Human pyriform sperm heads in infertile men and assisted reproductive technology outcomes: a retrospective cohort study based on propensity score matching

Infertility is characterized by the inability to conceive despite regular unprotected sexual activity for a year or longer. 1 Infertility affects 8%–12% of couples globally in the reproductive age group 2 and from 30%–50% of cases are attributed to male factors. 3 Teratozoospermia is characterized by structural abnormalities of spermatozoa, such as head defects, and ranks as one of the significant forms of male infertility. 4 The presence of spermatozoa with normal morphology below the lower reference limit (<4%) is referred as teratozoospermia. 5 According to the latest 6 th edition of World Health Organization (WHO) Laboratory manual for the examination and processing of human semen, 5 sperm heads may exhibit several abnormal phenotypes, including pyriform, tapered, round, amorphous, vacuolated acrosomal/post-acrosomal areas, and small/large acrosomal areas. Precisely identifying and understanding the etiology of specific sperm defects will lead to a better understanding of spermatogenesis. 6 The pyriform head spermatozoa is a typical deformed human spermatozoon listed in the latest 6 th edition of WHO manual; 5 it has distinct pear-shaped morphological features and was originally classified as a human sperm morphological abnormality by Eliasson 7 in 1971. Later, Blom 8 reported that pyriform sperm heads were a subfertility factor in bovine husbandry. Rousso et al . 9 indicated that the pyriform-headed spermatozoa was a frequent but little-studied morphological abnormality. Human zona pellucida (ZP) are highly selective for spermatozoa with normal morphology, and the rate of binding of pyriform sperm heads to ZP is negatively related to fertilization rates in vitro . 10 However, there is limited research on the occurrence of pyriform-headed spermatozoa in males, and neither is there any evaluation of its potential correlation with other semen parameter values. Some studies have shown a correlation between morphological sperm defects and indices of sperm function (the DNA fragmentation index [DFI], sperm nuclear maturity, sperm acrosome function [acrosome reaction, integrity, and acrosin activity], mitochondrial membrane potential [MMP], and reactive oxidative stress, which may negatively impact fertility potential). 11 12 13 Several studies have investigated the effects of teratozoospermia on pregnancy outcomes and assisted reproductive technology (ART) success, 14 15 16 as well as the predictive role of sperm morphology on pregnancy rates after assisted reproduction. 17 Despite the growing body of literature on teratozoospermia as a cause of male infertility, only a limited number of publications have specifically examined the role of pyriform sperm heads in male infertility. The presence of pyriform sperm heads in semen is one of the causes of teratozoospermia. But they have never been considered as the sole factor causing male infertility. Therefore, this study is of the clinical importance of pyriform sperm heads in male infertility, including their characteristics, basic sperm parameters and function, and ART ( in vitro fertilization [IVF] and intracytoplasmic sperm injection [ICSI]) outcomes (fertilization, high-quality embryo, clinical pregnancy, early pregnancy loss, and live birth rates) for male patients with pyriform spermatozoa. This study was conducted to investigate the importance of pyriform sperm heads by using propensity score matching (PSM), 18 with the aim of providing clinical values for pyriform sperm heads useful in diagnosing male infertility and selecting treatments in IVF procedures. The study also mentions the causes and molecular mechanisms of pyriform sperm heads.

FPL led the study and all authors contributed to the study design. TTY, QYC, and YBW collected the data. YZ and LY analyzed and interpreted the data. YLJ drafted the paper and LY provided statistical knowledge. FPL critically revised the paper. All authors read and approved the final manuscript.

In this single-center retrospective cohort study, a total of 62 077 patients were recruited from the Andrology Clinic of West China Second University Hospital (Sichuan University, Chengdu, China) between January 2022 and July 2024. A total of 6944 patients were included after applying the selection criteria; the exclusion criteria were: (1) males aged 45 years, and female aged 40 years; (2) azoospermia, retrograde ejaculation, any hormonal stimulation treatment, recent fever, and incomplete ejaculate; (3) couples with chromosomal abnormalities or monogenic diseases; (4) reproductive system malformations, including female unicornuate uterus, didelphic uterus, septate uterus, endometriosis, adenomyosis, male external genital abnormalities, cryptorchidism, and varicocele; or (5) patients with chronic diseases (liver/renal disease, hypertension, and diabetes), a history of cancer, diagnosed autoimmune disorders, treatments that may alter spermatogenesis, or the necessity for chronic drug consumption, as well as surgical treatments within 3 months before enrollment. Males with no spermatozoa with pyriform heads in the semen and ≥4% normal sperm forms, according to the strict classification method (the same as the WHO 6 th edition manual), were selected as control males (C-male group) after being screened on the exclusion criteria ( Supplementary Figure 1 ). 