In the current study, 12 patients had a p-lactate greater than or equal to 9 mmol/L, the cutoff considered for when hyperlactatemia is driven by an anaerobic metabolism (23, 25). Thus, in these patients, ECMO may have helped alleviate the tissue oxygen deficiency by increasing oxygen delivery.
Of the 10 patients treated with venovenous-ECMO, hence not receiving any cardiac support, six patients still survived the septic shock episode. This could be pertained to the fact that venovenous-ECMO supplies oxygen. Venovenous-ECMO helps resuscitate a deficient oxygenation capability by the lung, but it will not support perfusion directly. However, it may do so by increased oxygenation in the coronaries leading to improved cardiac function.
In the few studies on patients with septic shock and ECMO treatment, the survival rates range from 15% to 71%. Two studies included ECPR patients making them hard to interpret (10, 11). In the study by Bréchot et al (8) on 14 septic shock patients with cardiac failure, 12 (86%) came off ECMO and survival after 13 months was 71%. Their results were comparable to the LVF subgroup in the current study. Their conclusion that ECMO was useful in patients with septic shock, and myocardial dysfunction is confirmed by our study. Vogel et al (26) reported 75% survival in patients with septic cardiomyopathy.
A retrospective study concerning patients treated at our institution between 1995 and 2013 found an overall survival of 64% (27). The same study showed that patients who survived the first 90 days after discharge had a 5-year survival rate of 87%. In the same material, there was no reduction of cognitive function among patients with prolonged hypoxemia (28).
Of the 17 patients in this study with distributive shock, 70.6% survived to discharge from hospital. Especially, venoarterial ECMO could be considered, not only for patients with myocardial dysfunction but also for distributive septic shock patients with preserved myocardial function. However, despite these findings, we believe that a randomized controlled trial with randomization between venoarterial ECMO and conventional intensive care should be performed before drawing any definite conclusions. Furthermore, our cohort of patients contained a mix of both venovenous and venoarterial ECMO patients. In retrospect, these patients may confuse the numbers especially because the majority of them later were converted to venoarterial ECMO. However, the current analysis was based on the initial ECMO mode which in its turn is based on the attending physicians’ experiences and context when selecting ECMO mode. Still conversion to venoarterial from venovenous was based on clinical assessment that venovenous-ECMO was not sufficient support for survival. Therefore, these patients were converted to venoarterial ECMO. These difficult decisions require both experience and thorough knowledge in treating this patient group. Another limitation of the study was the single high-volume center retrospective design making it difficult to draw general conclusions applicable to other ECMO practitioners. However, septic shock patients should only be treated at high-volume centers with experience and capability to perform all modes of ECMO (venovenous, venoarterial, and veno-venoarterial) according to patient’s need. These patients are complex to treat not only from an experience point of view but also concerning resource demand. Still no randomized control trials have been published on sepsis and ECMO. Although multicentric randomized studies are needed, these are difficult to conduct due to low numbers and treatment heterogeneity between centers (29). Larger well-designed prospective single-center studies would also add to our community’s knowledge.
The results from this mixed population of both distributive septic shock and septic shock with cytotoxic cardiac failure with high risk for mortality indicate that peripheral cannulation ECMO support may be beneficial for both hospital and long-term survival. It cannot be emphasized enough that experience from septic shock patients is fundamental and that ECMO should be initiated by and/or at least continued at a high-volume ECMO center with experience in both venovenous and venoarterial ECMO for a dynamic approach to patient need. Further randomized controlled trials are warranted to investigate the effect of venoarterial ECMO in septic circulatory failure.
