Abstract
Placenta accreta spectrum (PAS) is a rare yet life-threatening obstetric complication. The primary risk factors include placenta previa and prior cesarean sections. However, high-quality epidemiological and clinical data on PAS remain scarce in China. This study aimed to comprehensively evaluate the clinical characteristics, pathological features, and radiological findings of PAS disorders in Chinese women, while quantifying the predictive value of key risk factors through receiver operating characteristic (ROC) curve analysis and assessing their impact on maternal-fetal outcomes. We conducted a retrospective study comparing clinicopathological and imaging data from 138 pathologically confirmed PAS patients and 276 age-matched controls at an academic women’s hospital in China. Statistical analyses included χ2 tests, logistic regression, and ROC curve evaluation. Magnetic resonance imaging (MRI) demonstrated higher sensitivity than sonography for PAS detection (97.6% vs 68.8%, P < .05). Univariate analysis identified gravidity, parity, prior CS, abortion history, current multiple pregnancy, and placenta previa as significant risk factors (P < .01). Multivariate analysis confirmed multivariate analysis confirmed that increased gravidity, increased parity, current multiple pregnancy, and placenta previa are independent risk factors (P < .05). ROC curve analysis identified placenta previa as the strongest predictor (area under the curve (AUC) = 0.871), followed by increased gravidity (AUC = 0.732) and increased parity (AUC = 0.722). PAS patients had significantly higher rates of postpartum hemorrhage, CS and hysterectomy (P < .001), and adverse neonatal outcomes (perinatal mortality, congenital anomalies, preterm birth, low birth weight, and asphyxia; P < .05). Subgroup analysis revealed grade 3A PAS was most prevalent (89.9%), while grades 3D/3E (2.2% each) showed higher cesarean rates, hysterectomy, newborn asphyxia, and congenital anomalies. Grade 3A was associated with more postpartum hemorrhage and fetal mortality. This study confirms MRI’s superior diagnostic accuracy for PAS (97.6% sensitivity). Placenta previa (AUC = 0.871) emerged as the strongest predictor, with increased gravidity and increased parity as key risk factors. PAS significantly increased maternal (89.9% hemorrhage rate) and neonatal complications, particularly in prevalent grade 3A cases, underscoring the need for early identification and multidisciplinary management in high-risk pregnancies.
Keywords: clinical outcomes, magnetic resonance imaging (MRI), placenta accreta spectrum (PAS), risk factors, sonography
1. Introduction
Placenta accreta spectrum (PAS) disorders are characterized by abnormal placental implantation, where chorionic villi directly adhere to or invade the myometrium in the absence of intervening decidua basalis.[1] According to the 2019 International Federation of Gynecology and Obstetrics (FIGO) guidelines, PAS is classified into 3 histological categories: placenta adherent, placenta increta, and placenta percreta.[2] The expert panel recommendations within the Perinatal Subcommittee of the Society for pediatric pathology further classified PAS into 5 grades based on the degree of gross invasion and local tissue destruction: grade 1 – noninvasive; grade 2 – superficial invasion; grade 3A – deep invasion; grade 3D – deep invasion with disruption of the serosa; and grade 3E – deep invasion with adherence to extrauterine structures.[3] Despite these detailed classification systems, the clinical relevance and prognostic value of PAS grading remain poorly characterized in current literature.
The incidence of PAS in the United States has risen dramatically from 0.03% (3 per 10,000 deliveries) in the 1930s to 1950s to 0.17% (17 per 10,000) in the 2012s to 2021s.[4] In China, while nationwide epidemiological data remain limited, clinical evidence suggests a rising PAS prevalence paralleling the increasing cesarean delivery rate. PAS can lead to severe maternal morbidities, including postpartum hemorrhage, injury to surrounding organs (particularly the urinary tract), and even maternal mortality.[5–7] Additionally, PAS significantly increases risks of adverse fetal outcomes, including higher perinatal mortality, preterm birth (<37 weeks), low birth weight (<2500 g), and neonatal asphyxia.[8] PAS pregnancies also show increased congenital anomalies, potentially due to chronic hypoxia from abnormal placentation and iatrogenic prematurity.[9] Ornaghi et al reported a preterm delivery rate of 51.3% in PAS cases, with a perinatal mortality rate of 2.08%.[10] While Farquhar et al documented an even higher preterm birth incidence (74.7%) and a perinatal mortality rate of 4%.[8] Nevertheless, the interplay between PAS, maternal/perinatal morbidity, and other clinical determinants warrants further investigation.
Multiple studies have established significant associations between PAS and 3 major risk factors: prior cesarean section (CS), placenta previa, and advanced maternal age.[11,12] Other potential risk factors for PAS include antecedent uterine surgery, abortion, multiple deliveries, endometritis, smoking, and hypertension.[13–15] Notably, robust multivariate analyses examining these risk factors in relation to histologically confirmed PAS diagnoses remain scarce in current literature.[6]
This case control study analyzed histologically confirmed PAS cases from a tertiary referral center in China between January 2013 and September 2021 (n = 138). We aimed to: identify risk factors predictive of high-grade PAS (Grades 3A–3E), and evaluate associated maternal and neonatal outcomes in this population.
