Fertility, sexuality and testicular adrenal rest tumors in adult males with congenital adrenal hyperplasia

Subjects

CAH males, ≥19 years of age, were mainly recruited from the two participating university hospitals. The diagnosis was confirmed by mutation analyses of the CYP21A2 gene (31 patients) or the HSD3B2 gene (one patient).

The data were divided into subgroups according to age as <30 years and older and according to the three most prevalent CYP21A2 mutations as null, I2splice, and I172N. In compound heterozygotes, the mildest mutation defines the genotype group. Null is associated with the salt-wasting (SW) phenotype, I2splice is most often associated with the SW phenotype but some patients may escape SW, and I172N typically leads to simple virilizing (SV) CAH. Matched controls born on the same date as the patients were recruited from the National Population Registry. The only exclusion criterion was severe mental or psychiatric disturbance with inability to consent to the study. Age-matched national fertility data were also used (children/father) (14).

This study was approved by the Ethics Committees of the Karolinska Institute, Stockholm, and the University of Gothenburg, Göteborg, Sweden. All participants gave their written informed consent.

Study protocol

Patients and controls were examined as outpatients at Karolinska University Hospital (22 patients and 22 controls), Stockholm, or Sahlgrenska University Hospital (ten patients and ten controls), Göteborg, Sweden. Serum and plasma samples were collected in the morning after an overnight fast before morning medication. All participants answered questionnaires concerning social and sexual issues, fertility, and fecundity. Fecundity problems were defined as attempting to become a father for >1 year. Patients underwent testicular ultrasound imaging and were also asked to deliver a semen sample. Published data from this cohort on body composition measured by dual-energy X-ray absorptiometry and cardiovascular risk factors (15) were used to evaluate whether differences could be found in these parameters between patients with normal and patients with pathological semen.

Testicular examination

The testicles were only examined at one of the participating centers. This was done by manual palpation and ultrasound. An orchidometer was used to estimate the volume. The ultrasound examinations were performed by one physician (S Granberg) using a Voluson Expert 730 machine equipped with a 6–16 MHz real-time three-/four-dimensional linear probe (GE Healthcare Austria GmbH & Co. OG, Zipf, Austria). Total testicular, TART, and functional (total testicular−TART) volumes were estimated using the formula for a prolate ellipsoid (maximal length×maximal width×maximal depth×0.523) (16).

Semen analysis

Seminal fluid was collected after 1–4 days of ejaculatory abstinence. The analyses included assessment of semen volume, sperm concentration, total sperm count, motile and immotile spermatozoa, and morphology. Results were evaluated according to the World Health Organization (WHO) standard (17).

Biochemical assays

Serum testosterone, sex hormone-binding globulin (SHBG), and dried blood spot 17-hydroxyprogesterone (17OHP, measured at 0800, 1400, 1900, 0100, and 0600 h), and 24-h urinary pregnanetriol were measured as described previously (15). Bioactive testosterone was calculated from total testosterone, SHBG, and albumin (18). Serum FSH, LH, estradiol (E₂), and total prostate-specific antigen (PSA) and free PSA were measured by fluoroimmunoassay (AutoDelfia, PerkinElmer, Waltham, MA, USA). FSH and LH had normal ranges of 1.0–10 U/l. Serum inhibin B was measured by ELISA (Beckman Coulter, Toronto, ON, Canada; normal range was 25–325 ng/l and detection limit was 10 ng/l). Plasma ACTH was measured on an advantage automatic immune analyzer and plasma renin by immunoradiometric assay (both from Nichols Institute Diagnostics, San Clemente, CA, USA).

Statistical analysis

SigmaStat for Windows (Jandel Scientific, Erkarath, Germany) was used for data analysis. Results are presented as mean±s.d. if not otherwise stated. An unpaired t-test was used to compare two groups when values were normally distributed. Otherwise, the Mann–Whitney rank sum test was used and in these cases the median and range were reported. One-way ANOVA was used when comparing three groups with continuous variables and if normally distributed, followed by the post-hoc Bonferroni t-test; otherwise, the Kruskal–Wallis test was performed, followed by the post-hoc Mann–Whitney rank sum test and Dunn's method. χ² test or Fisher's exact test was used in frequency table calculations. All proportions were calculated discounting missing values. Spearman's correlation coefficient was used for correlation analyses. Statistical significance was set at P<0.05 and tendency at 0.05–0.10.

