Nutritional status and cytokine concentration during chemotherapy in Mexican children: A longitudinal analysis.

OBJECTIVE: The aim of this study was to assess whether the nutritional status of children with cancer is influenced by variations in cytokine concentrations observed during chemotherapy. We also evaluated whether this relationship could be modified by nutritional status at diagnosis and type of cancer. METHODS: Mexican children with lymphoma or solid tumors were evaluated at diagnosis and at 2- and 6-mo follow-up visits. Blood samples were obtained to determine serum prealbumin, tumor necrosis factor (TNF)-α, interleukin (IL)-6, leptin concentrations, and hemoglobin. Children were classified as undernourished (UN) or well nourished (WN), according to prealbumin concentration. The influence of each cytokine on prealbumin concentration was analyzed by time-series regression model. RESULTS: Fifty patients (ages 2-17 y) were enrolled. There were 17 children with lymphomas and 33 with solid tumors. At baseline, 56% were UN and 26% presented anemia; the frequencies of UN children were higher for those with lymphoma than for those with a solid tumor (P = 0.003). By nutritional status, UN children presented lower leptin (P = 0.002) but higher IL-6 concentrations (P = 0.009) than the WN group. Children with lymphoma presented lower prealbumin (P = 0.003), but higher TNF-α (P = 0.001) and IL-6 (P = 0.011) concentrations than those with solid tumors. At follow-up, the concentration of prealbumin increased and IL-6 decreased in children with lymphoma. Multivariate analysis demonstrated that decreases in prealbumin concentration at the end of follow-up were associated with increases in IL-6 and TNF-α concentration during chemotherapy. CONCLUSIONS: These results suggest that the cytokine responses during chemotherapy are related to nutritional status at the end of 6 mo of treatment regardless of the initial nutritional status and the type of cancer.

Utilidad de la proteína C reactiva para el diagnóstico de infección bacteriana en el paciente pediátrico con cáncer, fiebre y neutropenia / Usefulness of C-reactive protein in the diagnosis of bacterial infection in the pediatric patient with cancer, fever and neutropenia

Introducción. El diagnóstico de infección bacteriana en el paciente con cáncer, fiebre y neutropenia se dificulta debido a una pobre respuesta inflamatoria. Se han realizado evaluaciones con reactantes de fase aguda, como la proteína C reactiva, con resultados variables. El objetivo de este trabajo fue calcular la sensibilidad, especificidad, valores predictivos positivos y negativos y razones de verosimilitud de la proteína C reactiva para el diagnóstico de infección bacteriana en pacientes con cáncer y neutropenia febril. Métodos. Se realizó el estudio de la prueba diagnóstica. Se incluyeron pacientes pediátricos con cáncer, y neutropenia (<500 NA/mm³). La proteína C reactiva se cuantificó por nefelometría. Los episodios se clasificaron en cuatro grupos: grupo I, infección microbiológicamente documentada; grupo II, infección clínicamente documentada; grupo III, fiebre por otras causas; y grupo IV, pacientes con neutropenia sin fiebre. Se realizó el cálculo de sensibilidad, especificidad, valores predictivos positivos y negativos, curvas operantes del receptor y razones de verosimilitud. Para la comparación de variables cuantitativas se emplearon la U de Mann-Whitney y Kruskal-Wallis y para variables cualitativas, χ². Resultados. Se incluyeron 127 episodios que se distribuyeron en: 29, 47, 20 y 31 episodios para los grupos I, II, III y IV, respectivamente. Las medianas de la proteína C reactiva fueron 282 mg/L para el grupo I, 205 mg/L grupo II, 27.3 mg/L grupo III y 5.1 mg/L para el grupo IV (p < 0.001). Con la proteína C reactiva de 60 mg/L se obtuvo una sensibilidad de 94%, especificidad de 94%, valor predictivo positivo 96% y valor predictivo negativo 92%; razón de verosimilitud para un resultado positivo 15.6 y de 0.06 para resultado negativo. Conclusiones. La proteína C reactiva es una prueba útil y económica para el diagnóstico de infección bacteriana en el paciente con cáncer, fiebre y neutropenia.

Background. Diagnosis of bacterial infection in the patient with cancer, fever and neutropenia is difficult due to the poor inflammatory response. Several evaluations of acute phase reactants such as C-reactive protein (C-RP) have been performed with diverse results. The aim of this study was to calculate the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV), and likelihood ratios (LR) for C-RP in the diagnosis of bacterial infection of patients with cancer, neutropenia and fever. Methods. We carried out a diagnostic test study. Pediatric patients with cancer and neutropenia (<500 NA/mm³) were selected. C-RP was determined by nephelometry. Episodes were classified into the following groups: group I: microbiologically documented infection; group II: clinically documented infection; group III: fever of unknown origin; group IV: patients with neutropenia without fever. Sensitivity, specificity, PPV, NPV, receiving operating curves (ROC) and LR were calculated. Mann-Whitney U test and Kruskal-Wallis test were used for comparison of quantitative variables. For qualitative variables, χ2 test was used. Results. There were 127 episodes distributed as follows: 29, 47, 20 and 31 for groups I, II, III and IV, respectively. Median of C-RP values were 282 mg/L for group I, 205 mg/L group II, 27.3 mg/L group III and 5.1 mg/L group IV (p <0.001). With a C-RP value of 60 mg/L, we obtained a sensitivity of 94%, specificity 94%, PPV 6% and NPV 92%. LR for a positive test was 15.6 and LR for a negative test was 0.06. Conclusions. C-RP is a useful and economically feasible test for diagnosis of bacterial infection in patients with cancer, neutropenia and fever.