Abstract
Objective
The purpose of this study is to evaluate the diagnostic performance of MRI parameters to predict adverse maternal peripartum outcomes in pregnant females at high-risk for placenta accreta spectrum (PAS) disorder.
Methods and materials
This retrospective study evaluated 60 pregnant females who underwent MRI for placental assessment. MRI studies were reviewed by a radiologist blinded to all clinical data. MRI parameters were compared with five maternal outcomes: severe bleeding, cesarean hysterectomy, prolonged operation time, need for blood transfusion, and need for intensive care unit (ICU) admission. The MRI findings were associated with pathologic and/or intraoperative findings for PAS.
Results
The study identified 46 cases of PAS disorder and 16 cases of placenta percreta. The agreement between the radiologist impression of PAS disorder and the intraoperative/histological findings was substantial (0.67, p < 0.001), and almost perfect for the presence of placenta percreta (0.87, p < 0.001). The presence of a placental bulge was highly associated with placenta percreta, with sensitivity of 87.5% and specificity of 90.9%. The MRI signs that associated with more maternal outcomes were myometrial thinning, with significant odds ratio for severe blood loss (20.2), hysterectomy (4.0), need for blood transfusion (4.8) and prolonged surgery time (4.9), and uterine bulging, with significant odds ratio for severe blood loss (11.9), hysterectomy (34.0), ICU admission (5.0), and need for blood transfusion (4.8).
Conclusion
MRI signs significantly correlated with invasive placenta and were independently associated with adverse maternal outcomes. The presence of a placental bulge was highly accurate in predicting placenta percreta.
Advances in knowledge
First study to evaluate the strength of the association between individual MRI signs and five adverse maternal outcomes. Conclusions support published MRI signs associated with placental invasion, especially regarding the value placental bulging in predicting placenta percreta.
Introduction
Placenta accreta spectrum (PAS) is defined as the abnormal placental adhesion to the uterine wall, being classified according to the degree of invasiveness as: placenta accreta, when it adheres to the myometrium without interposing decidua; placenta increta, when placental tissue invades the myometrium; and placenta percreta when it penetrates through the uterine serosa, possibly invading adjacent organs. 1–5
PAS is a relatively new pathology, described by the first time in 1937, with a progressive increase in its incidence in recent decades, a fact attributed to the increase in cesarean section rates, a procedure that has transitioned from urgent and highly lethal to a safe and elective surgery. 3 The incidence of PAS in the last years has increased more sharply in developed countries; in the USA it has risen from 1 case per 4.027 births in the 1970s to 1 case per 533 births in the 2000s. 4,5
PAS is associated with significant morbidity and mortality and has replaced uterine atony as the leading indication for cesarean hysterectomy in developed countries. It carries increased risk of massive peripartum bleeding and related complications such as hypovolemic shock, consumptive coagulopathy, multiorgan failure, and even death. 5,6 The reported blood transfusion rate is 82% in placenta percreta and increta and 71% in cases of placenta accreta, with the need of cesarean hysterectomy in up to 89% of the patients with PAS disorder. 7
This disorder requires accurate antenatal diagnosis and a multidisciplinary approach in specialized centers. 8 Ultrasonography remains as the primary diagnostic modality for antenatal diagnosis, given its low cost and high availability. Placental Magnetic Resonance Imaging (MRI) is emerging as complementary diagnostic tool when ultrasound evaluation is inconclusive, especially in cases of posterior placenta and in obese patients. 9–11 MRI also offers better performance predicting the depth and topography of placental invasion, allowing better treatment planning. 11,12
It must be noted, however, that the accuracy of placental MRI depends on the levels of experience of the radiologists, 13 a conclusion supported by a recent study from our practice, showing that specific experience in placental MRI significantly improved diagnostic performance. 14
In this study, we retrospectively investigate the association between MRI parameters of placental invasion and adverse events in pregnant females at high risk for PAS disorders, aiming to identify the signs that most commonly predict those outcomes, and also to facilitate the creation of a standardized template for MRI reports.
Methods and materials
Study population
The study is designed as a retrospective, bicenter observational study of 68 females who consecutively underwent MRI for placental evaluation between 2016 and 2020 in a public university hospital and a private hospital, both tertiary centers with specialized care for high-risk pregnancies. Ethical approval was obtained from Research Ethics Committee at HCPA and HMV (CAAE 19927919.0.0000.5327 and 36987320.6.0000.5330, respectively). Informed consent was waived because of the retrospective nature of the study and the analysis used anonymous clinical data. Written confidentiality agreement was obtained from all researchers involved.
The indications for MRI evaluation in our series included those cases with technically difficult ultrasound examinations (such as obese patients and posterior placenta), patients with placenta previa with indeterminate US findings and patients with sonographic findings suspicious for placental invasion (to evaluate the extent of placental invasion and the relationship of the placenta to surrounding structures). Eight patients were not included in the analysis due to lack of adequate surgical description or histopathological study.
MRI protocol
MRI scans were performed using a 1.5 T unit (Magnetom Aera, Siemens Healthcare, Erlangen, Germany or Philips Achieva, Philips Healthcare, Best, The Netherlands). All patients were instructed to have a partially filled bladder before entering the scanning room. No IV contrast medium was administered.