Characteristics of the patients and controls

The characteristics of the patients have been described earlier (15). In brief, 32 males with CAH, aged 19–67 years, were examined. Thirty-one had 21-hydroxylase deficiency. Eighteen had been diagnosed with the SW phenotype (genotype groups: null, n=7; I2splice, n=9; I172N, n=1; and C75R, n=1). One patient had 3-beta-hydroxysteroid dehydrogenase (3β-HSD) type II deficiency (genotype: C75R). Twelve patients had been diagnosed with the SV form (I172N, n=8; I2splice, n=3; and P453S, n=1). One SV patient had karyotype 46,XX (I2splice). Two patients had the non-classic (NC) phenotype with normal baseline ACTH levels (mildest mutation V281L and P105L+P453S). The patients with 3β-HSD deficiency and 46,XX were excluded from further analysis of the data but are separately described at the end of this section. Thus, 30 patients were included in the statistical analysis.

Glucocorticoids were taken by 93% (28/30) of the patients. One recently diagnosed 29-year-old man (I172N) took no medication and another with the NC phenotype (V281L) only used prednisolone when acutely ill. The most frequently used preparations were prednisolone (62%, 18/29) and hydrocortisone (17%, 5/29). The mean dose expressed in glucocorticoid equivalents was 32.7±12.4 mg using the conversion previously described (19). Urine pregnanetriol and blood 17OHP diurnal curves were similar for the different genotypes and age groups (15).

Most patients (87%, 26/30) received fludrocortisone, which was also received by most of the SV patients (82%, 9/11) and one of the NC patients. The SV and NC patients used fludrocortisone to minimize the glucocorticoid dose. None of the CAH males had experienced prostate problems. In the control group, a 67-year-old man had antiandrogen therapy after prostatectomy for cancer and was excluded in all calculations concerning testicular function, a 58-year-old man was operated on for benign prostatic hyperplasia, and a 45-year-old man had recurrent prostatitis.

Fertility, fecundity, and sexuality

On comparing the number of biological children of CAH patients with age-matched national data, the fertility was low (in all patients, 0.9±1.3 vs 1.8±0.5; <30 years, 0.1±0.3 vs 1.3±0.4; ≥30 years, 1.2±1.4 vs 2.0±0.3; children/father, P<0.001 for all comparisons). In patients with classic 21-hydroxylase deficiency, the corresponding frequency was 0.9±1.3. The reduction of fertility was similar in the three genotype groups (null, I2splice, and I172N) compared with age-matched national data (P<0.001). One NC man had three while the other had no children. However, compared with our investigated controls, fertility and the age at the time of birth of the first child (not shown) were similar although more patients than controls had experienced fecundity problems (Table 1). Three patients had attempted to father but never succeeded, compared with one control, but one of them fathered a child 1 year after the investigation. The rate of marriage/cohabitation was similar in the entire patient group compared with controls. A propensity to increased frequency of marriage/cohabitation in older compared with younger subjects was only found in controls. A similar percentage of patients and controls had no partner (Table 1). One control had two adopted children.

The age at sexual debut and frequency of intercourse were similar between the groups (not shown), but the lifetime number of partners was lower in the total group of patients and in older patients compared with controls. One control declared that he was homosexual compared with none in the patient group. No bisexuality was reported. As shown in Table 2, there was no difference between the genotype groups regarding fertility, fecundity, and sexuality. However, compared with controls, the null group patients have had fewer partners, fewer I2splice patients were married/cohabitants, and fecundity problems were more common in the I172N group patients.

Biochemical tests

Inhibin B was lower and FSH higher in all groups of CAH males compared with controls (Tables 1 and 2). The values were negatively associated in the entire patient cohort (r=−0.514, P<0.001). Three patients had subnormal inhibin B and four had elevated FSH concentrations. No differences in LH levels were found. No subject displayed subnormal FSH or LH.

As expected, both older patients and controls had lower bioactive testosterone than the younger ones. An overall tendency to lower bioactive testosterone and higher E₂ in patients than in controls resulted in a significant decrease of the testosterone/E₂ ratio in all and older patients, with the latter also having a lower ratio than their younger counterparts (Table 1). SHBG was similar in all groups (not shown). Total PSA and free PSA concentrations were similar in all comparisons (not shown).

Testicular examination

At manual palpation performed in 70% (21/30) of patients and controls, smaller testicular volumes were found in patients compared with controls (Table 1). Patients with the null mutation had smaller testicles than those with other genotypes or controls. A unilateral tumor was palpated in one patient and multiple bilateral tumors in another (Table 1). However, with the aid of ultrasound (completed in 70% (21/30)), 86% (18/21) were shown to have TARTs, which were bilateral in 72% (13/18) of cases. An image illustrating a TART from one of the patients is shown in Fig. 1. No significant difference was found in TART frequency and number, total TART volume, or functional testicular volume between genotypes (Table 3). Two patients with genotypes associated with very mild disease (V281L and P105L+P453S) had small TARTs (0.31–0.38 cm³).

