Optimization of spermatogenesis-regulating hormones in patients with non-obstructive azoospermia and its impact on sperm retrieval: a multicentre study


Alayman Hussein, Urology Department, Minia University Hospital, Kornish El-Nile Street, El-Minia 61111, Egypt. e-mail: alaymanh@hotmail.com


Study Type – Therapy (outcomes)

Level of Evidence 2a

What's known on the subject? and What does the study add?

Clomiphene citrate, hCG and human menopausal gonadotropin (hMG) are widely used in treatment of oligospermia, because they increase FSH and testosterone which are essential for spermatogenesis. Finding a sperm in non-obstructive azoospermia for intracytoplasmic sperm injection is a challenge and much effort is required to reach the optimum method of sperm retrieval.

The study shows that a new protocol of clomiphene citrate, hCG and hMG in the treatment of non-obstructive azoospermia achieves an increase in the levels of FSH, LH and total testosterone to the target levels that we set. Our target level of FSH was 1.5 times its initial level and for serum testosterone it was 600–800 ng/dL. Using our described medical treatment protocol in cases of non-obstructive azoospermia, sperm may be found in patients' ejaculate (∼11%) and if they remain azoospermic they will have a greater likelihood of sperms being obtained in testicular sperm extraction.


  • • To evaluate the effect of optimizing serum level of follicle-stimulating hormone (FSH), luteinizing hormone (LH) and testosterone on sperm retrieval for intracytoplasmic sperm injection.


  • • A total of 612 patients with non-obstructive azoospermia were evaluated with routine history, physical examination and hormonal assessment.
  • • Of these, 116 patients underwent microsurgical (micro)-testicular sperm extraction (TESE) without any medical treatment and formed the control group and the remaining 496 patients were administered clomiphene citrate in a titrated dose.
  • • Patients were classified into four groups according to their response to clomiphene citrate. Group 1: patients with an obvious increase in FSH and total testosterone (n= 372). Group 2: patients showing an increase in FSH with no or little increase in LH and total testosterone (n= 62). For these patients we continued with clomiphene citrate and added human chorionic gonadotrophin (hCG). Group 3: patients with no increase in the levels of the three hormones (n= 46). Group 4: included patients with continuously decreasing serum testosterone levels in response to the increasing dose of clomiphene citrate (n= 16). Accordingly, patients in groups 3 and 4 discontinued clomiphene citrate and started hCG and human menopausal gonadotropin (hMG).
  • • Semen analyses were performed periodically and, in patients who remained azoospermic, micro-TESE was performed.


  • • Sperm were noted in 54 patients (10.9%) in semen analysis after treatment in all groups (with no significant difference) at a mean (sd) concentration of 2.3 (4.1) million/mL.
  • • For the 442 patients who remained azoospermic after treatment, successful sperm retrieval was significantly higher (57%) compared with the control group (33.6%).


  • • For patients with non-obstructive azoospermia, clomiphene citrate, hCG and hMG administration, leading to an increased level of FSH and total testosterone, results in an increased rate of sperm in the ejaculate and increased likelihood of successful micro-TESE.

intracytoplasmic sperm injection


human menopausal gonadotropin


microsurgical-testicular sperm extraction


For men with non-obstructive azoospermia, in vitro fertilization with intracytoplasmic sperm injection (ICSI) is the most promising method of conception. As a single sperm is required for the procedure, a low sperm yield in the ejaculate may preclude invasive sperm retrieval techniques. To minimize surgical intervention and to maximize the sperm retrieval yield when surgery is necessary, stimulation of spermatogenesis in these patients is potentially highly beneficial.

Previous studies have reported the successful use of hCG and human menopausal gonadotropin (hMG) to induce spermatogenesis for non-obstructive azoospermia in men with hypogonadotropic hypogonadism [1]. Aydos et al.[2] used pure FSH in men with non-obstructive azoospermia and concluded that pure FSH treatment improves the success of testicular sperm extraction (TESE) for those men with normal FSH levels in all histopathological patterns.

Clomiphene citrate increases pituitary secretion by blocking feedback inhibition of oestradiol, thus increasing endogenous GnRH secretion from the hypothalamus, and FSH and LH secretion directly from the pituitary gland [3,4]. LH stimulates Leydig cells to secrete testosterone. Intratesticular testosterone plays a critical role in promoting spermatogenesis [5]. FSH is essential for the initiation and maintenance of spermatogenesis through stimulation of Sertoli cell maturation, spermatogonial division and possibly an anti-apoptotic action on spermatogonia and spermatocytes [6].

