IVF News

Microdissection testicular sperm extraction (micro-TESE) guided by an artificial intelligence (AI) microscope system was linked to higher sperm retrieval than conventional micro-TESE. The retrospective study included 1,286 men with non-obstructive azoospermia (NOA) in Chengdu, China (Asian Journal of Andrology).

It also reports shorter operations with fewer complications. After intracytoplasmic sperm injection (ICSI), fertilisation, blastocyst and live birth rates were higher (Asian Journal of Andrology).

The authors conclude that AI-assisted micro-TESE "is associated with significantly higher sperm retrieval rates, reduced surgical trauma and postoperative complication risks". They also report "improved subsequent ICSI laboratory and clinical pregnancy outcomes" (Asian Journal of Andrology).

In their view, AI assistance was "an independent predictive factor for both successful sperm retrieval and live birth". They add that prospective, multicentre randomised trials are "required to confirm broad clinical applicability and causal effects" (Asian Journal of Andrology).

The study covered January 2020 to December 2024 at Sichuan Provincial People's Hospital and Sichuan Jinxin Xinan Women's and Children's Hospital. In all, 542 men had AI-assisted micro-TESE and 744 had conventional micro-TESE. The same three urologists operated in both groups (Asian Journal of Andrology).

The AI system (AI-SpermScope, Wuhan Mutual Innovation Technology) projects overlays onto the microscope eyepiece, marking regions with a high probability of containing sperm. Which men had it depended "exclusively" on whether the AI microscope was available at the time, the authors write (Asian Journal of Andrology).

In the AI group, each biopsy specimen was also processed with the system's sperm detection tool. In the conventional group, an experienced clinical embryologist examined each specimen. Citing earlier work, the paper says the tool finds motile sperm in 2 to 3 minutes, against 10 to 15 minutes manually (Asian Journal of Andrology).

Sperm were retrieved in 68.1% of men (369 of 542) with AI assistance and 47.6% (354 of 744) without it (P < 0.001). In Sertoli cell-only syndrome, the rates were 47.2% and 22.7%. In maturation arrest they were 72.8% and 51.0%, and in hypospermatogenesis 94.2% and 79.3% (Asian Journal of Andrology).

In a propensity score-matched sensitivity analysis of 486 pairs, the rates were 67.9% and 48.1%. After multivariable adjustment, AI assistance was linked to higher odds of sperm retrieval (odds ratio 3.182, 95% confidence interval [CI] 2.374 to 4.265) (Asian Journal of Andrology).

Operations were shorter with AI assistance (mean 58.32 vs 87.65 minutes), with fewer biopsy sites (3.12 vs 6.87). Complications within three months were less frequent (3.3% vs 7.8%; P < 0.001). Testicular atrophy was recorded in no men in the AI group and four in the conventional group (Asian Journal of Andrology).

The 723 men with retrieved sperm had ICSI at the two centres, with the same embryology team throughout. Female partner age and numbers of oocytes and mature (MII) oocytes did not differ between groups (Asian Journal of Andrology).

With AI-assisted retrieval, normal fertilisation (two pronuclei, 2PN) per MII oocyte was 72.5% against 64.8%. Blastocyst formation per cultured day 3 embryo was 55.3% against 44.6%. Live birth per embryo transfer cycle was 38.5% against 26.8% (P < 0.001) (Asian Journal of Andrology).

After adjustment, AI assistance was linked to higher odds of live birth (odds ratio 1.876, 95% CI 1.325 to 2.656). A post hoc intention-to-treat analysis counted men treated elsewhere as having no live birth. It gave an odds ratio of 1.72 (95% CI 1.23 to 2.40) (Asian Journal of Andrology).

The authors offer possible explanations but say their observational study "cannot establish the causal mechanisms". Their listed limitations include the retrospective design, possible residual confounding and a narrow cohort (Asian Journal of Andrology).

Men with Klinefelter syndrome or other sex chromosome aneuploidies, or complete AZFa or AZFb microdeletions, were excluded. Sperm DNA fragmentation was not measured. Both authors declare no competing interests (Asian Journal of Andrology).

IVF.net notes: the methods describe two centres, while the limitations call it "a single-center retrospective study". The inclusion criteria require ICSI at the two centres, yet men without retrieved sperm are counted in the retrieval rate (Asian Journal of Andrology).

