Uso de diferentes marcadores inflamatórios para a predição da ocorrência de febre em pacientes neutropênicos portadores de doenças hematológicas

Abstract

Infection is the main complication in neutropenic patients. Fever, usually the only clinical manifestation of infection in these patients, might be a late sign, with consequent delay in the onset of antibiotic therapy. Our hypothesis is that the use of inflammatory markers may allow an earlier diagnosis of infectious complications and potentially improve the outcomes in neutropenic patients. In an observational prospective single centre study, we evaluate the role of serum interleukin-8 (IL-8), inducible protein-10 (IP-10), tumor necrosis factor alpha (TNF-a), soluble receptors sTNFRI and sTNFRII, monocyte chemotactic protein-1 (MCP-1), macrophage inflammatory protein (MIP-1a), procalcitonin and eotaxin in 32 afebrile neutropenia events occurring in 26 hematological hospitalized patients. Blood samples were obtained in three different clinical periods: in the first day of neutropenia, the day before fever and in the day of fever. Patients were followed up to the 28th day, hospital discharge or death. The median age was 36 (range: 18 to 69) year-old and acute myeloid leukemia (AML) was the most common diagnosis (9 patients). Ten patients undertook a bone marrow transplant. Eight patients did not become febrile during the study procedures. Taking the whole population (32 events of neutropenia), sTNFRI, sTNFRII and IL-8 levels showed aprogressive increase pattern during the follow-up. There were a significative increase in sTNFRI levels (p=0,029) just prior to the fever episode as compared to the patients that did not presented fever. In a subgroup of non-transplanted patients, there were a significative increase in sTNFRI and MCP-1 levels (p=0,041 and p=0,042, respectively) just prior to the febrile episode as compared to afebrileneutropenic patients. In the so-called day of fever, all samples were undertaken with a median time of 11 hours (range: 1 to 15 hours) before fever episode. Using ROC curves, a cut-off of 1514 pg/mL for sTNFRI in fever day was able to discriminate between neutropenic patients that will or will not present fever, with sensibility of 65%, specificity of 87%, positive and negative predictive value of 93% and 46%, respectively. In a subgroup of non-transplanted patients, the same cut-off for sTNFRI in fever day was able to discriminate between neutropenic patients that will or will not present fever with sensibility of 71%, specificity of 87%,positive and negative predictive value of 91% and 63%, respectively. TNF-a, sTNFRII, IL-8, MIP-1a, IP-10, eotaxin and procalcitonin were not able to discriminate among neutropenic patients that will present or not a febrile episode. Conclusion: Daily measurement of sTNFRI during neutropenia could predict occurrence of fever in onco-hematological patients. This could result in a shorter time to start antibiotic treatment and, consequently, improve prognosis. A large prospective trial is needed to confirm these data.