The Predictive Value Of Anti-müllerian Hormone On The Outcome Of Sperm Retrieval In Non-obstructive Azoospermia Patients: Diagnostic Accuracy Meta-analysis

Objective:To systematically evaluate the predictive effect of anti-Müllerian hormone（AMH）,as a potential spermatogenic marker,on the outcome of surgical sperm retrieval and its diagnostic value for spermatogenesis in patients with non-obstructive azoospermia（NOA）.Methods:Two investigators independently and systematically retrieval Pub Med,EMBASE,Web of Science,Cochrane Library,CNKI,Wan Fang Database,Chinese Science Journal Database and China Biomedical Database according to the pre-established search strategy and collect the relevant diagnostic accuracy test from inception to January 18^th,2021,the search language is limited to Chinese or English.Obtain target literature by following the inclusion/exclusion criteria,according to the data in the text directly or indirectly calculated to form four grid tables of true positives（TPs）,false positives（FPs）,false negatives（FNs）and true negatives（TNs）.The literature quality evaluated by the Quality Assessment of Diagnostic Accuracy Studies-2（QUADAS-2）and the methodological quality graphs and quality summaries produced by the Review Manager 5.3 software.Perform statistical analysis and draw a summary receiver operating characteristic curve（SROC）and forest plot by the MIDAS module in Stata MP14.0 software.The Spearman correlation coefficient was calculated by Meta-Disc 1.4 software then to detect the threshold effect.The Cochran’s Q test and Higgin’s I² statistic were used to detect heterogeneity.Sensitivity（SEN）,specificity（SPE）,positive likelihood ratio（PLR）,negative likelihood ratio（NLR）,area under curve（AUC）were used to evaluate the predictive accuracy of anti-Müllerian hormone.Perform subgroup analysis and univariate Meta-regression analysis on heterogeneity and covariates include race,country,research sample size,surgical procedure,best-cutoff value,etc.Fagan’s nomogram calculates the positive post-test probability and the negative post-test probability.Deeks’funnel plot asymmetry test is used to assess publication bias.Results:Nine articles containing 333 patients were included.The pooled sensitivity,specificity and area under the curve were 0.79[95%CI（0.70,0.87）],0.74[95%CI（0.64,0.82）]and 0.84[95%CI（0.80,0.87）],the positive post-test probability and the negative post-test probability is 73%and 20%,respectively.The sensitivity,specificity and area under the curve of the seminal plasma group were 0.76[95%CI（0.65,0.85）],0.77[95%CI（0.64,0.86）]and 0.81[95%CI（0.78,0.84）],the positive post-test probability and the negative post-test probability is 75%and 21%,respectively.Overall,no threshold effect and heterogeneity detected in the included studies.The diagnostic accuracy parameters have varying degrees of heterogeneity among different studies.The results of heterogeneity analysis showed that race,data source and sample size had a statistically significant influence on specificity（P<0.01）.No statistical evidence supporting publication bias.The serum group did not undergo Meta-analysis due to the insufficient number of studies.Conclusions:Anti-Müllerian hormone has moderate predictive accuracy for the outcome of surgical sperm retrieval in patients with non-obstructive azoospermia,and helpful to evaluate whether there is local spermatogenesis in the testicles,diagnostic specificity is higher in non-Caucasian populations.Due to the small number of related studies and the lack of a clear diagnostic threshold,only recommended to use anti-Müllerian hormone as an additional laboratory data to assist in the evaluation.