Effect of laser-assisted hatching on ectopic pregnancy after frozen-thawed embryo transfer in patients with a history of ectopic pregnancy: a large cohort study

Abstract

Background

Assisted hatching (AH) influences embryo implantation, whereas ectopic pregnancy (EP) is embryo ectopic implantation and AH may influence EP. Patients with EP history are at higher risk of EP resulting from IVF/ICSI (IVF-EP), and this is associated with the number of previous EP. However, the effect of laser-AH on IVF-EP in patients with EP history has not been studied so far, so we investigated the effect and considered possible different effects in populations with different numbers of EP.

Methods

This retrospective cohort study included all patients with previous EP history receiving frozen-thawed embryo transfer cycles at our center between August 2019 and August 2023. A total of 5129 cycles from 2888 patients were grouped according to AH or not. Subgroup analyses were performed based on the number of previous EP (One-EP and > = 2 EP). Binary regression analysis together with generalized estimating equation (GEE) after inverse-probability-of-treatment weighting (IPTW) were performed in all cycles, One-EP cycles, and > = 2 EP cycles respectively to clarify the effect of laser-AH. Stratified regression analysis was also performed based on embryo stage.

Results

A total of 1022 cycles were included in the AH group and 4107 cycles were included in the Non-AH group. No statistically significant difference was observed in IVF-EP rate between the two groups (1.4% vs. 2.1%, P = 0.113). In subgroup analyses, AH group had a lower IVF-EP rate than Non-AH group in the One-EP subgroup (1.0% vs. 2.2%, P = 0.028); whereas there was no statistically significant difference in the > = 2 EP subgroup (2.4% vs. 1.9%, P = 0.612). In GEE after IPTW, AH did not influence the IVF-EP rate in the entire population (aOR: 0.888, 95% CI: 0.414–1.904, P = 0.760) and One-EP population (aOR: 0.418, 95% CI: 0.154–1.131, P = 0.086), while increased risk of IVF-EP was observed of AH in > = 2 EP population (aOR: 2.438, 95% CI: 1.509–3.939, P = 0.023). Similar regression results were obtained in cleavage-embryo cycles.

Conclusions

Laser-AH had no effect on IVF-EP after FET in patients with EP history. However, laser-AH was associated with an increased risk of IVF-EP in patients with recurrent EP, suggesting that laser-AH warrants caution in this specific population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13048-026-02015-y.

Background

Ectopic pregnancy (EP) is a serious clinical gynecological emergency. EP rupture can cause hypovolemic shock, which is the primary cause of early pregnancy mortality [1, 2]. Although embryos in in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) are directly transferred into the uterine cavity without passing through the fallopian tubes, theoretically preventing ectopic pregnancy, the incidence of EP in IVF/ICSI has been reported to be approximately 2–8%, which is four times higher than that in spontaneous conception [3, 4].

In recent years, advances in cryopreservation and embryo thawing techniques have greatly improved the outcomes of IVF/ICSI [5–8]. However, the in vitro incubation and freeze-thawing processes that cryopreserved embryos must undergo are believed to induce alterations in the glycoprotein matrix, leading to the sclerosis of the zona pellucida (ZP) [9]. The assisted hatching (AH), initially described by Cohen et al. in the 1980s involving artificially disrupting ZP through chemical, laser or mechanical manipulation to facilitate hatching and implantation [10], has been proposed as a complementary technique to improve pregnancy outcomes in frozen-thawed embryo transfer (FET) cycles. Theoretically, AH may accelerate embryo hatching and result in early contact between embryo and endometrium, thereby enhancing embryo developmental potential and optimizing the synchronization between the embryo and the endometrium [11, 12]. In addition, the artificial gap created by AH acts as a conduit for the exchange of metabolites, growth factors and information between the embryo and the endometrium, also increasing the implantation potential [13].

Given the possible influence of AH on implantation, EP, which is actually an ectopic implantation of the embryo and is thought to be correlated with embryo preimplantation wandering and implantation timing, may also be influenced by AH. To the best of our knowledge, only a limited number of studies have evaluated the association between AH and EP after IVF/ICSI (IVF-EP). In fresh embryo transfers, a retrospective study showed that AH increased the risk of IVF-EP [14], while another randomized controlled trial (RCT) in young patients (< 38 years old) found that AH was not associated with IVF-EP [15]. In contrast, in FET cycles, AH has been demonstrated not to influence IVF-EP, with the IVF-EP rates comparable between groups with and without AH (P > 0.05) [16–18]. This discordance, in addition to being due to differences between fresh and frozen transfer cycles themselves, may also be attributed to the fact that some certain risk factors which have been proved to be correlated with IVF-EP (such as salpingectomy surgery history [19]) were not adequately considered in the previous research. Besides, since IVF-EP was not the primary outcomes, together with the limited sample size, the conclusions in the previous studies appear conflicting and less convincing.

