Immune Aspects of Infertility

Spermatozoa have an unusual relationship with the immune systems of both men and women. Although they are produced by men, they bear new developmental antigens to which the male immune system is not tolerant. In women, spermatozoa periodically invade their bodies as foreigners. Spermatogenesis begins during puberty, long after the immune system has learned to distinguish the antigenic expression of its own tissues and extracorporeal antigens. Despite the occurrence of new developmental antigens on sperm,¹ an autoimmune response against spermatozoa is uncommon, occurring only in approximately 7–8% of unselected men from infertile couples.²

After their intravaginal deposition, spermatozoa pass through various compartments of the female reproductive tract that are capable of mounting an immune response yet fail to elicit one.^{3, 4} Clues to both the male's and female's lack of immunologic responsiveness to sperm reside within the immune system itself. Immunosuppressive factors exist within seminal fluid, and a population of T-suppressor lymphocytes has also been detected both within seminal fluid and in the interstitium of the testis and the submucosal regions of the epididymis.^{5, 6}

It has been known since the late 19th century that spermatozoa are antigenic and that experimental animals can be immunized with sperm.^{7, 8} In nonhuman species, these immunized animals exhibit diminished fertility.^{9, 10, 11} Although preliminary evidence suggested an association between the presence of circulating antisperm antibodies in men and women and infertility, it has only been in recent decades that many of the concepts dealing with immune-mediated reproductive impairment have been clarified.

The concept of immunologic infertility stems from several key observations. It was demonstrated in the 1950s that autoimmune responses to sperm and testis antigens had immunologic consequences. Guinea pigs immunized with homologous testis antigens in complete Freund's adjuvant developed autoimmune responses to the antigens, leading to orchitis and aspermatogenesis.¹² The clinical relevance of antisperm autoimmunity became apparent in 1959, when spontaneous agglutination of ejaculated sperm and serum antisperm antibodies were observed in infertile men by Rumke and Hellinga.¹³ During the same period, Franklin and Dukes¹⁴ noted the presence of sperm agglutinating activity in sera of women with unexplained infertility. This was soon followed by the detection by Fjallbrant¹⁵ and Isojima and colleagues¹⁶ of serum antisperm antibodies that promoted complement-dependent sperm immobilization in infertile women. When endocrinologic problems of ovulation, sequelae of genital tract infections, and abnormalities of sperm production are excluded, there remain a substantial number of patients with idiopathic infertility. In this group, antisperm antibodies appear to have an important etiologic role.¹⁷

Proof that antisperm antibodies have a role in infertility requires documentation of their presence in sera of men and women, demonstration of these immunoglobulins bound to the sperm surfaces, as well as evidence of an alteration in the ability of such immunoglobulin-coated sperm to function – that is, to enter the female reproductive tract and either to reach the site of fertilization or to fertilize eggs.

Until the late 1970s, most laboratories attempted to diagnose immunities to sperm by using one of three major forms of serologic tests: sperm agglutination (either macroscopic or microscopic), complement-dependent sperm immobilization, or indirect immunofluorescence. Although several studies documented a clear difference in the prevalence of antisperm antibodies in sera of infertile couples as compared with fertile men and women, other studies could not make this distinction between groups. The range of positive results for agglutination assays has varied from 2% to more than 30%, with a large overlap between fertile and infertile groups (Table 1). Although complement-dependent immobilization provided more discriminating power, in that the incidence of false-positive results was far lower in sera of fertile men and women, this test suffered from an inability to detect noncomplement-fixing immunoglobulins, such as IgA.^{18, 19} Also important was the realization that naturally occurring antisperm antibodies exist in many species, including humans, at low titer. The sera of approximately 60% of men and women possess immunoglobulins that react with sperm, as detected by indirect immunofluorescence.²⁰ These naturally occurring antibodies are readily absorbed by spermatozoa and testicular extracts but not by other human tissue. Of importance, they do not stain the surface of viable sperm in suspension but rather are directed against intracellular antigens of methanol-fixed permeabilized sperm. In one study, sera possessing antiacrosomal antibodies, when absorbed with bacteria of several species, no longer reacted with sperm. These naturally occurring antisperm antibodies have been shown to be present at only low titer (only 3 of 80 sera were positive at 1:16 or greater).²¹ In contrast to their high incidence in the general population, antibodies directed against antigens of the sperm surface, as detected by agglutination tests, mixed agglutination reaction, and immunobead binding, are uncommon and present at high titer in less than 10% of infertile men and women.²²

* Numbers reflect range.
GAT, gelatin agglutination test; FD, Franklin-Dukes; TAT, tray agglutination test; SIT, sperm immobilization test.
(Adapted from Beer AE, Neaves WB; Antigenic status of semen from the viewpoints of the male and female. Fertil Steril 29:3, 1978. Reproduced by permission of the publisher, The American Fertility Society.)

Methods for Detecting Antisperm Antibodies on Sperm

Several methods are now clinically available to determine whether spermatozoa themselves are coated with immunoglobulins. These include a direct antiglobulin assay using ¹²⁵I radiolabeled heterologous antibodies, a direct enzyme-linked immunosorbant assay (ELISA), mixed agglutination reaction, and immunobead binding.^{23, 24, 25, 26} Although each of these tests allows one to determine, in a semiquantitative way, the extent of autoimmunity to sperm, immunobead binding in particular provides a measure of the proportion of spermatozoa in the ejaculate coated with each of three immunoglobulin classes (IgG, IgA, and IgM). The precise amount of immunoglobulin associated with an individual spermatozoan surface, however, still cannot be determined by current methods.

Immunobead binding uses micrometer-size plastic microspheres to which antibodies produced against human antibodies are chemically coupled.²⁷ The antihuman antibodies on the immunobead surface bind to the human antibodies present on the sperm surface. Hence, they act as antibody detector particles, which can be visualized with an ordinary microscope. This test can be used to analyze sperm recovered from semen specimens and can be used indirectly to study serum for the presence of antisperm antibodies. In the latter case, spermatozoa must first be documented free of antibodies on their surface, by a direct immunobead assay. They are incubated in serum at various dilutions and subsequently washed free and retested to determine if they have acquired antibodies on their surface during this time. Immunobead binding allows one to determine the proportion of sperm that is coated with antibodies, the region of the sperm surface to which antibodies bind (head versus tail), and the immunoglobulin class (IgG, IgA, or IgM) of antibody present.