The MRI protocols were similar in both centers and all patients were evaluated with pelvic phased-array coil. The image MRI protocol included axial, sagittal and coronal planes 2D T 2-weighted single-shot turbo spin echo (T2W - SSTSE) images and true fast imaging with steady-state precession (True – FISP) along the axial, sagittal, and coronal planes of the uterus. T 1-weighted (T1W) TSE sequence with fat suppression of the entire uterus was obtained in the axial plane. When placenta previa was detected, additional high-resolution axial, and sagittal T2W nonfat-suppressed sequences were performed with a smaller field of view to better evaluate the interface between the bladder and uterus. Axial Diffusion-weighted imaging (DWI) of the entire placenta was performed. The images were acquired at slice thickness of 3 or 4 mm.
Image interpretation
All MRI scans were stored in a DICOM viewer and were anonymized. The studies were evaluated by a radiologist with 8 years of experience in abdominopelvic imaging, including placental MRI, having reviewed around 200 placental cases at the time of the study. The examiner was blinded to all clinical data of the patients, including patient’s age and gestational age, as well as any ultrasound diagnosis, surgical and histopathological findings. The radiologist recorded the presence or absence of each the following imaging findings associated with PAS disorder, described in the published literature 12–22 :
Heterogeneous placenta - refers to overall disorder and inhomogeneity of placental tissue due to fibrin deposition, intraplacental hemorrhage and abnormal vascularity.
Irregular contours and rounded edges - rounded placental edges and irregular or lobulated placental contours along the placental-uterine interface are suggestive features of PAS disorders.
Low attenuation T2 bands – also described as fibrin deposition, consists of one or more tortuous lines of low T2 signal intensity, thicker than normal placental septa and crossing the perpendicular axis of the placenta, that remains hypointense on steady-state free precession images. (Figure 1)
Figure 1.
37-year-old gravida 3 para two who had undergone one prior cesarean delivery, with pathologically proven placenta percreta. (a) Sagittal T 2-weighted image shows thick and tortuous dark bands coursing through the placental parenchyma inferiorly and posteriorly (white arrows). The corresponding steady-state free precession image (b) reveals hyperintensity of the lower dark bands (void arrowhead), confirming the presence of ectatic intraparenchymal vessels, while the more posterior band remains hypointense (white arrowhead), consistent with fibrin deposition.
Intraplacental hypervascularity - dilated (> 0,6 mm) and tortuous flow-voids in T2 weighted sequences with high signal intensity on steady-state free precession sequences. (Figure 1B)
Abnormal vascularization of the placental bed - engorged vessels following the retroplacental contour, sometimes seen in conjunction with neovascularization in the adjacent organs.
Retroplacental myometrial line disruption - loss of the T2 hypointense linear interface between the placental bed and underlying myometrium. (Figure 2)
Figure 2.
DWI in a 43-year-old gravida one without accretism. (a) Axial T 2-weighted image shows loss of retroplacental hypointense line and placental bulge into the myometrium at right posterolateral aspect of uterus (white arrows), suspicious for myometrial invasion. (B) Corresponding axial DWI image clearly depicts normal regular contour of the placenta, with high-signal intensity in comparison to subjacent miometrium and no evidence of invasion.
Myometrial thinning - focal thinning of the myometrium under the placental bed to less than 1 mm or even becoming imperceptible. (Figure 3)
Figure 3.
40-year-old gravida 4 para 3, who had undergone three prior cesarean deliveries. Axial T 2-weighted image depicts outward bulge of placental tissue and absence of miometrial signal at right anterolateral aspect of uterus, suspicious for placental invasion. Placenta percreta was confirmed at delivery.
Uterine bulging - is defined by an abnormal outward bulge of placental tissue, most likely resulting from placental tethering which hinders the normal migration of placenta during pregnancy, and is associated with deeper myometrial invasion. (Figure 4)
Figure 4.
28-year-old gravida 5 para 4, who had undergone four prior cesarean deliveries, with placenta percreta and bladder involvement. Sagittal T 2-weighted image shows focal uterine bulge and disruption of hypointense signal of the anterosuperior bladder wall (white arrow), consistent with invasion. In contrast, the posterior bladder wall shows preservation of the normal hypointense signal (arrowheads).
The examiner also recorded a final conclusion as to whether PSA disorder was present or absent and also regarding the presence or absence of placenta percreta in each case.
Additional placental features were also analyzed, including the presence of placenta previa (when the placental tissue covered either partially or completely the internal cervical os) and intraplacental hemorrhage (evidenced by foci of hyperintense T1 signal inside the placental tissue).
Standard of reference
The surgical diagnosis of PAS was confirmed by reviewing the operative reports for findings of abnormal placental adhesion and/or evidence of gross placental invasion. Histological examination was performed in all hysterectomy cases, and histopathologic reports were evaluated for the presence or absence of invasion. In cases of disagreement between the surgical and pathological reports, the maximum depth of placental invasion was recorded.
Maternal outcomes
The following adverse maternal outcomes were recorded: surgical technique (conservative surgery or cesarean hysterectomy); severe intraoperative bleeding, defined as estimated blood loss >2000 ml, in accordance to the guidelines from the Royal College of Obstetricians and Gynaecologists (RCOG) 23 ; admission to intensive care unit (ICU); need for blood transfusion (number of blood units received); and prolonged operative time (defined as surgical time >60 min from delivery to skin closure).
