Biochemical Alterations as Markers of Dengue Hemorrhagic Fever Luis Angel Villar-Centeno,* Fredi Alexander Díaz-Quijano, and Ruth Aralí Martínez-Vega School of Medicine, and Centro de Investigaciones Epidemiológicas, Universidad Industrial de Santander, Bucaramanga, Colombia Abstract. We evaluated biochemical alterations as predictors of dengue hemorrhagic fever (DHF). Patients with confirmed infection with dengue virus were prospectively evaluated for the first seven days of disease to determine their final clinical outcome. Serum samples taken 48–96 hours after onset of fever were used for biochemical tests. Of 199 patients, 30 developed DHF. Cases of DHF had higher levels of lactate dehydrogenase (LDH), creatine kinase (CK), and aspartate aminotransferase, and lower levels of albumin, total cholesterol, and triglycerides. Multivariate analysis showed that early alterations of CK (hazard ratio [HR]  6.98, 95% confidence interval [CI]  2.34–20.85, P  0.001), LDH (HR  3.19, 95% CI  1.01–10.12, P < 0.05), and albumin (HR  2.54, 95% CI  1.09–5.92, P  0.03) were associated with DHF. Triglyceride levels > 160 mg/dL were negatively associated with developing DHF (HR  0.07, 95% CI  0.01–0.59, P  0.01). Early alterations of biochemical markers can predict DHF in patients with acute fever caused by dengue. INTRODUCTION Dengue is the arboviral infection with the largest incidence worldwide. 1 Clinical expression of dengue virus infection var- ies widely from no symptoms to dengue shock syndrome, but this infection more commonly causes dengue fever (DF) and dengue hemorrhagic fever (DHF). 1,2 Nearly 100 million cases of DF and between 250,000 and 500,000 cases of DHF are annually reported to the World Health Organization. 2,3 The most dramatic increase in dengue in the past decade occurred in South America, mainly in Colombia, Ecuador, Paraguay, Peru, Venezuela, and Brazil. 2–5 In Colombia, den- gue infection is endemic with cyclical outbreaks in almost all cities less than 1,800 meters above sea level with an at-risk population of 20 million persons. 5 Dengue hemorrhagic fever is characterized by thrombocy- topenia, spontaneous hemorrhages, and gradual plasma leak- age that can lead to shock. 1,2 Despite its clinical variability, the acute phase of dengue begins with fever that is indistin- guishable from the initial phase of other acute febrile infec- tious diseases. 1,2,6 Thus, acute dengue infection is often un- recognized until the appearance of the more severe forms of the disease. This observation leads to underestimation of the actual incidence, as well as inadequate or late treatment of a disabling and potentially lethal medical condition. 7 There is direct and indirect evidence of biochemical alter- ations related to severity of dengue. Studies have reported that patients with DHF have elevated serum levels of trans- aminases (aspartate aminotransferase [AST] and alanine ami- notransferase [ALT]), 6–11 amylase, 11–13 lactate dehydroge- nase (LDH), 10,13–15 and creatine kinase (CK). 16 Patients with DHF also have elevated levels of phospholipase A2, a protein whose concentration is correlated with that of C-reactive pro- tein (CRP). 17,18 Cross-sectional studies have shown differ- ences in serum levels of cholesterol and triglycerides associ- ated with severe forms of DHF studies. 11,12 However, these potential biochemical markers have not been evaluated pro- spectively in early stages of dengue. Thus, the utility of bio- chemical alterations for timely identification of patients who will develop DHF is unknown. 19 The objective of this study was to evaluate biochemical markers as predictors of dengue severity. MATERIALS AND METHODS Setting. This prospective cohort study was conducted in metropolitan Bucaramanga, which includes four municipali- ties in the Department of Santander in northeastern Colom- bia. The incidence of dengue in the area has increased in recent years, ranging between 113.4 and 268.7 cases per 100,000 persons. 5,20 The study protocol was reviewed and ap- proved by the Medical Ethics Committee of the Universidad Industrial de Santander. Study participants were more than five years of age and had acute febrile syndrome caused by dengue with a duration of symptoms less than 96 hours. After a physical examination was conducted and informed consent was obtained, blood sample were obtained to determine albumin levels, hemato- crits, and platelet counts. Patients with any of the following conditions were excluded: history of concomitant diseases such as diabetes, acquired immunodeficiency syndrome, he- matologic disorders, cancer, or cardiac disease, or, at baseline, DHF (see case definition), major bleeding albuminemia (< 3 g/dL), effusions, or shock. All patients were enrolled before the development of DHF. Participants were followed daily until the seventh day of disease. Data collected included signs and symptoms and daily microhematocrit determinations to facilitate recognition of DHF. We conducted an IgM enzyme-linked immunosor- bent assay, viral isolation, and biochemical tests 48–96 hours after onset of fever. Platelet counts were repeated daily in cases with previous counts less than 120,000/mm 3 or when the patient had spontaneous hemorrhages, signs of effusion, edema, or a change in the hematocrit greater than 10%. A convalescent blood sample was obtained 7–15 days after the onset of symptoms to detect IgM antibodies to dengue. Study definitions. We defined dengue virus infection as any of the following: viral isolation; a shift from a negative to a positive IgM test result, or a  4-fold increase in previously existing levels of antibodies to dengue virus. Patients with negative results in convalescent IgM tests were considered to have other febrile (non-dengue) illnesses. The DHF cases had a dengue infection and satisfied all of the following criteria: a platelet count  100,000/mm 3 , any * Address correspondence to Luis Angel Villar-Centeno, Centro de Investigaciones Epidemiológicas, Facultad de Salud, Universidad In- dustrial de Santander, Carrera 32 No. 29–31, Bucaramanga, Colom- bia. E-mail: luisangelvillarc@yahoo.com Am. J. Trop. Med. Hyg., 78(3), 2008, pp. 370–374 Copyright © 2008 by The American Society of Tropical Medicine and Hygiene 370