Patients with a previous history of EP were conceived to have an elevated risk of recurrent IVF-EP [20–22]. Therefore, narrowing down our study populations to such patients could help to clarify AH effect on IVF-EP (if AH does have some influence, then focusing on such populations might magnify its effect and make it more easily be detected). Besides, study in such a high-risk population is of more clinical significances by determining the protective or harmful effect of AH on IVF-EP. Furthermore, the possible benefits or risks observed in this population, in turn, might be helpful in exploring the underlying mechanisms why female with a history of EP are susceptible to IVF-EP. Of note, there are no previous studies on whether AH affects the risk of IVF-EP in people with a history of EP, leading to blinded clinical decision-making.

Therefore, in this study, we aimed to clarify the effect of AH on IVF-EP in patients with a history of EP in FET cycles to provide evidence for risk warning and strategy optimization in this population. We also performed further analyses in patients with different numbers of previous EP to explore whether AH has different effects on IVF-EP with the increase in previous EP times.

Materials and methods

Study design and participants

This was a retrospective cohort study conducted at the Department of Assisted Reproduction, Shanghai Ninth People’s Hospital, Shanghai JiaoTong University School of Medicine. We included patients who underwent FET cycles at our institution between August 2019 and August 2023. Only patients with a history of spontaneous EP were included, involving a total of 2888 patients and 5129 cycles. Incomplete documented cycles were excluded. IVF and ICSI cycles were both included in the study.

Data sources and grouping

The data was obtained from medical records, which have been described previously [23]. Cycles were grouped according to AH or not, with 1022 cycles undergoing AH (AH group) and 4107 cycles not undergoing AH (Non-AH group). AH or not was determined based on the preference of doctors in charge of the treatment.

Subgroups were divided according to the number of previous spontaneous EP, to determine whether different EP history had an impact on the results. Cycles with previous single EP were categorized into the One-EP subgroup, and cycles with previous recurrent EP were categorized into the > = 2 EP subgroup. 732 AH cycles and 2974 non-AH cycles were in the One-EP subgroup, and 290 AH cycles and 1133 non-AH cycles were in the > = 2 EP subgroup. The history of previous EP was obtained from patient reports and medical records.

The data processing and cohort divisions are shown in Supplementary Fig. 1.

This study was ethically reviewed by the Shanghai Ninth People’s Hospital institution on September 23, 2023 (SH9H-2023-TK109-18). Written informed consent was obtained from all participants for the permission to analyze their data when IVF/ICSI treatments started.

FET protocol

Insemination, embryo Culture, vitrification and warming

The retrieved oocytes were subjected to IVF or ICSI, according to routine laboratory procedures. The quality of embryos was observed on the third day after IVF or ICSI, and graded according to Cummins standards [24]. High quality (grade I and II) day 3 embryos were prioritized for vitrification freezing. Additional high-quality cleavage embryos, as well as poor-quality (grades III and IV) cleavage-stage embryos, were cultured to blastocysts. Blastocysts were graded on day 5/6 according to Gardner standards [25]. Blastocysts on day 5 or 6 were selected for vitrification freezing if they were at stage 3 or more, with at least one B rating in the inner cell mass or trophoblast cell grading.

Vitrification and warming were performed according to conventional methods, as described in previous study [23, 26].

Laser-assisted hatching

After warming, thawed cleavage stage embryos and blastocysts were transferred to G2 culture medium (G-2™ PLUS, Vitrolife Group) and incubated for 10–20 min. For cleavage stage embryos, ZP was thinned using a microscopic laser serial punching device; and for blastocysts, 1/4 to 1/3 of the ZP was eliminated in the same way. The embryos were then transferred to fresh droplets of G2 culture medium and waited for transplantation. Each surviving embryos was transplanted within 3 h after AH.

Endometrial preparation and frozen embryo transfer

Endometrial preparation and the FET were performed as described in our previous study. The endometrial preparation was performed under the natural cycle, the ovulation cycle, or the hormone replacement cycle, according to the patient’s condition. For all three protocols, embryos aged 3 or 5–6 days old were thawed for transfer on the third or fifth day of endometrium transform.