The paper says all 723 men with sperm had ICSI there, but also that 36 had ICSI elsewhere. Atrophy was confirmed by ultrasound at three months, yet the limitations say this was not systematic. The text does not give the number of transfer cycles (Asian Journal of Andrology).

Without sperm, conventional surgery ended after the whole exposed tissue had been examined, and AI surgery after all AI-flagged regions. The title and abstract say AI assistance "improves" outcomes, while the conclusion says "is associated with" (Asian Journal of Andrology).

For validation of the system, the authors cite a 2026 laboratory image study of an AI augmented reality microscope for micro-TESE. It reported operator sperm detection of 66.9% with the overlay and 30.8% without, on images rather than patients (Bioengineering).

For andrology and embryology labs, the AI group differed in the laboratory search for sperm as well as in the surgery. Labs asked about AI-guided retrieval may wish to keep their own retrieval rates by histological subtype, with time to sperm detection, as a baseline. These results come from two hospitals in one city, using one system, and await prospective trials (Asian Journal of Andrology).

Related reading: IVF.net's report Testicular sperm in cryptozoospermia linked to more blastocysts but no pregnancy or live birth advantage over ejaculated sperm. See also the education page Semen analysis to the WHO 6th edition.

Sources

6 November 2026. Asian Journal of Andrology

• Artificial intelligence-assisted microdissection testicular sperm extraction improves sperm retrieval efficiency and clinical outcomes in men with non-obstructive azoospermia: a retrospective cohort study

15 January 2026. Bioengineering

• AI-Based Augmented Reality Microscope for Real-Time Sperm Detection and Tracking in Micro-TESE

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

Embryos from follicles under 8 mm showed "an approximately 8-hour delay in blastulation" compared with embryos from follicles of 11 mm or more. Day 6 blastocyst rates and euploidy rates did not differ significantly, a small retrospective study from Istanbul reports (Frontiers in Endocrinology).

The authors conclude that their "exploratory findings suggest" delayed blastulation alone "should not be taken as evidence of impaired developmental potential" in these embryos. For labs, that bears on whether oocytes from small follicles are kept and how their embryos are ranked (PMC).

The study, at Acıbadem Maslak Hospital, analysed 62 blastocysts from 18 women having intracytoplasmic sperm injection (ICSI). Twenty came from follicles under 8 mm and 42 from follicles of 11 mm or more. Follicles of 8 to 11 mm were excluded (PMC).

Small follicles were aspirated into separate tubes after the larger ones. Seven women contributed embryos to both groups. Timings were analysed with mixed models and rates with patient-clustered models to allow for this (PMC).

Fewer oocytes from small follicles were mature (51.4% vs 74.3% metaphase II; p < 0.001). Fertilisation of mature oocytes after ICSI was higher in the small-follicle group (92.9% vs 75.3%; p = 0.005) (PMC).

Pronuclear timings and cleavage timings from two to eight cells (t2 to t8) did not differ significantly between groups. Start of blastulation came later in the small-follicle group (median 101.2 vs 95.3 hours; p = 0.001) (PMC).

Full blastocyst formation was also later (108.1 vs 104.3 hours; p = 0.004). In mixed models, the delays were 7.91 hours (95% CI 3.84 to 12.00) and 8.29 hours (3.80 to 12.80), both p < 0.001 (PMC).

Day 5 blastocyst formation per fertilised oocyte was lower in the small-follicle group (28.8% vs 49.3%; p = 0.040). Overall blastocyst formation (38.5% vs 56.5%; p = 0.105) and day 6 rates (9.6% vs 7.2%; p = 0.683) did not differ significantly (PMC).

All embryos from small follicles were cultured in the time-lapse incubator, so the culture system was partly linked to group. Restricted to time-lapse embryos, blastocyst rates no longer differed (35.7% vs 32.6%; p = 0.657). The timing comparisons all come from time-lapse culture (PMC).

Euploidy was 5 of 10 biopsied blastocysts (50.0%) from small follicles and 7 of 23 (30.4%) from larger follicles (p = 0.433). PGT-A was done only in seven patients with a clinical indication, and mosaic results were grouped with aneuploid ones (PMC).

Of the 23 biopsied blastocysts from larger follicles, 14 came from one patient. Excluding her gave 5 of 9 (small follicles) against 4 of 9 (larger follicles) (p = 0.637). The authors say the euploidy comparison "should be regarded as descriptive" (PMC).

