Proportion of Cystic Fibrosis Gene Mutations Not Detected by Routine Testing in Men With Obstructive Azoospermia FREE

Cystic fibrosis (CF) is the most common serious autosomal recessive condition in whites, with a prevalence of approximately 1 per 2500 and a carrier frequency of 1 per 25.¹ The gene responsible for CF, called the cystic fibrosis transmembrane conductance regulator (CFTR), encodes the cyclic adenosine monophosphate (cAMP)–dependent chloride channel found in the apical membrane of secretory epithelial cells.¹ Clinical features of CF include chronic pulmonary obstruction and infections, exocrine pancreatic insufficiency, neonatal meconium ileus, elevated sweat electrolytes, and male infertility.^1- 2 The vast majority (>95%) of men with CF are azoospermic (absence of spermatozoa in ejaculate) due to anomalies in Wolffian duct-derived structures, with the body and tail of the epididymis, vas deferens, seminal vesicles, and ejaculatory ducts being atrophic or absent.^3- 6

Studies involving DNA mutation analyses of the CFTR gene have demonstrated that infertile men with obstructive azoospermia from congenital bilateral absence of vas deferens (CBAVD), congenital unilateral absence of vas deferens (CUAVD), or epididymal obstruction have a considerably higher incidence of mutations in the CFTR gene compared with the general population.^7- 33 The CFTR gene alterations associated with the obstructive azoospermic conditions are described rarely within the classic CF population, or they are considered to be of a mild form, such as the 5-thymidine variant of the polythymidine tract (IVS8-5T) in the splice acceptor site of intron 8.^{13- 15,24,26- 27,29,31- 34} However, most CFTR mutation screening panels test only for the more common mutations associated with CF, thus potentially missing rare CFTR gene alterations or those that may confer or contribute to an obstructive azoospermia phenotype.

Another recent twist is that advances in assisted reproductive technologies with in vitro fertilization–intracytoplasmic sperm injection (in which a single sperm is injected into the cytoplasm of a mature oocyte to obtain a viable embryo) have made it possible for these men to father their own children.^35- 50 One major concern with this procedure is that the genetic defect inherent in the patient (ie, CFTR mutations) can be iatrogenically transmitted to the offspring.^51- 56

Many of the articles on the molecular genetics of obstructive azoospermia advocate the evaluation of individuals with this condition and possibly their female partners for mutations or variants in the CFTR gene. However, these studies vary in their methods for detecting mutations. Most investigators have tested for a set of common CFTR mutations associated with the CF phenotype, while others performed a complete screening of all CFTR exons and flanking introns followed by direct sequencing as indicated. The purpose of our study was to assess the significance of evaluating not only common CFTR gene mutations associated with CF but also other alterations or variants using a more extensive screening method to gain insight into whether there is any yield in using such a technique. As outlined above, this issue has important clinical implications in the present era of reproductive options for men with obstructive azoospermia, many of whom may now realize biologic fatherhood due to the advent of successful gamete micromanipulation.

One hundred sixty-three otherwise healthy male subjects with obstructive azoospermia and 63 otherwise healthy male subjects with nonobstructive azoospermia from primary testicular failure presenting consecutively to the Andrology Clinic at the Mount Sinai Hospital, Toronto, Ontario, between 1995 and 1998, were approached for CFTR genotype analysis after giving informed consent and receiving appropriate genetic counseling. One hundred forty-nine men with obstructive azoospermia and 49 men with nonobstructive azoospermia consented to participate in the study, which was approved by the hospital's institutional review board and the Human Subjects Review Committee of the University of Toronto. A subset of study subjects (n=46) was previously described in an investigation examining the association between anatomic anomalies and CFTR gene mutations.⁵⁷ Some subjects (n=62) were also previously described with regard to obstructive azoospermia and CFTR genotype.²¹ Clinical evaluation included history taking, physical examination, semen analysis, and serum hormonal profile (luteinizing hormone, follicle-stimulating hormone, testosterone, and prolactin). Additional investigations such as scrotal and/or transrectal ultrasonography, testicular biopsy, and scrotal exploration were performed as indicated for diagnostic confirmation. Of the 149 obstructive azoospermic subjects, 64 were diagnosed as having CBAVD (51 white, 1 black, 11 Asian, 1 interracial [white/black]), 10 as having CUAVD (7 white, 3 Asian), and 75 as having epididymal obstruction (53 white, 8 black, 13 Asian, 1 interracial [white/Asian]). Of the 49 nonobstructive azoospermic subjects, 41 were white, 1 was black, and 7 were Asian. The racial proportions of our study groups reflected those of the Ontario population.⁵⁸

Peripheral venous blood for CFTR genotype analysis was obtained from each of the 198 subjects. Genomic DNA was isolated from lymphocytes according to standard protocols,⁵⁹ and then subjected to the following molecular evaluations:

Differences in mutation frequencies between men with the various types of obstructive azoospermia (CBAVD, CUAVD, or idiopathic epididymal obstruction) and men with nonobstructive azoospermia were compared by the χ² statistic or Fisher exact test. All P values were based on 2-sided comparisons and P <.05 indicated statistical significance. Estimates of the proportion of mutations identified by each test were made with 95% confidence intervals (CIs) for proportions using a normal approximation.

