OBJECTIVE:
Because of the increasing rates of cesarean delivery, placenta accreta spectrum (PAS) now occurs in 1 of 272 pregnant people.1 The prenatal diagnosis of PAS enables multidisciplinary management for optimal outcomes.2 The Placental Accreta Index (PAI), established in 2015, provides a standardized evaluation for ultrasound prediction of PAS.3 The PAI score is derived from the weighed contribution of 5 parameters: history of ≥2 cesarean deliveries, sagittal smallest myometrial thickness (≤1 mm, >1 mm but ≤3 mm, or >3 mm but ≤5 mm), grade 2 or 3 placental lacunae, and presence of bridging vessels or an anterior placenta previa (Figure). The PAI was predictive with an area under the receiver operating characteristic curve (AUC) of 0.87 (95% confidence interval [CI], 0.80–0.95) and has since been prospectively validated.4,5 This study aimed to determine whether additional ultrasound variables could enhance the prediction of the PAI for the outcome of cesarean hysterectomy for PAS following the International Federation of Gynecology and Obstetrics (FIGO) clinical or pathologic criteria. We evaluated the value added of placenta previa position, presence of a placental bulge, number of bridging vessels, length of transverse irregularity of the uterine-bladder interface, and the largest transverse diameter of suspected PAS. The transverse plane was used in measurement as we anecdotally observed this to be the larger dimension and, as it followed the previous uterine scar, suspected it was more likely to be clinically significant.
FIGURE. PAI sonographic variables and those assessed for added value.
Anterior placenta previa with grade 3 lacunae on transabdominal sagittal (A) and transvaginal sagittal (B) imaging. Grading determination is based on the presence of many large and bizarre lacunae as described by Finberg and Williams. Image B demonstrates that there is no intervening myometrium between the placenta and the bladder such that the sagittal smallest myometrial thickness is <1 mm. Bridging vessels (arrows) are categorized as 1 to 5 (arrows) in image C and >20 mm in image D because of a large confluent area as demonstrated on a 2D image taken from the transvaginal transverse color Doppler cine with low-velocity settings. A placental bulge (arrow) is also observed in image D with length of transverse irregularity of the uterine-bladder interface (lines) at 92 mm. (E) The largest diameter of suspected placenta accreta spectrum (line) at 59 mm in a different patient as observed on a single image of the 2D transverse color Doppler cine with low-velocity settings.
2D, 2-dimensional.
STUDY DESIGN:
This was a single-center analysis of prospectively collected data from a nonreferred pregnant population obtained from the 28- to 32-week follow-up scans performed from September 2019 to June 2021. Pregnant people with at least 1 previous cesarean delivery, presence of placenta previa or low-lying placenta, and delivery at our institution were included. Per institutional protocol, transabdominal and transvaginal 2-dimensional (2D) sagittal and transverse grayscale and color cine sweeps with a full bladder are performed as part of evaluation for PAS. All images were interpreted prospectively by 1 of 5 physicians with expertise in maternal-fetal medicine or radiology.
The original 5 PAI variables and PAI were reported, including the following variables: placenta previa position, presence of a placental bulge, number of bridging vessels, length of transverse irregularity of the uterine-bladder interface, and the largest transverse diameter of suspected PAS involvement (Figure). Placenta previa position was categorized as anterior, symmetrical, and posterior with anterior extension referring to scenarios in which the bulk of the placenta was anterior, symmetrically extended anteriorly and posteriorly, or posterior with a small amount of anterior extension over the cervix, respectively. The number of bridging vessels was estimated in ranges of none, 1 to 5, 6 to 10, 11 to 15, 15 to 20, or >20. This number was determined by counting the number of suspected individual vessels as scrolled and assessed on either the 2D sagittal or transverse color cine sweep. If a large area of confluent vessels was observed, this was scored as >20 (Figure). The length of transverse irregularity of the uterine-bladder interface was based on the largest area of confluent bridging vessels in a single image of the 2D transverse color Doppler cine involving the bladder-serosal interface. The largest transverse diameter of suspected PAS was the largest diameter of confluent vessels on a single image observed in the 2D transverse color Doppler cine, including both the bladder and surrounding myometrium.