Outcome measures

The primary outcome was the IVF-EP rate (including heterotopic pregnancy). The secondary outcome indicators included the implantation rate, biochemical pregnancy rate, clinical pregnancy rate, ongoing pregnancy rate, live birth rate, and miscarriage rate after IVF/ICSI. Ectopic pregnancy was defined as a pregnancy in which the embryo was implanted outside the uterine cavity. Heterotopic pregnancy was defined as the coexistence of ectopic and intrauterine pregnancies. Biochemical pregnancy was defined as blood HCG > 25 U/L on the 14th day after embryo transfer. Clinical pregnancy was defined as a gestational sac observed by transvaginal ultrasound 35 days after embryo transfer. Ongoing pregnancy was defined as a pregnancy that persisted for 20 weeks or more. Live birth was defined as the delivery of one or more live births. Miscarriage was defined as spontaneous abortion or intrauterine demise before 24 weeks of gestational age. All outcome indicators were calculated based on the number of cycles. Due to the time required for follow-up, the cycles included in the calculation of the live birth rate were those that met the criteria from August 2019 to April 2023.

Statistical analysis

All the statistical analyses were performed with SPSS 26.0 and R 4.3.3. The normality of the quantitative data was tested by the Kolmogorov‒Smirnov and Shapiro‒Wilk tests. Normally and nonnormally distributed data are expressed as the mean ± SD and median (first quartile, third quartile), respectively. Qualitative data are expressed as n (%). Student’s t-test was used to compare normally distributed independent samples, while the nonparametric Mann-Whitney U test was used to compare abnormally distributed independent samples. For all comparisons, P value < 0.05 indicated statistical significance.

Subgroup analyses and binary logistic regression analysis were performed to determine whether there were differences in the effect of AH on IVF-EP among patients with different previous EP history. As several cycles of the same patient were included in the study, we used generalized estimating equation (GEE) for adjusting the regressions. Inverse-probability-of-treatment weighting (IPTW) adjustment was used to balance baseline covariates between groups and reduce confounding bias (especially physician selection bias). Propensisty scores (PS) were estimated using binary logistic regression, with AH as the outcome and potential confounders (including maternal age, maternal body mass index, parity, duration of infertility, infertility cause, endometrial thickness, number of embryos, embryo stage, previous EP history, and fallopian tubal status) as predictors. The weight for AH group was 1/ PS. The weight for non-AH group was 1/ (1-PS). The effectiveness of weighting was evaluated by standardized mean differences (SMDs). A SMD < 0.1 was considered indicative of adequate covariate balance. Weighted logistic regression was then performed to assess the association between AH and IVF-EP. Additionally, in subgroup analysis after stratification, where data was sparse and the number of outcomes was low, stabilized weights were adopted to reduce extreme values and Firth corrected regression model was applied to improve model stability. Stabilized weights were constructed based on the marginal distributions of exposure and confounding factors. In this occasion, the weight for the AH group was Pt/PS, and the weight for the non-AH group was (1-Pt)/(1-PS), where Pt represented the proportion receiving AH.

Results

A total of 5129 cycles were included in this study, with 1022 in the AH group and 4107 in the Non-AH group. The baseline characteristics of the two groups are shown in Table 1, with statistically significant differences in duration of infertility, body mass index (BMI), embryo stage of transfer and endometrial thickness. The AH group had a longer period of infertility (2.30 ± 2.71 vs. 2.06 ± 2.45) and a higher BMI (22.56 ± 3.12 vs. 22.27 ± 2.99), compared to the Non-AH group. The AH group had a higher blastocyst transfer rate of 60.1%; whereas the blastocyst transfer rate in the Non-AH group was 22.4%. The Non-AH group had a thicker endometrial thickness (10.27 ± 2.48 vs. 10.46 ± 2.42). Other baseline characteristics were not different between the two groups.

Table 1.

Baseline characteristics of all cycles in patients with EP history

Cycles/N	AH group (N = 1022)	Non-AH group (N = 4107)	P
Maternal age (years)	34.44 ± 4.80	34.28 ± 4.79	0.346
Duration of infertility (years)	2.30 ± 2.71	2.06 ± 2.45	0.007*
Maternal BMI (kg/m2)	22.56 ± 3.12	22.27 ± 2.99	0.007*
Pluriparous (n, %)	201 (19.7%)	819 (19.9%)	0.844
Infertility cause (n, %)			0.544
Tubal factor alone	782 (76.5%)	3194 (77.8%)
Mixed with male factor	22 (2.2%)	69 (1.7%)
Mixed with ovulatory dysfunction	8 (0.8%)	45 (1.1%)
Mixed with uterine factor	16 (1.6%)	61 (1.5%)
Mixed with multiple factors	194 (18.9%)	738 (18.0%)
Fallopian tubal status (n, %)			0.362
Unilateral tubal removal	540 (52.8%)	2264 (55.1%)
Bilateral tubal removal	123 (12.0%)	493 (12.0%)
Unresected	359 (35.1%)	1350 (32.9%)
Number of previous EP	1.28 ± 0.45	1.28 ± 0.45	0.614
Number of embryos transferred	1.57 ± 0.50	1.57 ± 0.49	0.396
Blastocyte transfer rate (n, %)	614 (60.1%)	919 (22.4%)	< 0.001*
Endometrial thickness (mm)	10.27 ± 2.48	10.46 ± 2.42	0.025*

