First-line infertility treatment in normal or subnormal sperm: Interest of a simplified pre-IMSI test.

In the field of male fertility exploration, a few “add-on” routine tests can be made to complete the basic semen examination [ 1 ], like high magnification motile sperm organelle morphology (MSOME) [ 2 ], sperm deoxyribonucleic acid (DNA) fragmentation [ 3 ] and sperm DNA base oxidation evaluations [ 4 ]. Sperm DNA integrity evaluations can only be used at diagnosis, not in real time during ART attempts. Evaluating spermatozoa at high magnification remains a good non-invasive strategy to complete the initial examinations even for couples with apparently normal/subnormal sperm. It could help embryologists choose the most suitable ART technique to use, particularly IMSI when this is required. MSOME is not a recent technique but, despite a significant number of studies on the subject, there is a lack of consensus on results and the ensuing decisions [ 2 , 5 – 10 ]. The origin of nuclear vacuoles in male infertility remains also widely debated [ 11 ]. Early reports suggest that sperm head vacuoles may be due to DNA fragmentation [ 12 – 17 ], DNA decondensation [ 18 – 20 ] or exposure to reactive oxygen species (ROS) [ 21 , 22 ]. Previous results [ 12 , 13 , 15 – 17 , 20 , 23 , 24 ] show that the DNA fragmentation rate may reflect the size and number of vacuoles, though the role of vacuoles in the process of DNA fragmentation is not yet fully understood. Moreover, spermatozoa with vacuoles have been shown to be associated with chromatin disorganization (abnormal chromatin condensation, aneuploidy, modification of spatial and chromosomal positioning) [ 2 , 13 , 18 , 25 , 26 ]. A potential association between spermatozoa genomic stability and vacuolar morphology and location was also published [ 27 ] as well as the potential link between vacuolated spermatozoa and some epigenetic marks involved in chromatin condensation [ 28 , 29 ]. Conversely, some studies argue that nuclear vacuoles result from a natural physiological process unrelated to DNA fragmentation [ 30 , 31 ]. In all cases, the use of high-magnification sperm selection to discard vacuolated sperm prior to microinjection might improve sperm selection and the outcomes of assisted reproduction techniques [ 32 – 34 ]. In practice, even today the indications for MSOME or IMSI are not clearly identified and these techniques are still considered as too time-consuming. Besides, the cut-off for classifying spermatozoa as “usable” or “non-usable” remains unclear and the classifications for normal or abnormal MSOME spermatozoa vary. Thus, assessing the size, number and location of vacuoles differs greatly from one study to another [ 17 ]. Recently, it has been suggested that the depth and location of vacuoles should also be included in the evaluation criteria [ 27 ]. The lack of consensus on the various classifications is also clear in the decision-making about the indications for MSOME and IMSI. In a recent review, Mangoli et al . [ 35 ] concluded that only patients with repeated implantation failures, severe male infertility factors or advanced male or maternal ages had higher chances of conceiving with IMSI. However, only a few articles in the literature focus on the use of MSOME and IMSI on normal/subnormal sperm [ 36 – 40 ]. In the light of all this, our study aimed to (1) develop and (2) evaluate a faster, simplified MSOME technique for routine use on the day of oocyte retrieval. We then (3) retrospectively confronted our test results with each ART technique outcome i.e. IUI, IVF, ICSI and IMSI, in patients with apparently normal/subnormal sperm.