Cesarean hysterectomy was the primary outcome. Multivariable logistic regression was used to assess the contribution of these additional variables to generate odds ratios (ORs). Moreover, AUCs were determined. A P value of <.05 was considered significant. All analyses were done in SAS (version 9.4; SAS Institute, Cary, NC). This study was approved by the institutional review board of the University of Texas Southwestern Medical Center. Informed consent was waived for this study.
RESULTS:
Of 55 pregnant people who met the study inclusion criteria, 22 (40%) required a cesarean hysterectomy for PAS following the FIGO clinical criteria. Of the 22 pregnant people, 20 had PAS confirmed by pathology. The 2 cases of PAS following the FIGO clinical criteria alone included 1 FIGO grade 1 where manual removal of the placenta resulted in significant placental site bleeding requiring hysterectomy and 1 FIGO grade 2 with placental bulge and hypervascularity of the lower uterine segment resulting in the decision for hysterectomy. In the final pathology in the latter case, there was marked myometrial thinning of the uterine wall, but a decidual layer could be identified.
Our study further validated the original PAI, finding an AUC of 0.85 (95% CI, 0.75–0.95) for cesarean hysterectomy. Of note, 4 of 5 original PAI variables (number of previous cesarean deliveries, grading of lacunae, sagittal smallest myometrial thickness, and presence of bridging vessels) were significantly associated with the surgical outcome (Table). Placental location did not add significance based on a more specified location.
TABLE. Demographics and ultrasound variables by outcome of cesarean hysterectomy and contribution of additional ultra- sound variables from multivariable logistic regression.
| Variable | No hysterectomy (n=33) | Hysterectomy (n=22) | P valuea | |
|---|---|---|---|---|
| Baseline characteristics | ||||
| Maternal age (y) | 33.0±5.5 | 33.0±6.6 | .99 | |
| Race and ethnicity | Non-Hispanic Black | 4 | 4 | .64 |
| Non-Hispanic White | 3 | 4 | ||
| Hispanic | 15 | 23 | ||
| Asian | 0 | 2 | ||
| Parity | 1 | 10 | 5 | .78 |
| 2 | 17 | 6 | ||
| >2 | 16 | 11 | ||
| BMI (kg/m2) | 34.3±8.7 | 35.0±6.7 | .76 | |
| Delivery GA (wk) | 36.4 (36.0–38.3) | 36.6 (35.4–38.4) | .69 | |
| Ultrasound variables | ||||
| PAI score | 2.8±2.1 | 6.3±2.5 | <.001 | |
| Sagittal smallest myometrial thickness (mm) | 2.1±2.5 | 0.2±0.4 | <.001 | |
| Previous cesarean delivery | 1 | 19 (79%) | 5 (21%) | .043 |
| 2 | 6 (50%) | 6 (50%) | ||
| 3 | 4 (33%) | 8 (67%) | ||
| 4 | 2 (50%) | 2 (50%) | ||
| 5 | 1 (100%) | 0 (0%) | ||
| 6 | 1 (100%) | 0 (0%) | ||
| 7 | 0 (0%) | 1 (100%) | ||
| Placental lacunae grade | 0 | 16 (84%) | 3 (16%) | <.001 |
| 1 | 11 (69%) | 5 (31%) | ||
| 2 | 6 (60%) | 4 (40%) | ||
| 3 | 0 (0%) | 10 (100%) | ||
| The largest diameter of suspected PAS (mm)b | 14.9±17.4 | 44.0±24.3 | <.001 | |
| Length of transverse irregularity of the uterine-bladder interface (mm)c | 5.7±12.0 | 36.2±28.9 | <.001 | |
| Placenta previa position | Posterior with anterior extension | 6 (55%) | 5 (45%) | .131 |
| Anterior | 24 (69%) | 11 (31%) | ||
| Symmetrical | 3 (33%) | 6 (67%) | ||
| Placental bulge | No | 31 (69%) | 14 (31%) | .01 |
| Yes | 2 (20%) | 8 (80%) | ||
| Bridging vessels | None | 1 1 (100%) | 0 (0%) | <.001 |
| 1–5 | 12 (75%) | 4 (25%) | ||
| 6–10 | 2 (50%) | 2 (50%) | ||
| 11–15 | 4 (57%) | 3 (43%) | ||
| 16–20 | 3 (75%) | 1 (25%) | ||
| >20 | 1 (8%) | 12 (92%) | ||
| Additional variables | ORd | P valuee | AUCf | |
| Largest transverse diameter of suspected PAS (per 10 mm) | 1.35 (0.92–1.96) | .12 | 0.86 (0.76–0.96) | |
| Length of transverse irregularity of the uterine-bladder interface (per 10 mm) | 1.35 (0.92–1.96) | .03 | 0.89 (0.80–0.97) | |
| Placenta previa position | 0.97 (0.51–1.86) | .94 | 0.85 (0.75–0.95) | |
| Placental bulge | 2.64 (0.41–17.10) | .31 | 0.86 (0.76–0.96) | |
| Bridging vessels | 1.49 (0.91–2.43) | .11 | 0.88 (0.79–0.97) | |
2D, 2-dimensional; AUC, area under the receiver operating characteristic curve; BMI, body mass index; GA, gestational age; OR, odds ratio; PAI, Placental Accreta Index; PAS, placenta accreta spectrum.