Abbreviations: AH assisted hatching, BMI body mass index

Data are presented as mean ± SD or number of cases (n) with rate (%). * P < 0.05

Pregnancy outcomes are demonstrated in Table 2. No statistically significant difference was seen in IVF-EP rate between the two groups (1.4% vs. 2.1%, P = 0.113). The rates of biochemical pregnancy and miscarriage were both higher in the AH group than in the Non-AH group (biochemical pregnancy: 54.7% vs. 50.7%, P = 0.022; miscarriage: 9.4% vs. 7.3%, P = 0.027). The proportion of singleton births in the AH was higher than that in the Non-AH group. There was no difference in other pregnancy outcomes.

Table 2.

Pregnancy outcomes of all cycles in patients with EP history

Cycles/N	AH group (N = 1022)	Non-AH group (N = 4107)	P
Ectopic pregnancy (n, %)	14 (1.4%)	88 (2.1%)	0.113
Implantation rate (n, %)	31.7% (510/1608)	31.1% (2029/6522)	0.638
Biochemical pregnancies (n, %)	559 (54.7%)	2082 (50.7%)	0.022*
Clinical pregnancies (n, %)	463 (45.3%)	1807 (44.0%)	0.452
Ongoing pregnancies (n, %)	353 (34.5%)	1418 (34.5%)	0.993
Live birth rate (n, %)	35.7% (261/732)	33.8% (1146/3389)	0.341
Singleton (% per live birth)	235 (90.0%)	941 (82.1%)	0.004*
Twins (% per live birth)	26 (10.0%)	205 (17.9%)
Miscarriage rate (n, %)	96 (9.4%)	301 (7.3%)	0.027*

Abbreviations: AH Assisted Hatching

Live birth rate was counted in the cycles from August 2019 to April 2023: 732 cycles in AH group and 3389 cycles in Non-AH group

Data are presented as number of cases (n) with rate (%). * P < 0.05

We then performed subgroup analyses based on the number of previous EP, as shown in Table 3. In the One-EP subgroup, there were a total of 732 cycles in the AH group and 2974 cycles in the Non-AH group. There was a statistically significant difference in the IVF-EP rate between the two groups, with the AH group having a lower rate than the Non-AH group (1.0% vs. 2.2%, P = 0.028). As for the > = 2 EP subgroup, there were a total of 290 cycles in the AH and 1133 cycles in the Non-AH group. There was no statistical difference in the IVF-EP rate between the AH and Non-AH groups (2.4% vs. 1.9%, P = 0.612). The remaining pregnancy outcomes are shown in the table. The respective baselines for the two subgroups are shown in Supplementary Table 1 (Duration of infertility, BMI, blastocyst rate, and endometrial thickness were not comparable between groups in One-EP cycles; while blastocyst rate was not comparable between groups in > = 2 EP cycles).

Table 3.

Pregnancy outcomes of cycles with different number of previous EP

One-EP Cycles/N	AH group (N = 732)	Non-AH group (N = 2974)	P
Ectopic pregnancy (n, %)	7 (1.0%)	66 (2.2%)	0.028*
Implantation rate (n, %)	31.8% (367/1153)	30.9% (1455/4713)	0.529
Biochemical pregnancies (n, %)	399 (54.5%)	1511 (50.8%)	0.073
Clinical pregnancies (n, %)	331 (45.2%)	1298 (43.6%)	0.442
Ongoing pregnancies (n, %)	250 (34.2%)	1011 (34.0%)	0.935
Live birth rate (n, %)	36.8% (189/514)	33.3% (818/2456)	0.131
Singleton (% per live birth)	171 (90.5%)	670 (81.9%)	0.004*
Twins (% per live birth)	18 (9.5%)	148 (18.1%)
Miscarriage rate (n, %)	74 (10.1%)	221 (7.4%)	0.016*
>=2 EP Cycles/N	AH group (N = 290)	Non-AH group (N = 1133)	P
Ectopic pregnancy (n, %)	7 (2.4%)	22 (1.9%)	0.612
Implantation rate (n, %)	31.4% (143/455)	31.7% (574/1809)	0.902
Biochemical pregnancies (n, %)	160 (55.2%)	571 (50.4%)	0.147
Clinical pregnancies (n, %)	132 (45.5%)	509 (44.9%)	0.856
Ongoing pregnancies (n, %)	102 (35.2%)	407 (35.9%)	0.812
Live birth rate (n, %)	33.0% (72/218)	35.2% (328/933)	0.552
Singleton (% per live birth)	64 (88.9%)	271 (82.6%)	0.192
Twins (% per live birth)	8 (11.1%)	57 (17.4%)
Miscarriage rate (n, %)	22 (7.6%)	80 (7.1%)	0.757