Antibodies that mediate sperm immobilization through their interaction with complement were initially described by Fjallbrant in 1968¹⁵ and Isojima and colleagues in 1968.¹⁶ In these assays, sperm are washed and incubated with serially diluted heat-inactivated test serum or secretions. A source of complement, usually guinea pig serum, is added. A time end point for immobilization of 90% of sperm or the percentage of motile sperm at a standard time is compared microscopically with sperm incubated with control sera and complement alone. Complement-dependent immobilization, although highly specific in that false-positive reactions are uncommon, does not detect the presence of non-complement-fixing immunoglobulins, such as IgA. The amount of antibody present in the sperm surface also appears to have a role in the extent of complement-dependent immobilization (Table 2).¹⁹ Hence, results obtained from these tests have a definite relationship to immunologic infertility and are highly specific. However, not all antisperm antibodies are detected. In addition, because seminal plasma contains complement inhibitors, complement-dependent cytotoxicity tests cannot be applied to detection of autoantibodies to sperm in semen.

* Expressed as mean ± standard error of the mean; following a 4-hour incubation with guinea pig serum
(Adapted from Bronson RA, Cooper GW, Rosenfeld D: Correlation between regional specificity of antisperm antibodies to the spermatozoa surface and complement-mediated sperm immobilization. Am J Reprod Immunol 2:222, 1982.)

Agglutination tests are a relatively insensitive means of detecting antisperm antibodies, because a relatively large number of immunoglobulins must be present on the spermatozoan surface to lead to cross-linking of highly motile sperm (Table 3).²⁸ The two main procedures for sperm immobilization are the gelatin agglutination tests of Kibrick and colleagues²⁹ and the microagglutination test of Friberg.³⁰ Failure to give careful attention to controls can ultimately lead to misleading results, and nonspecific agglutination can occur in undiluted or only minimally diluted serum.³¹

* Results in immunobead binding recorded to the highest binding observed for any immunoglobulin class.
(Adapted from Bronson RA, Cooper GW, Hjort T et al: Anti-sperm antibodies detected by agglutination, immobilization, microcytotoxicity and immunobead binding assays. J Reprod Immunol 8:279, 1985.)

Although a number of ELISAs had been developed to detect the presence of antisperm antibodies in serum, this approach has not proved satisfactory. A high incidence of naturally occurring sperm-reactive antibodies in the sera of fertile men and women of all ages has posed a major problem of distinguishing this immunologic background noise from those antibodies reacting with fertilization-related antigens on the sperm surface.³² The method of fixing sperm is critical in determining which antigens are presented to the test serum sample.³³ A marked variation in the ability of ELISA to detect sperm-reactive antibodies has been documented when sperm were fixed in various manners (Table 4). Because this assay lends itself clinically to rapid analysis of larger numbers of sera, it has been a favorite in commercial laboratories. Results, however, have not always been shown to correlate well with clinical status.³⁴

The immunoglobulin classes of sperm-reactive antibodies detected by immunobead binding and ELISA were comparable only when live sperm were incubated in test serum before fixation, indicating that fixation of spermatozoa had altered their antigenicity. +, detected; -, undetected.
(Adapted from Bronson RA, Cooper GW, Witkin SS: Detection of spontaneously occurring sperm-directed antibodies in infertile couples by immunobead binding and enzyme-linked immunosorbent assay. Ann NY Acad Sci 438:504, 1984.)

An important clue in understanding how immunities to sperm can affect reproduction was the correlation demonstrated between the titer of circulating antisperm antibodies and both the duration of time to conception and the chance of conception. Rumke and associates showed in monitoring untreated men with autoimmunity to sperm during a 15-year period that the chance of conception declined markedly when the titer of circulating antisperm antibodies rose in blood to 1:128 and fell to zero at titers of 1:1024 and greater.¹³ These observations suggested that as the concentration of circulating antibodies rose, the chance of their appearance in seminal fluid increased, consistent with the fact that immunoglobulins of the IgG class enter semen as transudates from serum. Autoimmunities to sperm then occur on a continuum. Their effect on an individual's fertility depends on the titer of circulating antibody, the immunoglobulin class of these antibodies, and the antibody specificity for specific antigens.

The amount of immunoglobulin bound to the sperm surface at the time of ejaculation depends on several factors: the concentration of antisperm antibodies within the prostate and seminal vesicle secretions; the local production of antibodies within the reproductive tract, compared with their transudation from blood; the binding of antibodies to sperm as they transit the epididymis before ejaculation or, conversely, when they mix with seminal fluid; the elasped time since the last ejaculation; and the affinity of different antibody molecules for various antigens on the sperm surface. Hence, the amount of immunoglobulin on the sperm surface reflects the final common pathway of several mechanisms of immunoglobulin secretion. Evidence supporting the importance of studying sperm in the ejaculate directly, in making the diagnosis of autoimmunity to spermatozoa, comes from a comparison of sperm antibodies detected in matched semen and serum specimens.^{35, 36} In approximately 15% of cases, antibodies have been detected in serum but not on the sperm surface. The majority of these circulating antisperm antibodies that failed to enter seminal fluid were of low titer and directed against the sperm tail tip. In addition, antisperm IgM does not enter the male genital tract secretions, even when present in high concentration in blood. This immunoglobulin class of antisperm antibodies is only rarely encountered in sera of heterosexual men, though it is more common both in homosexual men and sera of women.³⁷

Circumstantial evidence has accumulated suggesting the production of locally secreted antisperm antibodies, within the genital tract, despite their absence in blood. These immunoglobulins are primarily of the IgA class. Secretory IgA is the major immunoglobulin present in tears, saliva, and colostrum, as well as in respiratory, gastrointestinal, and reproductive tract secretions.³⁸ It is the product of two distinct cell types. Secretory IgA is synthesized by plasma cells, and epithelial cells produce secretory component, which act as a regulatory transport protein for IgA.³⁹ A membrane SC–IgA complex forms and is then internalized and transported to the apical region of epithelial cells. The SC–IgA complex is then released into the external secretions. Although little is known about the mechanism of IgA secretion within the male genital tract, clear evidence shows the local production of Escherichia coli-specific IgA in men with chronic prostatitis.⁴⁰ A local secretory system exists in the human female reproductive tract, as suggested by the prominence of IgA-producing plasma cells in the fallopian tubes, cervix, and vagina.⁴¹ In an immunofluorescence study, Kutteh and colleagues demonstrated that tubal segments obtained at sterilization contained IgA plasma cells in the subepithelial lamina propria.⁴² These considerations suggest that if one relies solely on serologic tests to diagnosis immunities to sperm, results would be misleading in a sizable proportion of cases. The data enforce the notion that the presence of humoral antibodies directed against sperm is not relevant to fertility unless the circulating antibodies are present within the reproductive tract. As a corollary, tests capable of detecting immunoglobulins on living sperm recovered from the ejaculate are the most direct way to detect whether autoimmunity to sperm exists and, if so, to determine its extent and type.

