Epigenetic sperm quality testing for predicting fertility treatment success: a real-world and multi-site analysis.

Of the 537 patients, 449 had a normal SpermQT result (83.6%) and 88 patients had an Abnormal SpermQT result (16.4%); 532 had male age data, 461 had female age data, and 375 had pre-wash TMC data. The mean male and female ages were 35.9 and 33.8 respectively, while the mean pre-wash total motile sperm count was 105.8 million (Table  1 ). No statistical differences were seen between the male and female ages of couples with Normal vs. Abnormal SpermQT results, nor between the total motile counts of couples with Normal vs. Abnormal SpermQT results (Table  1 ). Among the 375 men with available TMC data, 64 had an Abnormal SpermQT result and 311 had a Normal SpermQT result. Median TMC values were similar between groups, with 70.6 million in the Abnormal group and 78.3 million in the Normal group. The 75th percentiles were also comparable (143.8 versus 142.4 million) (Supplementary Table  1 ). The Abnormal group showed slightly lower values at the left tail of the distribution (10th percentile 3.7 versus 11.4 million) and a wider right-sided spread, including higher upper percentile values (90th percentile 312.7 versus 225.0 million) and a greater maximum TMC (861.2 versus 545.6 million) (Supplementary Table  1 ). Kernel-style density histograms and box plots (Supplementary Fig.  1 a and b) demonstrated substantial overlap between groups, with both showing a similarly right-skewed distribution. Consistent with these visual findings, the Kolmogorov–Smirnov test detected no statistically significant difference between the distributions of TMC comparing Normal and Abnormal ( D  = 0.119; p  = 0.408). Table 1 Distribution of patient data includes SpermQT results, male ages, female ages, and total motile count (TMC) across the total patient cohort and subgroups with Normal and Abnormal SpermQT results. Total patient numbers are included for each type of measure. p -values were calculated using t -tests and comparing the Normal and Abnormal cohorts for male age and female age Distribution of patient data includes SpermQT results, male ages, female ages, and total motile count (TMC) across the total patient cohort and subgroups with Normal and Abnormal SpermQT results. Total patient numbers are included for each type of measure. p -values were calculated using t -tests and comparing the Normal and Abnormal cohorts for male age and female age Across all patients, 202 couples completed at least one IUI and had a reported outcome and 90 couples had completed at least one IVF-ICSI ET and had a reported outcome. The remaining couples had not yet completed treatment, did not have documented outcomes, were lost to follow-up, or were pending outcomes. A total of 202 couples completed at least one IUI with a reported outcome. Documented pregnancy outcomes included both ultrasound-confirmed or hCG pregnancy (clinical data collected did not discriminate between hCG pregnancy and ultrasound pregnancy). The mean number of completed IUI cycles per patient was 2.8. There were 39 total pregnancies across all IUI cycles (19.3% pregnancy rate), with 19 pregnancies occurring in the first IUI cycle (9.4% pregnancy rate). A previous study focused on IUI outcomes reported a 10.9% pregnancy success rate for the first IUI cycle and a 19.4% success rate across three or more IUI cycles [ 11 ]. One hundred seventy-seven couples had Normal SpermQT results (87.6%) and 25 couples had Abnormal SpermQT results (12.4%). The overall IUI pregnancy rate among couples with Normal SpermQT results was 22% across an average of 2.8 IUI cycles. There were no IUI pregnancies in the couples with Abnormal SpermQT results across an average of 2.4 cycles and a statistically significant p -value of 0.005 when comparing pregnancy outcomes between Normal and Abnormal SpermQT results (Table  2 ). No statistical difference in sperm TMC, male and female ages, or IUI cycle numbers was seen between the Normal and Abnormal SpermQT groups (Table  2 ). Importantly, among men with an Abnormal SpermQT result who had available TMC data ( n  = 18), 72.2% (13 of 18) had a TMC greater than 20 million, indicating that the majority would likely have been classified as fertile based on conventional semen analysis parameters. Among the 147 men who underwent IUI and had available TMC data, 18 had an Abnormal SpermQT result and 129 had a Normal SpermQT result. Median TMC values differed between groups, with 39.9 million in the Abnormal group and 70.1 million in the Normal group, although both groups demonstrated broad overlap in their distributions. The 75th percentiles were also comparable (84.5 versus 136.2 million) (Supplementary Table  2 ). The Abnormal group showed lower values at the left tail of the