Abbreviations: AH assisted hatching

Live birth rate was counted in the cycles from August 2019 to April 2023: 514 cycles in AH group and 2456 cycles in Non-AH group with previous 1 EP; 218 cycles in AH group and 933 cycles in Non-AH group with previous >=2 EP

Data are presented as number of cases (n) with rate (%). * P < 0.05

As the two groups were not perfectly balanced at baseline, and to mitigate physician selection bias, we applied GEE after IPTW adjustment to clarify the effect of AH on IVF-EP, see Table 4. In the entire cycles, we found that AH did not influence the IVF-EP rate (aOR: 0.888, 95% CI: 0.414–1.904, P = 0.760). The risk of IVF-EP only decreased with an increase in endometrial thickness (aOR: 0.729, 95% CI: 0.589–0.902, P = 0.004).

Table 4.

Risk factors associated with EP rate by logistic regression analysis in all cycles after inverse-probability-of-treatment weighting

Cycles/N = 5129	aOR (95% CI)	P
AH or not
AH	0.888 (0.414,1.904)	0.760
No AH	Reference
Maternal age (years)	0.907 (0.818,1.007)	0.067
Maternal BMI (kg/m2)	1.044 (0.853,1.277)	0.678
Duration of infertility (years)	0.920 (0.758,1.117)	0.402
Pluriparous (n,%)	1.017 (0.265,3.906)	0.980
Infertility cause (n,%)
Tubal factor alone	27.044 (11.295,64.753)	< 0.001*
Mixed with male factor	/	/
Mixed with ovulatory dysfunction	1.189 (0.088,16.080)	0.897
Mixed with uterine factor	/	/
Mixed with multiple factors	Reference
Number of embryos transferred	1.074 (0.303,3.805)	0.911
Stage of embryos transferred
Cleavage stage	Reference
Blastocyte	0.470 (0.124,1.780)	0.266
Endometrial thickness	0.729 (0.589,0.902)	0.004*
Previous ectopic pregnancy history
EP history > = 2	1.311 (0.440,3.907)	0.627
EP history = 1	Reference
Fallopian tubal status
Bilateral tubal removal	0.993 (0.206,4.793)	0.993
Unilateral tubal removal	2.364 (0.689,8.104)	0.171
Unresected	Reference	/

Abbreviations: BMI body mass index, aOR adjusted odds ratio, CI confidence interval

The interaction term between AH and Stage of embryos transferred was not statistically significant.Inverse-probability-of-treatment weighting: The weight for AH group was 1/ propensity score. The weight for non-AH group was 1/ (1-propensity score). All SMD < 0.1

Data are presented as aOR with 95%CI. * P<0.05

We also performed GEE with IPTW in each subgroup to determine the effect of AH on IVF-EP, the results of which are shown in Table 5 (the results without IPTW shown in Supplementary Tables 3–4. The regression results for each group with and without IPTW were consistent). In One-EP subgroup, AH had no significant effect on the risk of IVF-EP (aOR: 0.418, 95% CI: 0.154–1.131, P = 0.086). However, in > = 2 EP subgroup, AH significantly increased the risk of IVF-EP (aOR: 2.438, 95% CI: 1.509–3.939, P = 0.023). Besides, in one-EP subgroup, endometrial thickness decreased the risk of IVF-EP. In > = 2 EP subgroup, blastocyst transfer reduced the risk of IVF-EP.

Table 5.

Risk factors associated with EP rate by logistic regression analysis in patients with different EP history after inverse-probability-of-treatment weighting

