Parâmetros nutricionais como marcadores prognósticos pós-operatórios no câncer de cabeça e pescoço

Abstract

Changes in nutritional status manifestthemselves in allkinds of cancer, its intensity varying according to the type as well as the location of the tumor. In head and neck cancer (HNC) the incidence of malnutrition amounts to 72%. Studies have been demonstrating the association between that clinical condition and postoperative complications and survival. That situation demonstrates the need to identify, among the available nutritional assessment methods, which is mostly associated with the prognosis of patients with HNC. Accordingly, this study aims to compare nutritional parameters as possible prognostic markers in recent postoperative periodsof HNC patients submitted to surgery as an initial treatment strategy.The research was a prospective cohort study in which patients undergone a preoperative nutritional assessment and later monitored from the surgical procedure up until the hospital discharge.The comparison of variables was accomplished with the use of the Mann-Whitney test and the χ2test whereas the identification and the assessment of predictive properties were accomplished through logistic regression analysis and the ROC curve, respectively.30 HNC patients who were candidates for surgeries associated with unilateral cervical emptying as the initial treatmentoption were included in the study. The sample was comprised of individuals with the following characteristics: mean age of 64.25 years, males (90%), low schooling (90%), low family income (≤1 minimum wage), alcoholics (86.66%), and smokers (90%). We found that the prevalence of serious postoperative complications amounted to 30%. The tricipital skinfold (TSF), the ideal weight percentage (IWP), the corrected arm muscle area (CAMA), the muscular arm circumference (MAC), the Malnutrition Universal Screening Tool (MUST), and the serum albumin level (SAL), all of themmeasured on the firstpostoperative day, demonstrated the predictive role of postoperative complications in the logistic regression model. On the first postoperative day, the ROC curve displayed an accuracy of75% regarding the TSF indicator (≤8mm) [95% CI: 0.56-0.94; p = 0.031] and 88% with respect to the albumin level (≤2.8g/dl) [95% CI: 066 - 0.99; p = 0.00] so as to predictpostoperative complications in patients during the hospitalization period.