5 19 After the andrology laboratory confirmed the presence of any percentage (≥1%) of pyriform-headed spermatozoa in semen, 2300 males were grouped into the pyriform spermatozoa head (P-male) group. With a full understanding of the effects of teratozoospermia on male fertility and after one-to-one counseling provided by professional reproductive medicine specialists, 59 couples chose to undergo ART. All procedures performed in studies involving human participants were in accordance with approval of the Ethics Committee of West China Second University Hospital Affiliated with Sichuan University (WCSUH-SCU; Chengdu, China; Approval No. IRB 2020-001), as well as with the 1964 Helsinki Declaration and its later amendments. This article does not contain any studies involving animals and was approved by the Ethics Committee of WCSUH-SCU (Approval No. IRB 2021-197). Written informed consent was obtained from all couples included in the study. Permission to obtain publicly published data has been granted. Semen was examined according to WHO guidelines 5 on ejaculates collected by masturbation after 2–7 days of abstinence. We informed the participants that samples would be discarded if they did not follow the suggested protocol, emphasizing the importance of accurately reporting the abstinence period. The samples were analyzed within 30 min after liquefaction. Semen volume was measured by weighing the specimen in a standard container. A Makler chamber (Sefi Medical Instruments Co., Haifa, Israel) was used to assess sperm concentration and round cell count (10 6 per ejaculate). Sperm motility was analyzed by a computer-aided sperm analysis system (Suiplus, Beijing, China). Sperm vitality (%) was assessed using eosin staining. After smearing on the slides, air-drying, and fixation in 95% ethanol for more than 15 min, the semen sample slides were then immersed in sequential grades of ethanol (50%, 75%, and 95%) before the Papanicolaou staining, followed by sequential immersion in Harris hematoxylin, acid ethanol, G-6 orange stain, and EA-50 green stain. Slides were mounted in neutral balsam after dehydration in increasing concentrations of ethanol. Images were captured by an Automated Sperm Morphology Analysis (ASMA) system (Puhua Corporation, Chengdu, China) with an oil immersion magnification of 1000× to assess sperm morphology. Images of sperm morphology were then evaluated by at least two of the three experienced technicians. All sperm images were classified as one of two categories: normal sperm heads or abnormal sperm heads (including pyriform, tapered, and round sperm heads). The sperm chromatin structure assay was performed as previously described. 20 After semen analysis, depending on sperm concentration, a sufficient volume of fresh semen was added to 100 μl of 0.01 mol l −1 Tris-HCl, 0.15 mol l −1 NaCl and 1 mmol l −1 ethylenediaminetetraacetic acid (EDTA), pH 7.4, to achieve a concentration of approximately 1 × 10 6 –2 × 10 6 sperm per ml. Thereafter, 200 μl acid detergent solution (0.15 mol l −1 NaCl, 0.08 mol l −1 HCl, 0.1% [ v / v ] Triton X-100, pH 1.2) was mixed with 100 μl sperm suspension, and after 30 s, 600 μl of fluorescent stain (6 μg ml −1 purified acridine orange in phosphate citrate buffer, pH 6.0) was used to stain the cells. Each sample was analyzed by flow cytometry (EasyCell; Wellgrow, Shenzhen, China), and 5000 spermatozoa per sample were collected. Green fluorescence emitted by acridine orange bound to double-strand DNA and red fluorescence emitted by acridine orange bound to single-stranded DNA were recorded in the fluorescence 1 (FL1) and fluorescence 2 (FL2) channels, respectively. The flow cytometric sperm chromatin structure assay measures (1) the percentage of spermatozoa in a sample with fragmented DNA, referred to as the DFI and (2) the percentage of spermatozoa with high DNA stainability (HDS >15%). A modified acidic aniline blue assay assessed sperm chromatin condensation. 21 Spermatozoa were washed twice in saline and centrifuged at 800 g for 10 min; then the supernatant was removed, and a 10-μl droplet of the pellet was smeared on a slide. After air-drying, the sample was fixed in 95% ethyl alcohol for 10 min at room temperature, rinsed with water, and stain with 5% aniline blue in a 4% acetic acid (pH 3.5) for 5 min. Then, the slides were rinsed in water and stained in 0.5% eosin (Sigma Chemical Co., St. Louis, MO, USA) for 3 min. The slides were examined by oil immersion at 1000× magnification and bright field microscopy (CX41; Olympus Corporation, Tokyo, Japan). Immature chromatin was characterized by lysine-rich nuclear histone proteins which were stained dark blue, whereas mature chromatin with cysteine- and arginine-rich protamine, was stained red or pink with the eosin counterstain. Chromatin maturity was calculated by dividing the number of red and pink spermatozoa by the total number analyzed and expressing as a percentage. At least 200 spermatozoa were analyzed on each slide. Sperm acrosin activity was analyzed with a commercial kit (Sinothinker, Shenzhen, China), which is a modification of previous methods. 22 A sample of 7 × 10 6 spermatozoa was mixed with normal saline, centrifuged at 1500 g for 10 min and the seminal plasma supernatant was removed. Proteinase inhibitors (Sinothinker, 100 μl) were added before 1 ml reaction buffer containing N-α-benzoyl-DL-arginine-para-nitroanilide HCl substrate, and incubation for 1h at 24°C, when reactions were stopped with 100 μl 83 mg ml −1 benzamidine solution, and centrifuged at 2000 g for 15 min. The absorbance of the supernatants was spectrophotometrically determined at 410 nm (Sinothinker). Acrosin activity was presented in μIU per 10 6 spermatozoa. The numbers of acrosome-intact spermatozoa and ionophore-induced acrosome-reacted spermatozoa were determined by using fluorescein isothiocyanate-cluster of differentiation 46 (FITC-CD46) antibody staining, as described with slight modifications. 