Embryos from larger follicles were preferred for transfer. Two euploid blastocysts from follicles under 8 mm were transferred singly after freezing, in patients with no alternative. Both led to a live birth, but the authors say these transfers are too few to judge pregnancy or live birth potential (PMC).

Among the limitations the authors list, the study had no power calculation and was not powered to show equivalence. Follicle sizes were recorded in categories, without matching each oocyte to its follicle. One embryologist annotated all timings without blinding to follicle group (PMC).

The authors also see a risk for selection algorithms calibrated largely on embryos from dominant follicles. If the delay reflects follicle of origin rather than impaired potential, such algorithms "could systematically deprioritize" embryos from small follicles (PMC).

They call for larger prospective series before the findings inform practice (PMC).

The study is small, but the authors raise the possibility that later blastulation in these embryos reflects follicle of origin rather than poor potential. Oocytes from small follicles are often discarded, although they may help patients with few oocytes. Recording the follicle size category with each oocyte, as this unit did, would let labs audit timings, scores and outcomes for these embryos.

Related reading: IVF.net's report Morphokinetic scoring may penalise euploid embryos from partially compacted morulae. See also the education page Blastocyst grading explained.

Sources

20 October 2026. Frontiers in Endocrinology

• Embryos derived from small follicles (<8 mm) exhibit delayed blastulation: a retrospective time-lapse analysis

20 October 2026. PMC

• Embryos derived from small follicles (<8 mm) exhibit delayed blastulation (full text)

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

Embryos from capacitation in vitro maturation (CAPA-IVM) showed a "distinct, non-uniform morphokinetic profile", a retrospective time-lapse study from Vietnam reports. The comparison was with embryos from stimulated intracytoplasmic sperm injection (ICSI) cycles (Journal of Assisted Reproduction and Genetics).

The authors say this highlights "the need for maturation pathway-specific interpretation of time-lapse data". For laboratories that offer in vitro maturation (IVM), it bears on which timing benchmarks apply to IVM embryos (PubMed).

CAPA-IVM is a biphasic IVM method that includes a pre-maturation (capacitation) step. The researchers are at My Duc Hospital in Ho Chi Minh City. They compared CAPA-IVM embryos with embryos from controlled ovarian stimulation and ICSI (COS-ICSI) cultured over the same period (PubMed).

Women were under 40, with no previous fertilisation failure, gamete donation, planned preimplantation genetic testing or significant male factor infertility. Cycles were matched 1:1 on a propensity score built from female age and the number of analysable embryos (PubMed).

In all, 232 couples contributed 876 embryos: 421 from CAPA-IVM and 455 from COS-ICSI (PubMed).

In crude analyses, CAPA-IVM embryos showed later pronuclear appearance (+0.71 h; p < 0.001) and earlier pronuclear fading (−0.92 h; p < 0.001). They also reached the three-cell stage (t3, −1.42 h; p < 0.001) and five-cell stage (t5, −1.40 h; p = 0.017) earlier (PubMed).

Compaction (+2.00 h; p < 0.01) and blastulation (+1.67 h; p < 0.01) came later with CAPA-IVM. The authors report that the pronuclear, t3 and t5 differences "remained statistically significant after adjustment". The abstract does not say whether the later compaction and blastulation did (PubMed).

A lower proportion of CAPA-IVM embryos were of good quality (G1 or G2): 29% against 39% (p < 0.01). Among good-quality embryos only, CAPA-IVM again showed later pronuclear appearance, with earlier pronuclear fading, t3 and four-cell stage (t4). The abstract gives no p values for this subgroup (PubMed).

IVF.net notes: only the abstract was available, and the study is retrospective and from a single centre. The outcomes are embryo timings and grades, not implantation or live birth (PubMed).

The abstract does not give the women's diagnoses, how the timings were annotated or which factors the adjusted models included. Nor does it say how time zero was set, or which grading system G1/G2 refers to. Matching used only female age and embryo number, and the good-quality subgroup analysis was crude (PubMed).

An earlier randomised trial in Vietnam, by some of the same authors, enrolled 546 women with a high antral follicle count. It compared one cycle of IVM with a pre-maturation step against conventional IVF. Its authors reported: "We could not demonstrate non-inferiority of IVM compared with IVF" (Human Reproduction).

In that trial, live birth after the first transfer occurred in 35.2% of women with IVM and 43.2% with IVF (Human Reproduction).

