The Evaluation Of HIV Antibodies Laboratory Testing Algorithm

BackgroundAIDS is an infectious diseases caused by the Human Immunodeficiency Virus with high infection rate and huge disease burden for patients. So far there is no effective drug which can cure the disease. Since America’s first AIDS patients was diagnosed in 1981 , HIV/AIDS infection rates has growing fast. There had reported 35 million people living with HIV,1.5 million people died of AIDS in 2013. Reportted by the end of October 2015, the existing patients living with HIV/AIDS virus were 575000 in China, accounting for 0.06% of the total population, namely every 10000 people there may be six people infected with the AIDS virus. To make a HIV laboratory testing strategy suitable for China’s national conditions is very important for the AIDS epidemic monitoring, diagnosis and treatment in order to effectively control the spread of AIDS.Currently, there are many methods used for the detection of HIV antibody, antigen and virus nucleic acid. The methods to detect HIV antibodies such as enzyme linked immunosorbent assay, rapid test, Western Blot, Recombinant Immunoblot assay, Chemiluminescence or immunofluorescence test and gelatin particle agglutination test. Enzyme linked immunosorbent assay also used to test P24 antigen. The methods to detect the HIV virus nucleic acid are Polymerase Chain Reaction and nucleic acid sequence amplification experiment, branch of DNA amplification, etc. Due to the sensitivity and specificity of detection methods are different, we need to combine different tests to achieve the optimal detection result.Objective1. To evaluate the quality of three kinds of HIV antibody assay diagnostic tests in finding HIV infection.2. Putting forward a series of HIV antibodies laboratory testing algorithms for different crowds.MethodsThis study was divided into two parts, the first part was to evaluate the quality of three kinds of HIV antibody assay diagnostic kits (HIV rapid testing methods, enzyme-linked immune detection methods, western blot detection methods) in finding HIV infection by using six kinds of blood panels (master panel, seroconversion panels, cross-reacting panels, Linear dilution series panels, precision quality product panels, College of American Pathologists proficiency testing panels). The evaluation index were sensitivity and specificity of diagnostic assays, analysis sensitivity and pecificity, and precision of those assays.The possible combination of HIV antibodies laboratory testing algorithms for the second part were proposed according to the first part of the assessment results. UNAIDS put forward the latest generation of ELISA and rapid test as reliable as WB for confirmation.Meanwhile, the HIV repaid test and enzyme-linked immune experiment were used widely. There are many advantages such as low requirements for personal skills and laboratory facilities, the prices are low. So we combine ELISA and RT for the second part evaluating those test strategies.ObjectsFour HIV ELISA test kits, two HIV rapid test kits and two HIV West Blot test kits were evaluated by six serum panels including master panel, seroconversion panels, cross-reacting panels, linear dilution series panels, precision quality product panels, College of American Pathologists proficiency testing panels.We selected HIV and HCV sentinel samples in different areas for strategy assessment, a total of 991 samples, which were 184 samples of MSM crowd in Gansu province,165 samples of drug users in Yunnan,178 samples of Anhui HCV antibody positive crowd, 398 samples of Fujian VCT crowd,66 samples of 10 recent infection patients.ResultsPart 1. The methodology evaluation of HIV antibody diagnostic tests1. The sensitivity of the six kits were all 40/40, and the specificity were between 39/40-40/40 evaluated by the master panel.2. The intraplate precision value of four ELISA assays was between 4.40%-24.98%. Quality product with high titer antibodies were between 6.27%-12.89%, the median titers were between 4.40%-11.29%, the low titers were between 4.74%-24.98%.3. In the seroconversion panels, the ELISA and RT tests were 7 to 11.8 days latter to diagnose HIV compared with nucleic acid testing, HIV-1/2 antigen/antibody combination immunoassay was the shortest(7days), WB test was the longest(11.8days).4. In the seroconversion panels, WB find 3 positive samples,14 uncertain samples and 49 negative samples,while RIB A find 11 positive samples,10 uncertain samples and 45 negative samples.We did matching X2 test for RIBA and WB, the results had statistically significant difference. The Kappa value was 0.70.5. The analytical specificity of the six kits were 40/40, evaluated by the cross-reacting panels.Part 2. The evaluation of HIV antibodies laboratory testing algorithms1. The single assay testing algorithms.991 samples were tested by five single assays which were RT1 (rapid test), RT2 (rapid test), RT3 (rapid test), ELISA3 (ELISA, HIV-1/2 antigen diagnostic assay), ELISA4 (ELISA, HIV-1/2 antigen/antibody combination immunoassay).