Prevalence of CBAVD in azoospermic men carrying pathogenic CFTR mutations ‐ Evaluated in a cohort of 639 non‐vasectomized azoospermic men

Abstract Background Men with obstructive azoospermia (OA) due to impaired development of the genital tract often carry at least one Cystic Fibrosis Transmembrane Conductance Regulator CFTR mutation. Objective To determine the frequency of Congenital Bilateral Absence of Vas deferens (CBAVD) in men with azoospermia carrying CFTR gene mutations. Materials and methods Non‐vasectomized men with azoospermia referred to our andrological center were consecutively included. All men underwent palpation of the scrotal parts of the Vasa deferentia, ultrasonography of the testicles and hormone profile, and genetic analyses. Testicular biopsy was usually performed. A panel of 32 of the most important CFTR mutations was examined from genomic DNA isolated from blood lymphocytes. Either multiplex PCR analysis or a next‐generation sequencing technique was performed. Results Among the 639 men with azoospermia, 69 (10.8%) had at least one CFTR mutation. Of the 43 patients with at least one of the two CFTR mutations, ΔF508 and R117H, 19 (44.2%) showed CBAVD, 2 (4.7%) Congenital Unilateral Absence of Vas deferens (CUAVD), and 22 (51.2%) presence of the scrotal parts of the Vasa deferentia. In contrast, only 1/21 men (4.8%) with an isolated IVS8‐5T variant showed CBAVD. Among the further 20 men with an isolated IVS8‐5T variant, 11 had a history of cryptorchidism. Among the 570 men without CFTR mutations, CBAVD was found in only two men and CUAVD in one. FSH level was higher and testicular volume lower in men with present Vasa deferentia compared to those without (P < .001; Student's t test). Thirty‐one men with either ΔF508 or R117H mutations, or both, had a testicular biopsy. Motile spermatozoa were found in 100% of 16 cases with CBAVD but in only 6 out of 15 cases with present Vasa deferentia (P < .01; Fisher's exact test). Discussion and conclusions CBAVD was found in ~ 44% of men with ΔF508/R117H mutations. The data may support that CFTR mutations might affect male fertility through other mechanisms than obstruction of the genital tract. For a practical, clinical purpose analysis for only ΔF508, R117H and IVS8‐5T seems sufficient until further research shows anything else.