A total of 49 subjects with nonobstructive azoospermia due to primary testicular failure were assessed for CFTR gene mutations (Table 1). Two CFTR mutations (ΔF508, G542X) were identified by analysis using the 31 CFTR mutation panel (Table 2). The IVS8-5T variant was identified in 5 alleles. Two mutations (the identical ΔF508 and G542X mutations detected by the 31 mutation panel) were identified through screening of all CFTR exons and splice sites. The allelic frequency of the IVS8-5T variant was not significantly different between nonobstructive azoospermic subjects (5.1%) and the general population (26 [5.2%] of 498 chromosomes based on data pooled from Kiesewetter et al,⁶³ Cuppens et al,⁶⁵ Dork et al,⁶⁶ and Chillon et al¹³). Also, the carrier frequency of CFTR gene mutations (2 [4%] of 49 subjects) did not differ significantly from the carrier frequency (4%) in the general population.¹

A total of 149 subjects with obstructive azoospermia (64 with CBAVD, 10 with CUAVD, and 75 with epididymal obstruction) were evaluated for mutations in the CFTR gene. The 31 common CF-associated CFTR mutations, the IVS8-5T variant, and the entire coding sequence of the CFTR gene were assessed (Table 1).

For those in the CBAVD group, analysis of the 31 most common CFTR mutations associated with CF led to the identification of 39 mutations in 128 alleles (P<.001 for comparison with the nonobstructive azoospermia group) (Table 2). The mutation most frequently identified was ΔF508 (23 alleles), followed by R117H (9 alleles). Others detected using the 31 mutation panel included W1282X (2 alleles), R334W (2 alleles), S549R (1 allele), R347P (1 allele), and N1303K (1 allele). Twenty-five chromosomes tested positive for the IVS8-5T variant (P<.001 for comparison with the nonobstructive azoospermia group). Forty-seven mutations were detected by screening all CFTR exons for sequence alterations, including the same 39 mutations identified by the routine 31 mutation panel (P<.001 for comparison with the nonobstructive azoospermia group). Of the 8 additional CFTR gene sequence alterations detected using extensive CFTR exon screening, 5 have been described rarely in the CF population (L206W [identified in 2 subjects], P67L, 1677delTA, R117L, and 4016insT).⁶⁰ One mutation, D979A, was previously identified in a Vietnamese CBAVD patient.⁶⁰ Interestingly, our CBAVD subject with D979A (also a carrier of IVS8-5T) was of Vietnamese descent as well. The M952T mutation is a previously unreported, novel mutation. Therefore, of the alleles with identifiable CFTR mutations in the CBAVD group, 39 (54%), 25 (34%), and 47 (65%), respectively, were detected by the routine mutation panel, IVS8-5T allele-specific oligonucleotide hybridization, and extensive screening of all exons.

In the CUAVD group, of 10 subjects 2 had the ΔF508 mutation, detected using the 31 mutation panel analysis and also using the extensive CFTR screening, while 2 had the IVS8-5T variant (Table 2). Thus, for the CUAVD group, 50% of the mutations were identified by the 31 mutation panel (2/4), allele-specific oligonucleotide hybridization for IVS8-5T (2/4), and screening of all CFTR exons (2/4).

