Impact of IVF/ICSI on grades of placenta accreta spectrum disorders and pregnancy outcomes.

A total of 974 eligible patients were enrolled in the study based on predefined inclusion and exclusion criteria. Among them, 820 (84.2%) patients belonged to the spontaneous conception group, with 154 (15.8%) the IVF/ICSI group. After propensity score matching, each group is comprised of 123 patients. The matching rate for the IVF/ICSI group is 79.87% (Fig.  1 ). Fig. 1 Flowchart of patient selection Flowchart of patient selection Before propensity score matching (PSM), the IVF/ICSI group exhibits significantly higher age ( P  = 0.001), a greater number of advanced age cases ( P  = 0.001), and increased progesterone supplementation in miscarriage prevention ( P  < 0.001) compared to the spontaneous conception group. Conversely, the spontaneous conception group demonstrates significantly higher gravidity ( P  < 0.001), parity ( P  < 0.001), cesarean sections ( P  < 0.001), induced abortions ( P  < 0.001), anterior placenta ( P  < 0.001), and placenta previa ( P   0.05) (Table  1 ). Table 1 Baseline characteristics before and after PSM Characteristics Before PSM After PSM Spontaneous conception ( n  = 820) IVF/ICSI pregnancy ( n  = 154) P  value Spontaneous conception ( n  = 123) IVF/ICSI pregnancy ( n  = 123) P  Value Age [Year,M(P25,P75)] 34 (30,37) 35 (33,39) 0.001 35 (32,37) 35 (33,38) 0.403 Advanced age [n(%)] 370(45.1) 91(59.1) 0.001 64 (52.0) 72 (58.5) 0.097 Body mass index [kg/m 2 ,n(%)] 0.469 0.825  < 18.5 88(10.7) 13(8.4) 13 (10.6) 11 (8.9)  18.5 ~ 24 555(67.7) 103(66.9) 87 (70.7) 88 (71.5)  24 ~ 28 140(17.1) 33(21.4) 18 (14.6) 21 (17.1)  ≥ 28 37(4.5) 5(3.3) 5 (4.1) 3 (2.4) Gravidity [M(P25,P75)] 4(3,5) 3(2,4) < 0.001 3(2,4) 3(2,4) 0.989 Parity [M(P25,P75)] 1(1,2) 0(0,1) < 0.001 0(0,1) 0(0,1) 1.000 Number of vaginal deliveries [M(P25,P75)] 0(0,0) 0(0,0) 0.332 0(0,0) 0(0,0) 0.704 Number of cesarean deliveries [n(%)] < 0.001 0.899  0 253(30.9) 108(70.1) 79 (64.2) 78 (63.4)  1 386(47.1) 36(23.4) 35 (28.5) 35 (28.5)  2 167(20.4) 9(5.8) 7 (5.7) 9 (7.3)  ≥ 3 14(1.7) 1(0.6) 2 (1.6) 1 (0.8) Number of induced abortions [n(%)] 0.001 0.693  0 429(52.3) 106(68.8) 70 (56.9) 79 (64.2)  1 208(25.4) 30(19.5) 33 (26.8) 28 (22.8)  2 116(14.1) 13(8.4) 13 (10.6) 11 (8.9)  ≥ 3 67(8.2) 5(3.2) 7 (5.7) 5 (4.1) Other etiologies of accreta placentation [n(%)] 303(37.0) 64(41.6) 0.279 48 (39.0) 49 (39.8) 0.896 Progesterone supplementation in miscarriage prevention [n(%)] 156(19.0) 61(39.6) < 0.001 41 (33.3) 41 (33.3) 1.000 Cesarean scar pregnancy[n(%)] 61(7.4) 8(5.2) 0.319 7 (5.7) 6 (4.9) 0.776 Placental location [n(%)] < 0.001 0.258  Anterior wall 462 (56.3) 66 (42.9) 61 (49.6) 59 (48.0)  Posterior wall 221 (27.0) 66 (42.9) 36 (29.3) 46 (37.4)  Others 137 (16.7) 22 (14.3) 26 (21.1) 18 (14.6) Vaginal bleeding [n(%)] 245(29.9) 43(27.9) 0.626 42 (34.1) 33 (26.8) 0.213 Placenta previa [n(%)] 531(64.8) 53(34.4) < 0.001 43 (35.0) 48 (39.0) 0.509 Advanced age: age ≥ 35 years Other etiologies of accreta placentation: uterine curettage, manual delivery of the placenta, IUD, postpartum endometritis, hysteroscopic surgery, myomectomy for endometrial injury, etc. Baseline characteristics before and after PSM Advanced age: age ≥ 35 years Other etiologies of accreta placentation: uterine curettage, manual delivery of the placenta, IUD, postpartum endometritis, hysteroscopic