The diagnosis of clinically relevant immunity to sperm in women is difficult, given our current inability to adequately sample secretions of the uterus and fallopian tubes. In addition, immunoglobulin secretion within each of the reproductive compartments (cervix, uterus, fallopian tubes) is under hormonal control and exhibits different mechanisms in the regulation of antibody transport. As an example, estradiol lowers the content of immunoglobulins within cervical mucus while stimulating the active transport of IgA and transudation of IgG into the uterine lumen.⁴³ To add to the complexity, men may secrete blood group substances in their ejaculate, and these adsorb to spermatozoa. Antiblood group antibodies present in a woman's serum could in theory bind to these spermatozoa, giving a false-positive result. Indeed, we have presented evidence that certain antisperm antibodies of the IgM class are reactive with oligosaccharides common to the blood group substances.⁴⁴ In our study of sera from known fertile women, supplied by the World Health Organization (WHO) reference bank, 40% contained immunoglobulins that reacted with spermatozoa,⁴⁵ usually the tail end piece. These results suggest that there is a continuum in the extent of immunity to sperm and that those mechanisms in women that prevent immunization to paternally derived antigens are imperfect. Hence, care must be exercised in distinguishing between a 'positive' result and a clinically significant result, whether based on immunobead binding or any antisperm antibody assay. It is clear that results of these tests should not be interpreted in the absence of clinical correlates. This is true despite the development of better antisperm antibody assays and increasing laboratory evidence that these phenomena may lead to infertility.⁴⁶

Both spermatozoa and the mature oocytes are short-lived cells that only transiently make their appearance, in a periodic manner, within the female reproductive tract. Their interaction, if successful, may lead to the development of a zygote, initiating a potential chain of events resulting in successful reproduction⁴⁷. The evanescent nature of gametes, however, makes them especially susceptible to immunologically mediated damage. Antisperm antibodies (ASA) have been shown to be directed against several different antigens, and each would be expected to have different effects on sperm functions. At least six different major immunodominant antigens have been detected on human sperm, recognized by human sera containing antisperm antibodies.⁴⁸ Indeed, several studies have shown various effects of antisperm antibodies at the level of the zona pellucida and the egg surface. Hence we, and Aitkin and colleagues, have shown that antisperm antibodies might either inhibit, promote, or be neutral in their effects on the ability of human sperm to penetrate zona-free hamster eggs (Table 5).^{49, 50} In addition, different sperm head-directed antibodies exhibit various effects on the ability of human sperm to penetrate the human zona pellucida. Using a hemizona assay, Mahony and associates assessed the effects of labeling sperm from known fertile men with antisperm antibodies.⁵¹ The power of this approach is to eliminate the variation in sperm binding between men and zonae.⁵² Immunobead binding was used to confirm that at the serum dilution chosen which minimized sperm agglutination, nearly all sperm were labeled with immunoglobulin over the heads. Salt-stored hemizonas from the same egg were inseminated with antibody-free or antibody-labeled sperm from the same donor. A wide range in effect was observed, several sera markedly lowering the number of tightly bound sperm observed after serial zona washing, whereas other sera were completely without effect (Table 6). These results further emphasize that the functional effects of antisperm antibodies in different individuals may vary despite their same regional localization of binding on the spermatozoan surface. These observations emphasize the need for more specific tests that allow one to determine the antigenic moieties against which antisperm antibodies are directed.

All experiments were performed with spermatozoa from the same known fertile donor. Each ASA-positive serum (Br, Eb, Gr) was judged against the ASA-negative control. ASA were transferred to donor sperm in vitro, and then antibody-labeled spermatozoa were washed free of serum before overnight incubation and insemination.
(Adapted from Bronson RA, Cooper GW, Phillips DM: Effects of antisperm antibodies on human sperm ultrastructure and function. Hum Reprod 4:653, 1989.)

Hemizona Index = (Number of antibody-coated sperm bound to hemizona / Number of antibody-free control sperm bound to hemizona ) x 100.
(Adapted from Mahony MC, Blackmore PF, Bronson RA, Alexander NJ: Inhibition of human sperm-zona pellucida tight binding in the presence of antisperm antibody positive polyclonal sera. J Reprod Immunol 19:207, 1991.)

The proportion of ejaculated sperm that is coated with immunoglobulin varies markedly among men.⁵³ For instance, in 154 men found to have autoimmunity to sperm as judged by direct immunobead binding, 52% had greater than 90% of their sperm bound, 23% had 50–90% bound, and the remaining quarter had less than half of their sperm coated with antibodies. Because sperm that are antibody bound over most of their surfaces are unable to enter cervical mucus⁵⁴ (antibody binding to the sperm tail tip being a possible exception⁵⁵), yet they remain completely motile in semen, several studies have shown a relationship between antisperm antibodies and impaired results of postcoital tests.^{56, 57} We have found an inverse correlation between the proportion of sperm that is coated with immunoglobulin and the number of sperm present within the cervical mucus after sexual relations (Table 7). When all sperm are coated with immunoglobulin, it is rare to find one to two sperm per high-power field within well-estrogenized cervical mucus, despite the presence of hundreds of millions of motile spermatozoa in the ejaculate. However, as the proportion of antibody-coated sperm declines below 50%, as judged by immunobead binding, the numbers of motile sperm observed in cervical mucus increase. Hence, it appears that men who have high levels of autoimmunity to sperm, as reflected in the proportion of immunoglobulin-coated sperm in their ejaculates, appear to be functionally oligospermic. That is, their sperm cannot enter the reproductive tract, and the chance that they will reach the environs of the egg is diminished.

*Cervical mucus was examined within 48 hours preceding the thermal shift 8–12 hours after coitus. Wives were free of sperm-directed antibodies.
†Spermatozoa in cervical mucus are listed as the average observed or as a range in high-power field (x400).
(Bronson RA, Cooper GW, Rosenfeld DL: Autoimmunity to spermatozoa: Effect on sperm penetration of cervical mucus as reflected by postcoital testing. Fertil Steril 41:4, 1984. Reproduced with permission of the publisher, The American Fertility Society.)

This impairment of cervical mucus penetrating ability appears to be mediated through the Fc portion of the immunologlobulin molecule. Hence, sperm exposed to Fab preparations of antisperm IgG are able to swim through cervical mucus, whereas those labeled with intact antibody do not.⁵⁸ Similarly, immunoglobulins of the IgA class bound to the sperm surface can be degraded by an IgA₁ protease derived from Neisseria gonorrhoeae that cleaves the heavy chain at amino acid bond 235–236 of the hinge region.⁵⁹ In this manner, the Fc portion of IgA is liberated from the sperm surface. These protease-treated sperm, although still coated with IgA Fab, showed an improved ability to penetrate into and sustain motility within cervical mucus (Table 8). On this basis, we have postulated that a solid-phase component of cervical mucus possesses a yet unidentified receptor for a ligand on the Fc portion of the immunoglobulin molecule. That this effect is not species specific was demonstrated in a study of the effects of a number of antisperm monoclonal antibodies, raised in mice, on the ability of human sperm to penetrate bovine cervical mucus in vitro.⁶⁰ Those monoclonal antibodies directed against epitopes present on the living sperm surface (as detected by immunobead binding) impaired sperm penetration through a column of bovine cervical mucus, whereas monoclonal antibodies directed against subsurface epitopes (detected by indirect immunofluorescence with permeabilized sperm) did not (Table 9).