Epididymal obstruction can be caused by genitourinary surgery or infection and Young syndrome.⁶⁷ Young syndrome is characterized by chronic sinopulmonary infections and epididymal obstruction.⁶⁸ This condition may have an acquired origin⁶⁹ and does not appear to be associated with CFTR gene mutations.^26,70 Of the 75 subjects with epididymal obstruction, 8 had a history of genitourinary surgery, 10 had a history of genitourinary infection, and 1 had a history consistent with Young syndrome. The remaining 56 men were considered to have idiopathic epididymal obstruction. None of the 19 subjects with nonidiopathic epididymal obstruction were found to have CFTR mutations using the routine panel or by gene scanning. Furthermore, the IVS8-5T allele was not detected in any of these 19 subjects. Of the 56 subjects with idiopathic epididymal obstruction, analysis using the 31 mutation panel resulted in the identification of 5 mutant alleles: ΔF508 (2 alleles), W1282X (2 alleles), and R117H (1 allele) (Table 2). The IVS8-5T variant was found in 12 alleles. Evaluation of all CFTR exons led to the identification of 12 sequence alterations (P=.01 for comparison with the nonobstructive azoospermia group), including the same 5 mutations detected by the 31 mutation panel. Of the 7 additional mutations, G544S, 2423delG, V754M, and −741T→G are associated with the CF phenotype and R258G with the CBAVD phenotype, while R75Q (identified in 2 subjects), previously thought to be a benign polymorphism, may in fact confer phenotypic features of CF.² Therefore, of the 24 alleles with CFTR mutations in men with idiopathic epididymal obstruction, 5 (21%) were identified by routine CFTR mutation analysis, 12 (50%) by IVS8-5T allele-specific oligonucleotide analysis, and 12 (50%) by complete analysis of all CFTR exons.

In our subjects with CBAVD, CUAVD, and epididymal obstruction, an additional 46% (95% CI, 34%-58%), 50% (95% CI, 0%-100%), and 79% (95% CI, 63%-95%) of alleles with mutations or variants, respectively, would have escaped detection if assessment for IVS8-5T and an extensive screen of CFTR exons had not been performed (Table 2). It has been shown that the single most common CFTR sequence alteration in men with obstructive azoospermia is the IVS8-5T variant,^{13- 15,24,26- 27,29,31- 34} and it is likely that most centers will routinely test for this allele. Nevertheless, even if IVS8-5T were included as part of the routine screening process, comprehensive analysis of all CFTR exons led to the identification of an additional 8 (17% [95% CI, 6%-28%]) of 47 sequence alterations in the men with CBAVD and an additional 7 (58% [95% CI, 30%-86%]) of 12 sequence alterations in those with epididymal obstruction (Table 2).

To date, all published reports of CFTR mutation analysis in men with obstructive azoospermia have either used a screening panel of the common CFTR gene mutations associated with the classic CF phenotype or examined all CFTR exons and immediate flanking introns. To our knowledge, the present study reports, for the first time, a comparative analysis of both methods of CFTR gene mutation testing in the same study population. We found that a substantial proportion of CFTR gene mutations in men with obstructive azoospermia consists of sequence alterations not detected by a routine panel designed for the white classic CF population.

More than 800 mutations in the CFTR gene have been reported.⁶⁰ These mutations may be classified as follows: defective protein production, class 1; defective protein processing, class 2; defective channel regulation, class 3; decreased channel function, class 4; and decreased protein synthesis, class 5 mutations.² In general, class 1, 2, or 3 mutations are expected to have more serious phenotypic consequences than class 4 or 5 mutations. The CFTR gene alterations identified in our subjects with Wolffian duct abnormalities can be divided into 4 groups. The first group consists of mutations that have been identified in classic CF patients. These severe CFTR gene mutations are associated with pancreatic insufficiency and are generally class 1 through 3 mutations: ΔF508, W1282X, N1303K, S549R, 1677delTA, R117L, 4016insT, G544S, 2423delG, V754M, and 741T→G. The second group consists of mutations found in more benign presentations of CF. These mild CFTR gene mutations are associated with pancreatic sufficiency and tend to be class 4 through 5 mutations: R117H, R334W, R347P, L206W, and P67L. The third group consists of mutations identified exclusively in some men with obstructive azoospermia; however, because these sequence alterations are extremely rare, it is only speculated that they contribute to this phenotype.^7,10,12 These CFTR gene sequence changes include D979A, R258G, and M952T. As mentioned, D979A has been previously detected in a Vietnamese man with CBAVD,⁶⁰ and interestingly, our CBAVD subject with D979A was also of Vietnamese descent. The fourth group consists of CFTR gene alterations that were formerly considered to be benign sequence variations.^7,12,14 These include R75Q and IVS8-5T. In a substantial proportion of our patients with congenital obstructive azoospermia (36/130 [28%]), only 1 mutation or sequence alteration (including IVS8-5T) was identified. Not unexpectedly, the men with CBAVD were more likely to have 2 detectable CFTR mutations compared with those with the milder phenotype of idiopathic epididymal obstruction. Except for 1 subject who had the ΔF508/S549R genotype, all 32 (25%) of 130 men with 2 mutant CFTR alleles carried at least 1 mild class 4 or 5 mutation.