5 A highly motile sperm population was prepared from semen by the swim-up procedure with fresh spontaneous Eagle’s balanced salt solution (sEBSS) medium. Control and samples were prepared with each containing approximately 1 ml suspension with 1.0 × 10 6 motile spermatozoa. Sperm suspensions were incubated for 3 h at 37°C in an atmosphere of 5% ( v / v ) CO 2 in air to induce capacitation, when 10 µl of A23187 stock solution (1 mmol l −1 ; Sailenna Sperm Acrosome Reaction Assay Kit, Chengdu, China) was added to a final concentration 10 µmol l −1 , and 10 µl dimethyl sulfoxide (DMSO) was added to the controls. All samples were incubated at 37°C for 15 min, and acrosomal integrity (spermatozoa with intact and induced acrosome reactions) was calculated from the acrosomal status of the control and experimental samples. To analyze the sperm MMP, lipophilic cationic dye, 5,5’,6,6’,tetrachloro-1,1’,3,3’-tetraethylbenzimidazolocarbocyanine iodide (JC-1; Sigma-Aldrich, St. Louis, MO, USA) solution was applied to samples. After semen parameters were analyzed and seminal fluid was removed by centrifugation (800 g for 10 min), aliquots of 2 × 10 6 spermatozoa were incubated in 1 mL phosphate-buffered saline (PBS) containing 5μg ml −1 JC-1 at 37°C in the dark for 20 min. The stained spermatozoa were analyzed in a flow cytometer. JC-1 spontaneously forms J-aggregates in spermatozoa with a high MMP and exhibits orange-red fluorescence, which was measured in the FL2 channel; JC-1 remains in the monomeric form in spermatozoa with a low MMP and exhibits green fluorescence, which was measured in the FL1 channel. Semen samples were collected and spermatozoa were washed three times with 1×PBS (pH 7.4), fixed in 3.0% ( v / v ) glutaraldehyde (Sinopharm Group Chemical Reagent Co., Ltd., Shanghai, China) at 4°C overnight. For TEM, spermatozoa were post-fixed in 1% osmium tetroxide, dehydrated with gradient acetone solutions and embedded in Epon 812 (Zhongjingkeyi Technology Co., Ltd., Beijing, China) and cut into ultrathin (60–90 nm) sections in an ultramicrotome (EM UC7rt, Leica Microsystems, Wetzlar, Germany). After evaluation of semi-thin sections to locate the spermatozoa in an optical microscope, the ultrathin sections were double-stained with lead citrate and uranyl acetate, and sections were observed and imaged with a TEM (JEM-1400FLASH, JEOL, Tokyo, Japan) at an accelerating voltage of 80 kV. Patients underwent standard ovarian stimulation through the administration of gonadotrophins using either the gonadotropin releasing hormone agonist in long or short protocols. The ovarian stimulation protocols and the daily dose of follicle-stimulating hormone injection were determined from the patient’s age, ovarian reserve, and previous responses to ovarian stimulation. Patients were monitored by serial ultrasonography until at least two leading follicles reached 18 mm in diameter. Oocyte maturation was induced with human chorionic gonadotropin (hCG; Lizhu Pharmaceutical Trading, Zhuhai, China) and oocytes retrieved 36–38 h later via vaginal ultrasound-guided follicular puncture. All IVF/ICSI procedures were performed according to our routine clinical management. The variation in time intervals between different steps of the IVF/ICSI cycle reflected differences in the daily departmental workload. After oocyte retrieval, they were denuded by brief exposure (60 s) to hyaluronidase (Irvine Scientific, Santa Ana, CA, USA). Depending on sperm indices and the couples’ reproductive histories, the oocytes were inseminated using conventional IVF, and cumulus oocytes were incubated with 60 000 spermatozoa per oocyte in IVF supplemented with human serum albumin-IVF-plus medium (HSA-IVF-plusmedium; Vitrolife, Goteborg, Sweden). ICSI was performed by observation with an Olympus IX73 microscope (Olympus Corporation, Shinjuku, Japan) with a micromanipulator (MTK-1, Narishige Co., Ltd., Tokyo, Japan). Embryos were transferred on day 3 or day 5 and all were at least four cells with even blastomeres and <40% fragmentation. All patients were given luteal phase support via the intramuscular injection of progesterone (60 mg per day) or vaginal progesterone gel (90 mg per day) plus oral dydrogesterone (20 mg per day). Two weeks after embryo transfer, the pregnancy was assessed from serum β-hCG levels. Pregnancy was confirmed by transvaginal ultrasound 4 weeks after embryo transfer. Serum β-hCG levels >50 IU l −1 were regarded as a biochemical pregnancy, and the presence of the gestational sac was regarded as a clinical pregnancy. Cleaving embryos were evaluated for cell number, blastomere appearance (the presence of vacuoles, multinucleation, cytoplasmic darkness or inclusions, and uneven size of blastomeres), and fragmentation rate. A high-quality embryo was defined as follows: 4–6 cells on day 2, ≥8 cells on day 3, <10% fragmentation, and equally sized mononucleated blastomeres. The higher quality embryos were selected for transfer. Supernumerary good-quality embryos were frozen. For each couple, one or two embryos were transferred by ultrasound guidance and serum hCG levels were measured two weeks after embryo transfer. The fertilization rate was defined as the percentage of pronuclear oocytes relative to the number of mature (or injected) oocytes. The quality of the embryo was evaluated in an inverted microscope on the second day of development, according to a previously reported method. 