distribution (10th percentile 5.4 versus 14.2 million) and a narrower right-sided spread, while the Normal group exhibited higher upper percentile values (90th percentile 222.6 versus 166.4 million) and a greater maximum TMC (544.6 versus 387.8 million) (Supplementary Table  2 ). Kernel-style density histograms and box plots (Supplementary Fig.  2 a and b) again demonstrated substantial overlap between groups, with both displaying similarly right-skewed distributions typical of TMC data. Consistent with these visual findings, the Kolmogorov–Smirnov test detected no statistically significant difference between the distributions of TMC comparing Normal and Abnormal SpermQT results ( D  = 0.253; p  = 0.221). Table 2 Analysis of pregnancy outcomes for patients who completed an IUI and had documented pregnancy outcomes ( N  = 202). Total pregnancies and pregnancy percentages were calculated across all IUI cycles (average 2.8 cycles) and are provided for the total population and by SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using Fisher’s exact test, and p -values for male age, female age, and total motile count (TMC) comparisons were calculated using t -tests Analysis of pregnancy outcomes for patients who completed an IUI and had documented pregnancy outcomes ( N  = 202). Total pregnancies and pregnancy percentages were calculated across all IUI cycles (average 2.8 cycles) and are provided for the total population and by SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using Fisher’s exact test, and p -values for male age, female age, and total motile count (TMC) comparisons were calculated using t -tests To achieve statistical significance with a distribution ratio of 12.4% Abnormal SpermQT results, with the achieved pregnancy rates (22% vs. 0%), we required at least 167 patients (assumes a type-1 error rate of 0.05 and a type-II beta error rate of 0.2), which was exceeded in this analysis. Using the Haldane–Anscombe correction (adding 0.5 to all cells to account for the zero cell in the Abnormal group), the crude risk ratio (RR) for pregnancy among patients with a Normal SpermQT result was estimated to be 11.5 (95% CI = 0.73 to 182.1). The absolute risk difference (RD) remained 22.0 percentage points (95% CI = 15.9% to 28.1%), indicating a substantial observed difference in pregnancy outcomes between groups. To evaluate whether provider-level variation confounded the association between SpermQT result and IUI pregnancy, we performed a physician-stratified analysis using the Cochran–Mantel–Haenszel method. Of the 28 physicians represented in the dataset, 12 had at least one patient in both SpermQT results, allowing them to contribute to the stratified analysis (135 total patients) (Supplementary Table  3 ). Across these providers, all documented IUI pregnancies occurred in the Normal SpermQT group, with no pregnancies observed among men with Abnormal SpermQT results. After stratifying by physician to adjust for provider-specific variability in patient selection and treatment practices, the association between SpermQT result and IUI pregnancy remained statistically significant (CMH χ 2  = 6.31, p  = 0.012). Stratification by clinic produced a similar pattern. Of the 10 clinics, nine had patients with both Abnormal and Normal SpermQT results (200 total patients analyzed), and no IUI pregnancies occurred among men with Abnormal SpermQT results (Supplementary Table  4 ). After adjusting for clinic-level variation, the association between SpermQT result and IUI pregnancy again remained statistically significant (CMH χ 2  = 6.82, p  = 0.009). These findings indicate that the reduced IUI success among men with Abnormal SpermQT results is consistent across physicians and clinics and is not explained by provider- or site-specific differences in patient selection or treatment practices. When analyzing the pregnancy success from just the first cycle of IUI across all 202 couples, the pregnancy rates for the first IUI cycle in these patients were 9.4%, n  = 19 pregnancies. All pregnancies in the first IUI cycle were from patients with a Normal SpermQT result (Table  3 ). To achieve statistical significance with a distribution ratio of 12.4% Abnormal SpermQT results and with the achieved pregnancy rates of 10.7% vs. 0%, the analysis requires at least 390 patients (assumes a type-1 error rate of 0.05 and a type-II beta error rate of 0.2). In the case of 202 patients, the p -value comparing pregnancy outcomes between Normal and Abnormal SpermQT results is p  = 0.14 (Table  3 ). Additional patient outcomes will be needed for a statistically powered analysis of the first IUI cycle. Table 3 Analysis of pregnancy outcomes for the first IUI cycle. Total pregnancies and pregnancy percentages were calculated for the first IUI cycle ( N  = 202) and