The application of a comprehensive screen for CFTR mutations such as gene scanning (eg, multiplex heteroduplex analysis, denaturing gradient gel electrophoresis, single-strand conformation polymorphism analysis) followed by direct DNA sequencing in men with obstructive azoospermia should obviate testing for a set of common CFTR gene mutations since the former method would lead to detection of the mutations identified using such a panel and to the identification of rare or private mutations. Such a method should be more sensitive than others that directly screen for a limited number of mutations, as offered by some commercial laboratories such as Genzyme Genetics (Framingham, Mass), which tests for 70 CFTR mutations including the 31 mutations detected with the routine panel used in the present study. This 70 mutation panel would have detected no additional mutations not already identified by the 31 mutation panel in our subjects, thus emphasizing again the rarity of many of these CFTR gene mutations found in the obstructive azoospermic population. However, it must be stressed that screening for mutations in all coding regions and their immediately flanking intron sequences does not rule out mutations within the promoter region or introns of the CFTR gene. In addition, the absence of detectable sequence alterations in CFTR alleles in some subjects suggests that other genetic and/or environmental factors may contribute to the obstructive azoospermia phenotype.

The multiplex heteroduplex analysis used in the present study can detect approximately 95% of all known CFTR mutations, with a false-positive rate of about 2%.⁶⁴ We estimate that the cost of this method is around US $100, which is comparable to that of a mutation panel (approximately US $50-$150 per test for 6-72 CFTR gene mutations). The estimate for the CFTR gene scanning method does not include the cost of direct DNA sequencing.

The practice of routine screening of men with obstructive azoospermia for CFTR gene mutations that are commonly associated with the classic CF phenotype may involve the erroneous assumption that the types of mutations and their frequencies in men with obstructive azoospermia are the same as those in men with classic CF. If one were to adopt the strategy of screening obstructive azoospermic men with a limited panel of CFTR gene mutations, then it would be necessary to first determine which mutations are the most common in these men. In other words, one would need to design a panel of CFTR mutations that is specific for the obstructive azoospermia phenotype. The selection of specific CFTR gene mutations for screening infertile males should be based on the same assumptions that have been used for the classic CF population, which would take into consideration the frequencies of the different mutations or variants commonly associated with the isolated obstructive azoospermia phenotype and with the ethnicity of the study population. At the present time, except for the IVS8-5T variant, it is not known with certainty which CFTR mutations constitute a significant proportion of those carried in the obstructive azoospermia population. Nevertheless, as CFTR gene mutations from larger groups of infertile men are analyzed, one could potentially develop a panel of mutations specific for the isolated obstructive azoospermia phenotype.

Another pertinent issue that warrants discussion is the screening for CFTR gene mutations in the female partners of men with obstructive azoospermia. For the couple considering assisted reproduction, we and others have advocated testing of the spouse in the situation in which her partner was found to carry at least 1 CFTR gene mutation, regardless of its class severity.^55- 56,71 However, since a false-negative rate exists with any mutation screening method, the spouse may wish to be tested at the same time as her partner. Of course, the spouse would also be tested as soon as possible if there were a personal or family history suggestive of CF. What is unclear at the present time is the type of screening method one should use for these women. Compounding the problem is the lack of an absolute genotype-phenotype correlation with respect to CFTR gene mutation status and the as yet uncertain consequences of inheriting at least 1 mild mutation in the offspring, male or female, produced via in vitro fertilization–intracytoplasmic sperm injection. Further complicating the picture is the increasing evidence that CFTR gene mutations may contribute etiologically to certain monosymptomatic disorders, such as isolated nasal polyposis,⁷² disseminated bronchiectasis,^73- 74 allergic bronchopulmonary aspergillosis,⁷⁵ or chronic pancreatitis.^76- 77 Therefore, appropriate genetic counseling is required for these couples, and consideration should be given to prenatal or preimplantation diagnosis if both partners are carriers of mutant CFTR alleles, especially severe class 1 through 3 mutations. These recommendations on CFTR genotyping and genetic counseling are consistent with those outlined in the recent National Institutes of Health Consensus Statement on Genetic Testing for Cystic Fibrosis, which include offering CFTR gene mutation analysis to all couples who are considering pregnancy, regardless of infertility history.⁷⁸

In summary, most laboratories screen otherwise healthy men with obstructive azoospermia for the more common mutations associated with the classic CF phenotype. However, studies performing a more exhaustive search for CFTR gene sequence alterations have demonstrated that many individuals with isolated obstructive azoospermia have rare or private mutations. Since the obstacle to conception for these men has been overcome by in vitro fertilization–intracytoplasmic sperm injection, decreasing the risk of passing pathogenic CFTR mutations on to the progeny is of paramount importance. This consequence may be avoided through clinical strategies such as genetic counseling and preimplantation or prenatal diagnosis. Further identification of mutations related to isolated obstructive azoospermia and the determination of genotype-phenotype associations may contribute to enhanced clinical management of these individuals.