23 The high-quality embryo rate was calculated as the percentage of type I or II embryos to the total number of pronuclear oocytes. Clinical pregnancy was defined as the observation of the gestational sac on vaginal ultrasonography 6–7 weeks after embryo transfer. Pregnancy rate was calculated as the percentage of cycles resulting in pregnancy from the total number of IVF/ICSI cycles. Descriptive statistics and statistical analysis were performed using SPSS version 25.0 (IBM, Armonk, NY, USA). GraphPad Prism version 6 (GraphPad Software, La Jolla, CA, USA) was used to prepare the graphs. Whether the baseline data of patients conformed to a normal distribution was assessed from the Shapiro–Wilk test or histograms. The mean ± standard deviation (s.d.) was used to describe data with a normal or approximately normal distribution. The median (IQR) was used to describe non-normally distributed data. Comparisons were made using the Mann-Whitney U test. A bilateral P < 0.05 was considered statistically significant. This study assembled propensity score-matched pairs of pyriform spermatozoa head (P-male) group and control (C-male) group to minimize the imbalance from measured baseline and demographic covariates between patients. The distribution of variable characteristics in the matched experimental and control groups tends to be balanced, reducing the impact of confounding factors. 24 The caliper matching method, an application of the nearest-neighbor’s matching method, was used to match the two groups. The distribution of propensity scores differs significantly between the experimental and control groups. The nearest-neighbor matching method fails to ensure the quality of matched subjects in this situation. To address this issue, our method sets a caliper value. The caliper value refers to the permissible margin of error allowed when matching research subjects on the basis of their propensity scores. In this study, the caliper value was set to 0.02. 25 26 Clinically descriptive baseline variables (age, body mass index [BMI], abstinence delay days, and semen volume) and main parameters that have a significant impact on semen quality (sperm concentration, sperm vitality, progressive motility, total motile sperm count [TMSC], and normal sperm morphology percentage) between groups were included in the PSM, with a matching tolerance of 0.02 and 1:1 match ( Table 1 and Supplementary Figure 1 ). Spearman’s correlation analysis was used to investigate the correlation between the proportion of pyriform spermatozoa and semen parameter values in the P-male group (the presence of ≥1% pyriform-headed spermatozoa in semen; Table 2 ). The intraclass correlation coefficient (ICC) values for intraobservers and interobservers were determined to assess the reliability of the three experienced technicians. At least 200 spermatozoa per semen sample were randomly captured. The obtained morphometric values of pyriform sperm heads were described by nine features: head length (L), width (W), length/width ratio (R = L/W), head girth (HG), ellipse intersection over union (EIoU), girth intersection over union (GIoU; Supplementary Figure 2 ), short-axis symmetry (SAS), long-axis symmetry (LAS), and head area (HA). The measurement data are presented as the mean ± s.d., the 2.5 th to 97.5 th percentiles, and the 95% confidence interval (CI), as shown in Table 3 . Characteristics of pyriform spermatozoa head (P-male) group and pre-pregnancy examination (C-male) group before and after propensity score matching Data are presented as the median (IQR), and comparisons between the pyriform spermatozoa head (P-male) group and the pre-pregnancy examination (control) group were determined by the nonparametric Mann–Whitney U test for continuous data and the Chi-squared test for categorical variables. TMSC=ejaculate volume×sperm concentration × percentage of progressive motile spermatozoa (PR%). Normal sperm morphology (%) = (normal sperm morphology/the sum of all evaluated spermatozoa) ×100%. TZI (%) = (the sum of all abnormalities/the sum of abnormal spermatozoa) ×100%. SDI (%) = (the number of defects/the total number of spermatozoa) ×100%; not only the abnormal spermatozoa. A value of P =0.05 was used as the limit of statistical significance in all the analyses. |SMD| <0.1 indicates good balance between groups; 0.1≤ |SMD| <0.2 indicates moderate imbalance; |SMD| ≥ 0.2 indicates severe imbalance. SMD: standardized mean difference; PR: the percentage of rapid and slow progressive sperm (type A+B); NP: the percentage of non-progressive spermatozoa (type C); IM: the percentage of immotile spermatozoa (type D); BMI: body mass index; IQR: interquartile range; VCL: velocity along the curvilinear path (µm s −1 ); ERC: excess residual cytoplasm; ALH: the amplitude of the lateral displacement from the track midline to one side of the head (µm); VSL: velocity along the straight-line path (µm s −1 ); VAP: velocity along the average path (µm s −1 ); TMSC: total motile sperm count; TZI: teratozoospermia index; SDI: sperm deformity index Spearman correlation analysis between pyriform sperm percentage and basic semen parameter values in P-males ( n =1302) CI: confidence interval. Other abbreviations can be identified in Table 1 . Distributions of the morphometric parameters of pyriform sperm heads ( n =2805) The measurement of sperm head from acrosomal end to the nucleus end (long axis). EIoU (%) = a real sperm head area/a standard digital mimetic ellipse area ×100%. GIoU (%) = real sperm perimeter/a standard digital mimetic ellipse perimeter ×100%. s.d.: standard deviation; CI: confidence interval; L: head length; SAS: short-axis symmetry (the symmetrical degree of sperm head in short axis); LAS: long-axis symmetry (the symmetrical degree of sperm head in long axis); W: head width (the measurement of sperm head short axis); ratio (R): head length/head width ratio; HG: head girth (the circumference length of sperm head); HA: head area (the area of sperm head); EIoU: ellipse intersection over union; GIoU: girth intersection over union