for each SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using Fisher’s exact test. All other descriptive statistics are identical to Table  2 Analysis of pregnancy outcomes for the first IUI cycle. Total pregnancies and pregnancy percentages were calculated for the first IUI cycle ( N  = 202) and for each SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using Fisher’s exact test. All other descriptive statistics are identical to Table  2 Due to the absence of pregnancy events in the Abnormal group, multivariable regression modeling was not feasible. While these findings suggest a strong association between epigenetic sperm quality and IUI pregnancy success, the confidence interval for the risk ratio is wide due to the small sample size in the Abnormal group. Ninety couples had pregnancy outcomes available and completed at least one IVF-ICSI ET. Patients completed an average of 1.6 ET cycles. There were 51 total pregnancies across all ET cycles (56.7% pregnancy rate), and 32 pregnancies in the first ET cycle (35.6% pregnancy rate). Sixty-eight couples had Normal SpermQT results (75.6%) and 22 couples had Abnormal SpermQT results (24.4%). Thirty-six pregnancies (52.9%) were achieved in couples with Normal SpermQT results across a mean of 1.6 ETs and 21 pregnancies (30.9%) were achieved in their first ET cycle. Fifteen pregnancies (68.2%) were achieved in couples with Abnormal SpermQT results across a mean of 1.7 ETs and 11 pregnancies (50%) were achieved in their first ET cycle. No statistical differences were observed comparing pregnancy rates between the Normal and Abnormal results, p  = 0.21 and p  = 0.10 respectively (Tables 4 and 5 ). No statistical differences were observed in male or female ages, TMC, or ET cycle numbers between Normal and Abnormal SpermQT results (Table  4 ). Across all ET cycles and within the first ET cycle, patients with an Abnormal result have higher pregnancy rates, a phenomenon that is worthy of further investigation and would require a patient number of 501 (assuming similar pregnancy outcomes and distribution of results). Table 4 Analysis of pregnancy outcomes across patients who completed an IVF-ICSI embryo transfer (ET) and had documented pregnancy outcomes ( N  = 90). Total pregnancies and pregnancy percentages were calculated across all ET cycles (average = 1.6 ETs) and are provided for the total population and each SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using chi-squared tests, and p -values for male age, female age, and total motile count (TMC) comparisons were calculated using t -tests Table 5 Analysis of pregnancy outcomes for the first IVF-ICSI ET cycle. Total pregnancies and pregnancy percentages were calculated for the first ET cycle ( N  = 90) and for each SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using chi-squared tests. All other descriptive statistics (male age, female age, and total motile count [TMC]) are identical to those presented in Table  4 Analysis of pregnancy outcomes across patients who completed an IVF-ICSI embryo transfer (ET) and had documented pregnancy outcomes ( N  = 90). Total pregnancies and pregnancy percentages were calculated across all ET cycles (average = 1.6 ETs) and are provided for the total population and each SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using chi-squared tests, and p -values for male age, female age, and total motile count (TMC) comparisons were calculated using t -tests Analysis of pregnancy outcomes for the first IVF-ICSI ET cycle. Total pregnancies and pregnancy percentages were calculated for the first ET cycle ( N  = 90) and for each SpermQT subgroup. p -values for pregnancy rate comparisons were calculated using chi-squared tests. All other descriptive statistics (male age, female age, and total motile count [TMC]) are identical to those presented in Table  4 Using a modified Poisson regression model with robust standard errors, we assessed whether SpermQT result was independently associated with pregnancy while adjusting for male age, female age, TMC, and number of ET cycles. Sample size was reduced to 57 to accommodate complete cases of all collected data. The analysis found that SpermQT result was not significantly associated with pregnancy outcome in the cases of IVF-ICSI (adjusted risk ratio (aRR) = 0.81, 95% CI = 0.53–1.22, p  = 0.31). None of the other covariates (male age, female age, TMC, number of ET cycles) was significant predictors of pregnancy either (Table  6 ). Table 6 A modified Poisson regression model with robust standard errors was used to evaluate the independent association between SpermQT result and clinical pregnancy, adjusting for male age, female age, total motile count (TMC), and number of embryo transfer (ET) cycles. The analysis was limited to 57 complete cases