| Schistosomiasis japonica is one of the major infectious diseases that still threaten people's health in China.According to the report on national epidemiologic situation of Schistosomiasis japonica in 2006,there are 671,265 people infected by schistosomejaponicum(S.j),among those,28,939 are at the advanced stage of schistosomiasis,642,119 are chronic schistosomiasis and 207 are acute cases.Schistosomiasis control is still a great challenge in China.In schistosomiasis control programs,prevalence and intensity are the two basic and most important indexes to be monitored,based on which, epidemiologists evaluate the seriousness of disease,assess the effectiveness of control program,establish or adjust control strategies,and predict disease prevalence trend.Besides,these two indexes also influence government's budget for disease control programms.However,it is hard to acquire the exact prevalence data for absence of diagnostic method with high sensitivity and specificity for schistosomiasis.So far,definitive diagnosis is still made through the identification of eggs in stool for S.japonicum.But the utility and reliability of this method is largely questioned for its low sensitivity,considerable error by performers and poor compliance,especially being applied in low endemic areas where both prevalence and intensity have decreased significantly attributed to over 50 years of control activities in China.With high sensitivity and compliance,as well as its convenience to be performed,immune diagnostic methods(mainly serological methods)are being considered and gradually integrated into the national schistosomiasis control program.At present,S.japonicum specific antibody detection has been widely used in the field.Traditionally,cutoff value of specific antibody for S.japonicum diagnosis is determined using antibody level of non-infected population from unendemic areas as reference.For example,commercially available ELISA kits for S.japonicum specific antibody detection define the diagnostic cutoffs based on the average antibody level in non-infected population.It is either 2.1 times of the average,or 2 to 3 standard deviation greater than the average.Anyhow,antibody level distributions in 'active infections','cured infections' and 'noninfections' are highly overlapped due to the characteristics of the immune system as well as the heterogeneous in antibody production and clearance dynamics amoung different individuals.In order to explore appropriate system or method for schistosomiasis diagnosis and especially prevalence estimation in endemic areas,in this study,S.j soluble egg antigen specific antibody IgG and IgM were detected and analyzed for their distribution pattern in the whole population of endemic area.We established a serological model using EM algorithm to estimate the intensity at the population level.Quantitive S.j specific antibody level were obtained by indirect ELISA assay.All data analysis and model establishment were done with standardized results.Our study includes the following 3 contents:1.Establishment and proof of a novel standardization method 'I-STOD' for indirect ELISAThe results of indirect ELISA are not sufficiently consistent and thus hard to be compared on different multi-well plates.In our study,it takes at least a month to carry out the antibody detection work for the sample of large size in each study field,it is important to firstly decide an effective standardization method for ELISA results.Because the results of ELISA are affected by time, temperature,activity of enzymes,as well as reaction duration of each step, which makes ELISA results performed by different technicians in different time is not suitable to be compared directly.Though several standardization methods have been proposed and used,but method of both simple and applicable to large samples for antibody standardization is not available.We established a novel standardization method for indirect-ELISA results that we call 'improved standardization method for optical density'(I-STOD), which was derived from the reaction dynamics of the indirect-ELISA.This standardization method can transform the original OD values into the relative antibody concentrations.By using an appropriate serum as the reference on each multi-well plate to construct the I-STOD calibration curve,results on different plates can be standardized and compared.The efficiency of this standardization method was evaluated and compared with 5 other methods by results on 23 ELISA assay plates.The results revealed that calibration curve constructed by I-STOD methods fitted quite well with experimental data and its fitness is one of the best;CVs(coefficient of variation)of results for the same sample on different plates decreased significantly after standardization by I-STOD;by using quality control serum and control charts to control the assay results between normal limits,results on many plates appeared outside the normal limits,though the out of control rate of these plates fell to the smallest after standardization by I-STOD,which remarkablely improve the rate of results applicable.Therefore,I-STOD is a simple and effective standardization method for antibody level detected by indirect ELISA,standardization by which can decrease the variance between plates and improve the comparability of results on different plates.2.Evaluation of diagnostic efficiency for antibody detection by ELISA methods for Schistosomiasis japonicaOur two study fields are Jiahu country in Ruihong town,Yugan county, Shangrao city,Jiangxi Province and Mache country in Hejiashan town, Jiangshan city,Zhejiang province.Research study was carried out in year of 2005 and 2006 in the former field,which is a Schistosomiasis japonica endemic area.Rearch study was carried out in year of 2005 in the latter field, in which Schistosomiasis japonica transmission has been blocked for 10 years. Both feces and sera samples were collected from all participants aged from 5 years to 65 years in these two regions.Each participant was required to provide two stool samples on the first day and the last day of one week.Both IgG and IgM antibody level were detected by indirect ELISA in all the sera samples and all the results were transformed by I-STOD method into standardized antibody concentrations for further analysis.All infections with S.japonicum detected by Kato-Katz method in both years' studies in Jiangxi study region constitute the disease sample,while the whole population