This study enrolled 2300 patients (P-male group) with pyriform spermatozoa in their semen and 4644 pre-pregnancy examination males (C-male) group with no seminal pyriform spermatozoa. After being balanced for age, BMI, abstinence delay days, semen volume, sperm concentration, total sperm count, sperm vitality, progressive motility, TMSC, and normal sperm morphology percentage, the PSM resulted in 1302 pairs of patients with pyriform spermatozoa in semen, and controls lacking such cells. The baseline characteristics and semen analysis parameters of the study cohorts are presented in Table 1 for both the initial unmatched cohort and the PSM cohort. Before PSM, the average age of the P-males was significantly higher than that of the C-males (median [interquartile range; IQR]: 32 [29–35] years versus 31 [29–34] years, P < 0.001]. After PSM, 1302 pyriform sperm head patients were successfully matched to 1302 control males. Imbalances in the measured baseline semen parameters were minimized in PSM groups compared with the unmatched cohort. There was absolute standardized difference >0.05 for 11 covariates (male age, BMI, abstinence, semen volume, total sperm count, round cells, TMSC, velocity along the curvilinear path [VCL], velocity along the straight-line path [VSL], the amplitude of the lateral displacement of the head [ALH], and velocity along the average path [VAP]). A significant imbalance persisted for ten covariates in the P-male group and C-male group (sperm concentration, percentage vitality, total motility, progressive motility, non-progressive motility, immotile spermatozoa, normal sperm morphology, excess residual cytoplasm [ERC], teratozoospermia index [TZI], and sperm deformity index [SDI]). Values of these ten covariates for the P-males were significantly worse than those for the C-males (all P ≤ 0.05). Specifically, sperm concentration, percentage vitality, total motility, progressive motility, and normal sperm morphology were lower in P-males, while non-progressive motility, immotile spermatozoa, ERC, TZI, and SDI were higher. A total of 71.8% patients in the P-male group exhibited 20.1%–40.0% pyriform spermatozoa in semen samples; the remaining 28.2% of P-males had ≤20.0% or ≥40.0% pyriform spermatozoa ( Supplementary Figure 3 ). The intra- and interobserver correlation coefficients were considered good and excellent, respectively. The intraobserver ICC was 0.890 (95% confidence interval [CI]: 0.853–0.905), and the interobserver ICC was 0.902 (95% CI: 0.860–0.940). This assessment was conducted by three experienced laboratory technicians, each with at least five years of experience in evaluating sperm morphology. After image captured by the ASMA system, 2805 spermatozoa were considered typically pyriform on the basis of complete or partial agreement of three or two experts. Table 3 presents the morphological characteristics of 2805 pyriform sperm heads, including size, shape and area. Detailed descriptions of all the measured sperm parameters are provided in the legend of Table 3 . Compared with the normal sperm heads from fertile males ( Supplementary Table 1 ), four head morphometric parameters (L, R, HG, and HA) of pyriform heads were significantly higher, whereas the sperm head area parameters (EIoU, GIoU, SAS, and LAS) were markedly lower ( Supplementary Table 2 ). The characteristics of classic pyriform sperm heads are given in detail in Table 3 . Distributions of the morphometry in normal sperm heads from fertile males These data come from the former published article by the same research team. 54 HG: head girth; HA: head area; EIoU: ellipse intersection over union; GIoU: girth intersection over union; SAS: short-axis symmetry; SAS: short-axis symmetry; CI: confidence interval; s.d.: standard deviation Morphometric parameters of pyriform and normal sperm heads The abbreviations have been identified in Table 3 . IQR: interquartile range; HG: head girth; HA: head area; EIoU: ellipse intersection over union; GIoU: girth intersection over union; SAS: short-axis symmetry; SAS: short-axis symmetry Papanicolaou staining and SEM were performed to investigate pyriform sperm head morphology. Normal spermatozoa, from the C-male group, showed an oval head with a length-to-width ratio of (1.30–1.80):1 ( Figure 1a ). Typical pyriform spermatozoa, from the P-male group, exhibited an extremely elongated head with a length-to-width ratio of (1.56–3.72):1 ( Figure 1b ). Unlike the normal morphological characteristics of sperm heads, pyriform sperm heads exhibited an obvious deformation with a narrowed post-acrosomal area, a long head, and a larger head area ( Supplementary Table 2 ). A noteworthy morphological abnormality, an enlarged space between the acrosome and nucleus and a smaller coverage area of the acrosomes, was observed in the SEM images of pyriform sperm heads ( Figure 1c and 1d ). Papanicolaou staining and transmission electronic microscope (TEM) analysis of normal and pyriform spermatozoa head. ( a ) Normal sperm head using Papanicolaou staining (blue arrow). ( b ) Pyriform sperm head after Papanicolaou staining (red arrows). ( c ) TEM for normal sperm head. ( d ) The detached acrosome in the probands’ sperm. Red arrow points to acrosomes and asterisks indicate the enlarged space between the acrosome and nucleus. Ac: acrosome; Nu: nucleus. Papanicolaou staining scale bars = 10 µm. TEM scale bars = 500 nm. In the P-male group, as the proportion of pyriform spermatozoa increased, the percentage of normal sperm morphology, sperm