Infertility despite surgery for cryptorchidism in childhood can be classified by patients with normal or elevated follicle-stimulating hormone and identified at orchidopexy


J. Thorup, Department of Paediatric Surgery 4072, Rigshospitalet, University of Copenhagen, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark.



To analyse infertility despite orchidopexy in childhood.


The study comprised patients with cryptorchidism (70 bilateral and 65 unilateral) who had a simultaneous biopsy taken at orchidopexy in childhood, and in adulthood had analyses of semen and FSH. In adulthood 42 formerly bilateral cryptorchid boys had repeat testicular biopsies taken. Infertility was suspected in men with < 5 million sperm/mL in the best sample of semen and concomitant poor sperm motility, and who were classified by follicle-stimulating hormone (FSH) values. At orchidopexy the number of spermatogonia/tubule and the germ cell differentiation were measured. In adulthood the percentage of tubules with complete spermatogenesis, spermatogenic arrest and Sertoli-cell only status was assessed.


Infertility was suspected in 38 of 70 (54%) of formerly bilateral and six of 65 (9%) formerly unilateral cryptorchid patients. High FSH values were expected in these suspected infertile patients, but 15 of 38 (59%) formerly bilateral and five of six formerly unilateral cryptorchid patients had normal FSH values. These patients were identified in childhood at orchidopexy; those with bilateral cryptorchidism generally presented with germ cells, but the mean number of spermatogonia per tubule was < 30% of the lowest normal value, and the germ cells were seldom normally differentiated, whereas those with unilateral cryptorchidism generally lacked germ cells in the biopsies. No patients had a decreased FSH value.


Despite surgery for cryptorchidism, infertility was probable in a third (44 of 135) of the patients. We expected high FSH values in these patients, but in 45% (20/44) the FSH values were normal. These patients may have relative FSH deficiency. At orchidopexy these patients were identified to be bilaterally cryptorchid with few germ cells and those unilaterally cryptorchid had none in the biopsy. After orchidopexy in childhood, additional hormonal treatment, e.g. recombinant FSH or buserelin, may be indicated in these patients.


Cryptorchidism is associated with infertility; in general, men with persisting bilateral cryptorchidism are infertile. In contrast, after surgery in childhood, normal sperm counts are reported in> 20% of men with a history of bilateral and in ≈ 70% with a history of unilateral cryptorchidism [1]. In the background population normal sperm counts are found in ≈ 80% of men [1]. The recommended age at surgery for cryptorchidism has been decreased in an attempt to minimize the risk of infertility. The recommendation is primarily based on an examination of the number of spermatogonia and gonocytes, if any, per tubule (S/T) in biopsies at orchidopexy. Surgery for cryptorchidism is important before germ cells become scarce, i.e. before 15–18 months old, because infertility is associated with a lack of germ cells in a biopsy at orchidopexy. The risk of later infertility is 80–100% in bilateral cryptorchidism, depending whether no germ cells are found in one or both biopsies, and is ≈ 33% in unilateral cryptorchidism [1,2]. These findings have been reported in other studies [3,4]. Unfortunately, early orchidopexy may not guarantee later fertility. At ≈ 3 months old there is normally a so-called ‘mini-puberty’, with high levels of FSH, LH, testosterone and inhibin B [5–7]. During the first 6 months of life there is normally a transformation of the fetal germ cells, the gonocytes, to the adult reservoir of stem cells, the Ad spermatogonia [5–8]. If this transformation does not take place the risk of later infertility is ≈ 90%[8]. Consequently, besides early orchidopexy, more efforts may be necessary to preserve germ cell development and thus avoid subsequent infertility.

From analyses of semen and FSH it is not possible to clearly distinguish fertile and infertile men; the WHO has defined 20 million sperm cells/mL as the lowest normal value. Without assisted reproductive techniques (e.g. intrauterine insemination) a man can achieve paternity if the sperm count is> 5 million/mL and the sperm motility is normal. From 5–60 million/mL there is no correlation between sperm counts and paternity [1]. Hypergonadotrophic hypogonadism exists in cases of low sperm counts and elevated FSH levels, and a primary testicular disease must be suspected. In some cases there may be a chromosomal disorder, e.g. Klinefelter's syndrome or microdeletions in different parts of Yq, including known regions involved in spermatogenesis (DAZ, RBM, AZFa,b,c) [2,9]. In contrast, patients with normal FSH values and < 5 million sperm/mL may have partial obstruction of sperm, fewer late spermatids, spermatogenic arrest, or a relatively low level of FSH, based on a relative hypothalamic and/or pituitary defect [1,10].