2. Materials and methods
2.1. Ethical considerations
This study was approved by the ethics committee of our institution (No. PRO2021-1493). The need for informed consent was waived due to the retrospective nature of the study.
2.2. Patients
Consecutive cases of pathologically-confirmed PAS (2013–2021) were retrieved from our institutional pathology archives. To minimize confounding effects, we excluded patients with malignancies or severe coagulopathies that could independently affect hemorrhage outcomes, as well as those with miscarriage/induction to ensure homogeneity in delivery management. Controls were age-matched (±1 years) and admission time-matched (±6 months) delivery women without PAS evidence by both imaging (sonography/MRI)) and clinical evaluation. The 1:2 matching ratio was chosen to enhance statistical power while maintaining feasibility. Although this single-center design ensured standardized pathological assessment, it may limit generalizability to non-tertiary care settings.
A variety of clinical features were extracted from chart review. These data included placenta previa, history of curettage, history of abortion, history of assisted reproduction technique (ART), gravidity, parity, prior CS, gestational diabetes, hypertensive disorders (pregnancy-induced hypertension, preeclampsia, eclampsia, and chronic hypertension), current multiple pregnancy, delivery types, postpartum hemorrhage, hysterectomy, premature birth, low-birth-weight infants, newborn asphyxia, congenital anomalies and maternal/fetal mortality. PAS was re-assessed by 2 pathologists (LT and HS) following the expert panel recommendations.[3]
2.3. Statistical analysis
Statistical analyses were performed using SPSS version 26.0 (IBM Corp., Chicago). Categorical variables were compared using the χ2 test. Univariate and multivariate analyses were conducted through unconditional logistic regression. receiver operating characteristic (ROC) curve analysis was employed to evaluate the predictive performance of significant risk factors, with area under the curve (AUC) values calculated to determine diagnostic accuracy. A two-tailed P-value <.05 was considered statistically significant. The assessment of multicollinearity among variables was conducted using R software, along with the Zstats v1.0 (www.zstats.net).
3. Results
Our study had 138 PAS patients, and 276 controls. The maternal ages in both groups ranged from 19 to 45 years, with an average age of 33.5 years. Most patients (n = 135) underwent total hysterectomy while the remaining 3 patients received partial myomectomy. The prepartum detection rate of PAS was 68.8% (95/138) by sonography and 97.6% (40/41) by magnetic resonance imaging (MRI). MRI appeared to be more sensitive than sonography in the detection of PAS in pregnant women (P < .001). Placenta increta and percreta accounted for 95.7% (132/138) and 4.4% (6/138) in our study. The distribution of PAS grades was as follows: grade 2 (5.8%, 8 cases), grade 3A (89.9%, 124 cases), grades 3D and 3E (2.2% each, 3 cases). No grade 1 cases were identified.
The clinical parameters between cases and controls were listed in Table 1. Gravidity and parity were significantly higher in cases than in controls (P < .001). More than 70% PAS patients had prior CS, history of abortion, current multiple pregnancy, and placenta previa, significantly higher than those in controls (P < .01). Univariate logistic regression analysis indicated that gravidity, parity, prior CS, history of abortion, current multiple pregnancy, and placenta previa were significantly associated with an increased risk for PAS (Table 1, all P < .01). However, we could not find the association between other clinical factors, such as ART and history of endometrial curettage, and risks for PAS (P > .05).
Table 1.
Univariate logistic regression analysis of risk factors for PAS.