Cycles/N	One-EP cycles (N = 3706)		>=2 EP cycles(N = 1423)
	aOR (95% CI)	P	aOR (95% CI)	P
AH or not
AH	0.418 (0.154,1.131)	0.086	2.438 (1.509,3.939)	0.023*
No AH	Reference		Reference
Maternal age (years)	0.918 (0.833,1.012)	0.086	0.906 (0.747,1.098)	0.312
Maternal BMI (kg/m2)	1.028 (0.768,1.375)	0.853	1.100 (0.849,1.424)	0.471
Duration of infertility (years)	1.005 (0.859,1.177)	0.947	0.668 (0.396,1.127)	0.130
Pluriparous (n,%)	0.660 (0.081,5.368)	0.697	1.507 (0.295,7.702)	0.622
Infertility cause (n,%)
Tubal factor alone	32.001 (11.558,88.600)	< 0.001*	21.850 (6.825,69.954)	< 0.001*
Mixed with male factor	/	/	/	/
Mixed with ovulatory dysfunction	1.493 (0.068,33.014)	0.800	/	/
Mixed with uterine factor	/	/	/	/
Mixed with multiple factors	Reference		Reference
Number of embryos transferred	2.398 (0.479,12.003)	0.287	0.371 (0.102,1.356)	0.134
Stage of embryos transferred
Cleavage stage	Reference		Reference
Blastocyte	1.521 (0.358,6.471)	0.570	0.117 (0.016,0.884)	0.038*
Endometrial thickness	0.689 (0.519,0.913)	0.009*	0.735 (0.509,1.063)	0.102
Fallopian tubal status
Bilateral tubal removal	0.165 (0.034,0.812)	0.027	1.305 (0.152,11.189)	0.808
Unilateral tubal removal	2.262 (0.544,9/406)	0.262	2.489 (0.271,22.902)	0.421
Unresected	Reference		Reference

Abbreviations: AH assisted hatching, BMI body mass index, aOR adjusted odds ratio, CI confidence interval

The interaction term between AH and Stage of embryos transferred was not statistically significant.Inverse-probability-of-treatment weighting: The weight for AH group was 1/ propensity score. The weight for non-AH group was 1/ (1-propensity score). All SMD < 0.1

Data are presented as aOR with 95%CI. * P < 0.05

Given that the blastocyst transfer rate was higher in the AH group, we conducted a stratified GEE regression analysis after stabilized weighting + Firth correction based on different embryo stages and previous EP history, with results shown in Table 6. When transferring cleavage-stage embryos, AH significantly increased the risk of IVF-EP in cycles with > = 2 EP history (aOR: 2.217, 95% CI: 1.172–4.251, P = 0.014). Among blastocyst transfer cycle, the risk of IVF-EP after AH did not differ across different populations (e.g. >=2 EP cycles: aOR: 4.294, 95% CI: 0.533-558.357, P = 0.214). However, the low sample size in the recurrent EP cycles may have influenced the results. We also evaluating the influence of AH on other pregnancy outcomes in all groups. For One-EP subgroup, >=2 EP subgroup and the entire population with a history of EP, AH did not affect the rate of biochemical pregnancy, clinical pregnancy, live birth or miscarriage. Detailed results are shown in Supplementary Table 2.

Table 6.

The influence of AH on EP rate by logistic regression analysis in different cycles after inverse-probability-of-treatment weighting (stabilized weighting + Firth correction)

Blastocyst cycles/N	EP rate in AH group	EP rate in non-AH group	aOR (95% CI)	P
All cycles (N = 1533)	1.1% (7/614)	1.0% (9/919)	1.200 (0.421,3.419)	0.732
One-EP cycles (N = 1092)	1.1% (5/439)	1.4% (9/653)	0.854 (0.258,2.835)	0.797
>=2 EP cycles (N = 441)	1.1% (2/175)	0.0% (0/266)	4.294 (0.533,558.357)	0.214
Cleavage-embryo cycles/N	EP rate in AH group	EP rate in non-AH group	aOR (95% CI)	P
All cycles (N = 3596)	1.7% (7/408)	2.5% (79/3188)	1.036 (0.404,2.656)	0.942
One-EP cycles (N = 2614)	0.7% (2/293)	2.5% (57/2321)	0.306 (0.084,1.121)	0.074
>=2 EP cycles (N = 982)	4.3% (5/115)	2.5% (22/867)	2.217 (1.172,4.251)	0.014*

Abbreviations: AH assisted hatching, aOR adjusted odds ratio, CI confidence interval

Variables introduced into the analysis: maternal age, maternal body mass index, parity, duration of infertility, infertility cause, endometrial thickness, number of embryos, previous ectopic pregnancy history, fallopian tubal status and assisted hatching or not. Only the results of AH or not were displayed in this table, and No AH was set as the reference of AH in all logistic regression. The interaction term between AH and Stage of embryos transferred was not statistically significant. All stratified data underwent stabilized weight regression + Firth correction. The weight for the AH group was Pt/PS, and the weight for the non-AH group was (1-Pt)/(1-PS), where Pt represented the proportion receiving AH

Data are presented as number of cases (n) with rate (%) and aOR with 95%CI. * P < 0.05

Discussion

Given the increasing use of AH in FET cycles and its ability to advance embryo implantation timing and enhance implantation potential, it is crucial to investigate the relationship between AH and IVF-EP, which may be related to embryo migration and implantation time, particularly in patients with a history of EP. In the study, a total of 2888 patients with a history of EP and their corresponding 5129 cycles were included in our study, involving 1022 cycles with AH and 4107 cycles without AH. We found that the effect of AH on IVF-EP was different in different populations with a history of EP. In the entire population, AH showed no effect on IVF-EP. In contrast, after classification according to the number of previous EP, it was found that AH had no effect on IVF-EP in patients with single previous EP; while AH increased the risk of IVF-EP in patients with recurrent previous EP. Therefore, patients with recurrent EP may be contraindicated for laser-AH.