Statistical analysis
The normality of the variables was assessed using the Kolmogorov-Smirnov test. Quantitative variables are presented with mean and standard deviation. The values obtained were compared using the Student’s t test. Categorical data were expressed in frequencies and percentages. Pearson’s Chi-square test, Chi-Square with Yates correction and Fisher’s exact test were used to assess the relationship between the categorical variables, followed by the use of standardized adjusted residuals to detect categories with a frequency greater than expected. Univariate logistic regression was used to estimate the odds ratios (OR).
Simple κ statistics were calculated to determine agreement between the radiologist impression and surgical and/or pathologic reports regarding the presence of PAS disorder and placenta percreta. The diagnostic performance (sensitivity, specificity, positive predictive value, negative predictive value and accuracy) was calculated with the respective 95% confidence interval (95% CI). The threshold for statistical significance was p < 0.05. All statistical analyses were performed using the SPSS software version 20.0 for Windows (SPSS Inc.; Chicago IL, USA).
Results
Study group
Sixty-eight consecutive pregnant females who were submitted to placental MRI examination were initially included in the study, and eight patients were excluded from the analysis due to the absence of surgical impression and/or pathological description regarding the presence of PAS.
The study identified 46 cases of PAS disorder (76.7%) and a final diagnosis of placenta percreta was made in 16 cases (26.7%). In 14 cases, there was no evidence of invasive placenta in both the surgical and pathological reports. There was a statistically significant difference in maternal age between females diagnosed with PAS and those with no PAS disorder. The maternal age (mean, ± SD) in the group with no PAS disorder was 32.6 ± 4.8 years, while in the group with PAS disorder it was 36.0 ± 4.3 years (p = 0.013). No significant difference was observed regarding gestational age at the time of MRI examination, parity, number of c-sections or prior uterine surgery between both groups (Table 1 ).
Table 1.
Demographic and Clinical Characteristics of Women With and Without Definitive Diagnosis of PAS Disorder
| Characteristic | No PAS (n = 14) | PAS (n = 46) | p |
|---|---|---|---|
| Maternal age (y), mean ± SD | 32.6 ± 4.8 | 36.0 ± 4.3 | 0.013 |
| Gestational age at time of MRI (wk), mean ± SD | 30.9 ± 2.4 | 31.8 ± 2.6 | 0.234 |
| Parity: 1 2 ≥ 3 | 1 (7.1%) 4 (28.6%) 9 (64.3%) | 4 (8.7%) 12 (26.1%) 30 (65.2%) | 0.972 |
| Number of C-sections: 0 1 2 ≥ 3 | 4 (28.6%) 5 (35.7%) 2 (14.3%) 3 (21.4%) | 15 (32.6%) 15 (32.6%) 10 (21.7%) 6 (13.0%) | 0.828 |
| Prior uterine surgery (except C-section) | 2 (14.3%) | 16 (34.8%) | 0.192 |
| Hysterectomy | 1 (7.1%) | 30 (65.2%) | <0.001 |
| Blood loss ≥ 2000 ml | - | 14 (30.4%) | 0.026 |
| Need for ICU stay | - | 8 (17.4%) | 0.179 |
| Need for blood transfusion | - | 14 (30.4%) | 0.026 |
| Prolonged surgery time a | 6 (60.0%) | 31 (81.6 %) | 0.206 |
surgery time data was lacking for 12 patients
MRI evaluation
There was substantial agreement between the radiologist impression of PAS disorder and the intraoperative/histological findings (0.67, p < 0.001), and almost perfect agreement for the presence of placenta percreta (0.87, p < 0.001).
Table 2 shows the diagnostic performance for each MRI parameter. Of nine MRI parameters evaluated, five showed significant differences between the PAS and no PAS groups. Radiologist impression of invasion was the parameter most strongly associated with PAS disorder (p < 0.001). Uterine bulging showed sensitivity of 39.1% and specificity of 100.0% for the presence of PAS disorder. When compared to the subgroup of patients with placenta percreta, it showed sensitivity of 87.5% and specificity of 90.9%. Similarly, the radiologist impression of placenta percreta has a sensitivity of 87.5% and specificity of 97.7% for the diagnosis of placental percretism.
Table 2.