concentration, total sperm count, sperm vitality, total motility, progressive motility, TMSC, VCL, VSL, ALH, and VAP slightly decreased (all P < 0.01). The percentage of immotile spermatozoa, TZI, and SDI increased as the proportion of pyriform-headed spermatozoa increased (all P < 0.01). The detailed outcomes are shown in Table 2 . Advanced evaluations of sperm function were performed on some semen samples from the C-male and the P-male groups. In the C-male group, 542 semen samples were analyzed for DFI, HDS, and sperm nuclear maturity; 145 samples were examined for sperm acrosome integrity, acrosome reaction, and sperm acrosin activity; and 320 samples were evaluated for MMP. In the P-male group, 537 semen samples were studied for DFI, HDS, and sperm nuclear maturity; 163 samples were evaluated for sperm acrosome integrity, acrosome reaction, and sperm acrosin activity; and 334 samples were analyzed for MMP. Compared with the C-males, the sperm DFI and HDS were significantly higher in the P-males (both P < 0.001). The sperm nuclear maturity rate, sperm acrosome integrity rate, acrosome reaction rate, and sperm MMP in the pyriform group were lower than those in the C-male group (all P < 0.001). There was no noticeable difference in the sperm acrosin activity of the P-male group versus C-male group ( P = 0.403; Supplementary Table 3 ). Advanced evaluations of spermatozoa between the P-male group and C-male group DFI: DNA fragmentation index; HDS: high DNA Stainability, MMP: mitochondrial membrane potential; IQR: interquartile range To identify the appropriate values for the specific proportion of pyriform spermatozoa that may be predictive of sperm functions (DFI, HDS, sperm nuclear maturity, sperm acrosome integrity, sperm acrosome reaction, sperm acrosin activity, and sperm MMP), we used 15.0% for DFI, 15.0% for HDS, 70.0% for sperm nuclear maturity, 85.0% for sperm acrosome integrity, 15.0% for sperm acrosome reaction, 59.3% for sperm acrosin activity, and 52.0% for sperm MMP, as the reference values to draw receiver operating characteristic curves. According to these, the proportions of pyriform spermatozoa showed the best discriminatory performance in distinguishing patients with the above-mentioned abnormal sperm DFI, HDS, sperm nuclear maturity, sperm acrosome integrity, acrosome reaction, acrosin activity, or sperm MMP are given in Table 4 . Additionally, the cut-off value for the proportion of pyriform spermatozoa in clinical pregnancies was 38.5% ( Table 4 and Supplementary Figure 4 ). According to the clinical pregnancy cut-off value for pyriform spermatozoa, the study population was divided into two subgroups (Group A and B): Group A of 47 couples, with male partners showing a proportion of pyriform spermatozoa ≤38.5%, whereas Group B included 12 couples, with the proportion of pyriform spermatozoa >38.5%. Receiver operating curves of pyriform sperm proportion predicting sperm functions and assisted reproductive technology outcomes AUC: area under curve; DFI: DNA fragmentation index; HDS: high DNA stainability; MMP: mitochondrial membrane potential; ART: assisted reproductive technology As shown in Table 5 and Table 6 , IVF/ICSI (69/39 cycles) outcomes were analyzed. There were no differences in male age, female age, female BMI, serum anti-Müllerian hormone (AMH) level, duration of infertility (in year), sperm concentration, progressive motility, sperm vitality, or the percentage of normal-form spermatozoa between the two groups. The median percentages of pyriform-headed spermatozoa were 30.0% and 40.0% in Group A and B, respectively ( P < 0.001). The ICSI outcomes in Group A were significantly higher than those in Group B, including high-quality embryo (44.2% vs 15.3%; P < 0.001), clinical pregnancy (50.0% vs 41.7%; P <0.001), and live birth rates (25.0% vs 16.7%; P < 0.001). The IVF fertilization rate (Group A vs B: 87.5% vs 90.0%) did not differ between the two groups ( P = 0.758). The fertilization rate of ICSI cycles was significantly higher in Group A compared with Group B (90.5% vs 85.2%, P <0.001). The total early pregnancy loss rate was significantly higher in Group B compared with Group A (42.9 % vs 13.3%, P < 0.001). Comparison between subgroups of pyriform sperm in clinical outcome parameters PSP: proportion of pyriform spermatozoa; BMI: body mass index; AMH: serum anti-Müllerian hormone; IQR: interquartile range; -: no data on this item The ART outcomes in subgroups of pyriform sperm Early pregnancy loss was defined as miscarriage within 12 weeks of gestation. * Both IVF and ICSI cycles between Group A and B in variables are statistically significant. PSP: proportion of pyriform spermatozoa; BMI: body mass index; AMH: serum anti-Müllerian hormone; IVF: in vitro fertilization; ICSI: intracytoplasmic sperm injection; ART: assisted reproductive technology; -: no data on this item

This research has revealed that pyriform sperm heads influence basic semen parameters, advanced sperm function analysis, and ART results. We established a link between the percentage of pyriform spermatozoa and ART outcome, and propose introducing the percentage of seminal pyriform spermatozoa as a feasible diagnostic metric. The current study has several limitations. For example, the use of a single center may limit the use of our findings, and it does not address external factors, such as lifestyle, environment, or medical conditions, which may have introduced a subjective bias. Future studies should provide clear recommendations for treatment and clinical intervention for patients with pyriform spermatozoa to assess the predictive value of these cells for fertility or reproductive treatment success.