surgery, myomectomy for endometrial injury, etc. Before propensity score matching (PSM), the IVF/ICSI group shows a lower incidence of invasive placenta ( P  < 0.001), postpartum hemorrhage ( P  < 0.001), red blood cells transfusion ≥ 4U ( P  < 0.001), and preterm delivery ( P  < 0.001) compared to the spontaneous conception group. No significant differences in hysterectomy ( P  = 0.254), birth weight ( P  = 0.399), and the number of newborns admitted to the NICU ( P  = 0.301) between the two groups were detected. After PSM, the IVF/ICSI group exhibits a significantly higher incidence of postpartum hemorrhage ( P  < 0.001) and red blood cells transfusion ≥ 4U ( P  = 0.044). In contrast, there are no statistically significant differences in PAS grades ( P  = 0.290), hysterectomy ( P  = 0.424), gestational weeks ( P  = 0.927), birth weight ( P  = 0.816), and the number of NICU admissions ( P  = 0.301) between the two groups (Table  2 ). Table 2 Maternal and neonatal outcomes before and after PSM Outcomes Before PSM After PSM Spontaneous conception ( n  = 820) IVF/ICSI pregnancy ( n  = 154) P  value Spontaneous conception ( n  = 123) IVF/ICSI pregnancy ( n  = 123) P  value Maternal outcome  PAS Grades [n(%)] < 0.001 0.290  Placenta crete 313(38.2) 103(66.9) 74 (60.2) 82 (66.7)  Placenta increta 459(56.0) 50(32.5) 45 (36.6) 40 (32.5)  Placenta percreta 48(5.9) 1(0.7) 4 (3.3) 1 (0.8)  Postpartum hemorrhage [n(%)] 313(38.2) 25(16.2) < 0.001 3 (2.4) 23 (18.7) < 0.001  Red blood cells transfusion ≥ 4U [n(%)] 163(19.9) 13(8.4) < 0.001 5 (4.1) 13 (10.6) 0.044  Hysterectomy [n(%)] 159(19.4) 6(3.9) 0.254 9 (7.3) 6 (4.9) 0.424 Neonatal outcome  Birth weeks [wk,n(%)] < 0.001 0.927 ≥ 28–34 a 103(12.6) 22(14.3) 19 (15.4) 18 (14.6) 34–37 a 377(46.0) 37(24.0) 33 (26.8) 31 (15.2) ≥ 37 340(41.5) 95(61.7) 71 (57.7) 74 (60.2)  Birth weight [g,n(%)] 0.399 0.816 < 1500 37(4.5) 10(6.5) 8 (6.5) 9 (7.3) 1500–2500 224(27.3) 36(23.4) 29 (23.6) 25 (20.3) ≥ 2500 559(68.2) 108(70.1) 86 (69.9) 89 (72.4)  NICU [n(%)] 28(3.4) 2(1.3) 0.301 34 (27.6) 27 (22.0) 0.301 Postpartum hemorrhage, which is defined as blood loss of more than 1000 ml during a cesarean section or within 24 h after delivery of the fetus a Preterm delivery Maternal and neonatal outcomes before and after PSM Postpartum hemorrhage, which is defined as blood loss of more than 1000 ml during a cesarean section or within 24 h after delivery of the fetus a Preterm delivery We then carried out conditional logistic regression analysis and found no link between IVF/ICSI treatment and the grades of PAS (OR = 0.76, 95% CI: 0.45–1.27, P  = 0.290). In addition, IVF/ICSI did not increase the risk of hysterectomy (OR = 0.65, 95% CI: 0.22–1.88, P  = 0.427), preterm birth (OR = 0.90, 95% CI: 0.54–1.50, P  = 0.697), low birth weight infants (< 2500 g) (OR = 0.89, 95% CI: 0.51–1.54, P  = 0.673), or neonates admitted to the NICU (OR = 0.66, 95% CI: 0.11–4.03, P  = 0.654). Strikingly, however, IVF/ICSI was identified as a risk factor for postpartum hemorrhage (OR = 9.20, 95% CI: 2.68–9.20, P  < 0.001) and red blood cell transfusion ≥ 4U (OR = 3.71, 95% CI: 1.21–11.33, P  = 0.021 (Table  3 ). Table 3 Conditional logistic regression analysis Outcomes OR 95%CI P value PAS grades 0.76 (0.45 ~ 1.27) 0.290 Postpartum hemorrhage 9.20 (2.68 ~ 9.22) < 0.001 Red blood cells transfusion ≥ 4U 3.71 (1.21 ~ 11.33) 0.021 Hysterectomy 0.65 (0.22 ~ 1.88) 0.427 Birth week < 37 0.90 (0.54 ~ 1.50) 0.697 Birth weight < 2500 g 0.89 (0.51 ~ 1.54) 0.673 NICU 0.66 (0.11 ~ 4.03) 0.654 Conditional logistic regression analysis

