Rendimiento de ecuaciones para predecir riesgo cardiovascular en una población de Argentina / Performance of equations to predict cardiovascular risk in an Argentine population

Resumen El rendimiento de las ecuaciones existentes de predicción de riesgo cardiovascular (RCV) en población argentina es desconocido. Se comparó RCV estimado por dichas ecuaciones, con la ocurrencia de eventos cardiovasculares (ECV) en una población de pacientes sin enfermedad cardiovascular de un hospital argentino. Se incluyeron aleatoriamente adultos entre 40 y 70 años, excluyéndose quienes al momento del enrolamiento presentaban historia de ECV mayor, cáncer activo, o tratamiento hipolipemiante. Se calculó RCV a 10 años al momento de inclusión, utilizando ecuaciones de Framingham 2008, SCORE (para poblaciones de bajo y alto riesgo), ATP III, Organización mundial de la saludregión América B (OMS-B) y Ecuación de Cohorte Agrupada (ECA). El fin de seguimiento fue 10 años ± 6 meses, ocurrencia de infarto de miocardio fatal o muerte por cualquier causa. Se utilizaron curvas ROC para evaluar discriminación (ABC > 0.75 buena discriminación). La calibración se evaluó mediante chi-cuadrado de Hosmer Lemeshow (Chi > 20 o p < 0.05 pobre calibración). Incluimos 606 pacientes, 366 mujeres, edad promedio 56.7 ± 8.4 años. Se observaron 10 (1.7%) muertes de causa no cardiovascular, 5 (0.8%) causa cardiovascular. Se registraron 58 (9.8%) ECV no fatales. Hubo aceptable discriminación para ecuaciones de Framingham, ATP-III y ECA. La calibración global solo fue buena con las ecuaciones de ATP-III y ECA. La frecuencia observada de ECV fue baja, y hubo sobreestimación de RCV con todas las ecuaciones. Sin embargo, se podría sugerir la aplicación de las ecuaciones de ATP-III o ECA en esta población.

Abstract The performance of available risk scores to predict cardiovascular risk (CVR) in the Argentinian population is unknown. Our aim was to compare the CVR predicted by several equations with the occurrence of cardiovascular events (CVE) in patients without known cardiovascular disease in an Argentinian hospital. Adults between 40 and 70 years were randomly selected, excluding those with prior history of major CVE, active cancer, lipid lowering treatment and absence of follow-up data. Framingham 2008, SCORE (low and high-risk populations), ATP III, World Health OrganizationAmerican B region (WHO-B) and Pooled Cohort equations (PC) risk scores were used to calculate 10-y CVR at time of enrollment. End of follow-up was 10 years ± 6 months, occurrence of fatal myocardial infarction or death from any cause. We used ROC curves to assess discrimination (AUC > 0.75 good discrimination), and Hosmer Lemeshow chi-square to evaluate calibration (Chi > 20 or p value < 0.05 poor calibration). We included 606 patients in our study, 336 women, average age 56.7 ± 8.4 year. Of those, 10 (1.7%) non-cardiovascular deaths, and 5 (0.8%) cardiovascular deaths were observed. 58 (9.8%) a non-fatal CVE were recorded. There was acceptable discrimination for Framingham, ATP-III, and both PC equations. The global calibration was only good with the ATP-III and PC equations. The observed frequency of CVE was low, and the CVR was overestimated by all equations. However, applying ATP-III or PC equations to assess CVR could be considered in our population.

[Performance of equations to predict cardiovascular risk in an Argentine population]. / Rendimiento de ecuaciones para predecir riesgo cardiovascular en una población argentina.

The performance of available risk scores to predict cardiovascular risk (CVR) in the Argentinian population is unknown. Our aim was to compare the CVR predicted by several equations with the occurrence of cardiovascular events (CVE) in patients without known cardiovascular disease in an Argentinian hospital. Adults between 40 and 70 years were randomly selected, excluding those with prior history of major CVE, active cancer, lipid lowering treatment and absence of follow-up data. Framingham 2008, SCORE (low and high-risk populations), ATP III, World Health Organization- American B region (WHO-B) and Pooled Cohort equations (PC) risk scores were used to calculate 10-y CVR at time of enrollment. End of follow-up was 10 years ± 6 months, occurrence of fatal myocardial infarction or death from any cause. We used ROC curves to assess discrimination (AUC > 0.75 good discrimination), and Hosmer Lemeshow chi-square to evaluate calibration (Chi > 20 or p value < 0.05 poor calibration). We included 606 patients in our study, 336 women, average age 56.7 ± 8.4 year. Of those, 10 (1.7%) non-cardiovascular deaths, and 5 (0.8%) cardiovascular deaths were observed. 58 (9.8%) a non-fatal CVE were recorded. There was acceptable discrimination for Framingham, ATP-III, and both PC equations. The global calibration was only good with the ATP-III and PC equations. The observed frequency of CVE was low, and the CVR was overestimated by all equations. However, applying ATP-III or PC equations to assess CVR could be considered in our population.

