Platelet-activating factor acetylhydrolase and haemophagocytosis in the sepsis syndrome.

Sepsis syndrome (SS) is associated with depressed PAF acetylhydrolase, the enzyme responsible for the degradation of platelet activating factor. PAF acetylhydrolase is in a large part produced by macrophages, whose inadequate activation with haemophagocytosis is frequent in patients with SS. The aim of this study was to test the hypothesis that PAF acetylhydrolase levels could be affected in these critically ill patients, because of the large amounts produced by activated macrophages in vitro and in vivo in animal models. The levels of serum PAF acetylhydrolase were assessed in 90 SS patients, who were divided into three groups: patients with (n = 34) or without haemophagocytosis (n = 31), and patients without thrombocytopenia (n = 25) who were used as a control group. The number of organ dysfunctions was matched between patients with haemophagocytosis and controls. Normal reference values were obtained in 59 randomly selected blood donors. Circulating levels of PAF acetylhydrolase were significantly (p = 0.0001) decreased in patients with SS (57+/-3 nmol/ml/min, n = 90) when compared with healthy subjects (69+/-3 nmol/ml/min, n = 59). PAF acetylhydrolase levels were greater in the presence of a haemophagocytosis but without statistical significance (64.2+/-6.5 vs. 50.1+/-2.8:p = 0.25). Despite the fact that macrophagic activation stimulates the in vitro release of PAF acetylhydrolase, no difference was found between patients with or without haemophagocytosis. The mechanism and the role of the PAF acetylhydrolase reduction in SS patients remain to be determined.


Introduction
The sepsis syndrome (SS) is a complex inflammatory response to various acute insults. 1 Platelet-activating factor (PAF), one of the main mediators involved in the inflammatory response, is produced through the metabolic pathway of the phospholipase A2 from phospholipids of the cell membrane. 2 PAF plays a key role in the initiation and development of various critical diseases which are associated with increased levels of phospholipase A2, including septic shock, adult respiratory distress syndrome and multiple organ failure (MOF). 3 Blood concentrations of PAF are regulated by a circulating enzyme, the PAF acetylhydrolase, which converts PAF into inactive lyso-PAF. 4 Depressed PAF acetylhydrolase has been reported in association with severe sepsis 5 and MOF. 3 This disregulation of PAF/PAF acetylhydrolase may result in an amplified inflammatory response secondary to the increased half-life of circulating PAF. Monocytes and macrophages physiologically participate in the regulation of the balance of PAF/PAF acetylhydrolase. Stimulated monocytes release a large fraction of their newly synthesised PAF, 6 whereas macrophages are a major source of PAF acetylhydrolase, this process being a mechanism for the resolution of the inflammatory response. 7 Recently a study has demonstrated that the PAF acetylhydrolase in human plasma originates from haematopoietic lineage cells. 8 We have reported the frequent association between severe SS with thrombocytopenia and haemophagocytosis. 9 As a strong macrophagic differentiation occurs in these patients, 9 we hypothesized that PAF acetylhydrolase levels could be affected in these SS patients, because of the large amounts produced by activated macrophages in vitro and in vivo in animal models. 10 We then conducted the present study to compare the levels of serum PAF acetylhydrolase between SS patients with or without haemophagocytosis.

Patients
All aspects of the study protocol were approved by the committee on human investigations of our institution. During a 12-month period, all consecutive patients who fulfilled the following inclusion criteria were studied: the presence of (1) an SS according to Bone's criteria, 1 and (2) a thrombocytopenia (platelet count < 100 000/mm 3 ) of undetermined origin. Patients were excluded from the study when at least one of the following criteria were present: hematological disease, previous anticancer chemotherapy, major bleeding and/or prior transfusion of any blood component, disseminated intravascular coagulation defined as the presence of a prothrombin time of less than 50% of the control value together with either a positive ethanol test or D-dimers test, and previous administration of heparin. In each patient, the severity of the acute illness at the time of admission to the intensive care unit (ICU) was assessed using the APACHE II score. 11 The potential causes of the sepsis syndrome as well as the number and nature of acute organ system dysfunction were evaluated at the time of enrolment. 12 Patients were considered as infected if an ongoing infection was documented at enrolment. Bacteraemia was defined by at least two positive blood cultures with the same pathogen. Focal infections were diagnosed on the basis of both clinical and microbiological findings. In each patient, a bone marrow aspirate was performed at the day of enrolment to study the cause of thrombocytopenia. It allowed us to divide patients into two subsets, according to the presence or absence of cytological findings associated with a macrophagic activation. Haemophagocytosis was assessed by cytological examination of May Grünwald Giemsa stained bone marrow smears, using a previously described protocol. 9 Haemophagocytosis was diagnosed when the percentage of macrophages exceeded 2% of bone marrow nucleated cells, in conjunction with the presence of more than one haemophagocytic cell per field at ´200 magnification.
During the same period, consecutive patients with SS but no thrombocytopenia were also studied. Although the bone marrow was not studied for ethical reasons in this group, the absence of thrombocytopenia was considered as potentially predictive of the absence of haemophagocytosis. Patients in this group were matched in terms of number of organ dysfunctions, according to the ODIN score, 12 with those suffering from haemophagocytosis.