In a previous study, we concluded that the use of clomiphene citrate therapy in cases of non-obstructive azoospermia may result in sufficient sperm for ICSI, either through sperm being found to be present in the ejaculate or potentially through increasing the probability of successful microsurgical (micro)-TESE [7]. In that study, all patients first underwent diagnostic testis biopsy, and patients with Sertoli-cell-only syndrome were excluded.

The aim of the present study was to determine whether endocrine treatment before micro-TESE could benefit males with non-obstructive azoospermia in a modern paradigm without diagnostic biopsy. The largest group of patients studied were those who responded to clomiphene stimulation therapy with a discernible increase in testosterone, as this was the most commonly identified response. For those patients who did not respond accordingly, we investigated novel approaches to hormonal optimization, with human chorionic gonadotropin (hCG) and/or human menopausal gonadotropins (hMG) as adjunct therapy.


The study was conducted between June 2005 and September 2010 at the Infertility Research and Treatment Unit of El-Minia University and Minia National Hospital in El-Minia, Egypt. Patients included men with non-obstructive azoospermia who remained azoospermic for at least 12 months without intervention. At least three semen analyses, with centrifugation and examination of the pellet, were performed to confirm azoospermia. On enrollment in the study, men were re-evaluated with routine history, physical examination and scrotal ultrasonography with Doppler. Morning plasma testosterone, FSH and LH assays (ELISA) were assayed for each patient.

In the present study, 612 patients with non-obstructive azoospermia were included with no histopathological assessment. We excluded all factors that could possibly affect the action of the medications used, including history of testicular biopsy, testicular malignancy, varicocele, hyperprolactinaemia or thyroid dysfunction, and those receiving chemotherapy or radiotherapy. Also excluded were patients with contraindications to clomiphene citrate and patients with FSH levels elevated to >1.5 times their laboratory normal reference level.

The study design is shown in Fig. 1. It included initial treatment, laboratory findings and final medication regimens of each of the study groups. In all, 116 patients randomly chose to undergo micro-TESE without any medical treatment and formed the control group. The remaining 496 patients were administered oral clomiphene citrate at a starting dose of 50 mg every other day. After a minimum of 2 weeks, plasma testosterone was assayed. The dose of clomiphene citrate was increased in increments of 25 mg every other day until morning serum testosterone was 600–800 ng/dL, or until 3 months had passed. In cases where the serum testosterone was noted to be >800 ng/dL, the dose of clomiphene citrate was decreased to 50 mg once every 3 days.

Figure 1.

Study design.

During treatment with clomiphene citrate, we monitored the levels of FSH, LH and total testosterone every 2–4 weeks. Patients were classified into four subgroups according to their response to clomiphene treatment. Group 1 (n= 372) included patients with an obvious increase in serum FSH level (by >50% of the initial level before clomiphene administration) and an increased serum total testosterone to the target level of 600–800 ng/dL. Group 2 (n= 62) included patients with a satisfactory increase in the level of FSH with little or no increase in the level of LH and total testosterone. For these patients we continued clomiphene citrate treatment and added hCG 5000 IU, two injections every week with endocrine assessment, and the hCG dose was adjusted for a target testosterone level of 600–800 ng/dL. Group 3 (n= 46) included those with no obvious increase in the levels of the three endocrine assays after 3 months of treatment, despite adjustment of the dose of clomiphene citrate. Group 4 (n= 16 patients) included patients with paradoxically decreasing serum testosterone levels in response to the increasing dosage of clomiphene citrate. When we stopped clomiphene treatment in these patients, serum testosterone increased. Patients in the third and fourth groups were considered not to be candidates for clomiphene citrate therapy and were administered hCG 5000 IU and hMG 75 IU once weekly. hCG doses were consequently adjusted for a target testosterone of 600–800 ng/dL, and hMG doses were titrated for an FSH assay result of 1.5 times the initial level.

Semen analysis was performed every month for all patients. If at 9 months of treatment, sperm was not noted in the ejaculate, micro-TESE was performed in an attempt to obtain testicular sperm for ICSI. Patients who were observed to have sperm in the ejaculate underwent cryopreservation of semen.

