Extracorporeal cell therapy of septic shock patients with donor granulocytes: a pilot study

During a four-month period all patients of one medical and two surgical intensive care units of a tertiary care university hospital were screened to see if they fulfilled the parameters of severe sepsis and septic shock as defined by international consensus criteria [16]. Definitions of organ dysfunctions were adopted from the "Recombinant Human Activated Protein C Worldwide Evaluation In Severe Sepsis Study" (PROWESS Study) [17] with the difference being that liver failure was not an exclusion criterion in this current study. The exclusion criteria were age under 18 years, hepatitis C, active bleeding and HIV infection. Ten consecutive patients with septic shock were enrolled in the study.

The study flow is depicted in Figure 1. After inclusion of a patient, a healthy blood donor was identified and stimulated with corticosteroids (each 20 mg p.o. methylprednisolone, Sanofi-Aventis Deutschland GmbH, Frankfurt, Germany) 17 h, 12 h and 2 h before donation of an ABO-compatible granulocyte concentrate. Granulocytes were collected by extracorporeal density gradient centrifugation using hydroxyethylstarch (HES 200/0.5 6%, Fresenius Kabi AG, Bad Homburg, Germany) and citrate in a cell separator (COBE Spectra, Gambro BCT, Planegg-Martinsried, Germany) according to standard procedures. Because of the delay due to identification and stimulation of a compatible donor the first treatment of a patient was one day after inclusion in four cases, two days after inclusion in three cases, and three days after inclusion in two cases. Prior to treatment the inclusion criteria were re-confirmed. The whole extracorporeal system was first rinsed and prefilled with hemofiltration solution HF-BIC 35-410 with 4 mM potassium (Fresenius Medical Care, Bad Homburg, Germany). In mean 1.41 ± 0.43 × 10E10 donor granulocytes were delivered in donor plasma and were placed into the bioreactor compartment of the device prior to connection to the patient. An excess of hemofiltration solution during cell filling was discarded; therefore, no additional fluid was infused into the patient. The patients were treated for up to six hours with an extracorporeal method consisting of a plasma separation and plasma perfusion through the cell-compartment containing the donor cells. Blood access was veno-venous via a Shaldon-catheter. Plasma separation was carried out by a dialysis monitor (BM25, Edwards Lifesciences GmbH, Unterschleissheim, Germany) using a 0.5 μm pore-size plasma filter (PF 1000N, Gambro Hospal GmbH, Planegg-Martinsried, Germany). The plasma was infused into the continuously re-circulating donor cell compartment. A schematic view of the extracorporeal treatment device is shown in Figure 2. Plasma reflux to the patient was done through a second PF 1000N plasma filter to withhold the donor cells from being infused into the patient. Total extracorporeal volume was 400 ml. The blood flow rate was 150 to 200 ml/minute with a plasma separation rate of 16.7 to 33.3 ml plasma/minute using the BM 25 monitor. The MARS-Monitor 1 TC (Gambro Rostock GmbH, Rostock, Germany) was used for the re-circulating bioreactor circuit at a rate of 200 ml/minute and to maintain the temperature in the cell compartment at 37°C. Unfractionated heparin (20 IU/kg, Roche, Grenzach-Wyhlen, Germany) was given at the beginning of the extracorporeal treatment followed by a continuous infusion into the circuit. Heparin administration was adjusted to maintain activated clotting time (ACT) between 150 to 200 seconds. Following tolerability assessment of the first treatment, all patients were treated a second time 48 to 72 hours after the first treatment, again for up to 6 hours with granulocytes from the same donor.

We recorded basic demographic information, illness severity (Acute Physiology and Chronic Health Evaluation (APACHE) II, Sequential Organ Failure Assessment (SOFA), Multiple Organ Dysfunction Score (MODS), and Simplified Acute Physiology Score (SAPS) II scores), microbiological results, pre-morbidity, and clinical outcome for the study cohort (see Table 1). Patients were followed up for 28 days and hospital survival. At the days "inclusion", 1 to 8, 10, 12, 14, 21, 28 and before/after an extracorporeal bioreactor-treatment the patients were screened for clinical and immunological data: hemodynamic, inflammation, coagulation, hemolysis, temperature, organ function blood parameters, endotoxin, cytokines, complement (C3, C4), and the number of human leukocyte antigen DR (HLA-DR) molecules per monocyte surface. "Day 1" was defined as the day of the first bioreactor treatment. Viability and functionality of the donor cells were tested at the begin and end of the treatments by trypan blue test, phagocytosis by flow cytometry (Beckman Coulter Immunotech, Krefeld, Germany) with florescence-labeled E. coli and oxyburst both by flow cytometry with dihydrorhodamine 123 as well as in a luminometer (Thermo Labsystems, Waltham, MA, USA) with luminol and lucigenin.

Ten consecutive patients with septic shock were included in the study. Details concerning diagnoses, age, sex, relevant scores and survival are shown in Table 1. All patients had positive microbial tests with a mean of 4.7 ± 2.6 different microbial species per patient, predominantly candida, coagulase negative staphylococcus, enterococcus and E. coli.

