Monocyte

Immunological assessment is important to characterize the host defence response of trauma patients as infection is the most common cause of severe morbidity and late death. Sixty trauma patients were followed serially and divided into three groups: those with an uneventful recovery (n = 17), those with recovery after major sepsis (n = 27) and those who died (n = 16). The ability of peripheral blood monocytes to express the antigen HLA-DR was measured and compared to the results from 77 asymptomatic volunteers. After initial injury, there was a significant reduction.from normal in the three trauma groups. It took one week for HLA-DR antigen expression to return to the normal range in the first group, three weeks in the second group, and in the third group it never returned to normal. Monocyte HLA-DR antigen expression, after incubation with lipopolysaccharide, distinguished those patients who survived from those who died. There was no difference in HLA-DR antigen expression between a high transfusion group of 31 patients who received 10 or more units of blood and a low transfusion group of 29 patients who received less than 10 units. The ability of monocytes to express HLA-DR antigen correlated directly with the clinical course in these patients and its measurement identified a group ofpatients at high risk of injection and death following trauma.

Infection continues to be a problem in the surgical patient despite advances in operative technique and perioperative management'. The victim who has sustained major trauma or burn injury is at especially high risk of developing subsequent infection because of initial bacterial contamination, shock and a suppressed immune Virtually all aspects of the immune response have been reported to be depressed following injury and surgery, including macrophage, lymphocyte and neutrophil function, delayed hypersensitivity skin test response, immunoglobulin production and serum opsonic capacity5-' '. Few studies, however, have correlated these tests of immune function with eventual outcome in terms of mortality and infectious complications. We have found two measurements to be of particular importance in the prediction of outcome: the ability of peripheral blood monocytes to express the cell surface antigen HLA-DRL3, and the ability of serum to opsonize bacteria for phagocytosis by normal n e~t r o p h i l s '~.
Most investigators agree that the initial event in the generation of the specific immune response is uptake and degradation of foreign antigens by macro phage^'^^'^. The major histocompatibility (MHC) class 11 antigens (HLA-DR in humans) appear to be critical to the T cell response that follows because T helper cells will only recognize and respond to foreign antigens that are presented on the macrophage cell surface in proximity to HLA-DR antigen molecule^'^. The ultimate response is T cell activation with subsequent B cell immunoglobulin production and enhanced phagocytosis of opsonized bacteria. HLA-DR bearing monocytes, therefore, play a central role in the generation of the immune cascade and we report their behaviour in relation to clinical outcome in 60 severely injured patients.