The Department of Obstetrics and Gynecology at the Third Affiliated Hospital of Guangzhou Medical University serves as the critical care center for obstetrics and gynecology in Guangdong Province, providing comprehensive treatment to critically ill patients from across the region. We conducted a retrospective study on patients with PAS who underwent cesarean section delivery at our hospital between January 1, 2018 and March 31, 2023. The study adhered to the principles outlined in the World Medical Association Declaration of Helsinki and received approval from the Ethics Committee of the Third Affiliated Hospital of Guangzhou Medical University (approval number: 20221106). All participants provided informed consent during their hospitalization, granting permission for their clinical data to be used in research while safeguarding their privacy. A total of 1007 patients were diagnosed with PAS and underwent cesarean section delivery. The classification system for PAS conforms to the clinical diagnosis criteria issued by FIGO in 2019 [ 2 ]. Exclusion criteria included gestational age less than 28 weeks ( n  = 33) and incomplete clinical data ( n  = 0). A total of 974 patients meet the inclusion criteria and are included in the study. Through the “Perinatal Medicine Database” and “Reproductive Medicine Database” established by the Third Affiliated Hospital of Guangzhou Medical University, we collected basic clinical characteristics of enrolled patients including age, body mass index, gravidity, parity, number of vaginal deliveries, number of cesarean sections, number of induced abortions, other etiologies of accreta placentation, progesterone supplementation in miscarriage prevention, scar pregnancy, placental location, vaginal bleeding, and placenta previa; maternal outcomes including PAS grades, postpartum hemorrhage, red blood cells transfusion ≥ 4U, and hysterectomy; and neonatal outcomes including preterm birth, neonatal birth weight, neonatal admission to NICU. We used propensity score matching (PSM) to balance the clinical baseline characteristics between spontaneous conception group and IVF/ICSI group, thereby minimizing bias. The matching ratio is 1:1, with a caliper value of 0.02 SD. In the matching process, we took into consideration the clinical baseline characteristics, including age, body mass index, gravidity, parity, number of vaginal deliveries, number of cesarean sections, number of induced abortions, other etiologies of accreta placentation, progesterone supplementation in miscarriage prevention, scar pregnancy, placental location, vaginal bleeding and placenta previa. The SPSS 25.0 software was utilized for data analysis. Quantitative data with non-normal distribution was denoted as M(P25 ~ P75), and inter-group comparisons were conducted using non-parametric tests. Frequency and rate are reported for the counting data, with inter-group comparisons performed using the χ2 test or Fisher’s exact test. Conditional logistic regression analysis was utilized to calculate odds ratios (OR) and corresponding 95% confidence intervals (CI) to explore the impact of IVF/ICSI on the grading of PAS; and pregnancy outcomes. A significance level of P  < 0.05 is adopted.