Diagnostic Accuracy of Individual MRI Signs for Placental Adhesion Disorder
| MRI Parameter | p | Sen | Esp | PPV | NPV | Accuracy (95% CI) |
|---|---|---|---|---|---|---|
| Heterogeneous placenta | 0.100 | 19.6 | 100.0 | 100.0 | 27.5 | 38.3 (26.1–51.8) |
| Irregular contours and rounded edges | 0.734 | 76.1 | 28.6 | 77.8 | 26.7 | 65.9 (51.6–76.9) |
| Low attenuation T2 bands | 0.009 | 97.8 | 28.6 | 81.8 | 80.0 | 81.7 (69.6–90.5) |
| Abnormal vascularization of the placental bed | 0.002 | 89.1 | 35.7 | 82.0 | 50.0 | 76.7 (62.1–85.3) |
| Intraplacental hypervascularity | 0.107 | 100.0 | 28.6 | 82.1 | 100.0 | 83.3 (71.5–91.7) |
| Retroplacental myometrial line disruption | 0.001 | 89.1 | 57.1 | 81.6 | 45.5 | 81.7 (69.6–90.5) |
| Myometrial thinning | 0.068 | 58.7 | 71.4 | 87.1 | 34.5 | 61.7 (48.2–73.9) |
| Uterine bulging | 0.006 | 39.1 | 100.0 | 100.0 | 33.3 | 90.0 (79.5–96.2) |
| Radiologist impression of invasive placenta (any degree) | <0.001 | 100.0 | 57.1 | 88.5 | 100 | 90.0 (79.5–96.2) |
| Uterine bulginga | <0.001 | 87.5 | 90.9 | 77.8 | 95.2 | 90.0 (79.5–96.2) |
| Radiologist impression of placenta percreta a | <0.001 | 87.5 | 97.7 | 93.3 | 95.6 | 95.0 (86.1–99.0) |
correlation with placenta percreta
Maternal outcomes
A statistically significant difference was identified in relation to the need for cesarean hysterectomy, severe blood loss and need for blood transfusion between patients with and without invasive placenta. In our series, thirty (65.2%) females with PAS disorder were submitted to cesarean hysterectomy, while only a single patient (7.1%) with no PAS disorder underwent the same procedure (p < 0.001). Severe blood loss (>2000 ml) was found in 14 patients (30.4%) with PAS disorder, and in no patients without this diagnosis (p = 0.026). Blood transfusion was made in 14 patients (30.4%) with PAS and in no patients without invasive placenta (p = 0.026). There was no significant difference in relation to the need for ICU stay and prolonged surgery time between both groups.
Association of the MRI parameters recorded with adverse maternal peripartum events is detailed in Table 3. Severe blood loss and the need for blood transfusion during surgery showed significant association with three MRI signs: heterogeneous placenta, myometrial thinning and uterine bulging. Cesarean hysterectomy was significantly associated with heterogeneous placenta, low attenuation T2 bands, intraplacental hypervascularity, myometrial thinning and uterine bulging.
Table 3.
Association of MRI Parameters with Adverse Maternal Events
| N | Severe Blood Loss (> 2000 ml) - N (%) | p | Hysterectomy - N (%) | p | ICU stay- N (%) | p | Blood Transfusion - N (%) | p | Prolonged Surgery Time a - N (%) | p | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Placenta previa | |||||||||||
| Yes | 25 | 10 (40) | 0.023 | 16 (64.0) | 0.176 | 5 (20.0) | 0.259 | 10 (40.0) | 0.023 | 18 (81.8) | 0.709 |
| No | 35 | 4 (11.4) | 15 (42.9) | 3 (8.6) | 4 (11.4) | 19 (73.1) | |||||
| Intraplacental hemorrhage | |||||||||||
| Yes | 8 | 5 (62.5) | 0.013 | 7 (87.5) | 0.053 | 2 (25.0) | 0.288 | 3 (37.5) | 0.374 | 7 (87.5) | 0.661 |
| No | 52 | 9 (17.3) | 24 (46.2) | 6 (11.5) | 11 (21.2) | 30 (75.0) | |||||
| Heterogeneous placenta | |||||||||||
| Yes | 9 | 7 (77.8) | <0.001 | 8 (88.9) | 0.027 | 3 (33.3) | 0.090 | 5 (55.6) | 0.025 | 6 (75.0) | 0.999 |
| No | 51 | 7 (13.7) | 23 (45.1) | 5 (9.8) | 9 (17.6) | 31 (77.5) | |||||
| Irregular contours and rounded edges | |||||||||||
| Yes | 45 | 13 (28.9) | 0.155 | 24 (53.3) | 0.881 | 7 (15.6) | 0.666 | 12 (26.7) | 0.483 | 29 (82.9) | 0.140 |
| No | 15 | 1 (6.7) | 7 (46.7) | 1 (6.7) | 2 (13.3) | 8 (61.5) | |||||
| Low attenuation T2 bands | |||||||||||
| Yes | 55 | 14 (25.5) | 0.329 | 31 (56.4) | 0.022 | 8 (14.5) | 0.999 | 14 (25.5) | 0.329 | 35 (79.5) | 0.221 |
| No | 5 | - | - | - | - | 2 (50.0) | |||||
| Abnormal vascularization of the placental bed | |||||||||||
| Yes | 49 | 14 (28.6) | 0.053 | 28 (57.1) | 0.145 | 7 (14.3) | 0.999 | 12 (24.5) | 0.999 | 32 (84.2) | 0.036 |
| No | 11 | - | 3 (27.3) | 1 (9.1) | 2 (18.2) | 5 (50.0) | |||||
| Intraplacental hypervascularity | |||||||||||