(Adapted from Bronson RA, Cooper GW, Rosenfeld DL et al: The effect of an IgA₁ protease on immunoglobulins bound to the sperm surface and sperm cervical mucus penetrating ability. Fertil Steril 47:987, 1987. Reproduced with permission of the publisher, The American Fertility Society.)

* A population of nearly 100% motile spermatozoa obtained by swim-up were incubated with monoclonal antibody then washed free of ascitic fluid or culture supernatant and exposed to immunobeads.
† Following 90 minutes incubation at 37°C.
(Adapted from Bronson RA, Cooper GW: Effects of sperm-reactive monoclonal antibodies on the cervical mucus penetrating ability of human spermatozoa. Am J Reprod Immunol 14:59, 1987.)

The presence of antisperm antibodies in women may also be associated with altered sperm motion within cervical mucus. Spermatozoa appear initially to gain entrance into the cervical mucus but then subsequently become immobilized, by either shaking in place, without forward progression, or being completely immobilized. The behavior of sperm within cervical mucus depends on the type of antibodies present within the mucus and their specificity for the sperm surface. Hence, high degrees of binding of non-complement-fixing antibodies to the sperm surface may result in sperm entrapment and shaking in place, whereas complement-fixing antibodies (when directed against the majority of the sperm tail) could also lead to immobilization. Levels of complement within cervical mucus are lower than those present in serum,⁶¹ and in fact, it may take as long as 6–7 hours for sperm immobilization to occur. Hence, overnight postcoital testing provides a clearer indication of antibody-mediated sperm damage than does a shorter interval (2 hours) after sex.

Current clinical assays for immunities to spermatozoa detect antidodies directed against epitopes of moieties on the spermatozoan surface. Recently, Veaute and associates developed an ELISA using truncated acrosin recombinant proteins as antigens for the detection of antibodies directed against acrosin, a trypsin-like protease present within the sperm acrosome ⁶². Anti-acrosin antibodies were detected in sera from 34 of 179 women (19%) who had consulted for infertility. Detection by immunobead binding of antibodies directed against sperm surface antigens resulted in a similar incidence (36 of 179), although only six of them showed correspndence between assays. Anti-acrosin antibodies inhibited proacrosin binding activity to a recombinant zona pellucida protein A in vitro, as well as acrosin activation.

During the onset of spermatogenesis, at puberty, new developmental antigens make their appearance on the sperm surface.⁶³ As immune tolerance for self-antigens is established in the neonatal period, these newly appearing sperm antigens may be immunogenic. It has been theorized that sequestration of developing sperm, behind the blood-testis barrier formed by tight junctions of Sertoli cells, prevents the generation of autoantibodies to sperm.⁶⁴ Additional evidence now indicates that some testicular autoantigens are accessible to circulating antibodies and to immune processing cells.^{65, 66} A population of suppressor T lymphocytes has been identified in the epididymis by immunoperoxidase staining with monoclonal antibody probes to T-cell surface antigens.^{5, 6} These putative T-suppressor lymphocytes may have an active role in preventing the development of autoimmunity to sperm.⁶⁷ (See section on Vasectomy and Autoimmunity to Sperm.)

Although women are regularly 'inoculated' intravaginally with spermatozoa during coitus, this event is usually not associated with the development of immunity to sperm. Yet the female reproductive tract is not an immunologically privileged site, as demonstrated by the presence of anti-Candida antibodies in women with yeast vaginitis.³ Experimental intravaginal inoculation with poliovirus in women has been shown to lead to the formation of locally produced antiviral antibodies in vaginal secretions.⁴ 

 Immunoinhibitory substances have also been detected and partially characterized in seminal plasma; they may protect sperm from immunologic damage and prevent sensitization of a woman to sperm antigens after coitus.^{68, 69, 70} A potent immunosuppressive agent, 19-hydroxy prostaglandin E, has been found in the seminal fluid of men and subhuman primates.⁷¹ Other possible immunosuppressive factors include polyamines,⁷² transglutaminase,⁷³ and high-molecular-weight Fc receptor binding protein similar to pregnancy-associated protein A.⁷⁴ Spermatozoa themselves have been shown to be immunosuppressive in rodents.⁷⁵ However, because semen samples in vasectomized males also exhibit immunosuppressive activity, this observation suggests that active components are derived not solely from testicular, epididymal, or spermatozoan origin. Extensive studies on fractions obtained by gel filtration of seminal fluid and chromatographic techniques support the view that the inhibitory effects of seminal plasma are due to a range of molecules of widely differing molecular weights and binding affinities for specific ligands. Research in this area has been difficult because of the diversity of immunomodulatory substances and the tendency of low-molecular-weight species to associate reversibly with high-molecular-weight components. The generation of cytotoxic polyamines following addition of calf serum to seminal plasma, in studies of immunosuppressive activity, has further confused the issue.⁷⁶ However, using serum-free culture conditions illustrates clear evidence of interference by seminal fluid in the immune function of T cells, B cells, natural killer (NK) cells, and macrophages.⁷⁷ The effects of human seminal plasma on immunologically active cells include a reduced ability to bind antigen and to differentiate or proliferate in response to mitogens, as well as a failure of phagocytosis in antibody-dependent cell lysis. Anticomplement activities have also been demonstrated.⁷⁸ Could nature then provide the means, through the common exposure at coitus, to seminal fluid-derived suppressors, as well as spermatozoa, to prevent the development of immunity of sperm in women? Conversely, would the lack of immunosuppressor activity of seminal fluid lead to the development of antisperm antibodies? These intriguing questions, unfortunately, currently have no answer.

Recent studies have provided clues as to how spermatozoa might escape recogntion and immune-mediated destruction within the female reproductive tract. It has been theorized that human spermatozoa are susceptable to lysis by NK cells present within the reproductive tract, as they appear to lack major histocompatibility class I molecules.⁷⁹ However, they may escape NK cell-mediated damage through the expression of Lewis^x  and Lewis^y  N-glycans on their surface, in a manner similar to tumor cells.⁸⁰ Human seminal plasma also contains prostasomes, membranous vesicles secreted by epithelial cells of the prostate gland that can fuse with the sperm plasma membrane.^{81, 82} They have been shown to possess the complement inhibitors CD46 and CD59.⁸³ In addition, when NK cells were co-cultured with purified prostasomes, their expression of CD244, an activating receptor, was diminished as shown by flow cytometry, as was their ability to degranulate and secrete interferon (IFN) gamma ⁸⁴.