The aim of the present study was to analyse infertile and formerly cryptorchid patients, and assess whether infertility is associated with normal FSH values. We analysed semen and FSH, and testicular biopsies taken at orchidopexy and in adulthood. To our knowledge, such data have not previously been reported.


The study comprised 135 patients who had testicular biopsies taken at orchidopexy, and in adulthood had semen analysed and FSH measured. Seventy men had had bilateral and 65 unilateral orchidopexy; their median (range) age at surgery and at follow-up was 12.8 (6.7–16.2) and 29.0 (19.9–34.7), and 10.8 (2.9–11.9) and 22.7 (18.4–32.7) years, respectively. At the end of the follow-up the patients agreed to a physical examination. No selection bias was detected in the patients who accepted participation [1,2]. Excluded at orchidopexy were patients with recognized additional abnormalities, including hypogonadotrophic hypogonadism and chromosome disorders, and those who underwent previous inguinal surgery; also excluded were those with a testicular germ cell tumour or carcinoma in situ at surgery, at the time of follow-up or any registered in the Danish Cancer Register for treatment of such neoplasms before 31 December 1994.

In adulthood, we requested two samples of semen; if the sperm count of the first was normal we did not insist on another. In all, 56 of the 70 (80%) formerly bilaterally cryptorchid and 35 of the 65 (54%) formerly unilateral cryptorchid patients offered two samples of semen. The semen was analysed in accordance with the WHO guidelines; these objective values were selected because most patients had no wish for paternity. FSH was analysed using a commercial immunometric monoclonal assay kit, as previously described [11]; for 18–50-year-old men the 2.5 and 97.5 reference percentiles are 1.1–7.9 IU/L. Infertility was suspected when the patient's sperm sample had < 5 million sperm/mL in the best sample and the sperm motility was poor [1,2]. Those with suspected infertility were analysed among those with normal or with elevated FSH values.

When hypogonadotrophic hypogonadism in adulthood was suspected clinically, then LH, testosterone, free testosterone, dihydrotestosterone, andostendione and dehydroepiandosterone sulphate were measured in blood samples, and the patients’ sense of smell evaluated.

From the biopsies at orchidopexy, the S/T was measured from at least 100 tubular cross sections and expressed as a percentage of the lowest normal S/T for age, as described previously [1]. The most differentiated germ cell was measured. The presence of spermatocytes in boys aged 4 years was taken as normal. The median age at spermarche is 13.5 years [12], and thus we took the presence of spermatids in boys aged 14 years as normal.

In adulthood, 42 of the 70 (60%) patients with a history of bilateral cryptorchidism agreed to have bilateral testicular biopsies taken, but one had a unilateral repeat biopsy only, as there was no contralateral testis. The testicular biopsies were taken under local anaesthesia, the tissue specimen dropped from the razor blade directly into freshly prepared Stieve's fixative. The biopsy was embedded in paraffin, 4 µm sections cut and stained with iron-haematoxylin and eosin, and examined by light microscopy. The percentage of tubules showing complete spermatogenesis, spermatogenic arrest and Sertoli-cells only, respectively, was calculated from ≥ 100 tubular cross sections of the biopsy specimens. The presence and frequency of the histological patterns were assessed [13]. In spermatogenic arrest, spermatogenesis is blocked and no sperm are formed, but primary spermatocytes, or less often, spermatogonia, are the most highly differentiated germ cells.

For the statistical analysis, in bilaterally cryptorchid patients with bilateral biopsies the mean value was used, such that every patient had one value in the statistical analysis. The Spearman rank, Mann–Whitney, Fisher's exact, Kruskal–Wallis or chi-square tests were used (two-tailed), with P < 0.05 considered to indicate significance. The study was conducted in agreement with the Helsinki II declaration and was approved by the local ethics committee.