All continuous variables are expressed as mean (sd) values. All the outcome variables were correlated to testicular volume, testosterone, FSH before and during treatment, dose of clomiphene citrate, hCG, hMG and duration of therapy. The Pearson correlation coefficient and independent sample t-tests were used for comparison of continuous variables when appropriate, and the chi-squared and Mann–Whitney U-tests were used to compare categorical variables. A P value <0.05 was considered to indicate statistical significance. spss® statistics software was used for statistical analysis.


The present study included 612 patients with non-obstructive azoospermia. Their mean (sd) age was 26.7 (4.9) years, mean (sd) right testicular volume was 14.6 (4.5) mL, mean (sd) left mean testicular volume was 14.4 (4) mL, and mean FSH was 6.41 mIU/mL. Figure 1 shows the number of patients and the type and duration at therapeutic levels for their final treatment regimen for each group. Initial total testosterone was below a threshold of 300 ng/dL in 140 patients (22.9%), and levels increased to the target level of 600–800 ng/dL in all patients in the four groups after treatment. FSH increased in all groups after 2–3 weeks of adjustment of the type and dose of the selected medications, increasing from 6.4 (2.4) to 11.3 (5.1) mIU/mL in group 1, from 6.1 (1.9) to 14.1 (2.4) mIU/mL in group 2, from 5.6 (2.8) to 13.2 (3.3) mIU/mL in group 3 and from 5.9 (2.6) to 14.2 (2.2) mIU/mL in group 4.

We noted the appearance of sperm in the ejaculate of 54 patients (10.9%) after medical treatment in all groups, with a mean (sd) concentration 2.3(4.1) million/mL and a mean (sd) motility 16.8 (11.3)%: in group 1, sperm were noted in the semen of 41 patients (11%); in group 2 in seven patients (11.3%); in group 3 in four patients (8.7%), and in group 4 in two patients (12.6%) (Table 1). The mean (sd; range) duration of treatment before observing sperm in the ejaculate was 5.4 (2.1; 3–9) months. We did not observe significant differences in age, testicular volume, initial serum total testosterone, or FSH in patients in whom sperm appeared in the semen analysis after treatment compared with those who remained azoospermic.

Table 1. Rate of sperm retrieval in ejaculate, at micro-TESE, and overall for each of the study groups
GroupEffect of Clomiphene n Final regimen**Sperm in semen, n (%)Sperm at micro-TESE (%)Overall sperm retrieval rate, (%) P (vs control)
ControlN/A116N/AN/A33.639/11) (33.6) 
1Increased testosterone(FSH)*372Clomiphene41 (11)191/331 (57.7)232/372 (62.4)<0.001
2No increase in testosterone62Clomiphene + hCG7 (11.3)31/55 (56.4)38/62 (61.3)<0.001
Increased FSH
3No increase in testosterone46hCG + hMG4 (8.7)22/42 (52.4)26/46 (56.5)0.01
No increase FSH
4Decrease in testosterone16hCG + hMG2 (12.6)8/14 (57.1)(10/16) 62.5<0.05

Table 1 describes the rates of successful micro-TESE and the overall rates of sperm availability for ICSI for each of the groups. For the 442 patients who remained azoospermic after medical treatment, sperm were retrieved from the testis in 252 patients (57.0%) with micro-TESE. As 54/496 (10.9%) patients manifested sperm in the ejaculate, the total number of patients with sperm acquired for ICSI was 306/496 (61.7%), compared with 33.6% in the control group (P < 0.001). Sperm retrieval at micro-TESE was successful in 191/331 patients (57.7%) in the group 1; 31/55 (56.7%) patients in the group 2; 22/42 patients (52.4%) in the group 3; and 8/14 patients in the group 4 (Table 1). We did not observe significant differences in age, testicular volume, initial serum total testosterone, or FSH in patients who had successful sperm retrieval as compared with those without sperm found at micro-TESE.


We evaluated the use of medication intended to stimulate spermatogenesis before TESE in men with non-obstructive azoospermia. Previously, hCG and hMG have been used to induce spermatogenesis for non-obstructive azoospermia in men with hypogonadotropic hypogonadism [1,8]. Schill et al.[9] reported a 15.3 million/mL increase in mean sperm concentration after 3 months of treatment with hCG and hMG in men with idiopathic oligospermia and normal levels of gonadotropins. Physicians have used clomiphene citrate, hCG and hMG as treatment for idiopathic oligospermia, and numerous studies have reported improvement in semen quality and increased pregnancy rates among the partners of men in whom these agents were administered [9,10].