Patients and methods
The study group consisted of 60 consecutive patients admitted to the trauma service at Ilniversity Hospital in Louisville, Kentucky, USA. All had sustained major trauma or burns and had an injury severity score (ISS) of at least 20. Patients with a previous medical history of myocardial infarction, organ failure, neoplastic disease or current immunosuppressive therapy (steroids or cytotoxic drugs) were excluded. Patients transferred from other centres after initial stabilization were accepted, but those transferred after prolonged care were not. Informed consent was obtained from the patient or relatives in all cases.
The ISS and %LD,, values were calculated for each patient using the abbreviated injury scale (AIS) table1s. The coded value for each of the six body compartments was squared and the sum of the three largest squared numbers was used as the ISS. The %LD,, value is simply the ISS corrected for age; this was derived by dividing the ISS by one of three factors dependent on the age of the patient".
Venous blood was taken within 24 h of admission and then serially throughout the patient's stay in hospital as well as in the follow-up trauma clinic. Expression of HLA-DR antigen on peripheral blood monocytes was identified by a previously described dual monoclonal antibody staining t e~h n i q u e '~. Briefly, 4 0~1 aliquots of buffy coat cells prepared from acidxitrate dextrose anticoagulated blood were stained for 20min with both fluorescent labelled antiMO2 (Coulter Immunology, Hialeah, Florida, USA) and antiHLA-DR (Becton-Dickinson, Sunnyvale, California, USA) monoclonal antibodies in the presence of 0.1 per cent sodium azide. Erythrocytes were removed by hypotonic lysis and the samples were preserved with 1 per cent paraformaldehyde. Sample analysis was performed on a cytofluorograph IIS flow cytometer (Ortho Diagnostics, Westwood, California, USA) with simultaneous two-colour discretion to detect both fluorescentlabelled antibodies. By this method, only cells that stained with the monocyte specific antibody, M 0 2 , and HLA-DR were analysed for the coexpression of HLA-DR antigen. In addition, paired aliquots of blood were incubated with lipopolysaccharide (LPS) (10pg/ml) for 2 h and were then analysed as described. At least 100 monocytes were analysed in each sample and complete white blood cell count and diKerential cell counts were carried out to calculate absolute monocyte numbers per millilitre of blood. Studies of 77 asymptomatic volunteer subject5 demonstrated highly reproducible results, showing no effect of age, sex race, time of day, week or year on the levels of HLA-DR expressior in any given subject. There was, however, a linear decrease to nc detectable level of antigen at 19 days of storage at 4'C, and so a1 samples were stained within 4 h of venepuncture.
Antigen expression in t r a u m a patients: M. J. Hershman et al.
The total number of units of blood transfused during initial resuscitation (within 48 h of injury) was also recorded. The majority of patients received units of packed red blood cells, each of a volume of 2W250ml. A few transfusions were given as whole blood units of 500ml each. The 60 patients were divided into a high transfusion group of 31 patients who received 10 or more units of packed red blood cells or whole blood and a low transfusion group of 29 patients who received less than 10 units. Serial samples from four units of packed red blood cells were withdrawn over a 19 day period for monocyte HLA-DR antigen measurement, with and without LPS incubation, to assess temporal changes in the expression of this antigen.
Results are expressed both in terms of the percentage of total M 0 2 staining monocytes that expressed HLA-DR and as mean fluorescence intensity of the gated M 0 2 positive cells; this is an expression of the density of HLA-DR expression. These results are compared with those derived from the control population of 77 uninjured volunteers, free of medical illness and ranging in age from 20 to 91 years. Results are expressed as mean (s.e.m.) values. The percentage data were transformed to the arcsines of their square roots to produce a nearly normal distribution before statistical analysis. When three or more groups were compared over the same time period, a 2-tailed, I-way analysis of variance was performed. Student's t tests for unpaired samples were then used to compare results and determine statistical sgnificance. Differences were considered significant at P < 0.05.

Results
Patients were divided into three clinical outcome groups: a good outcome group of patients who did not develop major sepsis (n = 17), a group who survived major sepsis (n = 27) and a group of patients who died (n= 16). Of the 16 patients in the third group 14 died of overwhelming infection with multiple system organ failure (MSOF), while the other two died from progressive hypoxia within 4 days of admission. Major sepsis was strictly defined as bacteraemia not associated with central lines, pneumonia, major wound or soft tissue infection, intra-abdominal abscess or other body cavity abscess, and the need for reoperation for sepsis, all of which prolonged hospital stay. The demographic make-up of each group, including ISS and %LD,,, is listed in Table 1 .
Relative monocytosis was seen in all patients, usually beginning by the third day after injury, becoming maximal on the 11 th day after hospital admission, and returning slowly to normal over several weeks ( F i g u r e l ) . The normal range, obtained from our cohort of 77 asymptomatic volunteers, was about I W O O monocytes/ml on peripheral smear. Interestingly, monocytosis was least marked in the group of patients who died, although differences did not reach statistical significance   admission. The two groups of patients who survived showed an eventual return of their mean values to within the normal range, but at differing times from initial injury. In the uneventful recovery group, the mean HLA-DR antigen expression returned to 70percent by 1 week and 85percent by 4 weeks. In the major sepsis group, mean HLA-DR antigen expression returned to 70 per cent after 3 weeks, but never reached 85 per cent during the period studied. Mean HLA-DR antigen expression steadily declined in the third group until the time of death.
A similar pattern of results was seen when the mean fluorescence intensity (HLA-DR density) data were analysed ( Figure 3). Depressed values returned to normal in the surviving groups but not in the group of patients who died. Figure4 shows a plot of HLA-DR antigen expression following 2 h incubation with LPS over time in the three clinical outcome groups. In both groups of surviving patients, the percentages of monocytes expressing MO2-DR were always within the normal range (>70percent). Patients who were destined to die exhibited monocyte HLA-DR antigen expression after LPS incubation below the normal range from day 1 until the time of death. Statistical significance was reached between the surviving groups and the dead group as shown on Figure4.
There was no significant difference in HLA-DR antigen expression between the 31 patients in the high transfusion group and the 29 patients in the low transfusion group, although the high transfusion group did have slightly higher values over the first 2 weeks after admission to hospital. To study the effect of There was a steady, linear decline in HLA-DR antigen expression over the period until none was detected on day 19. In contrast to the surviving patients, LPS incubation did not result in appreciably higher values at any time throughout the experiment and values were actually less after day 6.