| Yes | 56 | 14 (25.0) | 0.564 | 31 (55.4) | 0.049 | 8 (14.3) | 0.999 | 14 (25.0) | 0.564 | 36 (80.0) | 0.127 |
| No | 4 | - | - | - | - | 1 (33.3) | |||||
| Retroplacental myometrial line disruption | |||||||||||
| Yes | 47 | 13 (27.7) | 0.264 | 26 (55.3) | 0.445 | 7 (14.9) | 0.673 | 13 (27.7) | 0.264 | 31 (83.8) | 0.095 |
| No | 13 | 1 (7.7) | 5 (38.5) | 1 (7.7) | 1 (7.7) | 6 (54.5) | |||||
| Myometrial thinning | |||||||||||
| Yes | 31 | 13 (41.9) | <0.001 | 21 (67.7) | 0.020 | 6 (19.4) | 0.257 | 11 (35.5) | 0.046 | 24 (88.9) | 0.040 |
| No | 29 | 1 (3.4) | 10 (34.5) | 2 (6.9) | 3 (10.3) | 13 (61.9) | |||||
| Uterine bulging | |||||||||||
| Yes | 18 | 10 (55.6) | <0.001 | 17 (94.4) | <0.001 | 5 (27.8) | 0.045 | 8 (44.4) | 0.019 | 14 (87.5) | 0.293 |
| No | 42 | 4 (9.5) | 14 (33.3) | 3 (7.1) | 6 (14.3) | 23 (71.9) | |||||
| Radiologist impression of invasive placenta (any degree) | |||||||||||
| Yes | 52 | 14 (26.9) | 0.179 | 31 (59.6) | 0.002 | 8 (15.4) | 0.582 | 14 (26.9) | 0.179 | 36 (83.7) | 0.007 |
| No | 8 | - | - | - | - | 1 (20.0) | |||||
| Radiologist impression of placenta percreta | |||||||||||
| Yes | 15 | 8 (53.3) | 0.003 | 15 (100) | <0.001 | 5 (33.3) | 0.019 | 8 (53.3) | 0.003 | 13 (86.7) | 0.462 |
| No | 45 | 6 (13.3) | 16 (35.6) | 3 (6.7) | 6 (13.3) | 24 (72.7) |
surgery time data was lacking for 12 patients
The need for ICU stay only significantly correlated with uterine bulging. Prolonged surgery time was associated with abnormal vascularization of placental bed and myometrial thinning.
Table 4 shows the ORs and CIs of MRI parameters for prediction of adverse maternal events. Myometrial thinning, uterine bulging and the radiologist impression of invasion showed significant association with four of the five maternal outcomes evaluated, and heterogeneous placenta with three.
Table 4.
Odds Ratios (ORs) and CIs of Each MRI Parameter for Prediction of Adverse Maternal Events
| Variable | Severe Blood Loss (=>2000 ml) | Hysterectomy | ICU stay | Blood Transfusion | Prolonged Surgery Time a | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |
| Heterogeneous placenta | 22.0 | (3.8–128.2) | 9.7 | (1.1–83.7) | 4.6 | (0.9–24.3) | 5.8 | (1.3–26.1) | 0.9 | (0.1–5.1) |
| Irregular contours and rounded edges | 5.7 | (0.7–47.8) | 1.3 | (0.4–4.2) | 2.6 | (0.3–22.9) | 2.4 | (0.5–12.0) | 3.0 | (0.7–12.5) |
| Low attenuation T2 bands | 1.7 | (0.2–15.9) | 6.5 | (0.7–59.0) | 0.9 | (0.1–8.3) | 1.7 | (0.2–15.9) | 3.9 | (0.5–31.5) |
| Intraplacental hypervascularity | 1.3 | (0.1–12.9) | 5.0 | (0.5–47.2) | 0.7 | (0.1–6.8) | 1.3 | (0.1–12.9) | 8.0 | (0.7–98.4) |
| Abnormal vascularization of the placental bed | 4.4 | (0.5–37.4) | 3.6 | (0.8–15.0) | 1.7 | (0.2–15.1) | 1.5 | (0.3–7.7) | 5.3 | (1.2–24.3) |
| Retroplacental myometrial line disruption | 4.6 | (0.5–38.9) | 2.0 | (0.6–7.0) | 2.1 | (0.2–18.8) | 4.6 | (0.5–38.9) | 4.3 | (1.0–18.8) |
| Myometrial thinning | 20.2 | (2.4–168.2) | 4.0 | (1.4–11.7) | 3.2 | (0.6–17.6) | 4.8 | (1.2–19.4) | 4.9 | (1.1–21.8) |
| Uterine bulging | 11.9 | (3.0–47.6) | 34.0 | (4.1–282.2) | 5.0 | (1.0–23.9) | 4.8 | (1.3–17.1) | 2.7 | (0.5–14.6) |
| Radiologist impression of invasive placenta (any degree) | 2.9 | (0.3–25.7) | 11.8 | (1.4–101.5) | 1.5 | (0.2–13.3) | 2.9 | (0.3–25.7) | 20.6 | (2.0–212.7) |
| Radiologist impression of placenta percreta | 7.4 | (2.0–28.1) | 27.2 | (3.3–225.2) | 7.0 | (1.4–34.3) | 7.4 | (2.0–28.1) | 2.4 | (0.5–13.0) |
surgery time data was lacking for 12 patients
Discussion
The findings from this retrospective analysis showed that placental MRI is highly accurate in identifying placental invasion, with substantial (0.67) and almost perfect (0.87) agreement between the radiologist impression of PAS disorder and placenta percreta, respectively, and the operative and histopathological findings.
The present study supports some of the MRI signs associated with PAS disorder reported in previous studies, including low attenuation T2 bands (p = 0.009), abnormal vascularization of the placental bed (p = 0.002), retroplacental myometrial line disruption (p = 0.001) and uterine bulging (p = 0.006).