The abstract reports no test of a selection model and no link between timings and outcomes. Labs offering IVM may wish to review their IVM embryo timings separately before applying cut-offs set on stimulated cycles. Keeping the maturation pathway in the time-lapse record would make that audit easier.

Related reading: IVF.net's report Morphokinetic scoring may penalise euploid embryos from partially compacted morulae. See also the education page Oocyte maturity and morphology at denudation and ICSI.

Sources

5 November 2026. Journal of Assisted Reproduction and Genetics

• Morphokinetic characterization of human embryos derived from capacitation in vitro maturation versus controlled ovarian stimulation with intracytoplasmic sperm injection: a retrospective cohort study using time-lapse imaging

5 November 2026. PubMed

• Morphokinetic characterization of human embryos derived from capacitation in vitro maturation versus COS-ICSI (abstract)

1 November 2020. Human Reproduction

• In-vitro maturation of oocytes versus conventional IVF in women with infertility and a high antral follicle count: a randomized non-inferiority controlled trial

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

Patients whose anti-Müllerian hormone (AMH) and antral follicle count (AFC) were discordant had fewer oocytes and embryos than those with normal values on both. Pregnancy outcomes after a first transfer did not differ significantly between these two groups, a retrospective study from Ningbo, China, reports (Journal of Assisted Reproduction and Genetics).

For embryology laboratories, the results bear on expected first-cycle oocyte and embryo numbers, and on how yield audits group patients.

The study included 6,456 patients in their first IVF or intracytoplasmic sperm injection (ICSI) cycle between January 2018 and January 2024. Concordant normal reserve meant AMH of 1.2 ng/mL or more and an AFC of 5 or more. Concordant reduced reserve meant both below these values. Discordance meant one marker normal and the other reduced (Journal of Assisted Reproduction and Genetics).

In all, 816 patients (12.64%) had discordant markers. Against the concordant normal group, they had fewer oocytes after adjustment (adjusted beta coefficient, β, −4.04; 95% confidence interval [CI] −4.44 to −3.65). Each β is an adjusted difference in count, not a percentage (Journal of Assisted Reproduction and Genetics).

They also had fewer metaphase II oocytes (β −3.46) and two-pronuclear (2PN) zygotes (β −2.57). Transferable embryos (β −2.23) and high-quality embryos (β −1.73) were also fewer, all P < 0.001 (Journal of Assisted Reproduction and Genetics).

After adjustment, these outcomes did not differ significantly between the discordant and concordant reduced groups. Within the discordant group, normal AMH with reduced AFC was linked to higher oocyte and embryo yields than the reverse pattern. Pregnancy outcomes were comparable between the two patterns (Journal of Assisted Reproduction and Genetics).

Of the cohort, 6,188 patients had a first embryo transfer. Pregnancy outcomes did not differ significantly between the discordant and concordant normal groups (Journal of Assisted Reproduction and Genetics).

Against the concordant reduced group, discordant patients' adjusted odds of biochemical pregnancy were higher (odds ratio 1.38, 95% CI 1.03 to 1.86, P = 0.031). Their adjusted odds of live birth were also higher (odds ratio 1.40, 95% CI 1.01 to 1.95, P = 0.042). The authors call these odds "modestly higher" (Journal of Assisted Reproduction and Genetics).

IVF.net notes: IVF.net has seen the abstract only. It does not give the size of each concordant group, the factors adjusted for, or whether outcomes from later transfers were counted. Nor does it say whether transfers were fresh or frozen, or how transferable and high-quality embryos were defined. As a retrospective study, it shows associations rather than causes.

The finding within the discordant group matches a 2021 retrospective study of 19,239 first fresh cycles. After propensity score matching, normal AMH with low AFC was linked to significantly more oocytes and available embryos than low AMH with normal AFC. Reproductive outcomes did not differ significantly between these groups (Journal of Ovarian Research).

For laboratories, discordant markers may help explain a lower oocyte and embryo count in a first cycle. Labs that benchmark yield against ovarian reserve could consider recording both markers, and which one is reduced. That would keep discordant patients out of the normal reserve band in yield audits. The study's cut-offs may not match local thresholds. The findings may also help counselling teams set expectations.

Related reading: IVF.net's report Oocyte numbers needed for three euploid blastocysts rise steeply with age and the education page Oocyte maturity and morphology at denudation and ICSI.