During the first treatment performed in this study the heparin use was adjusted to a target ACT of 125 to 150 sec. After about 90 minutes the cell filter clotted and the treatment had to be terminated. Therefore, in all further treatments the heparin dosage was adjusted according to a target ACT of 150 to 200 sec. Except for Patient 6 where treatment #2 had to be terminated after five hours due to increased transmembranal pressure across the cell filter, all other treatments were carried out for six hours. Mean treatment time was 342 ± 64 minutes. Blood flow varied from 150 to 200 ml/minute depending on the patient's quality of blood access. The flow rate in the cell therapy circuit was 200 ml/minute. Plasma flow started with 16.7 ml/minute for the first 15 to 30 minutes and then increased to 33.3 ml/minute. A mean of 9.8 ± 2.5 liters of plasma were treated during each of the 20 treatments. To test whether the donor cells were still functional every two hours, cells from the cell circuit were evaluated for viability and functionality. For the whole treatment the cells showed a viability of more than 90% and unimpaired cellular functions like phagocytosis and oxidative burst.

During the extracorporeal procedures, no significant drop in mean arterial pressure was observed. All patients were on noradrenaline at the beginning of the first treatment and five of these patients also received it at the start of the second treatment. In 10 of the 20 procedures the noradrenaline dose could be reduced due to an increase in the mean arterial pressure. In five treatments the noradrenaline dose remained unchanged. Only in one case (Patient 4, second treatment) the noradrenaline infusion that had been turned off before the treatment was turned on again during the treatment, however, at a small dose (0.03 μg/kg/minute). Overall the Wilcoxon test showed a significant reduction in the noradrenaline dose (median from 0.06 to 0.035 μg/kg/minute; P = 0.016; Table 2) while the mean arterial pressure was stable during the bioreactor-treatment (median before 74, after 80 mmHg; not significant). Systemic vascular resistance index (SVRI) was not monitored in this study.

There was no significant change in mean platelet counts during the extracorporeal treatment (Table 2). D-dimers did increase significantly during the extracorporeal treatment from 752 ± 505 μg/l to 853 ± 450 μg/l but returned to 609 ± 381 μ/l within 12 hours. Antithrombin III concentration also changed significantly from 66 ± 17% at the beginning to 58 ± 15% at the end of the treatments, and improved slightly over the following 12 h to 61 ± 15%. Both activated partial thromboplastin time (aPTT) and prothrombin time (as International Normalized Ratio, INR) increased during the treatments due to heparin use but returned to pre-treatment values within 12 h after the extracorporeal circulation. No hemorrhages were observed.

No signs of hemolysis were observed. Haptoglobin remained within the normal range and no significant change in lactate dehydrogenase was seen during the treatments.

Moreover, no allergic reactions were recognized.

Expected in-hospital mortality based on the ICU entrance APACHE II (29.9 ± 7.2) and SAPS II (66.2 ± 19.5) scores were 69.1% and 71.5%, respectively [18‐20]. The observed mortality rate was 3 out of 10 within 28 days (on days 6, 9, and 18), and four during hospital stay (Patient 7 died on Day 40). Six patients could be discharged from the hospital in stable condition. No significant differences were seen between the survivors and non-survivors in the time at ICU before inclusion or the time between inclusion and first treatment.

The body temperature of the patients was stable during the treatments (Table 2). While creatinine did not show a significant change during the six-hour treatments there were small but significant increases in urea (Table 2), most probably due to interruption of dialysis in patients with renal failure. However, urea decreased again slightly within 12 h post treatment to 14.7 ± 8.4 mmol/l. No difference in PaO2 and FiO2 has been observed between start and end of the extracorporeal treatment (Table 2). Furthermore, no significant changes have been seen in PaO2 or FiO2 between the treatment day and the day after the treatment.

During the six-hour treatment a dramatic increase in white blood cell (WBC) counts was observed (Table 2). This increase was not due to changes in a particular subset of WBC, the ratio of segmented to banded neutrophils remained unchanged. Furthermore, there was a significant decrease in plasma endotoxin concentration from pre- to post-treatment (Table 2). In 11 of the tested cytokines a significant increase pre vs. post cell-bioreactor was observed (IL-2,-4,-8,-10,-1beta,-12, IP-10, Interferon gamma, Eotaxin, PDGF, RANTES). This resulted in significant increases pre- vs. post-treatment in the patients' plasma in 5 out of these 11 cytokines (IL-8,-10,-1beta, Eotaxin, RANTES) (Table 3). Moreover, there were significant decreases both in CRP as well as in PCT during the treatments (Table 2).

The statistical evaluation of the raw data showed improvements in several parameters evaluated during the 28 days of observation. The main findings include: significant reduction in CRP (Figure 3), PCT (Figure 4) and IL-8 (not shown); significant increase in HLA-DR on CD14-positive monocytes (Figure 5); significant increase in platelets and antithrombin III (not shown); significant reduction in noradrenaline use (Figure 6); significant reduction in alanine transaminase, aspartate transaminase and creatinine (not shown); and significant reduction in MODS and SOFA scores (not shown).

Out of these parameters PCT values, Noradrenaline dosage and SOFA score showed improvement already prior to the first treatment and further improved during the observation period.

In order to limit the bias due to the dropout of the non-survivors, an additional LOCF analysis was performed that also showed significant improvements for CRP, PCT, HLA-DR, noradrenaline, and creatinine.

Due to the large inter-individual differences no significant changes in leukocyte counts were seen except directly before and after treatment (see above).