In summary, IVF/ICSI conception does not influence PAS grades.While IVF/ICSI pregnancies adds to the risk of postpartum hemorrhage and blood transfusion, they do not increase hysterectomy rates or adverse neonatal outcomes. Clinicians should consider patients’ expectations, medical history, local delivery management, and assistance capabilities when assessing IVF/ICSI pregnancy prognosis.

Propensity Score Matching (PSM) effectively mitigates selection bias in observational studies, allowing a more reliable assessment of treatment or intervention impacts on study outcomes. Without PSM, we found that advanced age ( P  = 0.001) [ 12 ], progesterone supplementation in miscarriage prevention ( P  < 0.001) [ 13 ], gravidity ( P  < 0.001) [ 14 ], parity ( P  < 0.001) [ 14 ], cesarean sections ( P  < 0.001) [ 15 ], induced abortions ( P  < 0.001) [ 16 ], and anterior wall ( P  < 0.001) [ 17 ] and placenta previa ( P  < 0.001) [ 18 ] are associated with PAS. When these baseline features are statistically different and have impacts on PAS, it cannot effectively explain whether PAS is related to IVF/ICSI. By ensuring consistency in individual baseline features across different groups and accounting for the confounding variables, PSM enhances internal validity and credibility of the research results [ 19 ]. Over the past four decades, the global incidence of PAS has risen tenfold, with reported rates ranging from 0.01% to 1.1% [ 7 ]. Obstetricians have extensively investigated the risk factors associated with PAS and identified that the majority are iatrogenic in nature, including cesarean section, intrauterine surgeries causing endometrial injury, and the use of assisted reproductive technologies [ 7 , 20 ]. A multicenter retrospective study in the United States revealed that PAS incidence increases with the number of cesarean sections (ranging from 3 to 67% for 1 to 5 prior cesarean sections) [ 15 ]. IVF/ICSI is associated with an increased risk of PAS. The odds ratio (OR) of PAS in IVF pregnancies is roughly between 3 and 14 [ 13 ]. In particular, frozen embryo transfer (FET) with hormone replacement cycles seems to carry a higher risk compared to fresh embryo transfer (FET) or spontaneous conception [ 13 ]. The increased risk of PAS in IVF/ICSI pregnancies may be due to several factors. One possibility is that the procedures may cause subtle changes in the endometrial lining, affecting placental implantation [ 21 ]. Another factor is that IVF/ICSI pregnancies are often associated with other risk factors for PAS, such as advanced maternal age and a history of cesarean deliveries [ 13 ]. But does PAS necessarily cause adverse pregnancy outcomes such as hysterectomy and postpartum hemorrhage?Obstetricians have found through clinical observation that different grades of PAS can lead to different adverse pregnancy outcomes [ 9 , 16 ]. Therefore, determining the grades of PAS is a critical focus for obstetricians to address. However, why has PAS grades not garnered widespread attention? Before the establishment of clinical classification criteria for PAS, its classification relied exclusively on pathological findings. During placental dissection, adherent or implanted villi may remain embedded in the uterine muscle, making it challenging for pathologists to accurately assess the depth of implantation [ 2 ]. Additionally, only a small proportion of PAS patients undergo partial or total hysterectomy, and reliance solely on pathological diagnosis often results in a high rate of missed diagnoses for cases involving adhesions or implants. The introduction of clinical classification criteria for PAS not only facilitates more accurate diagnosis but also significantly reduces the likelihood of missed diagnoses. In this study, the impact of IVF/ICSI on PAS grading was evaluated based on clinical diagnostic criteria for PAS classification. Additionally, other confounding factors influencing PAS pathogenesis were balanced using the propensity score matching (PSM) method, as demonstrated in the Table  1 . The analysis revealed that IVF/ICSI does not significantly affect PAS grading ( P  = 0.290). Furthermore, conditional logistic regression analysis also shows that IVF/ICSI does not affect the PAS grades (OR = 0.76, 95% CI: 0.45–1.27, P  = 0.290). Although we accounted for the effects of estrogen on PAS during pregnancy, data on the use of artificial cycle hormones