with non-missing data for all covariates. SpermQT result was not significantly associated with pregnancy outcome (adjusted risk ratio [aRR] = 0.81, 95% CI = 0.53–1.22, p  = 0.31). None of the covariates were significantly associated with pregnancy (see table for full model estimates) Variable Risk ratio (RR) 95% CI lower 95% CI upper p -value Intercept 1.81 0.30 10.71 0.52 SpermQTResult (Normal vs. Abnormal) 0.81 0.53 1.22 0.31 Male age 1.01 0.98 1.05 0.45 Female age 0.96 0.89 1.03 0.24 TMC 1.00 1.00 1.00 0.17 ET cycles 0.96 0.75 1.24 0.78 A modified Poisson regression model with robust standard errors was used to evaluate the independent association between SpermQT result and clinical pregnancy, adjusting for male age, female age, total motile count (TMC), and number of embryo transfer (ET) cycles. The analysis was limited to 57 complete cases with non-missing data for all covariates. SpermQT result was not significantly associated with pregnancy outcome (adjusted risk ratio [aRR] = 0.81, 95% CI = 0.53–1.22, p  = 0.31). None of the covariates were significantly associated with pregnancy (see table for full model estimates) These results suggest that in IVF-ICSI, the pregnancy outcome does not appear to differ significantly based on SpermQT result, which is consistent with the hypothesis that ICSI may overcome epigenetic sperm quality deficiencies. Three hundred twenty-nine gene promoters were identified to be statistically enriched in men with an abnormal result by an FDR-corrected chi-squared test compared to Normal results. All 329 enriched gene promoters are listed in Supplementary Table  5 and had an odds ratio ≤ 0.43 and an FDR-correct chi-squared p -value ≤ 0.012. The amount of enriched dysregulated gene promoters reflects the biological complexity and heterogeneity between male infertility patients. Analysis of enriched gene promoters showed enrichment of 254 distinct molecular functions with the most significant enrichment in Binding (GO: 0005488) and Catalytic Activity (GO: 0003824). For biological function analysis, the top two functions included Biological Regulation (GO: 0065007) and Cellular Process (GO: 0009987). The gene promoters that showed the most significant p -values and lowest odds ratio among men with an Abnormal SpermQT result include C18orf63 , LCN9 , TAS2R5 , SERPINB13 , and MIR23A (refer to Table  7 for statistics associated with these genes). C18orf63 is an uncharacterized gene located on chromosome 18, with high expression in spermatogonia cells. It is believed to interact with LRGUK (leucine-rich repeat and guanylate kinase), a protein involved in various aspects of sperm assembly, including acrosome attachment, sperm head shaping, and early stages of axoneme development [ 12 ]. The LCN9 (Lipocalin 9) gene encodes a protein from the lipocalin family, which is thought to be involved in transporting small hydrophobic molecules, and shows high expression in the testes, however primarily in the epididymis [ 13 ]. TAS2R5 , a bitter taste receptor gene, is expressed in various tissues, including the testes. Research suggests that taste receptors like TAS2R5 may have roles beyond taste perception, potentially influencing male reproductive processes such as sperm function [ 14 , 15 ]. Specifically, bitter taste receptors in human sperm have been linked to sperm motility and overall fertility, indicating that TAS2R5 might help modulate sperm activity. SERPINB13 , also known as hurpin, is a serine proteinase inhibitor primarily expressed in the male reproductive tract. It functions as a scavenger for prematurely activated acrosin, a critical enzyme for sperm maturation and fertilization; by regulating acrosin activity, SERPINB13 protects sperm from excessive proteolytic damage, ensuring their viability and ability to fertilize an egg. MIR23A is a microRNA that plays a crucial role in regulating gene expression through post-transcriptional mechanisms. MIR23A RNA has been identified in mature sperm cells and is implicated in the regulation of sperm development and maturation, influencing processes such as spermatogenesis and sperm motility [ 16 ]. Table 7 Here we present the five most significantly enriched gene promoters in men with Abnormal SpermQT results compared to those with Normal SpermQT results. These genes were selected based on their lowest FDR-corrected chi-squared p -values and lowest odds ratios among all statistically enriched genes. Each gene is listed with its p -value, odds ratio, and prevalence among men with Abnormal SpermQT Most significantly enriched dysregulated gene promoter in men with Abnormal SpermQT results Gene Chi-square p -value (FDR corrected) Odds ratio Prevalence C18orf63 1.85E − 10 0.06 19.1% LCN9 3.18E − 09 0.06 16.9% TAS2R5 1.17E − 07 0.05 13.5% SERPINB13 2.53E − 07 0.03 11.2% MIR23A 8.37E − 07 0.06 12.4% Here we present the five most significantly enriched gene promoters in men with Abnormal SpermQT results compared to those with Normal SpermQT results. These genes were selected based on their lowest FDR-corrected chi-squared p -values and lowest odds ratios among all statistically enriched genes. Each gene is listed with its p -value, odds ratio, and prevalence among men with Abnormal SpermQT