Patients with the largest TARTs (n=8, volume >1 cm³) had similar age and fat mass but higher lean mass (19.5±2.4 vs 16.6 kg/m², P=0.025) and shorter height (162±8 vs 173±6 cm, P=0.011) than those with no or very small TARTs (n=7, volume <0.2 cm³). Poor treatment control, arbitrarily defined as urinary pregnanetriol ≥100 μmol/24 h and/or a 24-h median 17OHP value ≥100 nmol/l, was found in three of the eight patients with larger TARTs but in none of those with no or very small TARTs.

Two patients had undergone previous testicular surgery for suspected malignancy. In both cases, the histopathology was interpreted as benign Leydig cell tumors. One 41-year-old SW patient (I2splice) had a testicular tumor extirpated at 12 years of age. He now had one small tumor in each testicle (volume 0.11 cm³), normal semen, and two biological children. In a 34-year-old SW patient (I2splice) with two biological children, a tumor was removed before conception. He now showed multiple bilateral tumors. One 23-year-old SV patient (I2splice) with TARTs had been operated on for testicular retention at age 17. No testicular intervention was reported in any of the controls.

Semen analysis

Semen for analysis was obtained from 14 of the 30 CAH males. Fecundity problems tended to be more common in those who delivered a semen sample than in those who declined (5/14 vs 1/16, P=0.072). The mean ages were similar (37.7±12.7 vs 35.6±11.9 years).

Considering all WHO criteria, six of the 14 semen samples were pathological. The results were similar in younger and older patients (not shown). A lower sperm count, concentration, and motility in I172N compared with the I2splice group did not reach statistical significance in this limited material (Table 3). Only one patient in the null group had an analyzed sample and it was pathological. One 39-year-old patient with SV (I172N) had severe teratozoospermia with impaired motility and a previously performed intracytoplasmic sperm injection (ICSI) resulting in one child.

We found that inhibin B levels, regarded as a marker of spermatogenesis, were negatively correlated to body mass index (BMI; r=−0.509, P=0.004) in the controls and demonstrated a similar trend in the patients (r=−0.317, P=0.087). Hence, body composition data and other markers for cardiovascular risk were compared in males with pathological and normal semen. Males with pathological semen had a higher truncal and total fat mass, total fat/lean mass ratio, and heart rate, in spite of similar 17OHP levels and corticosteroid doses (Table 4).

Associations among TARTs, sperm number and quality, and the gonadotropin–gonadal steroid axis

As shown in Fig. 2, total functional testicular volume was positively correlated with inhibin B levels and sperm concentration. There was also a trend to a positive association with sperm volume (r=0.60, P=0.059) and a significant positive correlation with the testosterone/E₂ ratio (r=0.50, P=0.029). We also found the expected positive associations between inhibin B total functional testicular volume and the sperm count (r=0.64, P=0.014) and concentration (Fig. 2) and with normal morphology (r=0.48, P=0.061). FSH showed a negative correlation with sperm count (r=−0.53, P=0.047) and concentration (r=−0.70, P=0.004). The total TART volume tended to be negatively associated with height (r=−0.44, P=0.056) and was positively correlated with LH (r=0.41, P=0.046). There was no correlation with other hormones of the gonadotropin–testicular axis. No correlation was demonstrated between total TART volume or functional testicular volume and age, glucocorticoid dose, plasma renin, and ACTH concentrations (not shown).

Male with 3β-HSD type II deficiency

The patient with 3β-HSD deficiency was 21 years old, had male external genitalia with a micropenis, lived alone, and had no children. On testicular palpation, bilateral multiple tumors were found and ultrasound showed a very small total functional volume (1.9 cm³) due to large bilateral TARTs measuring 25.6 cm³. A tumor measuring 2.4×1.8×1.5 cm, with a weight of 3.2 g, was extirpated. He had no sperm in the semen sample, but a testicular biopsy was able to retrieve ∼100 severely abnormal sperms, which were saved for future ICSI. He had poor control with high DHEAS, and his FSH and LH were also high while inhibin B was low.

Male with karyotype 46,XX and 21-hydroxylase deficiency

The 46,XX CAH male was 61 years old and had a 4 cm phallos. He was diagnosed at age 7 when ovaries and a uterus had been found and was subsequently extirpated, following which he underwent glucocorticoid and testosterone replacement. He was married and had adopted two children.