There are no previous studies on the effect of AH on IVF-EP in patients with EP history. We are the first study to focus on this area. Our study found that performing laser-AH in patients with a history of EP had no effect on the risk of developing EP after IVF/ICSI, which is consistent with the findings of previous studies of AH in the whole population [16–18]. Our study also found a trend towards an opposite effect of laser-AH between the population with single previous EP and the population with recurrent EP. Though there were more cycles of patients with one previous EP, the trend towards lower risk of IVF-EP after AH did not meet the criteria for significance (P = 0.760); while patients with recurrent EP had a 1.438-fold increase in the risk of IVF-EP after AH (aOR: 2.438, 95% CI: 1.509–3.939, P = 0.023). Only when the two populations with different number of previous EP were combined to calculate, conclusion was obtained that AH did not affect the risk of IVF-EP in patients with a history of EP. Therefore, our study uncovered the specific effect of laser-AH on IVF-EP in different situations for the first time, and warns that patients with recurrent EP might be contraindicated for laser-AH. The difference in the result of IVF-EP between the > = 2 EP subgroup and the One-EP subgroup also suggested that patients with recurrent EP did have their uniqueness, which made them present different characteristics. In this view, our results on AH might shed some light on the causes of recurrent EP, with the underlying mechanisms to be explored.

There are two current views on the mechanism by which AH can improve embryo implantation as follows: (i) AH advances the time to implantation [11, 12]; and (ii) AH improves the implantation potential through enhanced messaging [13]. Our multifactorial regression of biochemical pregnancy rate found that AH had no effect on biochemical pregnancy rate, regardless of the population. This finding implies that the higher IVF-EP rate after AH in patients with recurrent EP is not achieved through a theoretically improved implantation potential of embryo. On the other hand, given that embryo implantation time may be the most important factor for the recurrent EP population, it is suggested that the occurrence of IVF-EP may be associated with early embryo wandering.

A previous experiment using 40 µL of radiopaque dye for simulated embryo transfer found that dye was observed in 38% of the fallopian tubes [27], suggesting that embryos are likely to wander into the fallopian tubes after being transferred into the uterine cavity. Since the IVF-EP rate is much lower than 38%, IVF-EP occurs not only because of the reach of embryos to fallopian tubes, but also because of the blockage of their movement back to the uterus. With the tendency for earlier hatching and implantation after AH, we make the following speculations: for patients with recurrent EP, the wandering time of embryos might extend and therefore embryos are more likely to implant exactly after they reach the fallopian tubes but before they return to the uterine cavity, during which the embryos have already returned to the uterine cavity for the rest of the population. The longer retention of embryos in the fallopian tubes in patients with recurrent EP could be explained from two aspects. First, as tubal diseases are the most common influencing factor preventing embryos to return to the uterus, previous EP history raises the risk of subsequent IVF-EP. Second, uterine and tubal signaling may also play a critical role, as suggested by the negative results in populations with single EP history. Specifically, the signals of uterine attracting for embryo movement and implantation may be weaker in the population with recurrent EP than in other populations. To further elucidate the potential biological mechanisms underlying the differential effects of AH between patients with a single EP history and those with recurrent EP, future studies are recommended to investigate the expression profiles of key regulatory factors involved in embryo implantation and localization. Specifically, comparative analyses should be conducted on the expression differences of signaling pathways critical for embryo positioning (e.g., the Wnt/β-catenin pathway) and cell adhesion molecules essential for embryo-maternal interaction (such as integrins and leukemia inhibitory factor) between these two patient populations. Such investigations may uncover valuable mechanistic clues explaining how AH impact IVF-EP across different EP history subgroups.

Our study also found that endometrial thickness affected the risk of IVF-EP in the overall population with a history of EP, but the effect was not statistically significant when subdivided according to the number of previous EP, which may be due to insufficient sample size to test for differences. In addition, the miscarriage rate in the AH group was slightly higher, especially in the group with single EP history (Table 3). The underlying mechanism likely involves the differential impact of AH: while it may promote trophoblast-mediated implantation (manifested by the trend toward increasing biochemical pregnancy rate, P = 0.073), its lack of beneficial influence on inner cell mass (ICM) development impairs the embryo’s ability to sustain pregnancy beyond the early stages. Thus, despite no significant difference in later-stage ongoing pregnancy rates, the inability to enhance ICM competence may contribute to the increased incidence of miscarriage in the AH group.