The sperm count was low (< 5 million sperm/mL) in 38 of the 70 (54%) men with a history of surgery for bilateral and in six of the 65 (9%) men with unilateral cryptorchidism (Fisher, P < 0.001). The respective median (range) counts were 0.3 (0–4.5) and 0.01 (0–5) million sperm/mL. The sperm motility was poor in all patients. Elevated FSH values were expected in these patients suspected of being infertile, but values were normal in 20 of 44 (45%) assessed. In all, 20 of 135 (15%) had normal FSH values and suspected infertility, detected more often in men with previous bilateral (15/70, 21%) than in men with unilateral cryptorchidism (five/65, 8%; Fisher, P < 0.05). No FSH value was below the lowest normal limit, and consequently none had hypogonadotrophic hypogonadism. The one patient who was suspected clinically of having hypogonadotrophic hypogonadism had normal values for the androgens assessed and a normal sense of smell.

In the formerly bilaterally cryptorchid men the median S/T and the median age-related S/T at orchidopexy correlated with the sperm count, FSH level and total testicular volume in adulthood (Spearman rank (all P < 0.005). FSH levels correlated with the sperm count and total testicular volume (Spearman rank, both P < 0.001).

The bilaterally cryptorchid men were divided into those supposedly fertile or infertile, the latter being again divided into those with a normal or high FSH level in adulthood (Table 1). The 23 suspected infertile men with high FSH levels, and consequently possible hypergonadotrophic hypogonadism, in adulthood all had Sertoli cells only (chi-square, P < 0.05; Table 1). At orchidopexy these patients predominantly had no germ cells (< 1% of the lowest normal S/T; chi-square, P < 0.001). In contrast the 15 suspected infertile patients with normal FSH levels in adulthood all had spermatogenic arrest (chi-square, P < 0.05). At orchidopexy, biopsies from these patients mostly had germ cells, but the number was severely reduced (mean age-correlated S/T < 30%; chi-square, P < 0.001; Table 1). The 32 supposedly fertile men in adulthood most often had complete spermatogenesis (chi-square, P < 0.001). At orchidopexy these patients had many germ cells (mean S/T> 30% of the lowest normal S/T; chi-square, P < 0.001), and predominantly normal germ cell differentiation (chi-square, P < 0.05; Table 1).

Table 1.  The fertility potential in adulthood of patients with a history of bilateral (70) or unilateral (65) surgery for cryptorchidism and simultaneous testicular biopsy
Variable Median (range), n (%) or n/N ‘Fertile’ ‘Infertile’P
normal FSHElevated FSH
  • *

    P from Kruskal–Wallis test;

  • P from chi-square test. LNV, lowest normal value.

No. of patients321523 
Age at surgery, years*12.7 (6.7–16.2)12.8 (11.5–15.4)12.9 (10.6–15.1)1.00
Testes at the external inguinal ring30 (94)13 (87)19 (83)0.61
Unsuccessful hormonal treatment24 (75)12 (80)13 (57)0.22
Bilateral biopsies at surgery26 (81)12 (80)17 (74)1.00
Mean S/T*  0.63 (0.02–2.68)  0.305 (0–2.92)  0.015 (0–1.20)< 0.001
Age-correlated mean S/T, %*35 (1–149)17 (0–162)  1 (0–66)< 0.001
≥  1 S/T of < 1% of LNV  4 (13)  2 (13)19 (83)< 0.001
Germ cells present, mean S/T < 30% of LNV10 (31)  9 (60)  0< 0.001
Mean S/T> 30% of LNV18 (56)  4 (27)  4 (17)< 0.05
Normal differentiated germ cells21 (66)  5 (33)  4 (17)< 0.005
Age at follow-up, years*28.6 (19.9–34.5)28.9 (21.9–33.9)28.9 (20.6–33.7)1.00
Maximum sperm count, million/mL*17.5 (5.8–85)  1.5 (0–4.5)  0.025 (0–2)< 0.001
Two samples of semen  9 (72)14 (93)11 (73)0.08
FSH, IU/L*  4.3 (1.2–22.6)  5.3 (2.6–7.4)16 (8.4–28.4)< 0.001
Total testicular vol. in adulthood, mL*33 (22–50)30 (25–45)22 (12–45)< 0.001
Both testes in scrotum30 (94)15 (100)20 (87)0.37
Patients with secondary testicular biopsies19 (59)10 (67)13 (57)1.00
Complete spermatogenesis in all tubules  7 (37)  0  0< 0.001
Tubules with complete spermatogenesis*84 (17–100)55 (8–92)13 (4–49)< 0.005
The pattern of spermatogenic arrest  9 (47)10 (100)10 (77)< 0.05
Tubules with spermatogenic arrest*12 (4–49)29 (4–92)  6 (5–80)0.08
Pattern of Sertoli-cell-only11 (58)  7 (70)13 (100)< 0.05
Tubules with Sertoli-cell-only*13 (1–77)20 (3–96)77 (1–100)< 0.005
No. of patients59  5  1 
Age at surgery, years*10.8 (3–11.9)10.8 (2.9–11.9)  6.40.61
Testes at the external inguinal ring50 (85)  5 (100)  1 (100)0.61
Unsuccessful hormonal treatment24 (41)  1 (20)  00.64
Mean S/T*  0.36 (0–3.61)  0.01 (0–0.25)  0< 0.05
Age-correlated mean S/T, %*20 (0–201)  0.6 (0–74)  00.08
S/T < 1% of LNV10 (17)  4 (80)  1 (100)< 0.005
Normal differentiated germ cells19 (32)  1 (20)  00.22
Age at follow-up, years*22.2 (18.4–32.7)24.9 (19.9–32.3)19.70.22
Maximal sperm density, million/mL*46.4 (5.5–292.5)  3.9 (0.01–5)  0< 0.001
Two samples of semen30 (51)  4 (80)  1 (100)0.37
FSH, IU/L*  3.7 (1.6–9.7)  3.6 (2.1–4.4)170.14
Total testicular volume in adulthood, mL*40 (25–62)40 (25–50)320.37
Both testes in scrotum57 (97)  5 (100)  1 (100)1.00