Evidence that spermatozoa share antigenic specificities with fertilized ova or cleaving embryos was initially provided experimentally in female animals immunized with sperm or testis cells. At serum dilutions permitting nearly normal fertilization rates, antisperm antiserum impaired embryo survival.⁸⁵ Postfertilization effects of antisperm antiserum have also been demonstrated by the reduced survival of fertilized eggs transferred from oviducts of nonimmune rabbits to those of immunized pseudopregnant females.⁸⁶ Uterine implantation rates of these embryos were one third to one half those in nonimmunized controls. In rabbits, antibodies raised against murine sperm have also been shown to react with early cleavage-stage mouse embryos⁸⁷ and a number of antisperm monoclonal antibodies have been shown to react with antigens present on trophoblast cells.⁸⁸

In theory, three mechanisms could account for the effects of experimentally induced antisperm antibodies and preimplantation embryo survival. Eggs may display sperm-derived antigens on the oolemma after fertilization.⁸⁹ Embryonic antigens cross-reacting with sperm may be expressed during early embryonic development.⁹⁰ Antisperm antibodies may indirectly affect embryonic development by stimulating the production of cytotoxic lymphokines from activated lymphocytes in the immune system of women sensitized to sperm.⁹¹

Although an anecdotal association between an increased risk of miscarriage and the presence of antisperm antibodies in women has been reported,⁹² there is at this time no clear evidence in humans that antisperm immunity causes abortions in clinically diagnosed pregnancies. For example, the risks of miscarriage in women who have antisperm antibodies and who conceive after intrauterine insemination (IUI) have been found to be no greater than in the general infertile population.⁹³ However, these observations do not address the question of whether antisperm immunity can cause abortion during the interval between ovulation and implantation. Because pregnancy would not be clinically diagnosed in these cases, the manifestations of this event would be occult and would present as unexplained infertility rather than recurrent abortion. A single study of the likelihood of successful preimplantation embryonic development and subsequent pregnancy, in women with immunologic infertility undergoing in vitro fertilization (IVF), again raises this issue (see below).

Reproductive tract tissues contain diverse lymphocyte and macrophage populations, which can be activated and may alter reproductive functions. Increasing evidence suggests that soluble products of these cells can either inhibit or promote various endocrine, gamete, and nidatory events.⁹⁴ Cell-mediated immunologic responses involve thymus-derived lymphocytes that are activated by antigenic stimulation and immunologic cytokines. Cytotoxic T cells can recognize and kill cells bearing foreign antigens by direct contact and by release of cytotoxic cytokines such as IFN gamma. Both the endometrium in the late luteal phase of the menstrual cycle and the decidua of early pregnancy contain numerous lymphocytes and macrophages.^{95, 96} Several studies have reported antisperm cell-mediated immune responses in infertile women. One study of endometrial T-cell subpopulations in infertile patients has revealed large numbers of activated T cells in some of these women.⁹⁷ Patients with endometriosis have been found to have large numbers of activated macrophages and lymphocytes in their peritoneal fluid.

Soluble products of activated lymphocytes and macrophages, including the lymphokine IFN gamma and the monokine tumor necrosis factor (TNF), have been shown to affect human sperm motility and fertilization, as measured by the hamster egg penetration test.⁹⁸ Secretion of these products and others within various regions of the female reproductive tract could affect sperm function. Various lymphokines and monokines have also been reported to exert adverse effects on the development of early mouse embryos in vitro.

Given evidence that impaired infertility can be induced experimentally by either active or passive immunization with zona pellucida proteins,^{99, 100, 101} several groups have attempted to determine whether such antibodies occur spontaneously in women.^{102, 103, 104, 105} Unfortunately, problems of methodology have continued to result in a failure to prove this thesis convincingly. Because the zona pellucida consists of a glycoprotein matrix capable of trapping immune complexes, the use of immunofluorescence to detect specific antibodies to intact zona antigens may lead to nonspecific reactions. Dunbar has shown that the production of immune complexes by freezing and thawing serum can give false-positive reactions from previously negative serum.¹⁰⁶ Unfortunately, the majority of studies of humans have also used relatively undiluted serum, without concern for the specificity of the assay. Clinical studies have been diverse and contradictory (Table 10). Dunbar studied sera from many infertile women using a sensitive radioimmunoassay with well-characterized, purified zona antigens and failed to detect any antizona antibodies.¹⁰⁶ Although it could be argued that these purified zona antigens have undergone alterations in their configuration such that native epitopes were lost, the presence of antizona autoantibodies in humans has not been confirmed.

Table 10. Incidence of anti-zona pellucida antibodies in sera of fertile and infertile women and men

Obstruction of sperm egress has been associated with development of autoimmunity to spermatozoa. Vasectomy, by inhibiting sperm exit from the epididymis and proximal vas deferens, results in sperm antigen leakage and antisperm antibody production in approximately 70% of men.^{108, 109} Because unique auto-antigens are expressed on cells of the basal compartment of the seminiferous tubule, most investigators have speculated that autoimmunity to sperm after vasectomy broaches the sequestration of antigens as well as local immunoregulatory mechanisms. The presence of these antibodies within reproductive tract secretions, noted at the time of vasovasostomy, occurs infrequently and has been correlated with an impaired chance of subsequent fertility.¹¹⁰ Men with bilateral congenital absence of the vas deferens, epididymis, or seminal vesicles,¹¹¹ as seen in cystic fibrosis,¹¹² also are found to be at risk for immunity to sperm. Our finding that autoimmunity to sperm does not develop in these men until after puberty (Table 11)¹¹³ suggests that the immune system may become exposed to developmental antigens expressed on spermatozoa and spermatids to which it is not tolerant after activation of the pituitary–testicular axis and the initiation of spermatogenesis. Developmental abnormalities of the formation of the blood–testis barrier, its traumatic disruption, or unilateral focal cryptic intratesticular obstruction at the level of the seminiferous tubules, could therefore lead to antisperm antibody formation.