In the present study, the majority of patients responded to clomiphene citrate treatment with an increase in serum testosterone, but in 16/496 patients who received clomiphene citrate, testosterone paradoxically decreased. This finding is consistent with that of Pasqualotto et al.[11], who described three cases of azoospermia in patients with oligospermia after clomiphene citrate treatment.

While the exact physiological effects of changes in testosterone and FSH induced by clomiphene citrate have not been fully elucidated, it seems likely that increased testosterone production in the testicle supports sperm production, while increased FSH from the pituitary stimulates Sertoli cells' function in spermatogenesis. Thus, those patients in whom clomiphene citrate did not result in an increase in testosterone and FSH, groups 2–4, were treated with hCG and hMG. We chose FSH rather than inhibin as an assay, as FSH is currently in more common clinical use.

In a previous study, for patients with non-obstructive azoospermia, the use of clomiphene citrate was correlated with the appearance of sperm in the ejaculate in 64.3% of patients [7]. Successful sperm retrieval for ICSI was achieved in all treated patients who remained azoospermic after therapy. It is difficult to directly compare the previous study with the present study because of notable differences; in the previous study, patients with Sertoli-cell-only histology were excluded, all of the patients were treated with only clomiphene citrate, and a substantial increase in testosterone was observed in all patients.

In the literature, the mean sperm retrieval rate for micro-TESE in non-obstructive azoospermia is 48% [12]. In the present study, 41 (11%) of the 372 patients that responded to clomiphene citrate with a significant increase in testosterone and FSH, group 1, ultimately had sperm observed in their semen, and these patients did not require micro-TESE. The 331 remaining patients underwent micro-TESE, and sperm was found in 191 (57.7%). Thus, in total, 232/372 patients patients (62.3%) had sperm available for ICSI, as opposed to 33.6% in the group that underwent micro-TESE immediately without medication (P < 0.001).

For the patients who did not respond to clomiphene citrate as sole therapy and were managed with addition of hCG with or without hMG, identified as groups 2, 3 and 4, 74/124 patients (60%) ultimately had sperm available for ICSI. This arm of the study was not powered to detect differences in sperm retrieval rates between the groups. We did observe that these patients had an approximately doubled sperm retrieval rate compared with the control group, and the increased availability of sperm from each group of medically treated patients compared with the control group was significant (P < 0.05).

The observation of a 10.9% rate of sperm appearing in the ejaculate with initial medical therapy is notable when compared with conventional TESE, an invasive procedure for which success rates range from 17 to 38.2% [13–16]. For patients who remained azoospermic after a trial of medical treatment, the sperm retrieval rate of micro-TESE was greater than the control group (57% vs 33.6%) that underwent immediate micro-TESE without medical therapy.

The present study suggests that a trial of medical therapy before micro-TESE, if followed by an obvious increase in FSH and testosterone, has the potential advantages of obviating surgery by making available ejaculated sperm and increasing the likelihood of successful sperm extraction with micro-TESE. A disadvantage of instituting medical therapy before TESE is that of delaying ICSI and could be particularly concerning when female partners are of advanced age. Physicians should consider all factors in treatment, and render decisions in joint discussion with an informed patient and spouse.

In conclusion, using our described medical treatment protocol in cases of non-obstructive azoospermia, ∼11% of patients may be expected to manifest sufficient sperm in the ejaculate for use in Assisted Reproduction Technique (ART), obviating the need for scrotal surgery. In the remaining patients for whom TESE is required, a greater likelihood of obtaining sperm for ICSI is achieved.

Using our protocol, we found no significant difference in the outcome of patients treated with different types of drugs stimulating spermatogenesis, as long as an obvious increase in the levels of FSH, LH and total testosterone was observed. We recommend starting with the most inexpensive of these drugs, clomiphene citrate, and adding, or replacing with, hCG and hMG according to each patient's endocrine response. Our target level of FSH was 1.5 times its initial level and our target level of serum testosterone was 600–800 ng/dL. We recommend micro-TESE if the patient remains azoospermic in the periodical semen analysis after 9 months of treatment with achievement of the target levels of FSH and serum testosterone.


None declared.