Međudjelovanje dijagnostičkih biljega endemske nefropatije [Interaction of markers endemic nephropathy]

Endemic nephropathy (EN) is a chronic tubulointerstitial disease. Different hypothesis and possible etiologic agents for development of EN have been investigated. Recent results from Croatia suggest that aristolochic acid (AA) is a major risk factor for EN. According to Ivic hypothesis, the most likely route of human exposure to AA is via-ingestion of home-baked bread prepared from flour contaminated by seeds of Aristolochia clematitis. One of the specifics of EN is that different countries use different diagnostic criteria. There is no specific diagnostic marker for EN. EN is a rare disease considering its prevalence and there is a problem of questionable diagnostics. The postulate for standardization of diagnostic markers is the analysis of its relationships. In 2005. a field survey was conducted in 3 endemic villages and 1 control village, which comprised 1081 examinees. They respondents were given the questionnaire. The clinical examination and anamnesis were done also. According to WHO criteria, the examinees were classified in 4 groups: diseased, suspected, at the risk and the others. The diagnosis of EN was confirmed in 23 examinees after surveillance from 2005-2010. Demographic and epidemiological characteristics of examinees, considering their exposure to AA, as well as identification of possible markers of EN, were analyzed by logistical regression (odds ratio) and by their 95% interval of reliability between positive and negative groups. The negative group was the one with examinees without EN. The positive group was defined by 2 models. According to the 1st model, the diseased and the suspected were in positive group, and according to the 2nd model, the diseased, the suspected and those under the risk were in positive group. Multidimensional relations among diagnostic markers were analyzed by CoPlot method. The ratio between the diseased and the others, the diseased and those under the risk, the diseased and suspected was demonstrated. In those CoPlot maps, WHO criteria were used (7 variables) and all available epidemiological and clinical-diagnostic criteria (18 variables). The prevalence of the diseased in the examined specimen is 1,32%, suspected – 3,6%, those at the risk – 8,88%. The results show that, considering the exposure to AA, the chance (probability) that examinees encountered AA on their fields 20-30 years ago was 2,5x (1st model), and 2x (2nd model) higher among examinees in positive group than in negative group. According to the 1st model, the chance that they sometimes encountered AA on their fields 20-30 years ago was 2x higher among examinees in positive group than in negative group. The chance that they always encountered AA seeds amongst the seeds intended for flour 20-30 years ago, was 2,6x (1st model) and 2x (2and model) higher among examinees in positive group than in negative group. CoPlot map shows the accepted goodness of fit measure tresholds in both reports (with exception of the diseased and suspected with 18 variables). This method identifies the examinees which with their characteristics, contrast the group they belong to. The best correlation in CoPlot presentation with WHO criteria, have alfa1microglobulin and creatinin, and the worst – age (CoPlot solution of the diseased and the others and the diseased and at risk) and positive family anamnesis (CoPlot solution of the diseased and suspected). The best correlation in CoPlot presentation with 18 variables have hemoglobin, hematocrit and eritrocites. The worst correlation in CoPlot solution have the diseased and the others – gender and systolic pressure, the diseased and those at the risk – systolic pressure and ratio of alfa1 microglobulinuria and albuminuria, the diseased and suspected – family anamnesis and gender. The following variables show redundancy: time spent in the family, in the village, in the region and age. They are negatively correlated with positive family anamnesis, which points the fact that certain genetic predisposition could implicate shorter time from exposure to the diseases. Eritrocites, hematocrit and hemoglobin also show redundancy. In CoPlot maps which show the diseased and the others, alfa1 microglobulinuria, creatinin, albuminuria, proteinuria and family anamnesis of EN show positive correlation with the diseased, hemoglobin, hematocrit, erotrocites, glomerular filtration calculated by Cockroft-Gault and MDRD formulas, systolic and diastolic pressure show negative correlation, and ratio of alfa1 microglobulinuria and albuminuria, time spent in the family, in the region, in the village, age and gender show dis-correlation. In CoPlot solution of the diseased and those at the risk, the diseased and suspected, alfa1 microglobulinuria, albuminuria, proteinuria and creatinin show positive correlation with the diseased. In all 3 presentations, time spent in the families, in the region, in the village, age, gender, alfa1 microglobulin/albumin are not correlated neither with the diseased nor with the rest of the groups according to WHO criteria. The 5 main diagnostic markers are: positive family anamnesis, glomerular filtration calculated by Cockroft-Gault, proteinuria, hematocrit i alfa1 microglobulinuria. Considering that hematocrit, hemoglobin, hematocrit and eritrocites give redundant information we can use one of them. According to WHO criteria glomerular filtration estimated by Cockroft-Gault formula has higher correlation then creatinin. These results show that CoPlot methodology for multidimensional relations is adequate for the analysis of rare diseases in heterogeneous groups, when it is not possible to have a large number of the examinees.