Table 11. Detection of antisperm auto-antibodies in 15 males with cystic fibrosis and their pubertal status

¹¹³)

The presence of circulating antibodies to continuously produce sperm antigens has raised the concern that vasectomy might result in immune complex disease.¹¹⁴ Immunologically mediated injury occurs when preformed immune complexes are deposited either in renal glomeruli or blood vessel walls. That an immune response need not be mounted for damage to occur has been shown when arteritis, endocarditis, and glomerulonephritis developed in normal animals infused with immune complexes. The formation of circulating immune complexes is dependent on both antigen and type of antibody (numbers of combining sites, size of immunoglobulin molecule, and molecular weight and charge of antigen). Large complexes formed in antibody access are rapidly removed, primarily by the Kupffer's cells of the liver. In contrast, complexes formed in antigen access will be of smaller size and may remain in the circulation. These may become widely disseminated and activate the complement cascade.¹¹⁵

Alexander has shown that antigen production continues after vasectomy and that entrapped sperm are engulfed by macrophages within the vas deferens.¹¹⁶ Sperm antigens may also leak from the reproductive tract. The efferent ducts of the testis in monkeys vasectomized one year earlier were found to have a significant increase in thickness, and 33% of these animals exhibited immune complexes within the thickened basement membrane.¹¹⁷ Granular deposits of both IgM and IgG, as well as the complement component C3, have been found in renal glomeruli of vasectomized animals. Vasectomized men have also been shown to have a higher incidence of circulating immune complexes than age-matched controls.¹¹⁸ Although some monkeys fed diets high in cholesterol after vasectomy developed atherosclerosis,¹¹⁹ several epidemiologic studies in humans have failed to substantiate an increased risk.^{120, 121}

Effective treatment of immunities to sperm rests on the accuracy of diagnosis. Four approaches have been used to treat couples with antisperm antibodies. Condom 'therapy' is mentioned only to emphasize its ineffectiveness, in the presence of more active treatment approaches.¹²² Its value was never clearly documented in the past, and the psychological burden of using contraception when desiring pregnancy is great. The remaining approaches have included corticosteroids, intrauterine insemination (IUI), IVF, and gamete intrafallopian transfer (GIFT). However, the risk:benefit ratio in the use of corticosteroids is still not well known for either men or women. Preliminary evidence suggests that this treatment approach is relatively ineffectual. IUI, when carefully timed to follicular maturation in superovulated and hormonally and sonographically monitored cycles, results in an increased chance of pregnancy, but success rates at best are no greater than 35% within six treatment cycles. Although technically intense and expensive, if IUI fails, IVF offers the greatest likelihood of achieving fertilization, in the presence of both autoantibodies to sperm in men and circulating antisperm antibodies in women. Early evidence indicates that GIFT may also be effective in treating couples in whom the man has developed antisperm immunity.¹²³ These results with the newer assisted reproductive techniques suggest that immunities to sperm impair the ability of spermatozoa to reach the site of fertilization within the fallopian tubes to a far greater extent than their ability to penetrate eggs, should the gametes meet in the presence of these antibodies.

Why treat in the absence of proof? The accumulated evidence from laboratory-based studies provides circumstantial evidence that immunities to spermatozoa can potentially impair processes leading to fertilization. However, there are currently no prospective studies demonstrating a decreased fecundity in those couples in whom antisperm antibodies are detected, when compared with couples in the absence of immunities to sperm. While we have maintained that such studies are needed as proof of immunological infertility,¹²⁴ these data are not likely to be available in the near future. Given the low incidence of significant immunities to sperm in men and women (which are found in approximately 3–5% of unselected infertile couples), many centers will need to participate in a prospective study of the effects of antisperm antibodies on fertility, and patient acquisition will be slow. In addition, we have performed a retrospective analysis of pregnancies in women treated for infertility, whose husbands were found to exhibit an autoimmunity to sperm but were not themselves treated.¹²⁵ Pregnancy rates varied from 15.3 to 66.7%, depending upon the proportion of spermatozoa coated with immunoglobulin, as judged by immunobead binding. The results suggest that the number of cases needed to obtain sufficient power to detect differences in fertility between antibody-positive and antibody-negative groups will be large.

Currently utilized tests, including immunobead binding, also do not identify those sperm-associated antigens to which antisperm antibodies are directed. Antibodies directed against different antigens would be expected to play differing roles in impairing processes leading to successful fertilization, and this appears to be the case. As noted earlier, when transferred to sperm of known fertile men, antisperm antibodies detected in sera of infertile couples have varying effects on gamete interactions in vitro, as judged by both hemi-zona assays and sperm penetration of zona-free hamster eggs. It is then currently difficult to select homogenous groups of couples with immunities to sperm for study. For these reasons, a prospective analysis of pregnancy rates of men or women with immunities to sperm is not possible at this time. I would argue, however, that one need not wait for clinical proof through prospective analysis of couples with immunities to sperm that they cause infertility, to recommend treatment. There is sufficient evidence garnered through laboratory investigation to suggest that these antibodies directed against sperm may alter their ability to enter the female reproductive tract and to successfully fertilize. The use of the immunobead binding test also provides useful information that allows the laboratory to optimize the preparation of spermatozoa for IUI (see below). 

A judgment as to whether an individual should be treated for immunity to sperm is far easier to make for men than for women. Spermatozoa are easily accessible for study. The degree of their impairment of penetrating ability of cervical mucus is directly related to the extent of autoimmunity. Men whose spermatozoa are all or nearly all (>70%) antibody-coated by at least one immunoglobulin class and fail to penetrate cervical mucus require treatment. When coated over the sperm head, not only do these sperm encounter difficulty penetrating cervical mucus, but should they reach the ampulla of the oviduct, fertilization may be impaired. Conversely, when less than 50% of sperm are antibody-bound, the number of sperm seen at postcoital testing is often no different from that of men without autoimmunity to sperm, and other causes of infertility should be investigated.

The diagnosis of clinically relevant immunity to sperm in women is more difficult, given our inability to adequately sample secretions of the uterus and fallopian tubes. Immunoglobulin secretion within the female reproductive tract is under hormonal control, and each of the reproductive compartments (cervix, uterus, fallopian tubes) exhibits different mechanisms in the regulation of antibody transport.¹²⁶ As an example, estradiol lowers the content of immunoglobulins within cervical mucus while stimulating the active transport of IgA and transudation of IgG into the uterine lumen. While it would be ideal to evaluate tubal fluid or cervical mucus for the presence of sperm antibodies, this is not often possible. Extraction of antisperm antibodies from cervical mucus, a 'hydrogel', is difficult and may result in damage to these immunoglobulins.

A high incidence of ‘immunological background noise’ is also present in women. In our study of sera from known fertile women, supplied by the WHO reference bank, 40% contained immunoglobulins that reacted with the tail end piece of spermatozoa.^{28, 45} These results suggest that there is a continuum in the extent of immunity to sperm and that those mechanisms in women that prevent immunization to paternally derived antigens are imperfect. Hence, care must be exercised in distinguishing between a positive result and a clinically significant result, whether based on immunobead binding or any antisperm antibody assay. Results of these tests should not be interpreted in the absence of clinical correlates.