The Wilcoxon rank-sum test and chi-square test were used for univariate analyses on the categorical and continuous variables, respectively
The largest transverse diameter of suspected PAS defined as the greatest length on 2D transverse cine of continuous abnormal bridging vessels or vascularity
Length of transverse irregularity of the uterine-bladder interface defined as the length on 2D transverse cine of continuous abnormal bridging vessels
The OR from multivariable logistic regression with the PAI as a covariate
The P value testing the null hypothesis of OR=1 with the PAI as a covariate
AUC from a logistic regression model with the corresponding additional variable + PAI.
Although significantly associated with the outcome of cesarean hysterectomy, most new variables were not additive to the AUC for prediction (Table). The length of transverse irregularity of the uterine-bladder interface was the only new variable with a significant additional contribution (P=.03) (Table). For every 10 mm increase, the odds of cesarean hysterectomy increased by 35% (OR, 1.35; 95% CI, 0.92–1.96). Multivariable logistic regression incorporating irregularity of the uterine-bladder interface and PAI yielded an AUC of 0.89 (95% CI, 0.80–0.97), a slight increase from the AUC generated from the original PAI parameters.
CONCLUSION:
The PAI’s ability to predict cesarean hysterectomy was excellent, with an AUC of 0.85. The length of transverse irregularity of the uterine-bladder interface slightly improved the prediction of cesarean hysterectomy when added to the PAI variables. In studies with larger cohorts using the PAI, adding the length of transverse irregularity of uterine-bladder interface may improve discrimination.
Acknowledgments
Dr. Herrera is supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development under Award Number K23HD103876. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
The authors report no conflict of interest.
SUPPLEMENTARY MATERIALS: Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ajogmf.2022.100832.
Contributor Information
Christina L. Herrera, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas TX.
Meredith J. Kim, University of Texas Southwestern Medical School, Dallas TX.
Yin Xi, Department of Radiology, University of Texas Southwestern Medical Center, Dallas TX.
Jodi S. Dashe, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas TX.
Catherine Y. Spong, Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas TX.
Diane M. Twickler, Department of Radiology, University of Texas Southwestern Medical Center, Dallas TX; Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas TX.
REFERENCES
- 1.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] [PubMed] [Google Scholar]
- 2.Shamshirsaz AA, Fox KA, Erfani H, et al. Outcomes of planned compared with urgent deliveries using a multidisciplinary team approach for morbidly adherent placenta. Obstet Gynecol 2018;131:234–41. [DOI] [PubMed] [Google Scholar]
- 3.Rac MW, Dashe JS, Wells CE, Moschos E, McIntire DD, Twickler DM. Ultrasound predictors of placental invasion: the placenta accreta Index. Am J Obstet Gynecol 2015;212. 343.e1–7. [DOI] [PubMed] [Google Scholar]
- 4.Happe SK, Yule CS, Spong CY, et al. Predicting placenta accreta spectrum: validation of the placenta accreta index. J Ultrasound Med 2021;40:1523–32. [DOI] [PubMed] [Google Scholar]
- 5.Abu Hashim H, Shalaby EM, Hussien MH, El Rakhawy M. Diagnostic accuracy of the placenta accreta index for placenta accreta spectrum: a prospective study. Int J Gynaecol Obstet 2022;156:71–6. [DOI] [PubMed] [Google Scholar]