[Is TNF-alpha "ripe" for routine diagnosis in sepsis?].

Mortality of the septic syndrome is around 40-60% and can rise to 100% if multiple organ failure (MOF) develops. It is generally assumed that the high mortality of sepsis can only be reduced by early diagnosis and prevention of subsequent MOF. The aim of our study was to investigate the validity of routine TNF-alpha determination for the diagnosis of septicemia and, in combination with clinical scoring systems [MOF score and Acute Physiological and Therapeutic Intervention Score (APATIS)], to define a "therapeutic window" during which an anti-TNF-alpha agent could be applied with the greatest chance of success. METHODS. TNF-alpha serum levels were measured and APATIS and MOF scores were calculated daily in 87 ICU patients. TNF-alpha serum levels were determined by means of an immunoradiometric assay (TNF-alpha IRMA, Medgenix, Belgium). Sepsis was diagnosed in 24 patients according to clinical criteria. To quantify the severity of sepsis, we set up the APATIS. The MOF score was used to assess the severity of MOF. Data were analyzed using the SAS software package (SAS Institute, Cary, N.C.) and are expressed as mean +/- SEM. RESULTS. The mean values of all sequential TNF-alpha determinations were significantly higher in the septic patients compared to the nonseptic patients (73.2 +/- 4.3 vs 8.5 +/- 0.4 pg/ml; P less than 0.01). Similarly, the maximum TNF-alpha values were significantly higher in the septic group (156.9 +/- 26.5 vs 20.1 +/- 1.3 pg/ml; P less than 0.01). To differentiate between sepsis and nonsepsis we set the cut-off point at a TNF-alpha serum level of 40 pg/ml and calculated a sensitivity of 70.8%, a specificity of 98%, and a diagnostic accuracy of 91.3%. None of the patients with a maximum TNF-alpha level above 250 pg/ml survived. Mortality was 80% above a maximum TNF-alpha serum concentration of 200 pg/ml, whereas only 40% of patients with a TNF-alpha maximum below 150 pg/ml died. The mean APATIS and MOF scores were significantly higher for septic than for nonseptic patients (APATIS: 20.3 +/- 0.5 vs 8.1 +/- 0.2 and MOF: 9.8 +/- 0.1 vs 4.6 +/- 0.1). To differentiate between survival and nonsurvival, we set the cut-off point at 25 for APATIS and calculated a sensitivity of 79% and a specificity of 93%. At a MOF score of 8, the sensitivity was 89% and the specificity 82%. In our series cumulative mortality at a maximum MOF of less than 8 was 4% and at MOF greater than 10, 68%. We found an interval of 2.9 +/- 0.9 days between the time TNF-alpha serum levels first exceeded 40 pg/ml and the development of severe MOF (MOF greater than 10) in 13 patients. CONCLUSION. Sequential TNF-alpha serum level determinations are useful for the diagnosis and prognosis of septicemia. We found an interval of 3 days between rising TNF-alpha serum levels and the development of severe MOF. This latency may represent the "therapeutic window" during which an anti-TNF-alpha-agent, e.g., a monoclonal anti-TNF-alpha-antibody, could be applied as a therapeutic consequence.