Sources

4 November 2026. Journal of Assisted Reproduction and Genetics

• Association of anti-Müllerian hormone and antral follicle count discordance with embryological and pregnancy outcomes in the first IVF/ICSI cycle

28 August 2021. Journal of Ovarian Research

• Efficacy of three COS protocols and predictability of AMH and AFC in women with discordant ovarian reserve markers: a retrospective study on 19,239 patients

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

Older paternal age was linked to lower cumulative live birth after IVF or ICSI mainly when the female partner was 40 or over. That is the finding of a retrospective cohort study from Shanghai. Among women under 40, the authors found no significant association (Reproductive BioMedicine Online).

In women aged 40 or over, fertilisation and high-quality embryo rates did not decline with paternal age (PubMed).

The study included 22,013 first oocyte retrieval cycles at a tertiary reproductive centre between 2016 and 2024. All used autologous oocytes and non-donor sperm, with IVF or intracytoplasmic sperm injection (ICSI). The first and last authors list Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine (PubMed).

The primary outcome was cumulative live birth within the first three embryo transfer attempts from one retrieval. Paternal age was analysed per 5-year increase, and the team formally tested whether maternal age changed its association. Models adjusted for female body mass index, duration and type of infertility, fertilisation method and major causes of infertility (PubMed).

Among women under 40, paternal age was not significantly associated with cumulative live birth. The adjusted odds ratio per 5-year increase was 0.98 (95% CI 0.94 to 1.02). Among women aged 40 or over, increasing paternal age was associated with lower cumulative live birth. The adjusted odds ratio per 5-year increase was 0.78 (95% CI 0.75 to 0.81). The interaction between paternal and maternal age was significant (P for interaction < 0.001) (PubMed).

In women aged 40 or over, older paternal age was also linked to a lower cumulative clinical pregnancy rate and higher miscarriage risk. Results were consistent when cumulative live birth was counted over the first five transfer attempts (PubMed).

The authors conclude that paternal age "was not uniformly associated" with cumulative live birth. Its adverse association "became clinically evident mainly" when the female partner was 40 or over. They say the findings support "joint consideration of paternal and maternal age when counselling couples" (PubMed).

IVF.net notes: IVF.net has seen the abstract only. It does not give the number of cycles in each maternal age group or the range of paternal ages. Nor does it say whether semen parameters were adjusted for. As a retrospective study from one centre, it shows associations rather than causes.

For laboratories, because fertilisation and high-quality embryo rates did not decline, fertilisation and embryo grading KPIs may not show this pattern. Labs that audit cumulative live birth by maternal age could consider adding paternal age to the analysis. Counselling teams may find the joint age framing useful.

Related reading: Sperm DNA fragmentation above 20% linked to poorer embryo development and the education page Semen analysis to the WHO 6th edition.

Sources

7 October 2026. Reproductive BioMedicine Online

• Couple age interaction and cumulative live birth rate after IVF/ICSI: a large cohort study

3 November 2026. PubMed

• Couple age interaction and cumulative live birth rate after IVF/ICSI: a large cohort study (abstract)

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

Preimplantation genetic testing for aneuploidy (PGT-A) may not improve live birth or pregnancy rates compared with IVF alone. It may make little to no difference. That is the conclusion of a health technology assessment (HTA) by Ontario Health. It adds that "the evidence is highly uncertain" (Ontario Health Technology Assessment Series).

Ontario Health is a government agency in the Canadian province of Ontario. In June it posted a draft recommendation, based on guidance from the Ontario Health Technology Advisory Committee. It "recommends against publicly funding PGT-A for people undergoing in vitro fertilization". The recommendation remains a draft: public feedback has closed and a final recommendation is still to be posted. The draft concerns public funding in Ontario, not whether clinics may offer PGT-A. The committee also weighed reviews of harms, patient preferences and ethics (Ontario Health).

The assessment included 6 randomised controlled trials of PGT-A with blastocyst biopsy and comprehensive chromosomal screening. Certainty was rated with GRADE, a standard grading system. The evidence is very uncertain about the effect of PGT-A on the cumulative live birth rate (GRADE: very low). PGT-A may result in little to no difference in live birth after the first planned embryo transfer, per person undergoing PGT-A (GRADE: low) (Ontario Health Technology Assessment Series).