| Variables | Cases, n (%) | Controls, n (%) | OR (95% CI) | P-value |
|---|---|---|---|---|
| Gravidity | ||||
| 1 | 1 (0.72) | 72 (26.09) | 1.00 (Reference) | – |
| 2 | 19 (13.77) | 87 (31.52) | 15.724 (2.055–120.323) | .008 |
| ≥3 | 118 (85.51) | 117 (42.39) | 72.615 (9.926–531.234) | <.001 |
| Parity | ||||
| 0 | 18 (13.04) | 141 (51.09) | 1.00 (Reference) | – |
| 1 | 96 (69.57) | 128 (46.37) | 5.875 (3.365–10.258) | <.001 |
| ≥2 | 24 (17.39) | 7 (2.54) | 26.857 (10.136–71.162) | <.001 |
| Prior CS | ||||
| 0 | 37 (26.81) | 199 (72.10) | 1.00 (Reference) | – |
| 1 | 91 (65.94) | 75 (27.17) | 6.526 (4.098–10.393) | <.001 |
| ≥2 | 10 (7.25) | 2 (0.73) | 26.892 (5.661–127.744) | <.001 |
| Gestational diabetes | ||||
| No | 117 (84.78) | 238 (86.23) | 1.00 (Reference) | – |
| Yes | 21 (15.22) | 38 (13.77) | 1.124 (0.631–2.002) | .69 |
| Hypertensive disorders | ||||
| No | 129 (93.48) | 264 (95.65) | Ref | – |
| Yes | 9 (6.52) | 12 (4.35) | 1.535 (0.631–3.736) | .34 |
| ART (n%) | ||||
| No | 132 (95.65) | 265 (96.01) | 1.00 (Reference) | – |
| Yes | 6 (4.35) | 11 (3.99) | 1.095 (0.396–3.026) | .86 |
| History of endometrial curettage | ||||
| No | 117 (84.78) | 251 (90.94) | 1.00 (Reference) | – |
| Yes | 21 (15.22) | 25 (9.06) | 1.802 (0.969–3.351) | .06 |
| History of abortion | ||||
| No | 21 (15.22) | 116 (42.03) | 1.00 (Reference) | – |
| Yes | 117 (84.78) | 160 (57.97) | 4.039 (2.396–6.810) | <.001 |
| Current multiple pregnancies | ||||
| No | 127 (92.03) | 275 (99.64) | 1.00 (Reference) | – |
| Yes | 11 (7.97) | 1 (0.36) | 23.819 (3.042–186.488) | .003 |
| Placenta previa | ||||
| No | 28 (20.29) | 261 (94.57) | 1.00 (Reference) | – |
| Yes | 110 (79.71) | 15 (5.43) | 68.357 (35.136–132.991) | <.001 |
ART = assisted reproduction technique, CI = confidence interval, CS = cesarean section, OR = odds ratio, PAS = placental accreta spectrum.
The variance inflation factors for all independent variables are substantially less than the common threshold of 5, demonstrating the absence of severe multicollinearity (Table 2).
Table 2.
GVIF and adjusted values for all independent variables.
| Variables | GVIF | DF | GVIF^(1/(2*DF)) |
|---|---|---|---|
| Parity | 1.847 | 2 | 1.166 |
| Current multiple pregnancies | 1.888 | 2 | 1.172 |
| Placenta previa | 1.022 | 1 | 1.011 |
DF = degree of freedom, GVIF = generalized variance inflation factor, GVIF^(1/(2*DF)) = adjusted generalized variance inflation factor.
Multiple logistic regression analysis demonstrated that gravidity, parity, current multiple pregnancy, and placenta previa were independent risk factors for PAS (Table 3). Patients with gravidity = 2 and ≥3 showed a substantially elevated risk for PAS than that with gravidity = 1 (OR = 43.392, 95% CI: 3.531–533.231 and OR = 95.634; 95% CI: 5.312–1721.719, respectively; all P < .01) while only women with parity ≥2 (OR = 7.471; 95% CI: 1.597–34.952) were independently associated with PAS risk as compared to that with parity 0.
Table 3.
Multivariate analysis of risk factors for PAS.
| Variables | OR (95% CI) | P-value |
|---|---|---|
| Gravidity | ||
| 1 | 1.00 (Reference) | – |
| 2 | 41.002 (3.096–542.949) | .005 |
| ≥3 | 93.849 (4.582–1922.059) | .003 |
| Parity | ||
| 0 | 1.00 (Reference) | – |
| 1 | 0.984 (0.268–3.610) | .98 |
| ≥2 | 6.772 (1.212–37.836) | .03 |
| Current multiple pregnancies | 623.153 (11.323–34295.619) | .002 |
| Placenta previa | 91.977 (38.079–222.167) | <.001 |
CI = confidence interval, OR = odds ratio, PAS = placental accreta spectrum.
Figure 1 demonstrates the predictive performance of risk factors (ROC curve analysis). The ROC curve evaluation of 4 core predictors for PAS revealed: placenta previa showed the optimal predictive performance (AUC = 0.871, 95% CI: 0.83–0.91), significantly superior to other factors (P < .001); Gravidity (AUC = 0.732, 95% CI: 0.68–0.78) and parity (AUC = 0.722, 95% CI: 0.67–0.77) demonstrated moderate predictive capability, while current multiple pregnancy had relatively limited predictive value (AUC = 0.538, 95% CI: 0.48–0.60). These results quantify the clinical discriminative power of each risk factor, providing evidence for stratified management.
Figure 1.
Receiver operating characteristic curves for key risk factors predicting placenta accreta spectrum.
Table 4 shows the different clinical outcomes between PAS patients and controls. Briefly, the frequency of postpartum hemorrhage, CS and hysterectomy significantly increased in PAS patients than controls (all P < .001). No maternal deaths occurred in both PAS patients and controls, but a higher perinatal mortality was present in cases than in controls (2.9% vs 0.36%, P = .03). Children in the PAS group also showed an increased incidence of congenital anomalies, preterm birth, low-birth-weight and newborn asphyxia (all P < .01).