The main characteristics of the patients under study are shown in Table 1 . Sperm concentrations and total sperm counts differed significantly between PIT score I and PIT score III subgroups even though these parameters were considered as normal according to the WHO 2010 laboratory manual [ 41 ]. We observed that the morphology was different between PIT score I and II subgroups, between PIT score I and III subgroups and between PIT score II and III subgroups. In PIT score I, the morphology score (1.5%) was considered as abnormal (normal if >4% of typical spermatozoa). Among the causes of infertility, we noted a significant difference in ovulatory factors between the three subgroups. Cycle characteristics are presented in Table 2 . The number of progressive spermatozoa inseminated was similar in each subgroup. We noted a significant difference in the number of ICSI cycles between PIT score II and III subgroups. No IMSI was proposed when the sperm was of PIT score III. Analysis of the number of embryos obtained and transferred showed that these were equivalent in each subgroup. Embryos were transferred between Day 2 and Day 5 according to our Center’s policy. Values are expressed as median and quartiles [q1;q3]. p<0 . 05 is considered significant. ICSI = intracytoplasmic sperm injection; IMSI = intracytoplasmic morphologically selected sperm injection; IUI = intrauterine insemination; IVF = conventional in vitro fertilization; MII = metaphase II oocyte; PIT = pre-IMSI test Clinical outcomes according to the three PIT subgroups were compared within the different ART techniques (Tables 3 and 4 ). The three PIT subgroups were compared with each other, using clinical pregnancy, miscarriage and live birth rates as reference points. For IUI, IVF and IMSI techniques, the PIT score subgroup appeared to have no effect on clinical outcomes (ongoing pregnancy and live birth). For ICSI, the effect of PIT score subgroup was significant for clinical pregnancies (p = 0.0054) and presented a statistical trend for live births (p = 0.0614). Miscarriage rates of IVF attempts were statistically different according to PIT score (p = 0.0348). In order to better interpret our study results, the last column of Table 3 shows the global percentage of all attempts at our ART Center during the study period (Feb 2017—Oct 2020). The only exclusion criterion of this reference population was women over 39 years of age. Values for ongoing pregnancies and live births are in percentages per transfer . Values for miscarriages are in percentages per ongoing pregnancy . ART = assisted reproductive technology; ICSI = intracytoplasmic sperm injection; IMSI = intracytoplasmic morphologically selected sperm injection; IUI = intrauterine insemination; IVF = conventional in vitro fertilization; NC = statistical non convergence; PIT = pre-IMSI test ART = assisted reproductive technology; CI95% = 95% confidence interval for OR; ICSI = intracytoplasmic sperm injection; IMSI = intracytoplasmic morphologically selected sperm injection; IUI = intrauterine insemination; IVF = conventional in vitro fertilization; OR = odds ratio; PIT = pre-IMSI test * = rates of ongoing pregnancies were statistically different overall between interventions in PIT score subgroups 2 and 3 (p<0.0001) a,b = OR on the same line with different letters are significantly different at the p<0.05 threshold Table 4 completes the results of Table 3 by adjusting clinical pregnancy results of PIT score subgroups on ART techniques and adding (1) two-by-two ART techniques comparisons in each PIT score subgroup estimating ORs and (2) comparisons of these different ORs of clinical pregnancy rates between the different PIT score subgroups. Results of the ICSI vs . IMSI comparison in PIT score I and II subgroups were significantly different and opposed. We observed that in PIT score II subgroup, clinical pregnancies rate appears to be higher with ICSI than with IMSI (OR = 2.69, CI95% = [0,53; 13.75]), whereas in the PIT score I subgroup there appears to be the reverse (OR = 0.385, CI95% = [0,10; 1.43]). In the PIT score II subgroup, more pregnancies occurred with ICSI than with IVF (OR = 4.65, CI95% = [1.69; 12.82]). The results of the ICSI vs . IVF comparison in the PIT score I (OR = 0.58, CI95% = [0.09; 3.64]) and III subgroups (OR = 1.19, CI95% = [0.71; 1.97]) were significantly different from PIT score II (OR = 4.65, CI95% = [1.69; 12.82]). The other comparisons (IVF vs . IUI, ICSI vs . IUI, IMSI vs . IVF and IMSI vs . IUI) gave results that were statistically equivalent between the three PIT score subgroups.

In this study we were able to validate a simplified pre-IMSI test, with international quality standards, in line with the scientific literature on the subject. The number of transfers in each ART and PIT score groups remains low and obliges us to remain cautious in interpreting these preliminary results. However, the results of this study appear to support the value of PIT for defining the choice of the most effective assisted reproduction technique in couples with normal/subnormal sperm. IMSI appears to be the technique of choice when sperm belongs to subgroup I, ICSI when it belongs to subgroup II and IVF when sperm is of subgroup III quality. If IUI is indicated, it is preferable to perform this technique mainly on subgroup III sperm. The initial reference thresholds we established have now become "decision thresholds" in our laboratory to avoid unnecessary IMSI for these couples. Today, we routinely use PIT in our ART laboratory for the following indications: isolated teratozoospermia, long periods of unexplained infertility, patients with failed IUI, two or more IVF attempts at other IVF Centers, fragmented embryos or lack of blastulation. The earlier PIT is performed, the more beneficial it is for patients. Ideally, PIT could be routinely performed as soon as the sperm is normal/subnormal or in cases of isolated teratozoospermia, before starting any first-line treatment. We believe that this strategy, which is routinely used in our laboratory, could avoid unnecessary treatment and shorten the time to delivery. These preliminary results need to be confirmed later by a prospective study (article in progress).