| INTRODUC TI ON
During more than 50 years it has been known that the majority (97%) of men with Cystic Fibrosis (CF) exhibit Congenital Bilateral Absence of Vas Deferens (CBAVD). 1 While only persons homozygous for two severe pathogenic Cystic Fibrosis Transmembrane conductance Regulator (CFTR) mutations develop the disease CF, men with CFTR compound heterozygoti or a single pathogenic CFTR exon mutation, in some cases combined with an intron 8 5T splice variant, may have CBAVD as an isolated phenotype without symptoms from the respiratory tract and gastrointestinal system. 2 CFTR mutations are also suggested to modulate the epididymal microenvironment and to inhibit spermatogenesis and capacitation, 3 but the exact mechanism is unclear.
The CFTR gene, located on chromosome 7 (q 31.2), consists of 27 exons and 26 introns. The CFTR gene encodes a channel protein that enables chloride passage through apical cell membranes. The well known CFTR mutations ΔF508 4,5 and R117H are located in exon 10 and exon 4, respectively. In addition to the exon mutations, intron mutations -or variants -are important. Thus, the polythymidine sequence located in the splicing acceptor site of intron 8 is found in three variants with 9, 7, or 5 thymidines: the IVS8-9T, IVS8-7T, and IVS8-5T, respectively. The 9T and 7T alleles predominantly generate normal mRNA transcripts, whereas the 5T variant affects exon 9 resulting in reduced levels of normal mRNA. The 5T allele is thus considered as a mild pathogenic mutation with incomplete penetrance.
The relative contributions of different CFTR mutations and variants show considerable variation between different populations, [6][7][8] the ΔF508 being the dominant CFTR mutation in Caucasians. 2 In Caucasians, CFTR exon mutations occur with a prevalence of around 4% in the background population. 9 Many studies have documented an extremely high prevalence of CFTR mutations in men with obstructive azoospermia (OA) and CBAVD. CFTR mutations occur in up to 40% of non-vasectomized men with OA. 10 In a metaanalysis including 38 studies, 78% of men with CBAVD carried one or two (a severe and a mild or two mild) CFTR mutations. 2 Conversely, the risk to develop CBAVD when carrying CFTR mutations is insufficiently studied. A reasonable cause may be that most andrologists and fertility clinics only perform CFTR gene analysis in men with OA and maybe even CBAVD/ Congenital Unilateral Absence of Vas deferens (CUAVD). Thus, the present cohort is unique since all men with azoospermia were tested for the clinically most relevant CFTR mutations, in addition to careful clinical examination, including palpation of the scrotal parts of the Vasa deferentia.
As it is unclear if CFTR mutations may also decrease fertility by mechanisms other than genital tract obstruction, 3 the first aim of this study was to estimate the prevalence of CBAVD in men with azoospermia carrying one or two CFTR mutations. While the prevalence of CFTR in CBAVD has been analyzed in other studies, the converse consideration thus is quite new. The second aim was to clarify if analysis for CFTR mutations should be performed in men with non-obstructive azoospermia (NOA) and OA without CBAVD/ CUAVD. This second aim is not quite new, 11,12 but conclusions are still very controversial. Thus, as mentioned above, CFTR mutations are suggested to inhibit spermatogenesis. 3 To answer this question, we compare FSH levels, testicular and ejaculate volumes, testicular histology, and rates of presence of motile, testicular spermatozoa between azoospermic CFTR carriers with and without CBAVD.
These data will be analyzed in relation to clinical data, to evaluate if other clinical conditions can explain possible differences.

| Patients
Non-vasectomized men with azoospermia, defined as at least two consecutive ejaculates without spermatozoa in the pellet after centrifugation, have been evaluated using an almost identical evaluation program since 1997. Until May 1., 2011, patients were examined in the Fertility Clinic, Braedstrup Hospital. 13 Hereafter patients were examined in the Centre of Andrology and Fertility Clinic, OdenseUniversityHospital, as the authorities selected this center to be one of two highly specialized centers of Andrology in Denmark. The included azoospermic men represent the majority of azoospermic men evaluated for infertility in Western Denmarkin the period mentioned.
Thus, since about 1% of the male population is expected to have azoospermia, 14 an incidence of almost 150 new cases is expected in Western Denmark every year. We have about 75 new cases referred every year. The explanations of the difference may be, that some may choose treatment with donor semen without further examination, choose to live without having children, and a few with a simple etiology may have been treated in a fertility clinic not specialized in andrology. In addition to vasectomized men, men with aspermia due to ejaculatory disorders such as anejaculation or retrograde ejaculation 15 were not included in this study.

| Clinical examination program
Men with azoospermia underwent a detailed history interview and careful clinical examination including scrotal palpation of the clinical purpose analysis for only ΔF508, R117H and IVS8-5T seems sufficient until further research shows anything else.

K E Y W O R D S
azoospermia, CFTR mutations, CBAVD, CUAVD, testis testicles, the Epididymides, and the Vasa deferentia. History interview included questions about cryptorchidism, genital tract infections, cancer, and other former or present diseases, trauma, irradiation, chemotherapy, working health, heated seats, sauna, medications, smoking, and alcohol use. 16 Additionally, ultrasonography of the scrotum, including measurement of the testicular volumes, was performed as described by Fedder. 16 These clinical examinations, including ultrasonography, were in all cases performed before genetic and hormonal analyses. Hormone profiles, including follicular stimulating hormone (FSH), luteinizing hormone (LH), testosterone, prolactine, and thyroid-stimulating hormone (TSH), karyotype analyses, and analyses for Y microdeletions and the clinically most relevant CFTR mutations were in every case performed. Furthermore, in recent years, inhibin-B, anti-Müllerian hormone (AMH), and serumestradiol were also determined.
Men with elevated FSH levels, small testicular volumes, and testicular histology showing hypospermatogenesis, maturation arrest, or Sertoli cell-only syndrome (SCO) were categorized as having non-obstructive azoospermia, while men with normal FSH levels, normal-sized testicles, and normal testicular tissue at histological examination were categorized as having obstructive azoospermia.