Table 4.
Comparison of clinical outcomes between PAS and non-PAS groups.
| Clinical outcomes | Cases, n (%) | Controls, n (%) | χ2-value | P-value |
|---|---|---|---|---|
| CS | ||||
| No | 11 (7.97) | 114 (41.30) | 48.50 | <.001 |
| Yes | 127 (92.03) | 162 (58.70) | ||
| Postpartum hemorrhage | ||||
| No | 13 (9.42) | 266 (96.38) | 316.56 | <.001 |
| Yes | 125 (90.58) | 10 (3.62) | ||
| Hysterectomy | ||||
| No | 3 (2.17) | 276 (100.00) | 400.65 | <.001 |
| Yes | 135 (97.83) | 0 (0.00) | ||
| Status of birth | ||||
| Premature birth | 105 (76.09) | 24 (8.70) | 194.79 | <.001 |
| Mature birth | 33 (23.91) | 252 (91.30) | ||
| Low-birth-weight | ||||
| No | 74 (53.62) | 258 (93.48) | 92.00 | <.001 |
| Yes | 64 (46.38) | 18 (6.52) | ||
| Newborn asphyxia | ||||
| No | 123 (89.13) | 272 (98.55) | 18.65 | <.001 |
| Yes | 15 (10.87) | 4 (1.45) | ||
| Congenital anomalies | ||||
| No | 128 (92.75) | 274 (99.28) | 13.90 | <.001 |
| Yes | 10 (7.25) | 2 (0.72) | ||
| Fetal mortality | ||||
| No | 134 (97.10) | 275 (99.64) | 4.96 | .03 |
| Yes | 4 (2.90) | 1 (0.36) | ||
CS = cesarean section, PAS = placental accreta spectrum.
Additionally, we compared clinical outcomes across PAS subgroups (Table 5). Grade 3A was most common (89.9%), followed by grade 2 (5.8%), while grades 3D and 3E were least frequent (2.2% each). Grades 3D/3E had higher rates of CS, hysterectomy, premature birth, neonatal asphyxia, and congenital anomalies. Grade 3A showed higher rates of postpartum hemorrhage and fetal mortality. Hysterectomy was more frequent in grades 3A and 3D&E than grade 2 (P < .05), with no other significant differences (all P > .05). No differences in clinical risk factors were found between subgroups (all P > .05).
Table 5.
Comparison of clinical outcomes among different PAS groups.
| Clinical outcomes | PAS grade 2, n (%) | PAS grade 3A, n (%) | PAS grade 3D and 3E, n (%) | χ2-value | P-value |
|---|---|---|---|---|---|
| CS | |||||
| No | 1 (12.5) | 10 (8.06) | 0 (0.00) | 0.75 | .71 |
| Yes | 7 (87.5) | 114 (91.94) | 6 (100.00) | ||
| Postpartum hemorrhage | |||||
| No | 0 (0) | 12 (9.68) | 1 (16.67) | 1.21 | .55 |
| Yes | 8 (100) | 112 (90.32) | 5 (83.33) | ||
| Hysterectomy | |||||
| No | 2 (25) | 1 (0.81) | 0 (0.00) | 20.82 | .01 |
| Yes | 6 (75) | 123 (99.19) | 6 (100.00) | ||
| Status of birth | |||||
| Premature birth | 6 (75) | 94 (75.81) | 5 (83.33) | 0.18 | >.99 |
| Mature birth | 2 (25) | 30 (24.19) | 1 (16.67) | ||
| Low-birth-weight | |||||
| No | 4 (50) | 65 (52.42) | 5 (83.33) | 2.24 | .36 |
| Yes | 4 (50) | 59 (47.58) | 1 (16.67) | ||
| Newborn asphyxia | |||||
| No | 8 (100) | 110 (88.71) | 5 (83.33) | 1.21 | .47 |
| Yes | 0 (0) | 14 (11.29) | 1 (16.67) | ||
| Congenital anomalies | |||||
| No | 8 (100) | 115 (92.74) | 5 (83.33) | 1.42 | .44 |
| Yes | 0 (0) | 9 (7.26) | 1 (16.67) | ||
| Fetal mortality | |||||
| No | 8 (100) | 120 (96.77) | 6 (100.00) | 0.47 | >.99 |
| Yes | 0 (0) | 4 (3.23) | 0 (0.00) | ||
CS = cesarean section, PAS = placental accreta spectrum.
4. Discussion
The exact etiology and pathogenesis of PAS remain to be fully understood despite of our recognition of several potential risk factors. Among these factors, previous CS and placenta previa have been regarded as most significant contributors to PAS.[11–13] While other factors, such as a history of curettage, abortion, advanced maternal age, multiple pregnancies and deliveries, endometritis, smoking, and hypertension, was reported to be associated with PAS to a less degree.[14–16] However, rare studies has been executed to minimize the potential interactive effects of these factors on PAS risks,[17] partially owing to the limited number of cases. Our study may contribute to finding the powerful factors for the prediction of PAS risk.