Although the association of placenta percreta with higher maternal morbidity in relation to the less invasive forms of acretism are well established in the literature, 23 few studies have focused on the correlation of MRI signs related to PAS disorder and maternal outcomes. 24–27 To our knowledge, our study is the first to evaluate the strength of the association between individual MRI signs and five adverse maternal outcomes.
Meta-analysis data by Familiari et al has shown that uterine bulge was associated with placenta increta and percreta, with ORs of 5.5 and 6.3, respectively, but was not associated with placenta acreta. 28 A study by Clark et al showed that the presence of a bulge had an OR of 7.4 for the need of cesarean hysterectomy. 24 Our results corroborate this evidence, with a sensitivity of 87.5% and specificity of 90.9% for the presence of placenta percreta. It was significantly associated with four of five outcomes analyzed, with an OR of 11.9 for severe blood loss, 34.0 for the need for hysterectomy, 5.0 for the need for ICU stay and 4.8 for the need for blood transfusion
Myometrial thinning was also associated with four of five outcomes analyzed, with an OR of 20.2 for severe blood loss, 4.0 for the need for hysterectomy, 4.8 for the need for blood transfusion and 4.9 for prolonged surgery time, a finding also supported by the study of Burgioti et al, where myometrial thinning was significantly associated with all maternal outcomes analyzed, but not significantly associated with neonatal events. 25 Because of the high interobserver variability and the physiological thinning of the myometrium that occurs as the gestation progresses, this sign should not be used in isolation, but in conjunction with other signs of PAS disorder. 20
In accordance with previous studies, 24,25 placental heterogeneity was a significant predictor of three maternal events, with OR of 22.0 for severe blood loss, 9.7 for the need of hysterectomy and 5.8 for the need of blood transfusion. Interestingly, placental heterogeneity has no clear definition in terms of imaging parameters, and refers to an overall disorganized and inhomogeneous appearance and can be composed of several intraplacental abnormalities such as low attenuation T2 bands, abnormal intraplacental hypervascularity and intraplacental hemorrhage. 11,17 Given the highly subjective nature of this finding and the interreader variability, a consensus threshold was not achieved for recommending its use in MRI reporting in a recent joint statement by the Society of Abdominal Radiology (SAR) and the European Society of Urogenital Radiology (ESUR). 20
Our results are consistent with the published literature showing that the presence of low attenuation T2 bands was the most sensitive sign for predicting PAS (97.8%), 20,28 but with a low specificity of 28.6%. However, the present study shows that this signal was not associated with any of the maternal outcomes evaluated, a finding contradictory to similar studies. 24,25,27
The main risk of PAS is massive hemorrhage during delivery. 1–5 The definition of clinically significant obstetric hemorrhage is variable in the literature. In our study, we applied the definition from the Royal College of Obstetricians and Gynaecologists (RCOG), which defines “severe” blood loss as >2000 ml. 29 Care was taken to assess blood loss during delivery, including weighing dressings soaked in blood and visually estimating nonmeasurable blood loss.
There are inherent difficulties in determining the degree of placental invasion. 20 Although histopathological examination is considered the definitive method for diagnosing PAS disorder, 30 it is unavailable when conservative management is chosen. Moreover, it is our understanding that poor sampling and specimens could impact the sensitivity of histopathological analysis, a concern shared by other authors. 24,31 On the other hand, the clinical diagnosis of PAS disorder suffers from the use of non-uniform criteria during surgery, an issue that has been addressed by the recent development of standardized clinical criteria for grading PAS by the International Federation of Gynecology and Obstetrics (FIGO). 32 In our series, there was disagreement between the surgical impression and the histopathological report in six cases, in which the highest degree of invasion was recorded.
In addition to its retrospective nature, this study has some limitations. First, the reduced sample size may have limited the power of our study to detect statistical significance in some of the associations evaluated, although the low incidence of this pathology must also be taken in consideration. Second, some patients underwent prophylactic hypogastric artery ballooning during delivery, which could affect the amount of bleeding observed. Third, the MRIs scans were reviewed by only one examiner, with specific experience in placental evaluation, which may not reflect standard radiology practice elsewhere. Finally, our sample’s high pretest probability of placental invasion creates a selection bias, even although our examiner was blinded to all clinical data, including ultrasound findings. This selection bias seems unavoidable in a restrospective study, given the the complementary role of MRI in the evaluation of PAS disorder. Finally, correlation with ultrasound images could possibly improve its performance in our analysis.
Conclusion
In summary, individual MRI signs were associated with placental invasion and independently associated with adverse maternal peripartum outcomes. The presence of a placental bulge was highly accurate in predicting placenta percreta. Prospective studies with larger populations and appropriate control groups are needed to further assess the currently established MRI criteria and to develop an index to predict clinically significant PAS based on MRI signs.
Footnotes
Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author disclosures: There are no conflicts of interest to declare.
Contributor Information
Cristiano Köhler Silva, Email: cristianokohlersilva@gmail.com.
Caroline Lorenzoni Almeida Ghezzi, Email: cghezzi@hcpa.edu.br.
Daniela Vanessa Vettori, Email: dvettori@hcpa.edu.br.
Gabriela Françoes Rostirolla, Email: gabirosti@gmail.com.
Janete Vettorazzi, Email: jvettorazzi@hcpa.edu.br.