When sperm-reactive antibodies are present in serum, the postcoital test is impaired, and other causes of hostile cervical mucus have been excluded (occult cervicitis, altered pH, or poor cervical mucus production due to an insensitivity of mucus-secreting cells to estrogen stimulus, as seen in prenatal diethyistilbestrol exposure), the diagnosis of immunological infertility is strongly suggested. The finding of antisperm antibodies at the site of fertilization, within peritoneal fluid and in uterotubal flushings retrieved at laparoscopy, or in vaginal secretions or cervical mucus, reinforces the diagnosis of immunological infertility. In these women sensitized to sperm, large numbers of spermatozoa are often observed within cervical mucus, either immobilized or displaying restricted motion.

Observations from recent clinical experience in the treatment of immunological infertility reinforce the notion that the major locus of action of antisperm antibodies is through their impairment of sperm transport to the site of fertilization and their shortened longevity within the female reproductive tract. In those couples with antisperm antibodies who proceed with IVF, in vitro fertilization rates have been found to be high in the presence of circulating antisperm autoantibodies.^{127, 128, 129, 130} Only when nearly all sperm (>70%) are coated over their heads with immunoglobulin, as reflected in mixed antibody reaction (MAR) or immunobead binding assays, is there a significant fall in the likelihood of fertilization.^{131, 132} In the latter situation, the possibility exists that fertilization-related antigens (likely to be specific gamete receptors and their ligands) may be the targets of these antibodies. As the etiology of autoimmunity to sperm in men remains in large part unknown, treatments must be empirical, being directed not against the cause, but rather against the abnormal response. Evidence suggests that the use of corticosteroids for immunosuppression is relatively ineffectual, benefiting only approximately 20% of treated men¹³³ (see below). When carefully timed to follicular maturation in superovulated, hormonally and sonographically monitored cycles, recent evidence suggests that IUI results in an increased chance of achieving pregnancy.

The rationale for IUI is to place within the uterine cavity a large population of living sperm that were excluded after coitus because of the presence of antisperm antibodies either on their surface or within cervical mucus. In theory, this would increase the likelihood that sperm might enter the fallopian tubes and reach the egg. Using careful sonographic and hormonal monitoring of follicular maturation, insemination can be timed to within a few hours of the expected ovulation. The accuracy of timing is important in that antibody-bound spermatozoa have a theoretically shortened survival time within the female reproductive tract, where they become opsinized by binding complement components, leading to either their phagocytosis by macrophages or their immobilization.¹³⁴

There are no controlled, prospective studies of the efficacy of IUI in the treatment of infertility mediated by antisperm antibodies, in men or women. However, retrospective case reviews suggest that IUI is beneficial in conjunction with the use of clomiphene citrate or gonadotropin. The large study of Margalioth et al.⁹³ illustrates this thesis. These authors reviewed the outcomes of IUI in a group of women with impaired postcoital tests who were also found to have antisperm antibodies in their sera. Monthly pregnancy rates in the first 3 months of treatment were 5% following IUI in the natural cycle, 9.7% following IUI in association with the use of clomiphene citrate, and 14.3% per cycle following gonadotropin stimulus and IUI. These differences were statistically significant (p <0.05). The majority of women conceived during three cycles of treatment, and 40% of women who failed to conceive in clomiphene-stimulated IUI cycles conceived with the subsequent use of gonadotropins. The likelihood of pregnancy was also lower for those women possessing sperm head-directed antibodies compared with those possessing antibodies directed against the spermatozoan tail. A recent large prospective study of infertile couples¹³⁵ provides supporting evidence of these trends, in that pregnancy rates were highest following IUI in gonadotropin-stimulated cycles when compared with both clomiphene-stimulated and unstimulated cycles. Although one could argue that the presence of antisperm antibodies in serum does not in itself constitute proof of an immunological basis of infertility in women, and that other factors could have introduced bias in the study of Margalioth et al.⁹³ the data do suggest the use of IUI in a limited trial in this clinical circumstance.

In a similar manner, IUI appears to be beneficial in men with autoimmunity to sperm, particularly when performed following the use of laboratory techniques in which ejaculation is performed directly into a buffer solution prior to sperm recovery. Sperm washing may provide some benefit, by eliminating low-affinity antisperm antibodies within seminal plasma that may have bound to spermatozoa during their residence within the vagina following coitus. Unfortunately, the affinities of immunoglobulins for antigens on the sperm surface are high, and once antibody binding to spermatozoa has taken place, simple sperm washing will not remove these antibodies from the sperm surface.¹³⁶ Techniques that lead to dissociation of antibody–immunoglobulin complexes (low pH or high ionic strength) are associated with irreversible loss of sperm motility.

Ejaculation directly into a washing buffer appears beneficial in terms of both maximizing sperm recovery and minimizing the amount of antibody coating sperm.^{137, 138, 139} The process of antibody coating of sperm within semen is complex. Witkin⁶ has shown, in rabbits, that immunoglobulins may enter the male reproductive tract via the epididymis. If this is the case in humans, antibody coating of sperm may occur prior to ejaculation as well as thereafter. Conversely, immunoglobulins are also present in prostatic secretions^{13, 140} and sperm become exposed to them only after semen liquification. Rapid dilution of the ejaculate and mixing of semen could, in theory, be beneficial on this basis. In addition, transportation of semen collected at home to the laboratory prior to sperm processing would allow time for postejaculatory sperm antibody coating and agglutination, which would be avoided by processing semen immediately after ejaculation.

Pregnancy rates associated with IUI, in the treatment of men with autoimmunities to sperm, have improved substantially since adopting washing techniques requiring ejaculation of semen into medium rather than into dry containers. While older reports describe 3–10% per cycle rates of conception associated with IUI,^{141, 142} a later study of men with high levels of antisperm antibodies,¹⁴³ and ejaculating into a buffered medium, described pregnancy rates of 64% following three cycles of IUI, with a 47% conception rate in the first cycle. Our own unpublished experience confirms these results.

In 1976, Shulman reported the successful use of corticosteroids to treat a man with autoimmunity to sperm.¹⁴⁴ Although various degrees of success have subsequently been described, most of these earlier reports suffer from their inability to document the change in antibody binding on the sperm themselves (Table 12).^{125, 145, 146, 147, 148, 149, 150} Success of treatment was often not judged on the basis of an observed quantitative change in the status of autoimmunity but rather on the rate of pregnancy following treatment. However, in retrospectively analyzing the pregnancy outcome of 108 women whose mates were found to have autoimmunity to sperm but were not treated during a 2-year period, spontaneous conception rates varied with the proportion of sperm antibody coated.¹²⁵ When more than 50% of sperm were bound with immunoglobulins, 22% of women conceived, whereas 45% conceived when fewer than 50% of sperm in the ejaculate were immunoglobulin coated. The results were even more distinct in couples whose sole cause of infertility was the man's autoimmunity to sperm. Here, pregnancy rates were 15.6% if the majority (more than 50%) of sperm were antibody coated, and rose to 63% if less than 50% of sperm were bound. Given these well-documented spontaneous conception rates, depending on the extent of autoimmunity to sperm, the use of pregnancy as a validation of treatment is misleading, without adequate placebo controls.