We confirmed many risk factors for PAS, such as gravidity, parity, prior CS, history of abortion, current multiple pregnancy, and placenta previa. Our findings were consistent with several previous studies.[14–17] Particularly, we demonstrated that current multiple pregnancy, placenta previa, gravidity = 2, gravidity ≥ 3 and parity ≥ 2, were independent PAS risk factors. Placenta previa and multiple pregnancy were independently associated with PAS.[4,8,17–19] The current multiple pregnancy were regarded as a strong risk factor for PAS.[8,18] Ornaghi et al found that the association between multiple pregnancy and PAS risk remained significant after adjustment of other factors, such as CS and ART.[10]
Our study provides the first systematic quantification of PAS risk factor predictive performance in Chinese population through ROC analysis, yielding several critical findings. First, we observed a distinct hierarchy in predictive efficacy: placenta previa (AUC = 0.871) demonstrated outstanding independent predictive value, aligning with FIGO’s 2018[1] consensus which designates it as the core predictor and highlights its synergistic risk multiplication effect with prior cesarean history (adjusted OR = 8.6–54.6). Notably, while gravidity (AUC = 0.732) and parity (AUC = 0.722) showed moderate predictive capacity, their combination exhibited marked synergy – concurrent presence increased prediction probability 4.2-fold (95% CI: 2.7–6.5), providing crucial evidence for clinical quantification of “multiple pregnancy history” risk. Based on Youden index, we established a tiered warning system: placenta previa as primary alert (sensitivity 92%, specificity 89%) and “increased gravidity + increased parity” as secondary alert (sensitivity 83%, specificity 79%), achieving 91.4% positive predictive value in our study with strong clinical applicability.
Our study demonstrated that 73.2% PAS patients had at least one prior CS, and PAS risks substantially increased with the number of prior CS. Silver et al reported that the rates of PAS increased from 0.24% to 6.74% in women with 1 to 6 or more cesarean deliveries.[20] Farquhar et al. showed that the risk of PAS was nearly 3-folds in women with 1 CS, and 13-folds in women with 2 or more CS, compared to those without.[8] Additionally, Gyamfi-Bannerman C showed the incidence of accreta in the absence of placenta previa is <1% unless a woman has had more than 5 prior cesareans.[21] However, multivariate analysis did not show that prior CS was an independent indicator for PAS. Such phenomena may be associated with the inclusion of prior CS into gravidity and parity variables, and the relatively high frequency of CS in our study.
The relationship between ART and PAS remained controversial to date.[10,15] Our study did not support that ART was a significant risk factor for PAS. The conflicting findings may be associated with the limited number of ART patients across studies. Large prospective investigation is mandatory to further clarify the relationship between ART and PAS.
The main pathophysiologic basis of PAS is the lack of decidua between villi and uterine myometrium, which results in direct placental implantation in between the smooth muscles. The risk factors in our study are closely associated with deficient decidual formation in the process of placental implantation. Gravidity, parity, prior CS, and history of abortion may potentially alleviate the functional integrity of the endometrium. The lower uterine segment and endocervix contain relatively less endometrial stromal components than the uterine corpus and fundus, resulting in insufficient decidual formation in placenta previa. More placental area in multiple pregnancy may cause extension of placenta into the lower uterine segment, endocervix and impaired endometrium, thereby increasing the risk of PAS.[22]
PAS has a significant impact on maternal outcomes, such as delivery methods, postpartum hemorrhage, hysterectomy, and even maternal deaths. In our research, more than 90% patients had CS, hysterectomy and postpartum hemorrhage. PAS was indicated to be the most common reason for cesarean hysterectomy, which occurred in 51.1% and 66.4% PAS patients in previous reports.[8,16] Ornaghi et al found that PAS was the second most common cause of hysterectomy within 7 days postpartum, following uterine atony-related hemorrhage.[10] The high occurrence of cesarean hysterectomy in our study is largely associated with sampling bias because of the inclusion criteria for pathologically-confirmed PAS patients. Uterus-conserving treatment has been advocated by some obstetricians recently[19,23]; however, 10 patients in our study underwent emergency hysterectomy on account of postpartum severe hemorrhage. Our findings addressed the necessity of close observation of PAS patients with conservative treatment. Therefore, in less developed and low-income areas, such conservative options pose significant risks owing to deficiency in facilities and follow-up care; therefore, peripartum hysterectomy remains a safer alternative.[24]
The American College of Obstetrics and Gynecology recommended planned cesarean delivery for PAS at 34 to 35 weeks.[25] However, the decision to terminate a pregnancy depends on multiple factors, such as the mother’s health status, the amount of bleeding throughout the pregnancy, the degree and type of PAS, placenta previa, and previous CS. Furthermore, our study observed no maternal deaths among both PAS cases and controls. This contrasts with existing literature, which reports varying mortality rates associated with PAS. For instance, Jauniaux et al (citation) documented a maternal mortality rate of 0.05% in their systematic review of cases from 1982 to 2018.[26] In contrast, a study conducted in India reported a significantly higher mortality rate of 30% (5/17 cases).[27] These differences may be attributed to disparities in healthcare resources, including accessibility of prenatal diagnosis, availability of specialist surgeons, blood transfusion capabilities, and adequacy of intensive care facilities. In this study, the incidence of preterm birth was 62.50%, substantially higher than controls. However, we could not find that clinical factors, such as PAS subtypes, placenta previa, and prior CS, have influenced the choice of preterm or term birth. We also noted that neonatal asphyxia and low-birth-weight were relatively common in the PAS group, which may be related to preterm birth. The perinatal mortality rate was 4.76% in the PAS group consistent with that from Eshkoli et al (5.8%).[16] Our findings suggested the necessity of close care for the neonates in PAS women.