After controlling confounding factors by using PSM method, this single-center retrospective non-intervention study of 1302 patients with previously detected pyriform spermatozoa in semen demonstrated a significantly increased risk of poor semen quality. Morphological parameters and SEM found that the pyriform sperm heads had noticeable structural deformation, characterized by increased length and reduced post-acrosomal space. A high proportion of seminal pyriform spermatozoa were associated with various sperm functions and ART results. Our current research showed that the presence of pyriform spermatozoa in semen is negatively related to sperm function and male fertility. The presence of pyriform-headed spermatozoa in semen indicates poor-quality spermatogenesis, affecting even spermatozoa with normal morphology. A previous paper showed that the average number of spermatozoa with a pyriform head was significantly higher in subfertile men. 9 The present study showed that the mean and median percentages of pyriform-headed spermatozoa in semen from subfertile men were 35.9% and 33.0%, respectively. These values are higher than that in a recent study reporting 22% for the mean proportion of pyriform head defects in subfertile males; 27 the difference may reflect the study population here coming from an andrology clinic and involving a larger number of potentially infertile male patients. A recent survey of the general male population found that the mean and median percentages of pyriform-headed spermatozoa were 9.1% and 6.0%, respectively. 28 Previous literature has reported that obese men have a higher proportion of pyriform-headed spermatozoa, and provided potential mechanisms through which obesity impairs sperm function. 29 Obesity might cause aromatase overactivity, affect the endocrinological properties of white adipose tissue, and induce oxidative stress and dysregulation of the hypothalamic–pituitary–gonadal axis. 29 In the current study, morphology-related endpoints (such as normal sperm morphology percentage, TZI, and SDI) of semen in P-male group remained significantly low even after PSM. The present data also showed that ERC defects were more frequently observed in the pyriform sperm head group than those in the control group. This result is similar to a recent study, showing that pyriform-headed spermatozoa are more commonly observed in men with abnormal semen results, and pyriform-headed spermatozoa have more frequent midpiece, tail, and ERC defects than spermatozoa with normal heads. 28 Spermatozoa may not be able to shed residual cytoplasm result in ERC after spermiogenesis, so ERC appearance can be indicative of abnormal sperm maturation; abnormal maturation of sperm leads to various sperm deformations, including pyriform sperm heads. 30 31 Blockage or abnormalities of sperm function and high levels of reactive oxygen species lead to potential male infertility. 32 In present study, the proportion of ERC, normal sperm morphology percentage, TZI, and SDI in men with pyriform spermatozoa in semen were markedly high, suggesting that pyriform spermatozoa reflect abnormal sperm development and male infertility. Sperm size and flagellum length vary considerably between species, while sperm head length varies to a lesser degree compared to sperm size and flagellum length. 33 34 However, sperm size is fairly uniform in many mammals, and large differences in sperm size can only be observed in marsupials and rodents group ( e.g ., marsupial spermatozoa are the longest among mammals, ranging in total sperm length from 80 μm to 350 μm). The present data show that in subfertile human males, the pyriform sperm head area and girth are considerably larger than those of normal sperm heads, and the symmetry of length and width are considerably lower than those of normal sperm heads. These findings indicate that the pyriform spermatozoon is a uniquely malformed cell. The heterogeneity of morphometric parameters in pyriform sperm heads increases the abnormal sperm characteristics of some infertile males. The cause of pyriform-headed spermatozoa is unclear. A recent report suggests that environmental pollution can increase the occurrence of pyriform-headed spermatozoa, especially with the intensification of urban particulate matter pollution, which has been proposed to significantly increase the incidence of pyriform-headed spermatozoa of men in urban areas. 35 Pyriform head defects usually result from abnormal testicular function, such as disturbances in intratesticular heat regulation or endocrine balance. 36 The percentage of pyriform heads is positively correlated with the percentages of elongated sperm heads, vacuolated heads, bent heads, and ERC defect. Ultrastructural analysis has shown that elongated sperm heads and pyriform sperm heads exhibit several distinctive features, including abnormal nuclear elongation, the presence of additional membranous layers between the outer and inner nuclear envelope membranes, and poorly condensed chromatin structure. From these results, it has been proposed that the noticeable extension of the sperm head and disrupted acrosome formation may be related to malfunction of the manchette and acroplaxome in spermatids, as well as the cytoskeleton of Sertoli cells. 37 An altered intratesticular microenvironment may also affect motor or signal proteins that are essential for these structures to operate. 35 Combined with the current data, these findings suggest that patients with a high proportion of pyriform-headed spermatozoa are more likely to have other sperm abnormalities, such as DNA, chromatin, and MMP defects, although a previous study found that spermatozoa with pyriform heads had a DFI similar to that of typical normal spermatozoa. 38 The sperm-zona pellucida (ZP) binding process preferentially selects sperm heads with a large anterior region that has relatively low optical density, high axial symmetry, and minimal neck abnormalities, but no bias against spermatozoa with pyriform morphology has been observed. 39 In the current study, the samples containing pyriform-headed spermatozoa still can bind to the ZP, possibly due to the acrosomal region of the pyriform spermatozoa is similar in shape to that of spermatozoa with normal morphology. Nonetheless, the ZP binding rate of spermatozoa in P-male group is lower than that of the spermatozoa with normal morphology. Moreover, a previous article reports that pyriform spermatozoa can bind to the ZP; however, it is still difficult for pyriform spermatozoa to achieve fertilization in vitro . 10 Pyriform-headed spermatozoa have a diminished capacity to bind to and penetrate the ZP, and zygotes resulting from the fertilization of oocytes by pyriform spermatozoa appear to have a reduced ability to initiate cleavage. 