1. Gaieski DF, Edwards JM, Kallan MJ, et al. Benchmarking the incidence and mortality of severe sepsis in the United States. Crit Care Med 2013; 41:1167–1174
2. Chockalingam A, Mehra A, Dorairajan S, et al. Acute left ventricular dysfunction in the critically ill. Chest 2010; 138:198–207
3. Dellinger RP, Levy MM, Rhodes A, et al.; Surviving Sepsis Campaign Guidelines Committee including the Pediatric Subgroup: Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock
: 2012. Crit Care Med 2013; 41:580–637
4. Bartlett RH, Roloff DW, Custer JR, et al. Extracorporeal life support: The University of Michigan experience. JAMA 2000; 283:904–908
5. Brierley J, Carcillo JA, Choong K, et al. Clinical practice parameters for hemodynamic support of pediatric and neonatal septic shock
: 2007 update from the American College of Critical Care Medicine. Crit Care Med 2009; 37:666–688
6. Carcillo JA, Fields AI; American College of Critical Care Medicine Task Force Committee Members: Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock
. Crit Care Med 2002; 30:1365–1378
7. Bartlett RH. Extracorporeal support for septic shock
. Pediatr Crit Care Med 2007; 8:498–499
8. Bréchot N, Luyt CE, Schmidt M, et al. Venoarterial extracorporeal membrane oxygenation
support for refractory cardiovascular dysfunction during severe bacterial septic shock
. Crit Care Med 2013; 41:1616–1626
9. Firstenberg MS, Abel E, Blais D, et al. The use of extracorporeal membrane oxygenation
in severe necrotizing soft tissue infections complicated by septic shock
. Am Surg 2010; 76:1287–1289
10. Huang CT, Tsai YJ, Tsai PR, et al. Extracorporeal membrane oxygenation
resuscitation in adult patients with refractory septic shock
. J Thorac Cardiovasc Surg 2013; 146:1041–1046
11. Park TK, Yang JH, Jeon K, et al. Extracorporeal membrane oxygenation
for refractory septic shock
in adults. Eur J Cardiothorac Surg 2015; 47:e68–e74
12. Combes A, Hajage D, Capellier G, et al. Extracorporeal membrane oxygenation
for severe acute respiratory distress syndrome. N Engl J Med 2018; 378:1965–1975
13. Broman LM, Holzgraefe B, Palmér K, et al. The Stockholm experience: Interhospital transports on extracorporeal membrane oxygenation
. Crit Care 2015; 19:278
14. Ericsson A, Frenckner B, Broman LM. Adverse events during inter-hospital transports on extracorporeal membrane oxygenation
. Prehosp Emerg Care 2017; 21:448–455
15. Rhodes A, Evans LE, Alhazzani W, et al. Surviving sepsis campaign: International guidelines for management of sepsis and septic shock
: 2016. Intensive Care Med 2017; 43:304–377
16. Chen YS, Ko WJ, Lin FY, et al. Preliminary result of an algorithm to select proper ventricular assist devices for high-risk patients with extracorporeal membrane oxygenation
support. J Heart Lung Transplant 2001; 20:850–857
17. Gaies MG, Gurney JG, Yen AH, et al. Vasoactive-Inotropic Score as a predictor of morbidity and mortality in infants after cardiopulmonary bypass. Pediatr Crit Care Med 2010; 11:234–238
18. Hou X, Yang X, Du Z, et al. Superior vena cava drainage improves upper body oxygenation during veno-arterial extracorporeal membrane oxygenation
in sheep. Crit Care 2015; 19:68
19. Lindfors M, Frenckner B, Sartipy U, et al. Venous cannula positioning in arterial deoxygenation during veno-arterial extracorporeal membrane oxygenation
: A simulation study and case report. Artif Organs 2017; 41:75–81
20. Malfertheiner MV, Broman LM, Belliato M, et al. Management strategies in venovenous extracorporeal membrane oxygenation
: A retrospective comparison from five European centres. Crit Care Resusc 2017; 19:76–81
21. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock
(Sepsis-3). JAMA 2016; 315:801–810
22. Bartlett RH, Gazzaniga AB, Jefferies MR, et al. Extracorporeal membrane oxygenation
(ECMO) cardiopulmonary support in infancy. Trans Am Soc Artif Intern Organs 1976; 22:80–93
23. Rimachi R, Bruzzi de Carvahlo F, Orellano-Jimenez C, et al. Lactate/pyruvate ratio as a marker of tissue hypoxia in circulatory and septic shock
. Anaesth Intensive Care 2012; 40:427–432
24. Suetrong B, Walley KR. Lactic acidosis in sepsis: It’s not all anaerobic: Implications for diagnosis and management. Chest 2016; 149:252–261
25. Friedman G, De Backer D, Shahla M, et al. Oxygen supply dependency can characterize septic shock
. Intensive Care Med 1998; 24:118–123
26. Vogel DJ, Murray J, Czapran AZ, et al. Veno-arterio-venous ECMO for septic cardiomyopathy: A single-centre experience. Perfusion 2018; 33:57–64
27. von Bahr V, Hultman J, Eksborg S, et al. Long-term survival in adults treated with extracorporeal membrane oxygenation
for respiratory failure and sepsis. Crit Care Med 2017; 45:164–170
28. von Bahr V, Kalzén H, Hultman J, et al. Long-term cognitive outcome and brain imaging in adults after extracorporeal membrane oxygenation
. Crit Care Med 2018; 46:e351–e358
29. Pappalardo F, Montisci A, Scandroglio A, et al. Veno-venous ECMO in Europe: Are we all speaking the same language? Minerva Anestesiol 2017; 83:424–425