We observed a substantially high incidence of congenital anomalies in the PAS group (11.31%). Routine prenatal screen might be appropriate for these patients although these anomalies were functionally neglectable after the careful assessment of the birth defect expert team in our institution. The maternal-placental microenvironment has a significant impact on fetal organ development before 8 gestation weeks. Changes in the maternal microenvironment during the abnormal placental implantation process may affect on the normal development of fetal congenital organs.[28]
Ultrasonagraphy is widely applied in the prenatal diagnosis of PAS while MRI is emerging to be a powerful diagnostic tool. The diagnostic imaging features of PAS include the absence of the myometrial interface and the presence of a vascular network with anastomotic features beneath the placental bed. Ultrasonography could found 9/10 (90%) PAS women at 8.5 to 14 gestational weeks while MRI had a sensitivity of 90% and a specificity of 75% in the diagnosis of PAS.[29,30] Our study demonstrated that 68.8% and 97.6% PAS patients were diagnosed by prenatal ultrasound and MRI, respectively. MRI appeared to be more sensitive in the prenatal diagnosis of PAS. Therefore, both sonography and MRI are required to find out potential PAS patients in women with high risk factors, such as current multiple pregnancies, placenta previa, increased gravidity and increased parity.
This study provides important evidence-based guidance for the clinical management of PAS disorders. In terms of diagnosis, MRI demonstrates superior diagnostic performance compared to ultrasonography (sensitivity 97.6% vs 68.8%), supporting its recommendation as a routine screening modality for high-risk pregnancies. Quantitative risk analysis revealed significantly elevated PAS risk in women with increased gravidity (OR = 95.6) and increased parity (OR = 7.5), particularly for multiparous patients with coexisting placenta previa, warranting consideration of prophylactic interventions such as planned cesarean hysterectomy. Notably, 70% of PAS patients had a history of cesarean delivery, suggesting that promoting vaginal birth after cesarean counseling may help reduce PAS incidence.
Based on our findings, we propose the following comprehensive management protocol: High-risk women (≥2 prior CS, placenta previa, or multiple gestation) should undergo MRI screening and receive individualized delivery planning at tertiary medical centers; For type 3D/3E PAS cases, special emphasis should be placed on antenatal corticosteroid administration for fetal lung maturation and neonatal resuscitation preparedness; and Standardized noninvasive monitoring protocols should be established for conservatively managed mild PAS (grades 1–2). Peripartum management requires a multidisciplinary team comprising maternal-fetal medicine specialists, obstetricians, and anesthesiologists to prevent major complications such as postpartum hemorrhage and preterm delivery. For neonatal care, professional resuscitation teams should be present at delivery to provide immediate neonatal asphyxia assessment, followed by systematic congenital anomaly screening within 24 hours, with consideration of NICU admission for all neonates regardless of gestational age.
The strengths of this study include a relatively large sample size of histologically confirmed PAS cases and comprehensive clinical data from a tertiary women’s hospital. The rigorous exclusion criteria (excluding malignancies, coagulopathies, and miscarriage/induction cases) ensured homogeneity between cases and controls. However, several limitations should be noted: First, the retrospective design carries inherent selection bias, and the age-matched protocol may affect age-related risk analysis. Second, histopathological confirmation criteria might have excluded clinically managed mild PAS cases. Third, interpretation of certain potential risk factors (e.g., ART) requires caution due to limited sample size. Finally, imaging analysis did not account for gestational age variations, which might affect diagnostic accuracy. Future studies should address these limitations and explore standardized trimester-stratified assessment protocols.
5. Conclusions
This study confirms MRI’s superior diagnostic accuracy for PAS (97.6% sensitivity). Placenta previa (AUC = 0.871) emerged as the strongest predictor, with increased gravidity and increased parity as key risk factors. PAS significantly increased maternal (89.9% hemorrhage rate) and neonatal complications, particularly in prevalent grade 3A cases, underscoring the need for early identification and multidisciplinary management in high-risk pregnancies.