El rendimiento de las ecuaciones existentes de predicción de riesgo cardiovascular (RCV) en población argentina es desconocido. Se comparó RCV estimado por dichas ecuaciones, con la ocurrencia de eventos cardiovasculares (ECV) en una población de pacientes sin enfermedad cardiovascular de un hospital argentino. Se incluyeron aleatoriamente adultos entre 40 y 70 años, excluyéndose quienes al momento del enrolamiento presentaban historia de ECV mayor, cáncer activo, o tratamiento hipolipemiante. Se calculó RCV a 10 años al momento de inclusión, utilizando ecuaciones de Framingham 2008, SCORE (para poblaciones de bajo y alto riesgo), ATP III, Organización mundial de la salud- región América B (OMS-B) y Ecuación de Cohorte Agrupada (ECA). El fin de seguimiento fue 10 años ± 6 meses, ocurrencia de infarto de miocardio fatal o muerte por cualquier causa. Se utilizaron curvas ROC para evaluar discriminación (ABC > 0.75 buena discriminación). La calibración se evaluó mediante chi-cuadrado de Hosmer Lemeshow (Chi > 20 o p < 0.05 pobre calibración). Incluimos 606 pacientes, 366 mujeres, edad promedio 56.7 ± 8.4 años. Se observaron 10 (1.7%) muertes de causa no cardiovascular, 5 (0.8%) causa cardiovascular. Se registraron 58 (9.8%) ECV no fatales. Hubo aceptable discriminación para ecuaciones de Framingham, ATP-III y ECA. La calibración global solo fue buena con las ecuaciones de ATP-III y ECA. La frecuencia observada de ECV fue baja, y hubo sobreestimación de RCV con todas las ecuaciones. Sin embargo, se podría sugerir la aplicación de las ecuaciones de ATP-III o ECA en esta población.

Distinctive colonic mucosal cytokine signature in new-onset, untreated pediatric Crohn disease.

OBJECTIVE: The aim of the study was to compare the colonic mucosal immune response in children with new, untreated Crohn disease (CD-New), CD in remission (CD-Remission), and unaffected children (CTRL [controls]). METHODS: We performed flow cytometry of mitogen-stimulated colonic lamina propria mononuclear cells isolated from colonic biopsies and 72-hour biopsy explant cultures, and analyzed the supernatant by an unbiased multiplex cytokine array of 45 analytes. RESULTS: Thirty-six children were studied (mean age 14 ± 3 years, 14 girls): 12 CD-New, 11 CD-Remission, and 13 CTRL. We found that stimulation of lamina propria mononuclear cells isolated from colonic biopsies induced comparable intracellular cytokine levels of interferon (IFN-γ), interleukin (IL)-17, and tumor necrosis factor (TNF)-α in T cells from CD-New, CD-Remission, and CTRL, suggesting that mucosal innate inflammation plays a larger role than activated T cells in CD-New. To measure factors released during the ongoing inflammatory response in CD-New, we cultured colonic biopsy explants and uncovered 13/45 factors that were significantly higher in CD-New versus CD-Remission, whereas 10 were increased in CD-New over CTRL. Ingenuity Pathway Analysis software revealed the anticipated interconnectivity of TNF-α, IL-6, and CSF-2 in CD-New of the colon. A novel subnetwork of chemokines was, however, evident, whereas IL-17a appeared as a peripheral factor. Principal component analysis and hierarchal clustering showed that CD-New and CD-Remission separated into distinct subgroups based on the 13 factors. CONCLUSIONS: At diagnosis of inflammatory bowel disease, the colonic cytokine response contains a predominance of innate immune factors, with chemoattractants and vascular adhesion molecules playing a central role.