prior to conception were unavailable due to limitations in the database.It has been proposed that the impact of IVF/ICSI conception on placenta accreta spectrum (PAS) grading may be associated with endometrial thickness and estrogen levels [ 13 ]. However, the underlying mechanisms require further investigation. In this study, we found that IVF/ICSI conception is associated with an increased risk of postpartum hemorrhage (OR = 9.20, 95% CI: 2.68–9.20, P  < 0.001) and red blood cells transfusion ≥ 4U (OR = 3.71, 95% CI: 1.21–11.33, P  = 0.021). These findings align with a retrospective cohort study that compared postnatal bleeding in 1064 IVF/ICSI pregnancies with 2059 spontaneous pregnancies. The incidence of severe postnatal bleeding is significantly higher in single pregnancies resulting from IVF/ICSI (aOR = 1.58, 95% CI: 1.12–2.24, P  = 0.010) compared to spontaneous conceptions [ 22 ]. While propensity score matching corrected for confounding factors related to postpartum hemorrhage (such as advanced age, placenta previa, and grades), some variables remain unaccounted for. Detailed records of IVF/ICSI treatment specifics (e.g., anticoagulant drug dosages and withdrawal times) [ 23 ] and reasons for using IVF/ICSI (e.g., endometriosis) are not available in the case records [ 24 ]. The surgical approach for placental removal in cases of invasive placenta accreta significantly influences postpartum blood loss [ 25 ]. Although standardized treatment protocols are implemented within the same institution, residual confounding factors related to procedural variations may still affect hemorrhage outcomes. These unmeasured factors may contribute to the increased risk observed. It is noteworthy that IVF/ICSI may contribute to postpartum hemorrhage and the need for blood transfusion without necessarily increasing the severity of PAS. Therefore, for IVF/ICSI patients, it is essential to enhance the management of hemoglobin levels during pregnancy and ensure adequate reserves of blood products in delivery hospitals. In cases where the supply of blood products is insufficient, patients should be promptly referred to a hospital equipped with the necessary resources and rescue capabilities. While this study provides valuable insights, it is essential to acknowledge its limitations. The data are derived from a single-center with a limited sample size. Although propensity score matching adjusts for numerous confounding factors, it only balances observed indicator variables. Additionally, SPSS propensity score matching achieves 1:1 matching between the experimental and control groups based on propensity scores but does not assess the balance of matched data. These limitations may introduce outcome bias. To validate the findings, prospective multi-center, large-sample case–control studies are warranted.

Placenta accreta spectrum (PAS) disorders refer to a group of conditions characterized by the abnormal invasion of placental villi into the uterine muscle to varying degrees [ 1 ]. According to the clinical diagnostic criteria established by the International Federation of Obstetrics and Gynecology (FIGO) in 2019, PAS can be classified into grades 1–3. Grade 1 is termed adherent placenta, indicating that the placental villi adhere to the surface of the uterine muscle, while grades 2–3 are referred to as invasive placenta, signifying that the placental villi implant and penetrate into the uterine muscle [ 2 ]. PAS can result in serious adverse outcomes such as preterm birth, postpartum hemorrhage, acute organ failure, hemorrhagic shock, disseminated intravascular coagulation (DIC), and even maternal mortality [ 3 ]. Accumulating evidence has established a firm link between the depth of placental villous invasion and poor prognosis [ 4 – 6 ]. PAS is associated with primary uterine abnormalities or secondary structural damage to the uterine wall [ 7 ]. Recent years have seen an increase in individuals conceiving through in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) [ 8 ]. Some studies suggest that these patients are at a higher risk to suffer (or develop) PAS [ 9 , 10 ]. However, whether invasive placenta is the cause remains an area requiring further investigation. Given the rising utilization of IVF/ICSI in China [ 11 ], it is clinically significant to explore the impact of these assisted reproductive technologies on grades of PAS and pregnancy outcomes.