Objective      To investigate whether reproductive hormone levels

testicular volume and other parameters can predict the existence of sperm in the patients with non-obstructive azoospermia (NOA)

and then be used as cutoff-point for the outcome of testicular sperm aspiration (TESA) so as to provid

e important clinical information of unnecessary surgical trauma.       Methods      A total of 133 NOA patients admitted in our outpatient center from January 2013 to January 2014 were enrolled in this study. They were divided into 2 group according to the results of TESA

that is

sperm group and non-sperm group. Reproductive hormone levels

testicular volume

seminal fructose

neutral-ɑ glucosidase

chromosome

azoospermia factor (AZF) were detected

and the results were compared between the 2 groups.       Results      There were significant differences in testis volume (left: 9.12±2.25 vs 6.25±2.63 mL

right: 9.23±2.02 vs 6.47±2.48 mL)

follicle-stimulating hormone (FSH

9.82±2.09 vs 22.62±15.00 mIU/mL)

total testosterone (TT

5.19±0.68 vs 4.15±1.82 ng/mL)

ratio of T/luteotropic hormone (T/LH

0.61±0.19 vs 0.57±0.35)

and ratio of T/estrogen (T/E

2

162.40±16.66 vs 141.40±93.40) between the sperm group and non-sperm group (all P0.01). The FSH level was significantly higher in the non-sperm group than the sperm group. When the cutoff-point of preferred FSH level was 13.31 mIU/mL

the sensitivity was 74.1% and specificity of96.2% when compared with the results by TESA

and the area under curve (AUC) of FSH receiver operating characteristic (ROC) curve was 0.850

indicating that the diagnostic accuracy was moderate.       Conclusion        When a single factor is concerned

FSH level is essential to predict whether there is spermatogenesis in the testis of NOA patients

and its value is superior to testicular volume. However

combination of the 2 factors is more significant in the prediction.