A total of 248 infertile couples from the IVF Center at Nîmes University Hospital, France, were finally included in the present study from February 2017 to October 2020. All patients eligible for inclusion in the study had received a letter of information and non-opposition to the use of their health data. The procedure was approved by the local institutional review board at Nîmes University Hospital (IRB n° 21.12.01). Only two authors (first and last) had access to information that could identify individual participants during and after data collection. These authors were also the embryologists of the patients included in this study. The data consulted for research purposes in this study was done throughout the study period, from the start-up of the database (November 2020) to the production of the final statistics (October the 26th 2023). We opted for an evolving methodology. An adapted version of the MSOME analysis called the pre-IMSI test (PIT), first based on well-known publications [ 5 , 39 ], was systematically proposed to all eligible patients (n = 402) during the study period when sperm was normal or subnormal according to the WHO laboratory manual (2010) [ 41 ]. Subnormal sperm may present one or more minor abnormalities in the spermogram-spermocytogram, but at least 1 million progressive spermatozoa were available after selection. All sperms were of sufficient quality to technically perform IUI, IVF, ICSI or IMSI, and all couples were eligible for each of these techniques. Indications for PIT were: having normal/subnormal sperm associated with unexplained infertility, failure of two or three IUI cycles, repeatedly fragmented embryos, embryonic development culture failures at blastocyst-stage, repeated miscarriages, a long period of infertility and two or more IVF attempts without pregnancy. We first validated the methodology to develop a simplified, efficient, reproducible technique based on 402 eligible patients. We then excluded 128 patients who had not yet done ART attempts, 26 women aged over 39 years old or patients with cryopreserved semen. Women aged 39 or over were excluded to avoid an oocyte quality bias and because, in our experience, IMSI does not offer better pregnancy outcomes (unpublished data). Only 248 couples were finally included in our study to compare PIT results and ART outcomes. We were able to individualize three PIT score patient subgroups (patients with ≤8% normal spermatozoa (score I), those with a 9 to 15% normal spermatozoa (score II) and those with a ≥16% normal spermatozoa (score III)). The choice of one or another of these ART techniques had only been made according to the usual clinico-biological criteria. Sperm samples were collected by masturbation after 2 to 7 days of sexual abstinence. Samples were analyzed by microscopy at x40 magnification after half to one hour of liquefaction. Sperm concentration (c-chip disposable hematocymeter [Malassez]), motility and morphology was assessed according to the WHO 2010 criteria [ 41 ]. Semen characteristics are summarized in Table 1 . Since 2019, all results of our ART laboratory are delivered under accreditation according to the ISO 15189 international standard. All sperms were prepared by the discontinuous density gradient technique. After liquefaction at room temperature, the semen was overlaid on a 40% and 80% SupraSperm ® (CooperSurgical, Denmark) gradient column and centrifuged at 350g for 20 minutes. The sperm pellet was washed at 600g in universal IVF medium (CooperSurgical, Denmark). The seminal fluid, the 40% phase, and the upper half of the 80% phase were gently aspirated. Then, 3 mL of universal IVF medium (CooperSurgical, Denmark) was gently layered over the lower half of the 80% phase and the sample was centrifuged 10 min at 600g. The supernatant was removed to obtain 0.2 to 0.4 ml of the preparation. Selected sperm preparation was store at 37°C. Values are expressed as median and quartiles [q1;q3]. p<0 . 05 is considered significant. AMH = anti-Müllerian hormone; BMI = body mass index; PIT = pre-IMSI test; WHO = World Health Organization For the PIT, a SpermSlow™ drop (CooperSurgical, Denmark) was placed in a glass-bottomed Fluorodish (WPI, USA). The drop was covered with sterile mineral oil (CooperSurgical, Denmark). The selected sperm preparation was then placed in the SpermSlow™ drop. Sperm cells were observed under an inverted microscope (Leica DMI8) equipped with Nomarski differential interference contrast optics. The images were captured by a TK-C1481BEG color video camera (JVC professional). Magnification ranged from x6600 to x12000. The sperm cell grading system used was adapted from the Vanderzwalmen classification [ 5 ]. It was based on the number and size of vacuoles ( Fig 1 ), without considering sperm morphology. Like Vanderzwalmen et al ., we first individualized four sperm grades and then, due to unreliable inter-operator CVs, we split them into two (grade A and grade B). Grade A sperm cells either presented no vacuole, a maximum of two small vacuoles or an intermediate vacuole. Grade A were considered as normal spermatozoa. Grade B sperm cells had either more than two small vacuoles or at least one large vacuole. Grade B were considered as abnormal spermatozoa. Evaluation of vacuole size (small, intermediate, large) was