In men with previous unilateral cryptorchidism the FSH levels correlated with the sperm count and the total testicular volume (Spearman rank, both P < 0.01). These men were also divided into three groups (Table 1). The supposed fertile patients had the highest S/T values, and the lowest risk of no germ cells (S/T < 1% of the lowest normal value; Kruskal-Wallis, P < 0.05 and chi-square, P < 0.005). Infertility was suspected in six (9%) of these patients. The absence of germ cells in the biopsy at orchidopexy (S/T < 1% of the lowest normal value) was predominantly in the patients who later were suspected to be infertile and with normal FSH levels (chi-square, P < 0.005; Table 1).

In all, the risk of later infertility and normal FSH levels was nine of 19 in those bilaterally cryptorchid (germ cells present, but the mean age-correlated S/T < 30%), and four of 15 in those with unilateral cryptorchidism (S/T < 1% of the lowest normal value). These patients identified at risk of later infertility and with normal FSH levels overall comprised 34 of the 135 (25%) patients.


Infertility was suspected in men with < 5 million sperm/mL; infertility might be a result of hypergonadotrophic hypogonadism, with high or normal FSH levels. The latter group comprised 15% of the patients, which is in accordance with reported values of 9.5%[14] and 17%[3] in previous patients with a history of cryptorchidism. The FSH levels were normal in 40% and 80% of the men who were suspected to be infertile, despite surgery for bilateral or unilateral cryptorchidism, respectively. At orchidopexy these bilaterally cryptorchid patients generally had germ cells in the biopsies, but at < 30% of the lowest normal value, and defective germ cell differentiation. In adulthood there was spermatogenic arrest. In those with unilateral cryptorchidism, at orchidopexy there was a general lack of germ cells in the biopsies.

FSH levels in men are usually inversely correlated with sperm counts and indicate the integrity of the seminiferous tubules [1,15]. In cases of suspected infertility and normal FSH levels the low sperm count could be accidental, but it is unlikely as two samples of semen were obtained from 90% (18/20) of the patients. Partial sperm obstruction is also unlikely, as complete spermatogenesis was not detected in the biopsies. In contrast, there was spermatogenic arrest, which has been reported in biopsy specimens from men with a history of undescended testes [13] or in retractile testes that descended spontaneously in boys aged 12–14 years [16]. Furthermore, spermatogenic arrest was found in the testes of normal men after nonsurgical testicular fixation close to the external inguinal ring [17]. This method was so effective that it was proposed as a means of contraception [17]. Consequently, spermatogenic arrest may result from an elevated testicular temperature, even in normal testes. However, in the present men, all testes of those suspected to be infertile and with normal FSH levels were in the scrotum in adulthood. Spermatogenic arrest may be found in hypogonadotrophic hypogonadism, and could be caused by a deficiency of GnRH [18]. None of the present patients had low FSH levels and in general hypogonadotrophic hypogonadism is present in only ≈ 2% of cryptorchid males [9]. Defective spermatogenesis may result from relatively low levels of FSH [19]. In females, relative FSH deficiency has been described in patients with polycystic ovary syndrome [20]. It is tempting to suggest that relative FSH deficiency might contribute in the present suspected infertile men with normal FSH levels and a history of cryptorchidism.