REFERENCES
- 1.Garmi G, Salim R. Epidemiology, etiology, diagnosis, and management of placenta accreta. Obstet Gynecol Int 2012; 2012: 873929. doi: 10.1155/2012/873929 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Khong TY. The pathology of placenta accreta, a worldwide epidemic. J Clin Pathol 2008; 61: 1243–46. doi: 10.1136/jcp.2008.055202 [DOI] [PubMed] [Google Scholar]
- 3.Jauniaux E, Jurkovic D. Placenta accreta: pathogenesis of a 20th century iatrogenic uterine disease. Placenta 2012; 33: 244–51. doi: 10.1016/j.placenta.2011.11.010 [DOI] [PubMed] [Google Scholar]
- 4.Mogos MF, Salemi JL, Ashley M, Whiteman VE, Salihu HM. Recent trends in placenta accreta in the United States and its impact on maternal-fetal morbidity and healthcare-associated costs, 1998-2011. J Matern Fetal Neonatal Med 2016; 29: 1077–82. doi: 10.3109/14767058.2015.1034103 [DOI] [PubMed] [Google Scholar]
- 5.Cahill AG, Beigi R, Heine RP, Silver RM, Wax JR, Society of Gynecologic Oncology, et al. Placenta accreta spectrum. Am J Obstet Gynecol 2018; 219: B2–16: S0002-9378(18)30892-5. doi: 10.1016/j.ajog.2018.09.042 [DOI] [PubMed] [Google Scholar]
- 6.Jauniaux E, Bunce C, Grønbeck L, Langhoff-Roos J. Prevalence and main outcomes of placenta accreta spectrum: a systematic review and meta-analysis. Am J Obstet Gynecol 2019; 221: 208–18: S0002-9378(19)30287-X. doi: 10.1016/j.ajog.2019.01.233 [DOI] [PubMed] [Google Scholar]
- 7.Panigrahi AK, Yeaton-Massey A, Bakhtary S, Andrews J, Lyell DJ, Butwick AJ, et al. A standardized approach for transfusion medicine support in patients with morbidly adherent placenta. Anesth Analg 2017; 125: 603–8. doi: 10.1213/ANE.0000000000002050 [DOI] [PubMed] [Google Scholar]
- 8.Collins SL, Alemdar B, van Beekhuizen HJ, Bertholdt C, Braun T, Calda P, et al. Evidence-based guidelines for the management of abnormally invasive placenta: recommendations from the international society for abnormally invasive placenta. Am J Obstet Gynecol 2019; 220: 511–26: S0002-9378(19)30433-8. doi: 10.1016/j.ajog.2019.02.054 [DOI] [PubMed] [Google Scholar]
- 9.Baughman WC, Corteville JE, Shah RR. Placenta accreta: spectrum of US and MR imaging findings. Radiographics 2008; 28: 1905–16. doi: 10.1148/rg.287085060 [DOI] [PubMed] [Google Scholar]
- 10.Rahaim NSA, Whitby EH. The MRI features of placental adhesion disorder and their diagnostic significance: systematic review. Clin Radiol 2015; 70: 917–25: S0009-9260(15)00155-5. doi: 10.1016/j.crad.2015.04.010 [DOI] [PubMed] [Google Scholar]
- 11.Azour L, Besa C, Lewis S, Kamath A, Oliver ER, Taouli B. The gravid uterus: MR imaging and reporting of abnormal placentation. Abdom Radiol (NY) 2016; 41: 2411–23. doi: 10.1007/s00261-016-0752-5 [DOI] [PubMed] [Google Scholar]
- 12.D’Antonio F, Iacovella C, Palacios-Jaraquemada J, Bruno CH, Manzoli L, Bhide A. Prenatal identification of invasive placentation using magnetic resonance imaging: systematic review and meta-analysis. Ultrasound Obstet Gynecol 2014; 44: 8–16. doi: 10.1002/uog.13327 [DOI] [PubMed] [Google Scholar]
- 13.Alamo L, Anaye A, Rey J, Denys A, Bongartz G, Terraz S, et al. Detection of suspected placental invasion by MRI: do the results depend on observer’ experience? Eur J Radiol 2013; 82: e51–7: S0720-048X(12)00409-3. doi: 10.1016/j.ejrad.2012.08.022 [DOI] [PubMed] [Google Scholar]
- 14.Ghezzi CLA, Silva CK, Casagrande AS, Westphalen SS, Salazar CC, Vettorazzi J. Diagnostic performance of radiologists with different levels of experience in the interpretation of MRI of the placenta accreta spectrum disorder. Br J Radiol 2021; 94(1128): 20210827. doi: 10.1259/bjr.20210827 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Srisajjakul S, Prapaisilp P, Bangchokdee S. Magnetic resonance imaging of placenta accreta spectrum: a step-by-step approach. Korean J Radiol; 2021; 22: 198–212. doi: 10.3348/kjr.2020.0580 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kilcoyne A, Shenoy-Bhangle AS, Roberts DJ, Sisodia RC, Gervais DA, Lee SI. Mri of placenta accreta, placenta increta, and placenta percreta: pearls and pitfalls. AJR Am J Roentgenol 2017; 208: 214–21. doi: 10.2214/AJR.16.16281 [DOI] [PubMed] [Google Scholar]