| Analyses of CFTR mutations
Genomic DNA was extracted from blood lymphocytes and analyzed according to routine analyses. During the entire period, all men with azoospermia were examined for the three exon CFTR mutations:

| Karyotyping and examination for Y microdeletions
These analyses were performed as described elsewhere. 10 After harvesting and culture of blood lymphocytes and blockade of mitosis, the metaphase chromosomes were stained with Giemsa (G). Based on the resulting G banding it was possible to identify each chromosome and construct the karyotype.
For the detection of Y microdeletions, based on extracted genomic DNA, at least two sequence-tagged sites (STSs) for each of the three regions: AZFa, AZFb, and AZFc were used for detection of deletions in the respective regions. As controls, STSs on Yp and Yq were used. 10

| Testicular biopsies and testicular histology
Unless the patient had Klinefelter's syndrome or primarily asked for treatment with donor semen, diagnostic testicular biopsy was usually performed. Sperm retrieval was usually performed as TEsticular

| Ultrasonography of the kidneys in azoospermic men with CUAVD
Since an increased frequency of renal aplasia has been suggested in the literature in patients with CUAVD, 22 the azoospermic men with CUAVD underwent ultrasonography to visualize the kidneys.

| Statistics
Comparisons on parametric data such as FSH, testicular volumes, and ejaculate volumes were calculated using Student's t test after verification of normality. Comparisons on categorical variables such as numbers with testicular spermatozoa or normal histology in relation to CFTR mutations or presence of CBAVD and implantation and pregnancy rates were performed using the Fisher's exact test.

| Approvals
The study was approved by the Danish Patient Safety Authority (journal nr 3-3013-2503/1) and the Danish Data Protection Agency (journal nr 18/18147).

| RE SULTS
Among this cohort of azoospermic men, more than 80% showed NOA. One-hundred and twenty-five men (19.6%) showed KS or Y microdeletions, and in addition 173 (27.1%) had a history of cryptorchidism. Less frequent occurring causes included hypogonadotrophic hypogonadism and chemotherapy due to malignancy.
As shown in Table 1 Table 2).
CFTR mutations other than ΔF508, R117H, and IVS8-5T were found in only 5 (1.1%) of the 444 azoospermic men exposed to the big panel analyzing for 32 CFTR mutations. Among these 5 men, only 1 showed CBAVD, while other explanations of the infertility should probably be found in the other 4 men.
In a couple of cases the family, including the parents of the man, was examined for the presence of CFTR mutations, and in these cases we are sure that the two CFTR mutations are located on two different alleles ("trans" position). In other cases, for example, men with both ΔF508 and R117H, we find it unprobably that both mutations should be located on the same allel ("cis" position). With the assumption that CFTR mutations in men with more than one CFTR mutation are in all cases located in a "trans" position (compound heterozygoti), the allel frequency of CFTR mutations (IVS8 variants excluded) can be calculated to 4.4% (ΔF508: 2.5%, R117H: 1.4%, other CFTR mutations: 0.5%).
The majority of the 69 non-vasectomized men with azoospermia carrying CFTR mutations were Danish; only 11 (15.9%) of the men were of other nationalities. Thus, of the 25 men carrying an isolated ΔF508 mutation with or without an accessory IVS8-5T variant (Table 3) (Table 2).
Remarkably, FSH levels were higher (P = .0003, t test; Table 3) and testicular volumes (both testicles) lower (P = .0000006, t test; These findings were furthermore supported by the histological patterns observed in testicular biopsies. Not all men had testicular biopsy, for example, two men with ΔF508 and present Vasa deferentia also had KS. In men with CBAVD, normal testicular tissue was found in all cases, while affected spermatogenesis (hypospermatogenesis, maturation arrest, or Sertoli cell-only syndrome) was found in 76% of 34 cases with presence of both Vasa deferentia (P = .00001, Fisher's exact test; Table 4). However, since normal testicular tissue was found in 8 men with both Vasa deferentia present, an undetected obstruction might be suggested in these cases.
While IVS8-5T was not associated to CBAVD, surprisingly 11 of 20 cases (55%) had an orchiopexy during childhood (Table 5; P = .004, chi-square test) suggesting a possible association between IVS8-5T and cryptorchidism. The coincidence between IVS8-5T and chemotherapy or KS (Table 5) in a few cases probably is random.
Coincidental diagnoses were relatively infrequent in CFTR exon mutation carriers. Two ΔF508 carriers with present Vasa deferentia had a KS karyotype (pure 47,XXY) and two CFTR carriers a Y microdeletion (AZFbc in a ΔF508 and AZFc in a R117H carrier). In addition, 3 ΔF508 carriers with present Vasa deferentia were treated for cryptorchidism in childhood (2 with orchiopexia), but also 2 CBAVD patients (1 with a R117H and 1 with a c3659del mutation) and 1 CUAVD patient (R117H + 394delTT carrier) were treated for cryptorchidism in childhood (2 with orchiopexia). In addition, one CBAVD patient carrying a ΔF508 mutation had taken anabolic steroids for  Table 3). Renal aplasia could not be detected with ultrasonography in any of the three men with CUAVD.