Generally the hormonal treatment of cryptorchidism is based on the hypothesis of gonadotrophin and/or androgen insufficiency manifesting during gestation. The normal surge of LH and FSH at age 2–3 months may be blunted [5,6]. In undescended testes the transformation from gonocytes to spermatogonia and meiosis is hampered; this transformation is essential for normal fertility. Despite orchidopexy before 2 years of age, sperm counts were abnormal in adulthood when the normal transformation of gonocytes into Ad spermatogonia did not occur [8]. Equivalently, at orchidopexy the later fertile patients had a higher frequency of normally differentiated germ cells than the later infertile patients. At orchidopexy lower FSH and LH levels were reported in unilaterally cryptorchid boys, with very few or no germ cells compared with patients with> 0.5 germ cells per tubule [5]. When a biopsy has an S/T of < 0.2 at orchidopexy in childhood, subsequent buserelin (a GnRH agonist) stimulation has been suggested [21]; this agonist primarily stimulates the release of LH and FSH. In adulthood, sperm counts were higher in treated patients than in those of equivalent age and low S/T values at orchidopexy without buserelin [21].

If hormonal therapy is used the age of the patient is very important, because of the large variation in the age-normal values for FSH, LH and testosterone [7]. In boys aged < 1 year the FSH, LH and testosterone values are usually rather high, and the combined administration of GnRH and hCG stimulated spermatogonial transformation and proliferation [6]. However, in 1–3-year-old cryptorchid patients there were fewer germ cells after hormonal therapy [6,7]. At 1–3 years old the levels of FSH, LH and testosterone are normally extremely low [7], and hormonal treatment may be hazardous to the germ cells [7].

The suspected infertile men with normal FSH levels in general had spermatogenic arrest in adulthood. FSH stimulates both germ cell meiosis and spermatogenesis during in vitro culture of human testicular cells. The possibility of using FSH in men with spermatogenesis arrest to induce in vitro spermatogenesis, and to use the resulting in vitro cultured germ cells for assisted reproduction, is under investigation [22]. In adults, treatment with FSH resulted in at least a doubling of the sperm count in those with oligospermia, normal FSH levels and a testicular cytological picture of hypospermatogenesis [18]. In young men with varicocele, FSH treatment resulted in higher sperm counts and better sperm motility [23].

The present patients were rather older at surgery for cryptorchidism than usual, as previously reported in Denmark. In those infertile despite surgery for cryptorchidism, the risk of FSH deficiency was ≈ 45%. Generally in patients with no germ cells in a biopsy taken at orchidopexy there is higher risk of subsequent infertility [1,2]. The risk of having no germ cells in a biopsy increases with increasing age at orchidopexy [2]; consequently, by decreasing the age at orchidopexy the risk of subsequent infertility may decrease. From the present results it is tempting to suggest that for the optimal treatment of cryptorchidism early surgery is an important first step. The second step, which may be more important as the age at surgery decreases, is hormonal-stimulation therapy in special groups of cryptorchid patients, e.g. those identified here, who comprised 25% of the patients.

In conclusion, infertility after surgery for cryptorchidism may be hypergonadotrophic, or with normal FSH levels. The latter group comprised ≈ 40% and 80% of men who were suspected to be infertile despite surgery for bilateral or unilateral cryptorchidism, respectively. These patients may have a relative deficiency in FSH. Those with bilateral cryptorchidism had germ cells in the biopsies at orchidopexy, but the mean S/T was < 30% of the lowest normal value, and there was defective germ cell differentiation, with spermatogenic arrest in adulthood. In unilateral cryptorchidism, such patients lacked germ cells at orchidopexy. Further studies are needed to investigate potential medical treatments, e.g. buserelin and/or recombinant FSH, to minimize the risk of subsequent infertility.


Dr Niels Erik Skakkebæk, Head of the Department of Growth and Reproduction, Rigshospitalet, University of Copenhagen analysed samples of semen, and provided a general description of the testicular biopsies taken from the adult patients. Dr Paul Erik Lebech, Head of the Department of Obstetrics and Gynecology, Frederiksberg Hospital, University of Copenhagen analysed samples of semen. The Gerda and Åge Haench Foundation gave financial support.


spermatogonia per tubule.