Discussion
Trauma induced deaths can be divided into those that occur early, usually due to ongoing haemorrhage or cerebral injury, and those that occur late, which are almost always associated with infe~tion'.~. We have found that the occurrence of sepsis is not entirely related to the severity of injury (as measured by 1%) or predicted by clinical assessment of the amount of bacterial contamination at the time of injuryI3. A variety of immunological measurements has been made in injured patients and an early depression of the overall immune response has been well documented' 12. The central role of the macrophage has become clearer through Faist's research, which has shown increased prostaglandin-E2 production, and reduced IL2 and interferon-gamma production, presumably as a result of inhibitory macrophages'. Our results show a relative monocytosis, confirming previous work'. ' ', but an overall reduction in monocytes that express surface HLA-DR. The smallest increase in peripheral monocytes occurred in the patients who died; they also had the lowest HLA-DR expression. This indicates that an increase in immature monocytes was not responsible for reduced HLA-DR expression. Furthermore, monocytes from surviving patients were able to express this antigen after incubation with LPS signifying a down regulation in HLA-DR expression after injury.
Although we have shown that M02-positive monocyte HLA-DR expression measurement, performed within 24 h of admission after injury, was reduced in virtually all patients, serial measurements gave more information. An upward trend in HLA-DR expression was highly predictive of recovery, whether or not infection supervened. Conversely, the patients who died showed a constant downward trend despite LPS incubation. Monocyte HLA-DR expression together with injury severity, age and bacterial contamination have been incorporated into an outcome predictive score which reliably predicts uneventful recovery from major infection or death in trauma patients".
The highest rate of infection was in burn patients since those who qualified for the study had at least 40 per cent second and third degree burns. In addition, the age and ISS of the patients who died were significantly greater than those of the patients who survived, although there was little difference between the group with major infection and those who recovered without sepsis. The major infection group was composed of slightly more patients with penetrating injury than the group without sepsis. This may suggest a greater bacterial contamination through colonic injury in the former group. Transfusion had no significant effect on HLA-DR antigen expression despite a linear decrease in the ability of monocytes from stored blood to express HLA-DR. The blood given to our patients was about nine days old on average and so values of HLA-DR antigen expression in this study reflect those of the patients' own monocytes rather than those of transfused monocytes. Controversy exists concerning whether massive transfusion predisposes to infection following trauma2'. We did not find an increased rate of major infection in the group that had more than 10 units of blood compared to the group that had less than 10 units.
The use of the monoclonal antibody antiM02 allows consistent identification of a similar population of monocytes since the M 0 2 antigen is not expressed on B, T or null lymphocytes, or neutrophils22. The test is simple to perform and requires only 4 h in centres equipped with flow cytometry. Measurement of monocyte HLA-DR antigen expression is a useful adjunct to overall clinical impression and helps to identify at an early stage those trauma patients who are at especially high risk of infection and death. This may permit earlier therapeutic intervention with new measures as they become available.