PGT-A may not improve, and may decrease, cumulative pregnancy rates (GRADE: low). PGT-A may reduce the cumulative frequency of clinical pregnancy loss among people who became pregnant (GRADE: low). The evidence is very uncertain about the effect after the first planned embryo transfer, per person (GRADE: very low). "No benefit in time to live birth was observed", the authors write, though they grade that evidence as very uncertain (GRADE: very low) (Ontario Health Technology Assessment Series).

The team did not carry out a cost-utility analysis "because of the uncertainty in the benefit of PGT-A". It estimated that public funding would add $22.2 million in year 1, rising to $33.5 million in year 5. The 5-year total is $139 million, in 2025 Canadian dollars, an estimate the authors describe as uncertain. PGT-A is not publicly funded in any Canadian province and is an add-on to IVF, the HTA notes (Ontario Health Technology Assessment Series).

The draft makes no recommendation on PGT for monogenic disorders (PGT-M) or structural rearrangements (PGT-SR). It finds "currently insufficient evidence to make a recommendation for or against publicly funding" them (Ontario Health).

The HTA says its findings align with a Cochrane review. They also align with a Swedish HTA published after its own review was completed (Ontario Health Technology Assessment Series). In September, IVF.net reported a Human Reproduction Update systematic review that found no evidence of PGT-A raising the live birth rate per woman (IVF.net).

The HTA notes that 2 studies did not test every embryo. That may not match Ontario practice, where PGT-A is most often offered for all, or most, good-quality embryos. Labs that biopsy embryos may wish to check that patient information gives per-person outcomes and states how uncertain the evidence is. The graded outcomes are a useful reference for that wording (Ontario Health Technology Assessment Series).

Related reading: IVF.net's report PGT-A shows no evidence of raising live birth rate per woman, review finds and the International IVF Initiative session PGT: 360 Degrees.

Sources

6 October 2026. Ontario Health Technology Assessment Series

• Preimplantation Genetic Testing for Aneuploidy: A Health Technology Assessment

18 June 2026. Ontario Health

• Preimplantation Genetic Testing for Aneuploidy, Monogenic Disorders and Structural Rearrangements (draft recommendation)

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

The International IVF Initiative's Session 179, Modern Thinking, takes place online on Tuesday 10 November 2026. It starts at 3pm EST / 8pm GMT / 9pm CET (Zoom registration).

Dr Isabella Spyropoulou and Wayne Bennett moderate three talks (International IVF Initiative):

• Dr Angie Beltsos: The Leaking Clinic. Sponsored by Cima Growth Solutions.
• Helen Hunter: Keep Calm and Culture On. Sponsored by FertilAI.
• Dr Vittorio Palmieri: Thinking Outside the Box: Rethinking Embryo and Gamete Transport.

A Q&A follows the talks.

The third talk may interest labs that move embryos or gametes between sites, or ship them to and from other clinics. Embryologists may wish to bring questions on transport practice to the Q&A.

Related reading: our earlier session article The Sum of Its Parts webinar, 3 November.

Register for the webinar on Zoom.

Sources

3 November 2026. International IVF Initiative

• Session 179: Modern Thinking (registration)
• International IVF Initiative upcoming webinars

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

In men with cryptozoospermia, intracytoplasmic sperm injection (ICSI) with testicular sperm was linked to more blastocysts than ICSI with ejaculated sperm. It was not linked to higher pregnancy or live birth rates, a retrospective study from Istanbul reports. With ejaculated sperm available, the authors conclude, routine surgical retrieval "does not appear to provide a clinical advantage" (Turkish Journal of Obstetrics and Gynecology).

The study included 127 ICSI cycles at Acıbadem Maslak Hospital's assisted reproduction centre between 2017 and 2026. Of these, 58 used ejaculated sperm and 69 testicular sperm: 51 by microdissection testicular sperm extraction (mTESE), 18 by testicular sperm aspiration (TESA). The primary outcome was pregnancy per embryo transfer (PubMed).

Baseline characteristics were similar, with two exceptions. The testicular sperm group had more previous IVF attempts (p = 0.032). The ejaculated sperm group more often had a history of miscarriage (15.5% vs 4.5%, p = 0.037) (PubMed).

With testicular sperm there were more day 5 blastocysts (median 5 vs 3, p = 0.004). The blastocyst formation rate was also higher (57.2% vs 34.3%, p < 0.001). However, there were more grade 3 blastocysts (p < 0.001) (PubMed).

