To determine the specific population of critically ill septic patients who benefit most from cytokine adsorption therapy with the HA-380 cartridge. Benefit of the treatment will be assessed on the basis of:

• The scope of the effect of cytokine adsorption therapy in this specific population of critically ill patients expressed by cytokine variability within the patients
• The scope of cytokine changes in passing the adsorption cartridge my measuring cytokine levels in the patient's blood directly before passing through the cartridge and directly after having passed through the cartridge.
• The scope of changes in organ dysfunction expressed by SOFA scores that are repeatedly calculated during the treatment with cytokine adsorption and then daily until day 7 of the ICU treatment.
• The scope of changes on cellular function on immune cells in serum samples taken before and after cytokine adsorption therapy.
• The scope of removal of anti-infective drugs from the blood in passing through the cytokine adsorption cartridge by measuring antibiotic drug levels in the patients blood during the cytokine adsorption therapy
• 30 day and 90 day mortality and location status in survivors

1. Medical Problem Sepsis, a syndrome of physiologic, pathologic, and biochemical abnormalities induced by infection. It is a major public health concern, accounting for more than $20 billion (5.2%) of total US hospital costs in 2011. Sepsis has shown an increase in incidence most likely due to a combination of aging patient populations with more comorbidities and greater recognition. Sepsis is by conservative estimates a leading cause of mortality and morbidity in critical illness with long term sequelae for survivors.

   Underlying to sepsis is a multifaceted host response to an infecting pathogen that may be significantly amplified by endogenous factors and is defined by a nonhomeostatic, dysregulated host response to the infection with a potential lethality that is considerably in excess of that straightforward infection. In the early phases, an overshoot of pro-inflammatory mediators is often observed, caused by pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs). High levels of pro- and anti-inflammatory cytokines are both associated with increased mortality. Blocking of overproduction of these cytokines using monoclonal anti- bodies targeting specific cytokines such as tumor necrosis factor-α (anti-TNF-α) have not demonstrated mortality benefits in larger trials.

   Therefore, an approach to decrease cytokine overproduction by using extracorporeal blood purification has previously been examined. Initial trials using Polymyxin B showed favorable results for markers of critical care treatment with a reduction of vasopressor dosage and higher ratios oxygen partial pressure to inspired oxygen. A larger randomized trial by Payenne however did not show any improvement of organ dysfunction or reduction in mortality by using Polymycin B hemoperfusion.

   A recent trial on the effect of the CytoSorb adsorption device using Interleucin-6 (IL-6) levels as marker of efficacy did also not find an effect of mortality, improved organ dysfunction or in fact a reduction of circulating IL-6 levels. However, in this trial the baseline cytokine levels were only moderately increased to begin with suggesting use of a device within a suboptimal patient population.

   As no more approved specific therapies to treat sepsis, exist following the withdrawal of recombinant activated protein C (Xigris; Eli Lilly, USA) from the market in 2011 following the results of the PROWESS-SHOCK trial and the pathophysiologic basis for mechanisms in sepsis due to an excess cytokine release are sound, the extracorporeal removal of cytokines remains a promising therapeutic target given the correct patient population.

2. Principal research question and novel aspects Within the trials addressing removal of excess cytokines, the selection of patients and timing of the intervention has been difficult, and results have not been conclusive. Data on the patient population which might benefit most in regard to the aspects of cytokine levels removed and organ failure or progressive organ dysfunction averted are sparse or non-existent. Even the largest randomised trial so far was active for 3 years in recruiting 100 patients out of 580 patients screened. This trial failed to show a positive effect on the primary endpoint of IL-6 removal as well as on secondary endpoints of morbidity or 30 day mortality. However, in this small trial there was a very low median cytokine level measured at baseline and randomisation which might be associated with the consecutive failure to detect a reduction in circulating cytokines. The question remains, whether this trial identified a patient population that was going to benefit from such a therapy to begin with. Post-hoc analyses of another trial, the Euphrates trial showed that there might be positive signal for reduced mortality in patients with an endotoxin activity that was in a mid-range and not at the highest end of the spectrum.

   We therefore aim to undertake an observational trial to examine the effect of treating sepsis according to the surviving sepsis campaign guidelines and using extracorporeal cytokine removal looking at circulating cytokine levels and development of new or worsening organ dysfunction in patients with sepsis or septic shock who show highly increased IL-6 levels.