In this analysis of real-world clinical outcomes, SpermQT was ordered for clinical use at 10 fertility clinics across the USA between May 2022 and December 2023. The selection of patients for SpermQT assessment was based on the clinical decision of the ordering provider; azoospermic men were not included. SpermQT was ordered at different times in the fertility journey for each patient, including prior to an IUI procedure and after a single or multiple IUI procedures. As is common with real-world data, this introduces inherent selection bias, as testing decisions were based on provider discretion rather than uniform eligibility criteria. A semen sample was collected either at the fertility clinic, andrology laboratory, or at home using an at-home collection kit. All semen samples were shipped directly to a centralized laboratory for epigenetic analysis and SpermQT report generation. Sperm DNA methylation has been shown to be thermally stable at high temperatures and can therefore be shipped at ambient temperature to the CLIA-certified laboratory without use of buffers or dry ice [ 9 ]. Following the completion of the epigenetic analysis, all SpermQT results were provided to the ordering provider for interpretation and patient consultation. All SpermQT testing was performed by Neogen Genomics, a CLIA-certified clinical laboratory, using the validated procedures previously described by Miller et al. [ 8 ]. Neogen Genomics functioned solely as a service provider and had no role in study design, data analysis, interpretation, or manuscript preparation. None of the authors has any financial or professional conflicts of interest with Neogen Genomics. SpermQT measures the epigenetic (DNA methylation) dysregulation in sperm at 1233 genes which are specifically important and known for male fertility [ 8 ]. The higher the number of dysregulated genes, the lower the quality of the sperm. Semen samples were classified as Excellent, Normal, or Abnormal based on the SpermQT assay, which quantifies DNA methylation instability across a predefined panel of 1233 sperm-specific promoter regions [ 7 , 8 ]. Promoters were defined as the two kilobase region centered on the transcription start site, and methylation within each promoter was summarized using M-values, the log ratio of methylated to unmethylated signal intensities [ 7 ]. During test development, fertile men with normal semen parameters and confirmed pregnancy outcomes were used to establish a reference profile of promoter stability. For each of the 1233 promoters, the standard deviation of M-values across fertile controls was calculated and a promoter-specific stability cutoff was defined as three times this value. For samples outside the fertile reference cohort, a promoter is classified as dysregulated when the standard deviation of its M-values exceeds its promoter-specific stability cutoff, and the total number of dysregulated promoters constitutes the continuous SpermQT score. Clinical SpermQT categories (Excellent, Normal, and Abnormal) reflect empirically defined ranges of this continuous score that were calibrated in an independent National Institutes of Health cohort by mapping dysregulated promoter counts to intrauterine insemination pregnancy and live birth rates [ 7 , 8 ]. The same pre-specified category boundaries were applied in the present study without modification. De-identified patient data was aggregated and limited to (1) male partner age, (2) female partner age, (3) pre-washed total motile sperm count (TMC), (4) number of IUI cycles completed, (5) IUI pregnancy outcomes, (6) number of IVF-ICSI embryo transfer (ET) cycles completed, (7) IVF-ICSI pregnancy outcomes, and (8) SpermQT results. Pregnancy outcomes included biochemical (hCG > 5 mIU/mL) and/or ultrasound-documented pregnancies (gestational sac and fetal pole present). Spontaneous pregnancy data was not collected for the analysis and conventional IVF data was not collected as IVF-ICSI was the dominant procedure utilized across the 10 fertility clinics. All standard semen analyses (sperm count and motility) were done onsite at the clinic and were not done by the SpermQT commercial laboratory. All aggregated analyses were done as de-identified secondary outcomes analysis with no interference with medical care and were considered exempt from institutional review board review by Category 4 of federal regulations