| D ISCUSS I ON
Traditionally azoospermia has been divided into obstructive (OA) and non-obstructive azoospermia (NOA). This division not always is quite unambiguous, since a few patients may be a little difficult to categorize, and some men may develop impaired spermatogenesis secondary to OA. 23 However, due to our usual criteria the proportion of azoospermic men with NOA exceeds 80% and may have slightly increased during the study period (Table 1). This may be due to referral of an increased number of men with KS (P = .002, χ 2 -test), while a trend for reduced proportions of azoospermic men with a history of cryptorchidism or CFTR mutations was observed (Table 1) One patient had testicular biopsy before referral to our department, and immediate examination for living spermatozoa (Table 3) was not performed in this case.
20-year long study period. We therefore suggest our cohort to be representative for men with azoospermia in general.

| Main findings/What is new?
In this study including 639 consecutive men with azoospermia, one or two CFTR mutations/variants were detected in 69 (10.8%). Of those only 21 (30.4%) had CBAVD and 2 (2.9%) CUAVD. CBAVD was found in ~ 44% of ΔF508/R117H carriers but only in ~ 5% of patients with an isolated IVS8-5T variant. CBAVD or CUAVD without an associated CFTR mutation was found in only 3 cases (0.5%) in this study. CFTR mutation carriers without CBAVD showed higher FSH levels, lower testicular volumes, and a lower frequency of testicular spermatozoa.
While the prevalence of CFTR mutations in men with OA or CBAVD has been analyzed in several studies, we demonstrate for the first time which prevalence of CBAVD can be expected in azoospermic men with CFTR mutations. The study also confirms that sperm retrieval is a successful treatment in men with CBAVD, since motile testicular spermatozoa were found in 100% of 16 men with CBAVD undergoing testicular biopsy. Since the prevalence of CFTR mutations other than ΔF508, R117H, and IVS8-5T seems extremely low, the study raises doubt about the benefit of extending the CFTR analysis from these three analyses to analysis of 32 mutations -at least from a simple clinical point of view.

| Strengths
One strength of this study is that it was performed on a consecu-

| Limitations
The inclusion of an increasing number of men with Klinefelter's syn- were examined in all cases.