Per embryo transfer, pregnancy rates were higher with ejaculated sperm than with testicular sperm (80.0% vs 54.5%, p = 0.006). So were clinical pregnancy rates (78.0% vs 47.3%, p = 0.001). Live birth rates did not differ significantly (48.0% vs 45.5%, p = 0.794) (PubMed).

The authors adjusted for female age, previous IVF attempts, embryo stage at transfer and embryo quality. Sperm source was then not independently associated with pregnancy (adjusted odds ratio 1.94, 95% confidence interval 0.71 to 5.26; p = 0.195). Blastocyst-stage transfer and grade 1 embryo quality were the independent predictors (PubMed).

IVF.net notes: IVF.net has seen the abstract only, and the study is small and retrospective. The abstract does not say how sperm source was chosen, how many transfers each group had, or whether only first transfers were counted. Nor does it report cumulative live birth per cycle, or say whether testicular sperm was used fresh or frozen.

A 2018 meta-analysis of six studies, with 578 men and 761 ICSI cycles, reached a different conclusion. It favoured fresh testicular sperm for embryo quality, implantation and pregnancy (pregnancy risk ratio 1.74, 95% CI 1.20 to 2.52) (Scientific Reports).

For andrology and embryology labs, the study suggests more blastocysts with testicular sperm may not mean a better chance of pregnancy per transfer. The testicular sperm group also had more grade 3 blastocysts. Labs may wish to record sperm source with blastocyst grade and outcome, per transfer and per cycle, to inform counselling about surgical retrieval.

Related reading: IVF.net's education page Semen analysis to the WHO 6th edition.

Sources

30 October 2026. Turkish Journal of Obstetrics and Gynecology

• Comparison of embryological and clinical outcomes of ICSI cycles using ejaculated versus testicular sperm (TESA/mTESE) in men with cryptozoospermia: A retrospective cohort study

30 October 2026. PubMed

• Comparison of embryological and clinical outcomes of ICSI cycles using ejaculated versus testicular sperm in men with cryptozoospermia (abstract)

18 May 2018. Scientific Reports

• Testicular sperm is superior to ejaculated sperm for ICSI in cryptozoospermia: An update systematic review and meta-analysis

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

Embryos from morulae in which some cells were excluded from compaction developed more slowly and were mostly placed in the lowest KIDScore category. Yet their euploidy rates were similar to those of embryos from fully compacted morulae, a retrospective study from Brazil reports. The authors suggest that speed-based morphokinetic scoring "may introduce bias in embryo prioritization" (Reproductive BioMedicine Online).

The study included 1,711 blastocysts from 611 intracytoplasmic sperm injection (ICSI) cycles cultured in a time-lapse system. Embryos were classified by KIDScore, morula compaction pattern and the Blastocyst Fragmentation Indicator (BFI), and chromosomal status was assessed by next-generation sequencing. The authors are based in Fortaleza, Brazil (PubMed).

The comparison was between fully compacted morulae (FCM) and partially compacted morulae with cells excluded from compaction from the outset (Exc-PCM). Exc-PCM embryos had significantly slower development from the 5-cell to the blastocyst stage. Their standardised developmental interval was 0.780, against -0.216 for FCM (P < 0.001) (PubMed).

KIDScore C, the lowest morphokinetic category in the study, included 73.9% of Exc-PCM embryos and 25.5% of FCM embryos. Euploidy rates were 43.7% and 43.4% respectively (PubMed).

In multivariable models, blastocyst fragmentation had the strongest association with euploidy, linked to lower odds. Fragmentation of both the inner cell mass and trophectoderm gave an adjusted odds ratio of 0.04 against no fragmentation (P < 0.001). After adjustment, Exc-PCM embryos had higher odds of euploidy than FCM embryos (adjusted odds ratio 1.56; P = 0.008) (PubMed).

Among KIDScore C embryos, those rated "no fragmentation" on the BFI had euploidy rates of 50.0% to 66.7%. The authors call these "relatively high" (PubMed).

The authors conclude that "structural integrity, rather than developmental timing, may better reflect chromosomal competence". They add that integrating structural assessment may improve embryo prioritisation (PubMed).

IVF.net notes: IVF.net has seen the abstract only, and the study is retrospective with euploidy, not implantation or live birth, as its outcome. Euploidy is reported only for blastocysts, so embryos that arrested earlier are not represented. The abstract does not give group sizes, confidence intervals, the factors adjusted for, or whether the models allowed for several blastocysts per cycle.