CD, cycle day.
† Greater than 50% of sperm antibody coated as determined by direct immunobead binding.
‡ Less than 50% of sperm antibody coated.

A single study has documented a variable suppression of antisperm antibodies within seminal plasma in some men treated with corticosteroids.¹⁵⁰ A single study has also reported changes in the amount of antibody bound to sperm.¹⁵¹ Whether the degree of suppression of autoimmunity in these cases would be sufficient to increase the number of antibody-free sperm in the ejaculate to a clinically significant level remains unproven. Given the side-effects of corticosteroids, such as mood changes, leg muscle cramps, hypertension, reactivation of ulcers, alterations of glucose tolerance, and rare but severe aseptic hip joint necrosis, one must remain conservative about their use pending further well-controlled studies.

A rationale for IUI is to place within the uterine cavity a large population of living sperm that were excluded after coitus because of the presence of antisperm antibodies either on their surface or within cervical mucus. In theory, this would increase the likelihood that sperm might enter the fallopian tubes and reach the egg. Using careful sonographic and hormonal monitoring of follicular maturation, insemination can be timed to within a few hours of the expected ovulation. The accuracy of timing is theoretically important in that antibody-bound spermatozoa have a shortened survival time within the female reproductive tract. Pregnancy rates have varied in different clinical series over the range of 20–40%. In theory, the triggering of ovulation by injection of recombinant hCG (once follicular maturation has been confirmed by measurement of serum estradiol and follicular diameters at endovaginal ultrasound scanning) will increase the accuracy of timing of IUI, although there are no current prospective studies to document a difference in conception rates. Conception occurred rapidly within four to six cycles, as in the case of women without antisperm antibodies,^{152, 153} and appeared to be higher in gonadotropin-stimulated cycles (Table 13).^{93, 154} These studies have been difficult to interpret, unfortunately, because the diagnosis of immune-mediated infertility is often unclear. Hence, the association of circulating antisperm antibodies in sera of women, in conjunction with impaired sperm survival within cervical mucus, is not necessarily proof of a cause-and-effect relationship. The clinical data do suggest, however, that this approach should be attempted before IVF. In addition, given the relatively low conception rates, these results suggest that antibody-coated sperm may not always reach the fallopian tubes after IUI, despite their placement high within the uterine fundus. Alternatively, they may be unable to fertilize, despite successfully meeting the egg at the site of fertilization, because of the presence of antisperm antibodies in tubal secretions.

Table 13. Conception rates per cycle following IUI in women with humoral antisperm antibodies and impaired postcoital tests*

* Fewer than 5 motile sperm per high-power field
(Adapted from Margalioth EJ, Sauter E, Bronson RA et al: Intrauterine insemination as treatment for antisperm antibodies in the female. Fertil Sterile 50:444, 1988. Reproduced with permission of the publisher, The American Fertility Society.)

If IUI fails, IVF currently appears to offer an excellent chance of conception in couples with documented immunities to sperm. Antisperm antibodies within follicular fluid can be removed by washing the cumulus–oocyte complex, and any residual immunoglobulins that may remain within the cumulus oophorus that surrounds the egg do not usually appear to significantly interfere with sperm penetration. While older studies, such as that of Yovich et al.(1984),¹⁵⁵ reported lower rates of fertilization in vitro when patient sera was used to supplement culture media, IVF in women with antisperm antibodies has been successful and rates in serum-free medium have been nearly comparable to those in their absence.^{156, 157, 158}

In men with autoimmunity to sperm, as previously noted, the diminished number of antibody-coated sperm entering cervical mucus after coitus markedly lowers the chances that the gametes will meet. IVF circumvents this problem of sperm transport and ensures the meeting of spermatozoa and egg (Table 14). Fertilization rates achieved by sperm obtained from men with autoimmunity to sperm have been high in several reports, suggesting that impaired sperm transport (functional oligospermia) has been the primary basis of their infertility.^{143, 159, 160, 161} In contrast to women with immunities to sperm, however, in whom follicular fluid containing antisperm antibodies can be washed from the egg, immunoglobulins in the ejaculates from men with autoimmunity to sperm remain bound to the sperm surface after their recovery from seminal fluid. Although antibodies present on the sperm tail do not significantly prevent fertilization in vitro, sperm head-directed antibodies have the potential to alter the spermatozoan egg-penetrating ability, as previously described in both the hemizona assay and the zona-free hamster egg penetration test. Fortunately, these effects may only become apparent when more than 70% of the sperm population used in IVF are coated with immunoglobulins.^{131, 132, 158, 162} In theory, results will also depend on the sperm antigen to which these antibodies are directed. Should they be directed against antigens that have no role in fertilization, the process of penetration by the spermatozoon would not be impaired. This has clearly been documented under experimental conditions, after the generation of antisperm monoclonal antibodies.¹⁶³ Unfortunately, no clinical test can predict this outcome before an actual attempt at IVF. Hence, when head-directed antisperm antibodies are detected by direct immunobead binding on all sperm, intracytoplasmic sperm injection (ICSI) should be performed, to ensure high rates of fertilization.^{131, 164, 165}

While the argument has been made that the routine testing for antisperm antibodies is not cost effective in the use of ART, given its low incidence and the high likelihood of fertilization, this is a value judgment. Should one wait for the occasional case of failed fertilization to perform tests for antisperm antibodies retrospectively, as has been suggested? How does one judge the emotional cost to the couple who have gone through these procedures and failed to conceive? Had an immunobead binding test been performed, for a very small individual cost relative to the total cost of an IVF cycle, these couples could have been placed in a high risk category for failed IVF and the need for ICSI discussed.

Table 14. Effect of antisperm antibodies on in vitro fertilization rates in 32 men with autoimmunity to spermatozoa

*As determined by direct immunobead binding.
(Adapted from DeAlmeida M, Gazagne I, Jeulin C et al: In vitro processing of sperm with auto-antibodies and in vitro fertilization results. Hum Reprod 4:49, 1989; and Clarke GN, Lopata A, McBain JC et al: Effect of sperm antibodies in males in human in vitro fertilization (IVF). Am J Reprod Immunol 8:62, 1985.)