- 17.Brown BP, Meyers ML. Placental magnetic resonance imaging Part II: placenta accreta spectrum. Pediatr Radiol 2020; 50: 275–84. doi: 10.1007/s00247-019-04521-2 [DOI] [PubMed] [Google Scholar]
- 18.Baughman WC, Corteville JE, Shah RR. Placenta accreta: spectrum of US and MR imaging findings. Radiographics 2008; 28: 1905–16. doi: 10.1148/rg.287085060 [DOI] [PubMed] [Google Scholar]
- 19.Azour L, Besa C, Lewis S, Kamath A, Oliver ER, Taouli B. The gravid uterus: MR imaging and reporting of abnormal placentation. Abdom Radiol (NY) 2016; 41: 2411–23. doi: 10.1007/s00261-016-0752-5 [DOI] [PubMed] [Google Scholar]
- 20.Jha P, Pōder L, Bourgioti C, Bharwani N, Lewis S, Kamath A, et al. Society of abdominal radiology (SAR) and European Society of urogenital radiology (ESUR) joint consensus statement for MR imaging of placenta accreta spectrum disorders. Eur Radiol 2020; 30: 2604–15. doi: 10.1007/s00330-019-06617-7 [DOI] [PubMed] [Google Scholar]
- 21.Sannananja B, Ellermeier A, Hippe DS, Winter TC, Kang SK, Lee SI, et al. Utility of diffusion-weighted MR imaging in the diagnosis of placenta accreta spectrum abnormality. Abdom Radiol (NY) 2018; 43: 3147–56. doi: 10.1007/s00261-018-1599-8 [DOI] [PubMed] [Google Scholar]
- 22.Fiocchi F, Monelli F, Besutti G, Casari F, Petrella E, Pecchi A, et al. MRI of placenta accreta: diagnostic accuracy and impact of interventional radiology on foetal-maternal delivery outcomes in high-risk women. Br J Radiol 2020; 93(1114): 20200267. doi: 10.1259/bjr.20200267 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Marcellin L, Delorme P, Bonnet MP, Grange G, Kayem G, Tsatsaris V, et al. Placenta percreta is associated with more frequent severe maternal morbidity than placenta accreta. Am J Obstet Gynecol 2018; 219: 193: S0002-9378(18)30380-6. doi: 10.1016/j.ajog.2018.04.049 [DOI] [PubMed] [Google Scholar]
- 24.Clark HR, Ng TW, Happe S, Dashe J, Hr C, Tw N, et al. Placenta accreta spectrum: outcomes of high-risk pregnancies. American Journal of Roentgenology 2020; June: 1–7. [DOI] [PubMed] [Google Scholar]
- 25.Bourgioti C, Zafeiropoulou K, Fotopoulos S, Nikolaidou ME, Theodora M, Daskalakis G, et al. Mri prognosticators for adverse maternal and neonatal clinical outcome in patients at high risk for placenta accreta spectrum (PAS) disorders. J Magn Reson Imaging 2019; 50: 602–18. doi: 10.1002/jmri.26592 [DOI] [PubMed] [Google Scholar]
- 26.Chen D, Xu J, Ye P, Li M, Duan X, Zhao F, et al. Risk scoring system with MRI for intraoperative massive hemorrhage in placenta previa and accreta. J Magn Reson Imaging 2020; 51: 947–58. doi: 10.1002/jmri.26922 [DOI] [PubMed] [Google Scholar]
- 27.Chen T, Xu XQ, Shi HB, Yang ZQ, Zhou X, Pan Y. Conventional MRI features for predicting the clinical outcome of patients with invasive placenta. Diagn Interv Radiol 2017; 23: 173–79. doi: 10.5152/dir.2016.16412 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Familiari A, Liberati M, Lim P, Pagani G, Cali G, Buca D, et al. Diagnostic accuracy of magnetic resonance imaging in detecting the severity of abnormal invasive placenta: a systematic review and meta-analysis. Acta Obstet Gynecol Scand 2018; 97: 507–20. doi: 10.1111/aogs.13258 [DOI] [PubMed] [Google Scholar]
- 29.Mavrides E, Allard S, Chandraharan E, Collins P, Green L, Hunt BJ, Riris S, Thomson AJ on behalf of the Royal College of Obstetricians and Gynaecologists. Prevention and management of postpartum haemorrhage.BJOG 2016;124:e106–e149
- 30.Aalipour S, Salmanian B, Fox KA, Clark SL, Shamshirsaz AA, Asl NM, et al. Placenta accreta spectrum: correlation between FIGO clinical classification and histopathologic findings. Am J Perinatol 2023; 40: 149–54. doi: 10.1055/s-0041-1728834 [DOI] [PubMed] [Google Scholar]
- 31.Thiravit S, Ma K, Goldman I, Chanprapaph P, Jha P, Hippe DS, et al. Role of ultrasound and MRI in diagnosis of severe placenta accreta spectrum disorder: an Intraindividual assessment with emphasis on placental bulge. AJR Am J Roentgenol 2021; 217: 1377–88. doi: 10.2214/AJR.21.25581 [DOI] [PubMed] [Google Scholar]
- 32.Jauniaux E, Ayres-de-Campos D, Langhoff-Roos J, Fox KA, Collins S, Duncombe G. FIGO classification for the clinical diagnosis of placenta accreta spectrum disorders. Int J Gynaecol Obstet 2019; 146: 20–24. doi: 10.1002/ijgo.12761 [DOI] [PubMed] [Google Scholar]