3. HA 330 and HA 380 Hemoperfusion Cartridge The HA type resin cartridge (Jafron Industries, China) is an extracorporeal hemoperfusion device that uses neutral microporous resin, and it has been proven to specifically absorb different media- tors such as bilirubin and cytokines.

The HA380 HA 330 and HA 380 resin cartridges haves the ability to absorb various medium-sized factors, including most inflammatory cytokines (IL-1, IL-6, IL-8, IL-10 and TNF-a) and haves therefore lately been used in the treatment of patients with sepsis in China with several case reports demonstrating promising results on the clinical course of treated patients. Furthermore, there pooled there is emerging data that this therapeutic concept is efficienct and helpful in adequately selected patient populations.

The HA 380 cartridge is the CE-labeled updated version of the HA-330 cartridge which features a larger surface area of the resin beads increasing the overall capacity and thus potential efficacy of the device.

The HA 380 Hemoperfusion Cartridge works using Van der Waals interaction between resin molecular groups and adsorbed substances. Unlike the CytoSorb device, these are incorporated into the resin molecule by a molecular sieve effect of the 3D network structure and the lipophilic hydrophobicity of the molecular group. Therefore, the pass-through effect in removing cytokines is expected to be larger and more sustained, as previously adsorbed cytokines will not again be released by continued blood flow or movement of the device. The range of adsorption is shown in figure 1. The device described has a CE label and is approved for treatment in a variety of clinically situations including but not limited to intra-operative use in cardiac surgery, intoxication, liver failure and sepsis. Compared to other available devices which could be used in the same settings this cartridge offers superior flow characteristics, less flow resistance and a higher capacity of cytokine removal. The effect of combining a standard treatment course of sepsis according to the Surviving Sepsis Campaign guidelines with extracorporeal cytokine removal however is not well examined overall.

In the investigator center, the investigators have therefore undertaken a retrospective pilot observational trial of 15 patients in which the using the HA 380 330 device was used as a rescue therapy in therapy in 15 patients with septic shock and multi-organ failure refractory to conventional treatment. In this trial of extremely ill patients the investigators found a significant difference between hospital survivors and non-survivors in circulating IL-6, IL-10 and TNF-alpha levels after 8 hours of treatment with the HLA 3380 cartridge. However, the overall variation of cytokine levels at baseline and the variation of changes from baseline during therapy were rather large. The sample size was too small to infer any causality or undertake multi-variate analysis.

Inclusion Criteria:

1. >18 years of age
2. diagnosis of sepsis or septic shock with at least 2 organ-dysfunctions defined by SOFA score
3. baseline serum IL-6 > 1000 pg/ml, IL-10 levels >250 pg/ml or TNF-alpha levels > 100 ng/l. The baseline IL-6 measurement must be on the same calendar day that the cytokine adsorption is started, the baseline IL-10 or TNF-alpha can be taken up to 36h before because of higher turn-around times for these tests.
4. need for CRRT or ECMO therapy

Exclusion Criteria:

1. Solid organ malignancies with or without treatment
2. Participation or enrollment in another clinical study within 30 day
3. Pregnant and lactating females If women of childbearing potential will be enrolled, a pregnancy test will be performed by assessing the value of ß-HCG in the routinely drawn blood and or urine.
4. Patient has been committed to an institution by legal or regulatory order
5. Participation in a parallel interventional clinical trial
6. The subject received an investigational drug within 30 days prior to inclusion into this project

We will perform an observational trial in a large cohort of patients with sepsis and septic shock who are treated with cytokine adsorption by hemoperfusion with a HA-380 cartridge. Sepsis will be defined as done in the Sepsis-3 definition. In addition to sepsis or septic shock patients must have a confirmed excess of IL-6, IL-10 or TNF-alpha or a combination of these with IL-6 levels > 1000 pg/l, IL-10 levels >250 pg/l or TNF-alpha levels > 100 ng/l. Patients must be in need of CRRT or ECMO therapy to perform the HA-380 therapy. Stand-alone hemoperfusion will not be performed.

Patients must be included within 24 hrs of diagnosing sepsis. The IL-6 measurement for inclusion must be from within 24 hrs, IL-10 or TNF-alpha measurements for inclusion can be from within 36 hrs before inclusion.