where the aggregated resulting dataset contains no information that can identify subjects. A total of 537 SpermQT results were generated across 28 ordering reproductive endocrinologists or reproductive urologists at geographically diverse locations across the USA. SpermQT was ordered at varying times during patient fertility evaluations. The semen samples tested with SpermQT were not the same samples utilized in the IUI or IVF-ICSI fertility procedures. The 537 patients likely represent a biased population of male patients who were considered ideal candidates for additional sperm testing, and do not represent the general population of patients entering a fertility clinic. All analyses of pregnancy outcomes were done comparing the cumulative pregnancy results between the individuals with a Normal and Excellent SpermQT result and individuals with an Abnormal SpermQT result. Due to the small sample size of Excellent results with IUI pregnancy outcomes ( n  = 14), this category was combined with Normal results. Cumulative IUI and IVF-ICSI ET outcomes were analyzed across all cycles and for the first IUI cycle and first IVF-ICSI ET cycle. Descriptive statistics were used to compare male and female age, TMC, number of IUI cycles, and pregnancy outcomes between groups with Normal and Abnormal SpermQT results. Chi-squared or Fisher’s exact tests were used to compare categorical variables (Fisher’s exact was used when any expected count was less than 5 or a zero cell occurred), and t -tests were used for continuous variables. To evaluate whether differences in TMC could account for observed associations with SpermQT results, we performed an exploratory analysis of TMC values among all men with available TMC data ( n  = 375 for all patients and n  = 147 for patients with IUI outcomes). We summarized TMC distributions for each SpermQT group using percentile values (10th, 25th, 50th, 75th, and 90th percentiles). To visualize potential distributional differences, we generated kernel style density approximations using fine bin histograms, as well as box plots for each SpermQT result. Distributional differences were statistically assessed using the two-sample Kolmogorov–Smirnov test. For pregnancy outcomes, crude risk ratios (RRs) and risk differences (RDs) were calculated to estimate the effect size of SpermQT result (Normal vs. Abnormal) on pregnancy. For the IUI group, where no pregnancies occurred in the Abnormal SpermQT group, we applied a continuity correction (adding 0.5 to the zero-event cell) to estimate the risk ratio and its 95% confidence interval (CI). The risk difference and 95% CI were calculated using standard epidemiological formulas based on pregnancy proportions. Due to zero outcomes in one comparison group, multivariable regression modeling was not performed for IUI outcomes. To assess whether provider-level or clinic-level variation could confound the association between SpermQT result and IUI pregnancy outcomes, we performed stratified analyses at both the physician and clinic levels. Physicians and clinics with at least one patient in each SpermQT result Normal versus Abnormal were included as independent strata for their respective analyses. For each analysis, the Cochran–Mantel–Haenszel (CMH) method was used to estimate the association between SpermQT result and IUI pregnancy while adjusting for variability attributable to physician or clinic, allowing us to determine whether the association remained significant after accounting for differences in practice patterns across providers and sites. The CMH analysis was performed by creating a 2 × 2 table for each provider and clinic comparing SpermQT result with IUI pregnancy outcome, and then combining these tables to estimate whether the association remained significant after adjusting for differences between clinics. Because no IUI pregnancies occurred in the Abnormal SpermQT group in any physician or clinic stratum, formal heterogeneity testing was not statistically reliable and was not performed. For the IVF-ICSI group, where pregnancy occurred in both SpermQT groups, a modified Poisson regression model with a log link and robust standard errors was used on complete cases ( n  = 57), defined as participants with non-missing data for all covariates (male age, female age, TMC, and number of ET cycles). To identify gene promoters associated with Abnormal SpermQT, we analyzed dysregulated genes across all patient samples. Using a chi-squared test adjusted by a false discovery rate (FDR) and odds ratio analysis, we determined which gene promoters were statistically enriched in men with Abnormal SpermQT compared to men with Normal SpermQT. Enriched genes were further analyzed using the Panther Classification System to assess common enriched molecular and biological functions [ 10 ].