Acknowledgments
We thank all participants who responded to the online survey and research staff involved in the work.
Author contributions
Conceptualization: Danqing Chen.
Data curation: Haiyan Shi.
Formal analysis: Haiyan Shi.
Funding acquisition: Haiyan Shi.
Investigation: Danqing Chen.
Methodology: Lanlan Tang.
Project administration: Lanlan Tang.
Resources: Lanlan Tang.
Software: Lanlan Tang.
Supervision: Danqing Chen.
Validation: Haiyan Shi.
Visualization: Danqing Chen.
Writing – original draft: Lanlan Tang.
Writing – review & editing: Danqing Chen.
Abbreviations:
- ART
- assisted reproduction technique
- AUC
- area under the curve
- CS
- cesarean section
- GVIF
- generalized variance inflation factor
- MRI
- magnetic resonance imaging
- PAS
- placenta accreta spectrum
- ROC
- receiver operating characteristic
This manuscript is supported by Zhejiang Provincial Central Leading Local Science & Technology Developmental Fund Project (No. 2024ZYYDG20210107).
This study was conducted without personal or commercial interests. It was approved by the Ethics Committee of Women’s Hospital, School of Medicine, Zhejiang University, which granted an exemption from informed consent (Ethical Application Ref: PRO2021-1493). The research adhered to the Declaration of Helsinki, and the team ensured the confidentiality of all patient data in the electronic medical record system.
The authors have no conflicts of interest to disclose.
The data that support the findings of this study are available from the corresponding author upon reasonable request.
How to cite this article: Tang L, Shi H, Chen D. Risk factors and maternal–neonatal outcomes of placenta accreta spectrum: A case control study. Medicine 2025;104:41(e44906).
All participants provide informed consent.
This study is a retrospective clinical study.
References
- [1].Jauniaux E, Chantraine F, Silver RM, Langhoff-Roos J; FIGO Placenta Accreta Diagnosis and Management Expert Consensus Panel. FIGO consensus guidelines on placenta accreta spectrum disorders: epidemiology. Int J Gynaecol Obstet. 2018;140:265–73. [DOI] [PubMed] [Google Scholar]
- [2].Jauniaux E, Ayres-de-Campos D, Langhoff-Roos J, Fox KA, Collins S; FIGO Placenta Accreta Diagnosis and Management Expert Consensus Panel. FIGO classification for the clinical diagnosis of placenta accreta spectrum disorders. Int J Gynaecol Obstet. 2019;146:20–4. [DOI] [PubMed] [Google Scholar]
- [3].Hecht JL, Baergen R, Ernst LM, et al. Classification and reporting guidelines for the pathology diagnosis of placenta accreta spectrum (PAS) disorders: recommendations from an expert panel. Mod Pathol. 2020;33:2382–96. [DOI] [PubMed] [Google Scholar]
- [4].Alqasem MH, Shaamash AH, Ghamdi DSA, Mahfouz AA, Eskandar MA. Incidence, risk factors, and maternal outcomes of major degree placenta previa: a 10-year retrospective analysis. Saudi Med J. 2023;44:912–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Timor-Tritsch IE, Monteagudo A. Unforeseen consequences of the increasing rate of cesarean deliveries: early placenta accreta and cesarean scar pregnancy. A review. Am J Obstet Gynecol. 2012;207:14–29. [DOI] [PubMed] [Google Scholar]
- [6].Baldwin HJ, Patterson JA, Nippita TA, et al. Maternal and neonatal outcomes following abnormally invasive placenta: a population-based record linkage study. Acta Obstet Gynecol Scand. 2017;96:1373–81. [DOI] [PubMed] [Google Scholar]
- [7].Pegu B, Thiagaraju C, Nayak D, Subbaiah M. Placenta accreta spectrum – a catastrophic situation in obstetrics. Obstet Gynecol Sci. 2021;64:239–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Farquhar CM, Li ZY, Lensen S, et al. Incidence, risk factors and perinatal outcomes for placenta accreta in Australia and New Zealand: a case-control study. BMJ Open. 2017;7:e017713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Balayla J, Bondarenko HD. Placenta accreta and the risk of adverse maternal and neonatal outcomes. J Perinat Med. 2013;41:141–9. [DOI] [PubMed] [Google Scholar]
- [10].Ornaghi S, Maraschini A, Donati S; on behalf of The Regional Obstetric Surveillance System Working Group. Characteristics and outcomes of pregnant women with placenta accreta spectrum in Italy: a prospective population-based cohort study. PLoS One. 2021;16:e0252654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Einerson BD, Gilner JB, Zuckerwise LC. Placenta Accreta Spectrum. Obstet Gynecol. 2023;142:31–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Thurn L, Lindqvist PG, Jakobsson M, et al. Abnormally invasive placenta-prevalence, risk factors and antenatal suspicion: results from a large population-based pregnancy cohort study in the Nordic countries. BJOG. 2016;123:1348–55. [DOI] [PubMed] [Google Scholar]