PAF acetylhydrolase measurement
Serum PAF acetylhydrolase levels were measured on the day of enrolment by assessing the degradation of 13 Briefly, 1 ´10 5 dpm of [ 3 H] acetyl-PAF, 10 Ci/mmol (NEN, Les Ulis, France), 0.1 mM PAF (Tebu, France), HEPES buffer (pH 7.8) in a final volume of 450 ml, and 50 ml of serum (1/50 dilution in HEPES buffer) were incubated for 15 min at 37°C. The reaction was stopped with 100 ml BSA (10%) and 500 ml trichloracetic acid (20%). Samples were centrifuged (1500 ´g , 15 min), supernatants were mixed with 2.5 ml of scintillation fluid (Opti-fluor, Packard, France) and counted in a liquid scintillation counter (Packard). Results were expressed as nanomoles of PAF degraded per ml of serum per min, as means of duplicate determinations. Normal reference values were established by measuring serum concentrations of PAF acetylhydrolase in 59 randomly selected healthy subjects.

Statistical analysis
Data were expressed as mean ± SEM. Continuous variables were compared between groups using either the Mann-Whitney U-test or the Kruskall Wallis one-way analysis of variance. Qualitative variables were compared using the chi-square test, with correction for continuity when appropriate. A p value of < 0.05 was considered statistically significant.

Results
Ninety patients (mean age: 58 ± 2 years, APACHE II: 23 ± 1) were enrolled into the study. Reasons for admission to the intensive care unit included multisystem trauma (n = 15), complicated postoperative course (n = 23), and severe medical illnesses (n = 52). Seventy-four patients (82%) sustained a multiple organ dysfunction and 31 (34%) died during their stay in the ICU. Sixty-five patients were diagnosed with thrombocytopenia. Megakaryocytic lineage was always normal. Macrophagic activation was identified on bone marrow smears in 34 cases (52%). When compared with patients without macrophagic activation, the presence of a haemophagocytosis was associated with a higher number of organ dysfunctions and greater infection rate (Table 1). Patients without thrombocytopenia (n = 25) had a similar number of organ dysfunctions and mortality rate, compared with those with haemophagocytosis (Table 1).

Discussion
In the present study, decreased levels of PAF acetylhydrolase were found in patients with SS when compared with healthy subjects. This finding is in agreement with previously reported results in patients suffering from sepsis 5 and MOF. 3 Reduction of the PAF acetylhydrolase in these settings is thought to increase substantially the half-life of PAF, which in turn may amplify the inflammatory cascade in response to an acute insult, leading to MOF. 14 Although lower PAF acetylhydrolase levels have been reported in patients who die from sepsis when compared with survivors, 5 they were not predictive of the outcome in our patients. In addition, we did not confirm the correlation between PAF acetylhydrolase levels and the number of organ dysfunctions previously reported. 3 Such a discrepancy may be explained by different study designs. In the present study, PAF acetylhydrolase levels were measured when MOF was already initiated, whereas Partrick e t a l. 3 compared the PAF acetylhydrolase determined at admission with subsequent occurrence of organ dysfunctions during the ICU stay. Patients with SS secondary to severe multisystem trauma had lower PAF acetylhydrolase levels than those with SS of medical or surgical origin. Direct tissue injuries observed in patients sustaining severe trauma may lead to extensive cellular damage and subsequent elevated phospholipase A2 activity. Increased release of PAF by injured tissue associated with reduced levels of serum PAF acetylhydrolase might amplify organ dysfunctions in this setting when compared with SS of other origin.