Κλινικοεργαστηριακή διερεύνηση της συστηματικής καντιτίασης σε τριτοβάθμιο νοσοκομείο και η αξιολόγηση μιας οροδιαγνωστικής μεθόδου για την έγκαιρη διάγνωσή της

Abstract

TitleClinical investigation and evaluation of a serological Candida antigen detection method inearly diagnosis of systemic Candidiasis, in a tertiary hospital.IntroductionSystemic candidiasis (SC) is associated with high morbidity and mortality between theimmunocompromised patients. The early diagnosis of systemic candidiasis remainsproblematic and the timing of antifungal therapy is critical. Since the conventional diagnosticmethods lack of sensitivity and positive blood cultures take a long period turn into positive,alternative non-culture techniques have been developed.Purpose of the studyThe purpose of the study is to assess the predictive value of a serological antigen detection testfor the diagnosis of SC, using the Candida Detection Cand-Tec* latex agglutination System .Material and methodThe study was prospective and 100 patients, hospitalized in a tertiary hospital, with Proven,Probable and Possible invasive candidiasis were enrolled in; 324 serum samples werecollected in total, 131 serum samples from 39 patients with proven and probable systemiccandidiasis (group A and B) and 193 from 61 patients with possible systemic candidiasis(group C). Besides, 100 healthy controls were enrolled in the study (group D).For the purpose of the study , the classification of the patients in proven, probable or possibleSC was based on the criteria of the original definitions of systemic fungal infections of theEORTC / MSG . The isolation and the identification of the candida spp, was performed bystandard techniques and the Candida antigens were detected in serum using the CandidaDetection Cand-Tec latex agglutination System (Ramco Lab., Inc. , Houston, Texas). Theassay was performed according to the manufacturer. Antigen titers of >=1:4 were correlatedwith SC.ResultsOverall, 100 patients were collected 324 serum samples, 131 of the 39 patients in groups A (7)and B (32) and 193 of the 61 patients of group C and 102 serum samples from 100 healthycontrols (group D) .Group A, patients with Proven SC (7 ): 2 patients (29%), with negative or marginal result<= ½, 4 patients (57%) with positive titer 1:4 and 1 patient (14%) with positive titer 1:8Group B, patients with Probable SC (32 )14 patients (44%) with negative or marginal titer <= ½, 11 patients (34%) with positive titer1:4 and 7 patients (22%) with titer 1:8.Group C, patients with Possible SC (61)45 patients (74%) with negative or marginal titer <= ½, 10 patients (16%) with positive titer1:4, 5 patients (8%) with titer 1:8, and a positive titer 1:16 in 1 patient (2%).Group D (healthy controls) (100): Two serum samples with borderline positive 1:2 wereconsidered to be negative after repeated the blood tests the week that followed. Conclusions1. The method presents a moderate sensitivity (56-72%) with an average of 64% andhigh specificity of 83% (76-89%) .2. The positive predictive value of the method, for marginal titles 1:2 is low to 26%(19-33%). However, for higher titers (> = 1:4) the positive predictive valueincreases to 56% (48-64%).3. The negative predictive value of the method for all the titles is very high andamounts to 86% (80-92%).4. The probability of finding a positive antigen titer ( > = 1:4 ) increases significantlyin cases of prior abdominal surgery or when the patient had passed throughmultiple colonization.5. Out of all the patients with proven , probable and possible SC, prior chemotherapyincreases the probability of finding a positive titer antigen by four times, previousabdominal surgery by six times , while the multiple colonization further increasesthe possibility of positive results also.6. Τhe results of our study regarding the sensitivity and specificity of the method areconsistent with the international literature.7. The implementation of this serologic method in the diagnosis of SC remainslimited, due to its low sensitivity. Ηowever, due to its high negative predictivevalue, the use of the method remains helpful for the exclusion of the SC.