This is the first multi-center clinical assessment of the predictive capability of the novel epigenetic sperm quality test, SpermQT. The data collected here support previously published retrospective data demonstrating that male partners with an Abnormal SpermQT result have a statistically lower likelihood of IUI-assisted pregnancy compared to those with a Normal result despite normal semen analysis parameters. These data additionally show that men with an Abnormal SpermQT result can still achieve pregnancy utilizing IVF-ICSI. There are unique limitations and strengths that come from the analysis of clinical real-world data. The most significant limitation is the inability to gather comprehensive clinical and phenotypic data. This analysis was limited to collecting ages, sperm TMC, procedure type (IUI or IVF-ICSI), number of treatment cycles, and procedure outcomes. Additional information regarding the female partner’s testing/health status (e.g., endometriosis, tubal patency, ovulatory status) was not available for this analysis and could have acted as confounders. Similarly, more details associated with the male patient’s health (e.g., full semen analysis, DNA fragmentation, hormone levels, and medical comorbidities) are necessary to rule out other male factors beyond SpermQT that could influence the chance of pregnancy. Importantly, unlike a blinded clinical trial, the analysis cohort of 537 male patients likely reflects a biased selection towards individuals deemed suitable by their physicians for advanced sperm testing. While this cohort likely reflects selection bias, given that providers ordered SpermQT based on clinical judgment rather than standardized criteria, it also reflects real-world clinical practice and decision-making patterns. Therefore, these findings should be interpreted within the context of this specific patient group, rather than generalized to the broader population. These results are particularly relevant for patients who benefit from additional male factor testing. Additionally, this data was limited to IVF-ICSI outcomes and did not contain any information regarding pregnancy success rates with conventional IVF across the SpermQT results. Understanding the role of conventional IVF will be crucial for better understanding the mechanism of the sperm epigenetic dysregulation measured by SpermQT and would help further refine the recommended treatment plans. Despite these limitations, this analysis shows the strong clinical utility of SpermQT in guiding treatment for couples where physicians deemed additional sperm testing as necessary. Interestingly, this analysis revealed a higher pregnancy rate among patients with an Abnormal SpermQT result undergoing IVF-ICSI. These results are not statistically significant ( p  = 0.21) but show a trend that is worth discussion and further investigation. Although our findings did not reach statistical significance, the effect estimates suggest potentially meaningful differences in pregnancy outcomes between Normal and Abnormal SpermQT groups. However, the study was underpowered to detect moderate associations, particularly within the IVF-ICSI subgroup, where only 57 complete cases were available for adjusted modeling. As such, wide confidence intervals should be interpreted with caution, and larger studies are needed to confirm these associations. Gene promoter analysis in individuals with an Abnormal SpermQT result reveals a broad spectrum of potentially dysregulated molecular and biological functions, consistent with findings from previous studies [ 8 ]. The significant enrichment and unique presence of certain genes, as indicated by low odds ratios, suggest that some dysregulated genes may be more prevalent than others. Continued data collection will allow for further refinement of SpermQT, facilitating the stratification of patients into more specific subgroups that could guide tailored treatment approaches. Moreover, the identification of these abundant genes may reveal therapeutic targets and intervention strategies aimed at improving sperm molecular function and ultimately enhancing pregnancy success rates. Given the influence of environmental exposures, diet, lifestyle, and medications on sperm DNA methylation [ 17 – 20 ], analyzing the impact of these factors on SpermQT should be the focus of future research. Efforts in this domain will play a crucial role in understanding how physicians can offer lifestyle recommendations to enhance the success of IUI as measured by SpermQT without resorting to more invasive and costly procedures such as IVF-ICSI. While additional blinded and observational studies are needed and will provide a controlled assessment of the ability of SpermQT to predict IUI outcomes, these data show the clinical utility of introducing SpermQT in the initial assessment of the male partner to guide fertility treatment recommendations in a real-world patient population.