| Prevalence
In this study, the prevalence of CFTR mutations and variants in men with azoospermia without CBAVD/CUAVD (7.5%) and of CFTR exon  The one with the chromosome 1 inversion mutations alone (4.2%, Table 1) only slightly exceeds the expected prevalence, since the prevalence of CFTR exon mutations in the general population is 3%-4%. 9 The ΔF508 mutation is the most prevalent mutation, particularly in Caucasians, 2 where it is the most prevalent CFTR mutation found in CBAVD. 24 Within 50 Iranian men with CBAVD, ΔF508 was found in 16% 6 and among 60 Indian CAVD patients (do not distinguish between CBAVD and CUAVD) ΔF508 was found in 26.7%. 7 In contrast, ΔF508 mutations were found in 9.4% of 53 Iranian men with non-CBAVD obstructive azoospermia 25 and in 11.6% of 60 Indian men with non-CBAVD obstructive azoospermia. 26 ΔF508 mutations are rare in China. 8,27 The R117H mutation is not very frequent globally 24 but, in accordance with this study, it may be more frequent in the Western world. 28 The IVS8-5T variant seems to be prevalent not only in Caucasians but also worldwide. 2

| Possible mechanisms
One explanation for the higher prevalence of CFTR mutations/variants in non-CBAVD men with azoospermia might be that the geni- It has been found that the CFTR gene, which encodes the chloride channel, can be detected in the epididymides and Vasa deferentia. 31 The CFTR protein is located in the apical membranes of the secretory epithelial cells. 31

| Possible effects on fertility
From minor studies, it may be suggested that testicular spermatozoa retrieved from men with NOA using microdissection TESE have a reduced potential in fertilizing oocytes compared to ejaculated donor spermatozoa. 32 The current data on fertility are very scarce, and this study has not sufficient power to conclude, that the fertilizing capacity of the spermatozoa from CFTR mutation carriers with present Vasa deferentia is reduced compared to CFTR mutation carriers with CBAVD. However, in an Australian study a 17% implantation rate was obtained after transfer of embryos fertilized with fresh, epididymal, or testicular spermatozoa retrieved from men with CBAVD. 33 Unfortunately, Phillipson et al 33 did not have sufficient power to distinguish between use of epididymal and testicular spermatozoa. Unless reconstructive surgery of the genital tract might be considered, it is always worth to consider retrieval of spermatozoa from the epididymis to avoid the risk of testicular traumas. 34 However, in patients with CFTR mutations and CBAVD the testicles use to be of normal size, and therefore, it is usually very easy to perform a testicular needle biopsy with a very low risk of complications.
In addition, in a clinical study in which epididymal spermatozoa from 63 patients with surgically confirmed congenital absence of Vas deferens (64% with a CFTR mutation) were used for in vitro fertilization of oocytes, CFTR mutations were associated with low fertilization and pregnancy rates. 35 Immediately, it may be suggested that the reason may be that the epididymal spermatozoa are "old," since they have not been able to leave the epididymis due to obliteration of the genital tract. 36 However, since the 21 men with an isolated ΔF508 mutation showed particularly low fertilization and pregnancy rates compared to other CFTR mutations, the effect of ΔF508 was suggested to be moderated by other CFTR mutations present in patients who are compound heterozygotes. Another explanation may be that patients with apparently isolated ΔF508 mutations might have another undetected CFTR mutation. 35 It is not possible to repeat this study to observe the ability of the spermatozoa to penetrate the oocyte investment and fertilize the oocyte, since the majority of patients currently undergo IntraCytoplasmic Sperm Injection (ICSI) using spermatozoa retrieved from one of the testicles.

| Association with other genes
As mentioned in the introduction the IVS8-5T variant shows variation in penetrance, which may explain the reasons for variation observed in spermatogenesis in this study. It appears that IVS8-5T may affect sperm production than rather resulting in CBAVD -at least in our population. However, the incomplete penetrance of the polythymidine allele 5T seems to be associated with other polymorphic loci. Thus, the number of TG repeats in the (TG)m locus located in intron 8 (like the polythymidine sequence), affects the amount of CFTR gene transcripts produced. 37 Furthermore, the M470V (1540A/G) polymorphism in exon 10 affects CFTR function. CFTR genes containing the V allele (V470) have lower functional ability compared to CFTR genes with the M allele (M470). Therefore, men with a 5T-TG12-V470 haplotype have a much higher risk of having CBAVD. 27,37 In this study, IVS8-5T appeared to affect spermatogenesis rather than resulting in CBAVD. Future research is required to examine if variations in the number of TG repeats in intron 8 or M470V polymorphism in exon 10 may explain this discrepancy.
In this study, only three men with azoospermia (~0.5%) were diagnosed with CBAVD or CUAVD without detection of a CFTR mutation.
In other studies, it has not been possible to detect CFTR mutations or variants in more than 20% of CBAVD patients. 2