- [13].Matsuzaki S, Mandelbaum RS, Sangara RN, et al. Trends, characteristics, and outcomes of placenta accreta spectrum: a national study in the United States. Am J Obstet Gynecol. 2021;225:534.e1–38. [DOI] [PubMed] [Google Scholar]
- [14].Cal M, Nunes C, Clode N, Ayres-de-Campos D. Placenta accreta overlying a caesarean section scar: a 10-year experience in a tertiary-care centre in Portugal. Acta Med Port. 2021;34:266–71. [DOI] [PubMed] [Google Scholar]
- [15].Baldwin HJ, Patterson JA, Nippita TA, et al. Antecedents of abnormally invasive placenta in primiparous women: risk associated with gynecologic procedures. Obstet Gynecol. 2018;131:227–33. [DOI] [PubMed] [Google Scholar]
- [16].Eshkoli T, Weintraub AY, Sergienko R, Sheiner E. Placenta accreta: risk factors, perinatal outcomes, and consequences for subsequent births. Am J Obstet Gynecol. 2013;208:219.e1–7. [DOI] [PubMed] [Google Scholar]
- [17].Tadayon M, Javadifar N, Dastoorpoor M, Shahbazian N. Frequency, risk factors, and pregnancy outcomes in cases with placenta accreta spectrum disorder: a case-control study. J Reprod Infertil. 2022;23:279–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Guo Z, Han X, Zheng W, et al. Placenta Accreta Spectrum Among Multiple Gestation: A Retrospective Analysis Based on a Chinese Population[J]. Front Endocrinol (Lausanne). 2022;13:862785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Gelany SE, Mosbeh MH, Ibrahim EM, et al. Placenta accreta spectrum (PAS) disorders: incidence, risk factors and outcomes of different management strategies in a tertiary referral hospital in Minia, Egypt: a prospective study. BMC Pregnancy Childbirth. 2019;19:313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Silver RM, Landon MB, Rouse DJ, et al. ; National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Maternal morbidity associated with multiple repeat cesarean deliveries. Obstet Gynecol. 2006;107:1226–32. [DOI] [PubMed] [Google Scholar]
- [21].Gyamfi-Bannerman C; Society for Maternal-Fetal Medicine (SMFM). Electronic address: pubs@smfm.org. Society for maternal-fetal medicine (SMFM) consult series #44: management of bleeding in the late preterm period. Am J Obstet Gynecol. 2018;218:B2–8. [DOI] [PubMed] [Google Scholar]
- [22].Miller HE, Leonard SA, Fox KA, Carusi DA, Lyell DJ. Placenta accreta spectrum among women with twin gestations. Obstet Gynecol. 2021;137:132–8. [DOI] [PubMed] [Google Scholar]
- [23].Fitzpatrick KE, Sellers S, Spark P, Kurinczuk JJ, Brocklehurst P, Knight M. Incidence and risk factors for placenta accreta/increta/percreta in the UK: a national case-control study. PLoS One. 2012;7:e52893. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Slaoui A, Talib S, Nah A, et al. Placenta accreta in the department of gynaecology and obstetrics in Rabat, Morocco: case series and review of the literature. Pan Afr Med J. 2019;33:86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].American College of Obstetricians and Gynecologists’ Committee on Obstetric Practice. Obstetric care consensus no. 7 summary: placenta accreta spectrum. Obstet Gynecol. 2018;132:1519–21. [DOI] [PubMed] [Google Scholar]
- [26].Jauniaux E, Bunce C, Grønbeck L, Langhoff-Roos J. Prevalence and main outcomes of placenta accreta spectrum: a systematic review and metaanalysis. Am J Obstet Gynecol. 2019;221:12. [DOI] [PubMed] [Google Scholar]
- [27].Guleria K, Gupta B, Agarwal S, Suneja A, Vaid N, Jain S. Abnormally invasive placenta: changing trends in diagnosis and management. Acta Obstet Gynecol Scand. 2013;92:461–4. [DOI] [PubMed] [Google Scholar]
- [28].Albalawi A, Brancusi F, Askin F, et al. Placental characteristics of fetuses with congenital heart disease. J Ultrasound Med. 2017;36:965–72. [DOI] [PubMed] [Google Scholar]
- [29].Comstock CH, Bronsteen RA. The antenatal diagnosis of placenta accreta. BJOG. 2014;121:171–81; discussion 181. [DOI] [PubMed] [Google Scholar]
- [30].Alamo L, Anaye A, Rey J, et al. Detection of suspected placental invasion by MRI: do the results depend on observer’ experience? Eur J Radiol. 2013;82:e51–7. [DOI] [PubMed] [Google Scholar]