In 2024, a team that included the same first author studied 835 blastocysts biopsied for preimplantation genetic testing for aneuploidy (PGT-A). In women under 38, aneuploidy in embryos with abnormal cleavage and partial compaction did not differ significantly from normally cleaved, fully compacted embryos. In women over 40 it was significantly higher. The authors proposed partial compaction might be "a potential correction mechanism" (Scientific Reports).

Some laboratories rank embryos by KIDScore or similar speed-based models. For them, the study suggests that a low score driven by slower development after partial compaction may not mean a lower chance of euploidy. Labs with time-lapse and PGT-A data may wish to compare how these embryos rank with their PGT-A results. Checking that fragmentation is recorded would also help before any change to selection rules.

Related reading: IVF.net's report Blastocyst contractions linked to poorer embryo quality and the education page Blastocyst grading explained. See also the International IVF Initiative session The future of embryology: technological advances in embryo selection.

Sources

8 October 2026. Reproductive BioMedicine Online

• Good embryos left behind: morphokinetic prioritization may penalize chromosomally competent embryos

Accessed 2 November 2026. PubMed

• Good embryos left behind: morphokinetic prioritization may penalize chromosomally competent embryos (abstract)

7 January 2024. Scientific Reports

• Association of early cleavage, morula compaction and blastocysts ploidy of IVF embryos cultured in a time-lapse system and biopsied for genetic test for aneuploidy

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.

When only day 6 blastocysts formed, transferring them first was not linked to higher cumulative live birth than transferring backup day 3 embryos first. The retrospective study is from Sun Yat-sen University in Guangzhou, China (Journal of Assisted Reproduction and Genetics).

In freeze-all cycles, the centre froze one or two good-quality embryos on day 3 as a backup. It cultured an equal or greater number of good-quality embryos to the blastocyst stage (PubMed).

The authors identified 275 patients whose cohorts gave only day 6 blastocysts, with no transferable day 5 blastocyst. They compared those whose backup day 3 embryos were transferred first with those whose top-graded day 6 blastocysts were transferred first (PubMed).

Cumulative live birth over 2 years from one oocyte retrieval was 50.6% with day 3 priority and 48.4% with day 6 priority. The difference was not significant (adjusted odds ratio 0.759, 95% CI 0.384 to 1.499; p = 0.427) (PubMed).

In the first frozen embryo transfer, top-graded day 6 blastocysts "did not yield a higher LBR" (live birth rate) than backup day 3 embryos. The abstract gives the rates as 36.3% vs 43.5% without labelling each group. The difference was not significant (adjusted odds ratio 0.601, 95% CI 0.299 to 1.208; p = 0.153) (PubMed).

Subgroup results showed no significant difference after single transfers (p = 0.907) or double transfers (p = 0.097). Clinical pregnancy, obstetric complications and neonatal outcomes in the first frozen transfer did not differ significantly (PubMed).

The authors conclude that in these cohorts blastocyst transfer "may not confer an advantage in achieving live birth sooner". Their comparison was with good-quality cleavage-stage embryos (PubMed).

IVF.net notes: IVF.net has seen the abstract only. It does not give the number of patients in each group or how the transfer order was chosen. Nor does it say what the odds ratios compare against or, for the first transfer and subgroups, which rate belongs to which group. The confidence intervals are wide, so a difference in either direction cannot be ruled out. As a retrospective study, it shows associations rather than causes.

For laboratories that extend culture, the study compared which embryo to transfer first, not whether to freeze a day 3 backup. It applies only to cohorts with no day 5 blastocyst. Labs that freeze a backup may wish to audit their own live birth rates in such cohorts before changing transfer order.

Related reading: IVF.net's report Day 7 blastocyst live birth linked to prior transfer history and the education page Day 2 and Day 3 embryo assessment.

Sources

29 October 2026. Journal of Assisted Reproduction and Genetics

• Does blastocyst transfer have a higher live birth rate than cleavage-stage embryo transfer in patients with delayed in vitro blastocyst development?

29 October 2026. PubMed

• Does blastocyst transfer have a higher live birth rate than cleavage-stage embryo transfer in patients with delayed in vitro blastocyst development? (abstract)

IVF.net Newsdesk articles are written from the original sources listed above. If you spot an error, please contact IVF.net and we will correct it.