Early prediction of Gram negative bacteraemia in febrile cancer patients: Correlation between some inflammatory mediators, exposure to gamma radiation and severity of infection

Bloodstream infections (BSI) have a significant impact on morbidity and mortality in the general population and in critically ill patients, it has worse outcomes as immunocompromised patients. The host response to such infection varies from clinical signs and release of certain inflammatory mediators, including Interleukin-6 (IL-6), Interleukin-8 (IL-8). Suggesting that inflammatory cytokines are already released in febrile patients before positive reports of microbiological cultures. The usefulness of these mediators, is to be used as early predictors of bacteraemia caused by Gram negative bacilli during onset of fever in cancer and non-cancer patients, in comparison to traditionally used markers as C-reactive protein (CRP) and complement C3 (C3) was reported in our study. One hundred twenty four feverish (cancer and non-cancer) in-patients were enrolled in the study. Serum samples were separated from collected blood samples at onset of fever for assay of inflammatory bio markers IL-6 and IL-8 (using (Enzyme linked immunosorbent assay ) ELISA technique), CRP (by Turbiquickreader) and C3 (by Diffu-plate). Plasma samples were separated for assay of total leukocytic count (using Beckman/Coulter semi automated). Blood samples were collected in blood culture bottels (BACTEC Peds Plus™/F) and cultured on MacConkey's agar No.3 for isolation of Gram negative bacilli which were identified by API 20E technique. Antimicrobial susceptibility test was performed (by disc diffusion method) using 14 antimicrobial agents with different mode of actions with and without exposure to gamma radiation at dose level of 24.4 Gy which is biologically equivalent to the fractionated multiple therapeutic dose used in the protocol of cancer therapy as well as lipase and protease enzymatic activities as virulence factors were performed (via tween and gelatin agar plates, respectively). Assay of serum IL-6 in ratss was also done using ELISA technique. Cesium 137 (137 Cs) Gamma cell 40 located at National Center for Radiation Research and Technology (Cairo, Egypt) was the irradiation source used in the study. Seventy cancer patients were representing 56% of all cases and 54 non-cancer patients were representing 4 4% were non-cancer patients. Positive blood culture samples for Gram negative bacilli in cancer and non-cancer patients were 49% and 54%, respectively. The most predominant Gram negative isolates in cancer patients were E. coli representing 41%, followed by Pseudomonas species as 27%, then K. pneumoniae as 23%. While, in non-cancer patients the most predominant isolates were K. pneumoniae representing 41%, followed by E. coli as 38%, then 4% of Pseudomonas sp. IL-6 and IL-8 serum levels were higher in feverish patients with Gram negative bacteraemia comparing to those with non-microbial fever for both groups (cancer and non- cancer) of patients. For cancer patients with Gam negative bacteraemia and those without there was significant difference in IL-6 and IL-8 serum levels (P=0.0001 and 0.0059, respectively). Similar results were also obtained for non-cancer patients (P=0.0288 and 0.0059). Moreover, serum levels of both mediators were higher in cancer patients with Gram negative bacteraemia than in non-cancer patients with Gram negative bacteraemia. The Cut-off levels to distinguish between bacteraemic (positive blood cultures) and non-bacteraemic (negative blood cultures) cases were determined using receiver operating characteristic curves (ROC): for CRP it was 29 mg/l for cancer patients, 119 mg/l for non-cancer patients and with 60% and 100% specificity and NPV 60% and 62.5%, respectively. While, the cut-off level of serum IL-6 was 398.6 pg/ml for cancer patients, 120.9 pg/ml for non-cancer patients and with 100% specificity and NPV 100% for both groups. So the efficacy of IL-6 as marker to discriminate between positive and negative blood cultures was higher than that of CRP. The ROC curve analysis showing diagnostic performance of CRP and IL-6 and their combination (via multi-ROC) could be used for discriminating patients with positive cultures from those with negative cultures (all cancer and non-cancer tested cases), We found that the best cut-off value of IL-6 was 120.9 pg/ml with 60% specificity, 100% sensitivity, NPV 100%, PPV 85.7%, Efficacy 88.2%, while for CRP the cut-off value was 85.9 mg/l with 50% specificity. Using the multi-ROC for both markers to improve the results for CRP , we found that CRP best cut-off value was 220 mg/l at IL-6 of 120.9 pg/ml with improved Sp, Sn, NPV, PPV and Efficacy of values 90%, 100%, 100%, 96% and 97.1%, respectively. The Area under the curve (AUC) also improved using the multi-ROC from 0.735 for CRP and 0.957 for IL-6 to 0.982 in combination with each other.The usefulness of both markers (IL-6 and CRP) was proved to distinguish between bacteraemic and non- bacteremic patients