40 The present research demonstrates that pyriform sperm heads display abnormal shape, with irregularities in overall structure, together with high DFI, indicative of genetic material damage. The increased presence of nucleoprotein abnormalities suggests potential disruption in the proteins crucial for sperm function. Pyriform-headed spermatozoa were found to have a decreased MMP, indicating impairment of sperm energy production. Acrosomal integrity of pyriform spermatozoa (a structural component of the sperm essential for fertilization) was lower than that of normal spermatozoa, with complications of fusion with an oocyte. The reduced acrosome reaction of pyriform spermatozoa suggests difficulties in their ability to initiate fertilization. These findings together indicate that the phenotypic morphological abnormalities of pyriform spermatozoa heads relate to the abnormal physical characteristics and functions of the spermatozoa. These abnormalities, including promoted DNA fragmentation, nucleoproteins, MMP, acrosome integrity, and induced acrosome reaction, have the potential to impact male fertility adversely. Abnormal morphological spermatozoa may not swim efficiently enough to achieve fertilization both in vivo and in vitro . 33 This inefficiency is related to hydrodynamic swimming properties and indicates problems in sperm formation that affect sperm signaling and bioenergetics. The current study noted fertilization, high-quality embryo, and clinical pregnancy rates were reduced with a high proportion of pyriform spermatozoa in semen related to poorer ART outcomes. This suggests that pyriform spermatozoa lead to problems in embryos and clinical outcomes. The extensively recorded link between chronological age and sperm methylome has facilitated the creation of epigenetic clocks, which determine the biological age of spermatozoa through DNA methylation, referred to as the epigenetic age of spermatozoa. 41 Research has indicated that epigenetic sperm age is positively correlated with prolonged time-to-pregnancy (TTP) for couples attempting IVF and negatively correlated with gestation time for their offspring, regardless of maternal contributions. 42 The Longitudinal Investigation of Fertility and Environment (LIFE) study showed a notable correlation between epigenetic sperm age and the increased occurrence of pyriform spermatozoa in semen, 41 which implies a link between pyriform sperm heads and prolonged time-to-pregnancy, highlighting the need for additional research. Many elements contribute to the formation of pyriform sperm heads, encompassing mutations in genes associated with sperm head formation. There are few reports on the pathogenic genes of pyriform-shaped spermatozoa. A screening of 82 Chinese such individuals showed that a homozygous mutation in coiled-coil and C2 domain containing 1B ( CC2D1B ; c.1277G>C [p.R426P]), in which it was proposed that CC2D1B plays a crucial role in the remodeling of the sperm nuclear envelope, which may promote a pear-shaped nucleus. 43 44 Another study detected CC2D1B in the germ cell nuclei of both humans and mice. This finding suggested that the gene is responsible for aberrant sperm head morphology. 44 However, this association, primarily from knockout studies, remains clinically unproven in men. 31 They also pinpointed two pathogenic variants of Septin12 linked to sperm head deformities and fractured acrosomes in mice; variants in Septin12 have been found in two individuals with abnormal sperm head shapes, predominantly pyriform, along with acrosomal defects. 44 A nonsense mutation in CAB CO-repressor containing 1 ( CABCOCO1 or KIAA1210 ; c.3418G>T [p.E1140*]) has been found in a family with two individuals affected with pyriform-headed spermatozoa and defective acrosomes. 44 The human KIAA1210 gene located on chromosome Xq24 and comprising 16 exons, is highly expressed in late spermatids, and the encoded protein is present in acrosomal vesicles, which localizes to the ectoderm specialization cell junctions. 45 However, Kiaa1210 knockout mice do not show any sperm head defects, which warrants further investigations. 31 The tight association of pyriform spermatozoa in semen with semen quality, sperm function, and the ART outcomes necessitate more research into the molecular mechanisms responsible for pyriform morphological anomalies. Research on knockout animals and human monomorphic teratozoospermia has revealed that mutations in certain spermiogenesis-regulating genes that may result in pyriform spermatozoa. 33 Spermiogenesis leading to pyriform sperm abnormalities may be affected by pathogenic genes and compromised testicular function. For severe cases of male infertility, traditional IVF or ICSI is an option, but even with their success, certain couples experience suboptimal results, possibly due to inferior oocyte or sperm quality. 46 Research has shown that ICSI cycles using sperm from men with severe oligoasthenoteratozoospermia result in lower embryo implantation and clinical pregnancy rates compared to those with normal semen parameters. 47 This finding highlights the significant impact of the paternal contribution on reproductive outcomes. It further emphasizes the necessity of selecting the most suitable spermatozoon for the ICSI procedure. The use of testicular spermatids is increasingly significant because of the lower DFI than that of ejaculated spermatozoa; 48 moreover, these bypass the sperm nuclear condensation process, avoiding abnormalities that may lead to pyriform spermatozoa. Lifestyle improvement is a widely used, non-invasive method to enhance semen quality. 49 This approach includes strategies such as weight reduction, physical exercise, and smoking cessation. These changes are associated with improvements in semen parameters, including sperm morphology. 50 51 52 53

All authors declare no competing interests.

Study participant selection flowchart. Schematic descriptions for sperm head shape parameters. Ellipse IoU = [Intersection/Union]. The ellipse drawn by the black line represents a standard digital mimetic ellipse. The drop-like figure drawn by the red line represents a real sperm figure. “Ellipse IoU = 1” means the real sperm figure is a standard ellipse. Girth IoU = [real sperm perimeter/ a standard digital mimetic ellipse perimeter]. The sperm head perimeter measurement is based on digital pixel measurement. The abbreviations have been identified in Table 3 . The pyriform sperm proportion (PSP) distribution in P-males group ( n = 1302). The patient numbers of the series PSP are 2, 61, 450, 485, 190, 63, 32, 13, and 6. PSP: pyriform sperm proportion. Receiver operating characteristic curve analysis. ( a ) ROC of assessment of sperm chromatin and ART outcomes (DNA fragmentation index: AUC = 0.57, P = 0.029. HDS: AUC = 0.65, P = 0.001. Sperm nuclear maturity: AUC = 0.64, P = 0.000. Clinical pregnancy: AUC = 0.59, P = 0.152). ( b ) ROC of assessment of Sperm acrosome integrity (AUC = 0.81, P = 0.000), acrosome reaction (AUC = 0.59, P = 0.193), acrosin activity (AUC = 0.564, P = 0.442 and sperm mitochondrial membrane potential (AUC = 0.63, P = 0.000). The best discriminating percentages are detailed in Table 4 . ROC: receiver operating characteristic curve analysis. AUC: area under the curve.