Infertility affects 1 in 6 couples [ 1 ], and infertility rates have been rising for several decades [ 2 ]. It is estimated that 40 to 50% of infertility issues are due to factors associated with the male partner [ 3 ]. Currently, diagnostic modalities for evaluating male fertility are severely limited. The gold standard for male factor infertility remains the semen analysis, which visually assesses sperm quantity, morphology, and motility. Although semen parameters provide important clinical information, their ability to predict fertility outcomes remains limited [ 4 , 5 ]. Reliance on semen analysis as the primary (and often exclusive) assessment of sperm leads to an incomplete understanding of the male partner’s fertility potential. This limitation contributes to unnecessary procedures, increased treatment costs, and delays in achieving pregnancy [ 3 ]. These burdens can lead patients to discontinue fertility care due to the emotional, financial, and psychological stress associated with repeated clinical visits and unsuccessful procedures. Approximately 65% of patients discontinue care before achieving pregnancy [ 6 ], often before reaching the treatments most likely to result in success. Improved diagnostic tools for male factor infertility are therefore needed to guide more effective and timely treatment and reduce time to pregnancy. An epigenetic assessment of sperm quality, the Sperm Quality Test (SpermQT), was made clinically available in 2022. SpermQT quantifies epigenetic instability across a curated panel of 1233 promoter regions that were identified by selecting the top tenth percentile of the most methylation-stable promoters in fertile men with normal semen parameters and confirmed pregnancy outcomes [ 7 ]. Promoters were defined as the two kilobase region centered on the transcription start site, and methylation levels within each promoter were evaluated using M-values, the log ratio of methylated to unmethylated signal intensities [ 7 ]. After establishing the stable promoter panel, promoter-specific stability thresholds were derived by calculating the standard deviation of M-values across the fertile cohort for each promoter and setting a cutoff at three times that value. For samples outside the fertile reference cohort, a promoter is classified as dysregulated when the standard deviation of its M-values exceeds its promoter-specific stability cutoff, and the total number of dysregulated promoters forms the continuous SpermQT score. Clinical category boundaries (Excellent, Normal, and Abnormal) were subsequently defined by mapping dysregulated-promoter counts in an independent NIH cohort to differences in intrauterine insemination pregnancy and live birth rates [ 7 , 8 ]. Initial clinical studies showed that men with an Abnormal SpermQT result had significantly lower pregnancy and live birth outcomes after intrauterine insemination (IUI) despite similar sperm motility and concentration and similar total motile counts [ 8 ], whereas outcomes after in vitro fertilization with intracytoplasmic sperm injection (IVF-ICSI) did not differ between SpermQT results, suggesting that ICSI can overcome epigenetic sperm quality impairments [ 8 ]. In the first retrospective study of 544 men with IUI outcomes, the Abnormal SpermQT result was enriched for dysregulated promoters in genes involved in spermatogenesis, acrosome biology, chromatin remodeling, and ion and small molecule transport, underscoring the biological relevance of promoter-level epigenetic disruption to impaired sperm function [ 7 , 8 ]. As all data supporting the development of SpermQT were from retrospective cohort studies, additional research is needed to independently confirm these associations. Here, we report a multi-site, real-world analysis of the relationship between epigenetic sperm quality and pregnancy outcomes among patients undergoing IUI and IVF-ICSI.

Below is the link to the electronic supplementary material. ESM 1 (DOCX 194 KB) (DOCX 194 KB)