(1) The sensitivity of the test results for the total goup was between 96.55%～99.31%, the specificity was between 98.84%-99.87%, and the coincidence rate was between 98.59%-99.35%. ELISA4 had the highest sensitivity (99.31%), RT3 had the highest specificity (99.87%).(2) The crowds (voluntary counseling and testing group, drug users group, hepatitis c antibody positive patients group, MSM crowd) count respectively, the sensitivity was between 80%-100%(4/5), the specificity was between 97.69%-100%, and the coincidence rate was between 97.58%-100%.2. Elisa for screening, rapid test for confirmation testing algorithmsELISA4�'T1, ELISA4 �'T2, ELISA4 �' RT3 testing algorithms were used respectively to test the samples.(1) The sensitivity of the test results for the total goup was between 96.55%-97.24%, the specificity was 99.87%, the coincidence rate was 99.46%.(2) The crowds (voluntary counseling and testing group, drug users group, hepatitis c antibody positive patients group, MSM crowd) count respectively, the sensitivity was between 80%-100%(4/5), the specificity was between 99.72%-100%, the coincidence rate was between 98.18%-100%.3. Rapid test for screening, elisa for confirmation testing algorithmsRT1�'ELISA4, RT1�'ELISA3 testing algorithms were used respectively to test the samples.(1) The sensitivity of the RT1 �' ELISA4 test results for the total goup was 97.24%, the specificity was 99.87%, the coincidence rate was 99.46%. The sensitivity of the RT1�'ELISA3 test results for the total goup was 96.55%, the specificity was 100%, the coincidence rate was 99.46%.(2) The crowds (voluntary counseling and testing group, drug users group, hepatitis c antibody positive patients group, MSM crowd) count respectively, the sensitivity of the RT1�'LISA4 test results was between 94.87%～100%, the specificity was between 99.72%-100%, the coincidence rate was between 98.79%-98.79%. the sensitivity of the RT1�'ELISA3 test results was between 94.87%～100%, the specificity was 100%, the coincidence rate was between 98.18%～100%.4. Rapid test for screening, another rapid test for confirmation testing algorithmsRT1�'RT2、RT1�'RT3 testing algorithms were used respectively to test the samples.(1) Two kinds of combination in the total group had the same sensitivity, specificity, and coincidence rate which were 96.55%,100% and 99.46%.(2) The crowds (voluntary counseling and testing group, drug users group, hepatitis c antibody positive patients group, MSM crowd) count respectively, the sensitivity was between 80%～100%(4/5), the specificity was 100%, and the coincidence rate was between 98.18%-100%.5. Elisa for screening, another Elisa for confirmation testing algorithmsThe ELISA4�' ELISA3 testing algorithms were used respectively to test the samples. The sensitivity of the ELISA4�'ELISA3 test results for the total goup was 96.55%, the specificity was99.74%, the coincidence rate was 99.24%.The crowds (voluntary counseling and testing group, drug users group, hepatitis c antibody positive patients group, MSM crowd) count respectively, the sensitivity was between 94.87%-100%, the specificity was between 99.42%-100%, and the coincidence rate was between 98.18%-100%.6. Three rapid test testing algorithmThe RT1�'RT2�'RT3 testing algorithm were used respectively to test the samples. The sensitivity of the RT1 �' RT2�'RT3 test results for the total goup was 95.86%, the specificity was 100%, the coincidence rate was 99.35%. The crowds (voluntary counseling and testing group, drug users group, hepatitis c antibody positive patients group, MSM crowd) count respectively, the sensitivity was between 80%(4/5)～100%, the specificity was 100%, the coincidence rate was 98.18%～100%.7. The recent infection samples testing algorithmsThe acute infection sample detected respectively by RT1, RT2, RT3, ELISA3, ELISA4. ELISA4 detected 22 positive samples, which had the highest sensitivity. ELISA4�'WB confirmed only three positive samples and 14 uncertain samples.While the ELISA4 �' PCR testing algorithm find 22 positive samples.Conclusions1. The sensitivity and specificity of the kits were both high. The antigen/antibody combination immunoassay sensitivity was relatively high, better for antibody screening.2. The precision of ELISA detections had large difference, especially for the low titers samples, may affect the results.3. Compared with West Blot, RIB A was better to detect the early HIV infection, uncertain rate of RIBA was low which reduced the follow-up rate caused by the uncertainty.4. There were false negative and false positive results for the single detection algorithms.5. The two detection algorithms’ results had increased the specificity to about 100%, but there had the false negative in voluntary consultation test samples.6. Compared with two detection algorithms, the three kinds of rapid test algorithm had no differences on specificity and sensitivity.7. Recent infection samples can use ELISA antigen and antibody detection assay combination nucleic acid detection, in order to detect more patients and shorten the window period.