in Zhejiang study region constitutes the negative control sample.Based on these two samples,diagnostic performance of different cutoff values for IgG and IgM antibody were analyzed and compared with that of the diagnostic criterion of the kits provided by the manufacturer(diagnostic cutoff for OD values on each ELISA plate is 2.1 times of OD value of the negative control on the same plate).IgM appears to be more valuable than IgG for Schistosomiasis japonica diagnosis.Diagnostic criterion of the kits provided by the manufacturer is not good.Combined diagnosis by IgG and IgM didn't improve diagnostic efficiency.Logistic model consisting of ln(IgM)and age exhibits slightly higher diagnostic efficiency than IgM being used singly.Thus,it is hard to improve the diagnostic efficiency of S.japonicum soluble egg antigen specific antibody IgG and IgM,the sticking point of which lie in the distribution pattern of antibody level which overlaps between infections and noninfections.3.Establishment of a novel serological model for Schistosomiasis japonica prevalence estimation Stool examination in diagnosising S.japonicum infection is highly specific but poorly sensitive,which makes it inapplicable for prevalence estimation. Several models based on distribution property of S.japonicum egg in endemic population have been reported,though none has been applied to field work for prevalence estimation.Among those models,the pocket chart derived from a stochastic model of S.japonicum egg count variation developed by Yu Jinming was designed to estimate the 'true prevalence' of Schistosomiasis through observed positive rate and infection intensity by single stool examination(2 kato slice).We evaluated the value of this pocket chart by using data obtained in our study.The results revealed that this pocket chart needs to be restudied and validated before being used in field work,because the 'true prevalence' obtained from the chart is not constant between studies carried out in the same population in different days within a week and is equivalent to the positive rate of twice stool examination(3 kato slice for each stool sample).S.japonicum specific IgG and IgM antibody level distributions of the whole population in Jiangxi study focus are both mixture models,consisting of component distributions of different subpopulations,i.e.current infections, past infections and noninfections,etc..To explore and identify the latent component distributions by analyzing the mixture model may serve as a more scientific and more intuitionistic method for prevalence estimation.Both IgG and IgM distributions of different populations in our study foci (population in Jiangxi focus,population in Zhejiang focus,population of S. japonicum infections detected by stool examination in 2005 and 2006)are all badly skewed.In order to analyze these distributions by theories of normal distribution,a natural logarithm transformation was done to all the original IgG and IgM concentration values.After transformation,IgG and IgM distributions of S.japonicum infections are nearly normal distributions.Both IgG and IgM distributions of Zhejiang population converged well,though slightly skewed.It is noted that IgG and IgM distributions of the whole population in Jiangxi focus are obviously not normal distributions,more subjects appearing on the right side of the main apex,indicating several component distributions of populations with different infection status in this mixture model.Analysis for S.japonicum specific antibody level distribution in S. japonicum infections detected by twice stool examination(3 kato slice for each stool sample)revealed a weak positive correlation between antibody level and infenction intensity(EPG)(the correlation coefficient is 0.3300 between IgG and EPG;is 0.2966 between IgM and EPG)and a shift of antibody level distribution towards lower level concomitant with the decrease of positive stool samples examined by Kato-Katz method as well as the decrease of positive kato slices,indicating that S.japonicum specific antibody level in at least some of the infections misdiagnosed by twice stool examination(3 kato slice for each stool sample)is lower than these infections detected by twice stool examination.Thus,we postulated that there is another lower antibody distribution for infections.Therefore,we made an assumption that both S.japonicum soluble egg antigen specific IgG and IgM antibody level mixture distribution models of the whole population in Jiangxi study focus consist of antibody distributions of 4 component normal distributions,among those,the distribution with lowest average antibody level includes noninfections and past infections with very low level of S.j specific antibody,the distribution with the second lowest average antibody level represents past infecitions,then the two distributions with high antibody levels both represent active infections,among them,the distribution with highest antibody level represents infections who are inclined to be detected by twice stool examination(3 Kato slice for each stool)and the other represents infections who are not prone be detected by twice stool examination(3 Kato slice for each stool).Both IgG and IgM antibody mixture distribution models are fitted by 4 component normal distributions using EM algorithm and prevalence of Schistosomiasis japonica was calculated accordingly.By IgM mixture models,prevalence of Schistosomiasis japonica in 2005 and 2006 in this region is 42.96%and 44.73%,respectively.By IgG mixture models,the prevalence in 2005 and 2006 is 37.56%and 42.45%, respectively.Besides,primary evaluation of these mixture models for prevalence estimation suggested that IgG mixture model failed to distinguish sub-populations with different infection status successfully,while IgM mixture model appears to possess the potential for prevalence estimation in Schistosomiasis japonica endemic areas.In summary,according to analysis on field data consisting results of both stool and serological examination,it is hard to obtain accurate prevalence of schistosomiasis by Kato-katz technique either by results of stool examination directly or using pocket chart of the stochastic model.Though specific antibody detection has some value in diagnosing Schistosomiasis japonica,S. japonicum specific antibody IgM antibody level mixture distribution model has the potential for more accurate prevalence estimation in Schistosomiasis japonica endemic areas.
|
PDF Full Text Request