Are there any predictive factors for successful testicular sperm recovery in azoospermic patients?

Recovery of testicular spermatozoa from azoospermic patients with testicular failure followed by intracytoplasmic sperm injection (ICSI) is a recent advance in the treatment of male infertility. This study aimed at investigating which parameter(s) may predict succesful testicular sperm recovery. We reviewed 395 testicular sperm recovery procedures and analysed the most frequently available parameters for clinical decision-making in azoospermic patients: (i) presence of at least one single spermatozoon in at least one preliminary semen analysis; (ii) maximum testicular volume; (iii) serum follicle stimulating hormone (FSH); and (iv) presence of spermatozoa in the histology of a randomly-taken testicular biopsy. Sensitivity, specificity, positive and negative predictive value, positive and negative likelihood ratio and accuracy were calculated for the above index parameters in different clinically relevant subgroups using receiver operating characteristic (ROC) curves whenever possible. Spermatozoa were always successfully recovered in patients with normal testicular histological findings (n = 173) or hypospermatogenesis (n = 16) but not in some patients with tubular sclerosis (seven out of 18), Sertoli cell-only pattern (55 out of 112) or maturation arrest (39 out of 76). Histopathology was the best test for predicting successful sperm recovery in the whole population (sensitivity: 86%, specificity: 93%, accuracy: 0.87). In patients with secretory azoospermia, histopathology was again the most accurate parameter (accuracy : 0.74), especially in patients showing Sertoli cell-only pattern (accuracy: 0.83) but not in patients showing maturation arrest (accuracy: 0.55). In patients with serum FSH concentrations >12 IU/I and maximum testicular volume <15 ml, histopathology was not found to be accurate. Semen analysis, maximum testicular volume and serum FSH were not highly predictive in all subgroups studied. Our analysis shows that no strong predictors for successful testicular sperm recovery are available except for testicular histopathology.