semi-quantitative, as described in Fig 1 . Grade A sperm cells either presented a maximum of two small vacuoles (a) or an intermediate vacuole (b). Grade B sperm cells had either more than two small vacuoles or at least one large vacuole (c). The PIT was done by two operators who each analyzed a total of two counts of one hundred motile sperm at high magnification. The PIT score corresponded to the average grade A sperm count of the two operators. In the event of unreliable inter-operator CVs, a third operator also performed two counts of one hundred motile spermatozoa. This technique was validated in 2019 by the French accreditation committee (COFRAC) according to ISO 15189 international standards. Finally, our PIT score interpretation was adapted from the Wittemer system [ 42 ], according to the percentage of grade A sperm cells. We first individualized four PIT scores; when we compared our PIT results and each ART technique outcome, there was no difference between the 2 intermediate groups. As a consequence, we decided to individualize three PIT scores: PIT score I corresponding to ≤8% of grade A sperm cells, PIT score II corresponding to 9 to 15% of grade A sperm cells and PIT score III corresponding to ≥16% of grade A sperm cells. The couples included were then separated into three subgroups according to their PIT score. At this stage, PIT score I was considered as the worst quality, PIT score II as of intermediate quality and PIT score III as the best. One to 20 millions prepared progressive spermatozoa were inseminated 36 hours after hCG administration. In IVF attempts, oocytes were inseminated approximately 2–3 hours after oocyte retrieval. During ICSI/IMSI attempts, selected spermatozoa were injected under x400 magnification. For ICSI, motile and morphologically normal spermatozoa were injected. For IMSI, spermatozoa were chosen according to sperm selection method of Berkovitz et al . [ 27 ]. Briefly, normal morphological spermatozoa with no vacuoles or a maximum of two small vacuoles were used. Spermatozoa that presented deep vacuoles and/or vacuoles either at the nucleus or equatorial segment were systematically discarded. Injected oocytes and embryos were cultured either in Global® medium (JCD Laboratoires, France) or Cleav™/Blast™ medium (CooperSurgical, Denmark) in tri-gas incubators. Assessment of normal fertilization was estimated 17±1 hours post-insemination (IVF) or post-injection (ICSI/IMSI) [ 43 ]. Early cleavage rate was checked 25 hours post-insemination or post-injection. Cleaved embryo quality was estimated 44 to 46 hours post-fertilization. The embryo score (Blefco classification [ 44 ]) was established on Day 2 (44 ± 1 hours post- fertilization) or at the 8-cell stage on Day 3 (68 ± 1 hours post- fertilization). Blastocysts were evaluated on Days 5 and 6 according to the Gardner classification [ 45 ]. Embryo transfers were also assisted by ultrasound control with Elios (Ellios Cat Tek, France), Set TDT® (CCD laboratory, France) or Echogyn Embryoview M-18cm (CCD laboratory, France) catheters from Day 2 to Day 5 post-oocyte retrieval. The number of embryos transferred was decided according to the consensus of the American Society for Reproductive Medicine [ 46 ]. Clinical pregnancy was defined as an ongoing pregnancy confirmed by the number of gestational sacs with embryo and heart activity using ultrasound. Early miscarriages or pregnancy arrest were defined as the termination of a pregnancy during the first trimester. Statistical analysis was performed at Nîmes University Hospital using Statistical Analysis System software (SAS Institute, Cary, NC, USA) version 9.4 without provision for replacing missing data. A 5% alpha level of significance was used for all tests. The initial data analysis was a description of the total population and per group. Statistical results were presented as means and standard deviations for quantitative variables with a Gaussian distribution, and medians and quartiles [Q1; Q3] for the other variables. For qualitative variables, the numbers and associated percentages were presented. The subgroup categories were compared according to PIT results. When the variables were quantitative, an Anova test was applied if the application conditions permitted it. A Kruskal-Wallis test was used where appropriate. For qualitative variables, a Chi-squared test was used when the application conditions permitted, otherwise a generalized Fisher’s exact test (Fisher-Freeman-Halton exact) was used. The clinical outcomes (ongoing pregnancy and live birth) of three PIT score subgroups for each of four different ART techniques were compared using mixed logistics models with the couple as a random effect. Miscarriages were expressed per ongoing pregnancy and compared using the Chi-squared test or Fisher’s exact test as appropriate. Another mixed logistics model with the couple as a random effect including PIT score subgroups and ART techniques was used to compare rates of ongoing pregnancies between ART techniques in PIT score subgroups. In addition, results of interaction between PIT score subgroups and ART techniques were used to estimate the odds ratio (OR) of two-by-two ART technique comparisons in each PIT score subgroup. These ORs were then compared between PIT score subgroups.

(XLSX)