| Relationship to urinary tract malformations
An association between CBAVD/CUAVD and urinary tract malformations has been described in literature. 40 Although renal abnormalities such as unilateral renal agenesis have been described in men with CBAVD but without detectable CFTR mutations, 40,41 renal agenesis is particularly frequent in men with CUAVD. 22,42 Thus, Casals et al 20 found renal anomalies in 41% 10 of 24 men with CUAVD compared to 5.4% of men with CBAVD. That is the reason that we only performed renal ultrasonography in the three patients with CUAVD, and renal abnormalities could not be detected in any of these three men. Since renal abnormalities occur less frequently in CBAVD/CUAVD patients with CFTR mutations compared to CBAVD/CUAVD patients without CFTR mutations, 22,24,40 it may be suggested that renal abnormalities coexisting with CBAVD/CUAVD in men without CFTR mutations may be due to a non-CFTR mechanism affecting major parts of the Wolffian ducts. In recent studies, it has been suggested that the polythymidine sequence and the numbers of TG repeats in exon 8 and the M470V polymorphism in exon 10, in addition to male reproductive ability, may also influence the risk of having renal abnormalities. 43 Furthermore, also EDNRA and TGF-β1 gene polymorphisms seem to affect the incidence of CBAVD as well as development of the kidneys, seminal vesicles, epididymides, and ejaculatory ducts in men with azoospermia carrying CFTR mutations. 7

| CON CLUS I ON S AND FUTURE DIREC TIONS
In conclusion, CBAVD was found in ~ 44% of a relatively unselected population of azoospermic men with ΔF508/R117H mutations. In addition, since normal testicular tissue was found in 8 men with present Vasa deferentia, it may be suggested that an undetected obstruction may be present in some cases. However, the present data also supports the hypothesis that CFTR mutations might affect male fertility through mechanisms other than obstruction of the male genital tract. Interestingly, we found a very high association between presence of the CFTR intron variant IVS8-5T and a history of cryptorchidism requiring orchiopexia, which has to be further investigated in up-coming studies. Furthermore, it will be interesting to study whether the number of TG repeats in exon 8 and the M470V polymorphism in exon 10 affects testicular histology and spermatogenesis in men with the IVS8-5T variant.
From a practical clinical point of view, the benefit by extending the CFTR analysis from ΔF508, R117H, and IVS8-5T to 32 mutations was very modest. Thus, it may be considered to apply the big panel with 32 mutations to only men with OA or maybe even only men with CBAVD/CUAVD, until further research shows if also other categories of infertile men should be examined for CFTR mutations and for which? Future research will determine whether analysis of sudden CFTR mutations (eg, the IVS8-5T mutation) should also be extended to other categories of infertile men, for example, men with severe oligozoospermia and a history of cryptorchism or patients with reduced ejaculate volume or very viscous semen.

ACK N OWLED G EM ENTS
We are thankful for the support and assistance provided from our technicians, andrology nurses, and colleagues from the fertility clinics in Braedstrup and Odense, Department of Clinical Genetics, Lillebaelt Hospital, and Department of Pathology, Odense University Hospital.

CO N FLI C T O F I NTE R E S T S
The authors declare no conflict of interest.

AUTH O R ' S CO NTR I B UTI O N S
JF was responsible for the concept, and he performed history interview and careful clinical examination including ultrasonography of the scrotum and testicular biopsy on all patients. He ordered blood samples and wrote the manuscript draft. MWJ performed CFTR mutation analyses, and BE performed histological examination of testicular biopsies. All authors approved the final version of the manuscript.