In summary, this study is a well-designed study with a large sample size. It is the first study to consider the effect of AH on IVF-EP in patients with a history of EP. Subgroup analyses were also done for patients with different numbers of previous EP, in which the outcomes of IVF-EP were different between recurrent EP population ( > = 2 EP) and populations with one previous EP. For the first time, the idea that recurrent EP may be inappropriate for AH was proposed. In addition, we inferred that there may be a time lag when embryo move back to the uterine cavity in the population with recurrent EP, which raised speculation about the possible causes of IVF-EP in the population with recurrent EP. Finally, we considered both previous salpingectomy and other EP risk factors in our study to minimize the interference of confounding factors, making the results more reliable.

However, we must acknowledge the limitations of this study. Currently, methods for AH include chemical AH, mechanical AH, laser-AH, or use of a piezo micromanipulator. The most used method is laser-AH. Although current guideline and large meta-analysis provisionally consider that there is no evidence that the effect of AH varies between AH methods [10], the fact that only laser-assisted AH was used in this study may have some limitations. The results may not be applicable to fresh embryo transfer, and further research is needed to validate the consistency of conclusions under different AH methods. Additionally, it is important to note that there was a statistical difference in blastocyst transferred rate between the two groups (Table 1: AH group 60.1% vs. Non-AH group 22.4%, P < 0.001), which itself may affect IVF-EP risk. To avoid the confounding effects of this result, we subsequently conducted stratified and regression analyses based on embryo stage (see Table 6). The results obtained from the blastocyst cycles and cleavage stage cycles were relatively consistent with those from the overall cycles, confirming the reliability of the conclusions drawn in this study. The interaction term between blastocyst transfer and AH was also tested in all GEE regression models, and no interaction was found. We also performed case-control matching between AH group and Non-AH group, with the matched results presented in Supplementary Tables 5–8. Under baseline-balanced conditions, the trend in IVF-EP rate results were consistent with the regression results in Tables 4 and 5 (with IPTW) and Supplementary Tables 3–4 (without IPTW). However, for strata with 0 events in one arm (e.g., the blastocyst > = 2 EP cycles), estimated results remain imprecise even after applying Firth correction and stabilized weighting. As such, stratified analysis results should be interpreted as exploratory and treated with caution to avoid over-inference. We also acknowledge that the non-randomized design of this study is a critical source of potential residual confounding, as AH use was determined by individual physician preferences rather than a standardized protocol (which might introduce unmeasured variables correlated with both treatment assignment and IVF-EP outcomes). Therefore, we cannot establish a causal relationship between laser-AH and IVF-EP, and further randomized controlled multicenter trials are needed to validate the result.

Conclusions

For the entire population with a history of EP, laser-AH was not associated with the risk of IVF-EP. After categorized according to the number of previous EP, it was found that in patients with one previous EP, laser-AH had no effect on the occurrence of IVF-EP; whereas in patients with recurrent previous EP, laser-AH was associated with an increased risk of IVF-EP. Notably, this study is retrospective and observational, so potential unmeasured confounders cannot be completely excluded. Thus, we cannot establish a causal relationship between laser-AH and IVF-EP based on the current data. Our retrospective findings required validation through well-designed multicenter RCT studies. Furthermore, the different mechanisms of its effects in single EP and recurrent EP population require further research. Therefore, the application of laser-AH in patients with recurrent EP warrants caution until more robust evidence is available.

Abbreviations

AH

Assisted Hatching

ART

Assisted Reproductive Technology

EP

Ectopic Pregnancy

FET

Frozen-thawed Embryo Transfer

GEE

Generalized Estimating Equation

ICSI

Intracytoplasmic Sperm Injection

IVF

In Vitro Fertilization

IVF-EP

EP resulting from IVF/ICSI

ZP

Zona Pellucida

Authors’ contributions

Study concept and design: S.J., and X.J. Acquisition of data: X.J., W.L. and X.S. Analysis and interpretation of data: X.J., J.Z., W.L., X.S., and Y.M. Drafting of the manuscript: X.J., J.Z., and S.J. Critical revision of the manuscript for important intellectual content: X.J., W.L., J.Z., X.S., Y.M., and S.J. Statistical analysis: X.J. and J.Z. Study supervision: W.L., S.J., and X.J.

Funding

¹This work was supported by grants from National Nature Science Foundation of China (82201886, 81901478).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was ethically reviewed by the Shanghai Ninth People’s Hospital institution. (SH9H-2023-TK109-18) approved the study. Written informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.