Is it possible to predict severe postpartum hemorrhage and need for massive transfusion in placenta previa cases?

doi:10.21203/rs.3.rs-3593784/v1

Download PDF

Research Article

Is it possible to predict severe postpartum hemorrhage and need for massive transfusion in placenta previa cases?

https://doi.org/10.21203/rs.3.rs-3593784/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose

We aimed to construct a trustable working model for the placenta previa (PP) patients that helps to predict the postpartum bleeding potential with the input gained from the antenatal imaging studies, namely ultrasound (US) and Magnetic Resonance Imaging (MRI).

Materials and Methods

Forty-three PP patients were evaluated first by US and then 3 Tesla MRI. Placenta Accreata Index (PAI) was utilized during US evaluation in order to define the risks. Uterine bulging, heterogeneous signal, dark placental bands, focal interruption of myometrium and, tenting of bladder wall were regarded as predictive criterions in MRI evaluations. The correlation between the findings gained from US and MRI studies and haemorrhage (> 1000 ml), severe haemorrhage (> 2000 ml) and, massive transfusion (> 5 units of RBC) need was investigated to build this predictive model. The findings from the imaging studies were also confirmed histopathologically

Results

In the multivariate analysis of data from patients in either group of classification according to set point of 1000 ml., none of MRI and ultrasound findings were found to be predictive. The multivariate analysis done in second classification group (bleeding > 2000ml or more) PAI values (OR:2.3(1.4–3.8)) and overall MRI reported PAS (OR:4.9 (1.8–12.9)) were detected as predictive findings. While MRI findings were not discriminative between transfusion groups, grade 3 loculation in ultrasound examination was found to be predictive for need of transfusion more than 5 units (OR: 67,5 (8,2-549,4)). We had no cases of hysterectomy.

Conclusions

The ultrasound and MRI findings of PP cases can be useful to predict postpartum bleeding.

Placenta previa

placenta Accreata spectrum

postpartum haemorrhage

placenta Accreata index

Placenta previa (PP) is defined as presence of low-lying placental tissue so as to close the internal cervical ostium. The sequela includes increased Caesarean section rates, and also potential of early labour and ante, intra and, postpartum haemorrhage. Large cross sectional studies and case analyses show that the prevalence of PP is 4/1000 live birth [1, 2].

The prevalence until 20th gestational age is twice higher than that of at term pregnancy which denotes the spontaneous resolution during the process of ongoing pregnancy. The risk factors include past uterine surgery, history of PP in previous gestations, increased parity and, maternal age, infertility treatment, history of uterine evacuation, maternal usage of cocaine, smoking, plural gestations, male foetus and, history of uterine artery embolization [3, 4, 5, 6].

Co-existence of Placenta Accreata Spectrum (PAS) probability should be considered when PP is diagnosed. PAS includes placenta accreata-increata-percreata terms which define invasion of trophoblasts into myometrium, uterine serosa and, beyond respectively. This condition with its serious maternal and foetal consequences seen in one in every 500 deliveries was a rare occurrence in the past [7]. Caesarean section, curettage, myomectomy is found to be present in the history of 80% of patients diagnosed with PAS [8]. Especially this condition should be considered when an anteriorly located placenta is found to be covering the hysterotomy incision line. A prospective study done by Silver et al., which included 723 patients reported that PAS rate was found to be increased from 3% up to 67% from first C-section to fifth one [9].

PAS can be predicted with ultrasonography or MRI studies in pregnancies with PP or an anteriorly located placenta with history of previous uterine surgery. Placental lacunae and, disruption of the line which delineates uterine serosa and bladder wall are the most trustable diagnostic markers [10]. Additionally loss of hypoechogenic zone under placenta, myometrial thinning, abnormal vascular pattern and exophitic mass image protruding from uterine contour are regarded as other sonographic findings [11, 12]. Turbulent lacunary blood flow, bridge vessels, diffuse or focal intra parenchymal flow, hypervascularity at bladder-serosa interface, and prominent subplacental venous complexes are valuable findings at colour Doppler studies [13].

MRI can be utilized when ultrasonographic examination is unsufficient, or if there is suspicion of PAS at posterior uterine wall, or to determine the extent of the invasion in myometrial or parametrial tissues. However, it is not clinically proven to be beneficial to include the findings gained by MRI in the patient management [14, 15]. Uterine bulging, interruption of the bladder wall, loss of retroplacental hypointense line on T2W images, abnormal vascularization of the placental bed, dark intraplacental bands on T2W imaging, myometrial thinning, focal exophytic mass are amongst the findings found at MRI to predict placental invasion. PAS probability is believed to be increased when the number of above mentioned findings increase [16, 17].

In our study we aimed to find a correlation between the antenatal ultrasound and, MRI findings and haemorrhage or massive transfusion rates which are causes of major morbidity secondary to placenta previa diagnosis.

In this study, the data gathered from the medical records of 43 patients diagnosed with total PP in a single centre between the years 2015 and 2022 were analysed retrospectively. Ethical approval to use the hospital medical records was obtained from local authorities (Decision Number: 11-09.2022).

All 43 patients with PP diagnosis were evaluated by US (3.5-5 and 7.5 MHz,Voluson 735 Expert Pro, GE, USA) at first, and then by MRI (3 Tesla magnet Achieva/Intera Release 2.6.1Series 12.6, Philips Medical Systems, Eindhoven, The Netherlands). The gestational ages were between 30 and 39 weeks. The mean age of the patients was 31.6 ± 6.4 (22–46). All other variables such as parity, birth weight, presence of gestational diabetes, operation time, use of blood products, and histopathological confirmation of PAS were also gathered from medical records. Histopathological diagnosis of PAS was done at hysterectomy specimens or if no hysterectomy was performed, at the placental tissues (The presence of myometrial fragments beneath trophoblasts were regarded as positive for PAS).

Ultrasound examinations were done by same perinatologist (EÇ) with the use 3.5-5 MHz abdominal and, 7.5 MHz vaginal probes. Patients were classified into groups with three different set points. First classification compared patients who bleed under 1000 ml with those of bleeding over 1000 ml [18]. Similarly second classification used 2000 ml as cut off¹⁹, and 3 rd classification compared patients according to their need for transfusion of either 5 units or less than 5 units of RBC or those of 5 units or more [20]. Each group was scored by Placenta Accreata Index (PAI). PAI utilizes parameters such as number of Caesarean sections, presence of lacunae, anteriorly localized placenta, presence of bridging vessels, and, myometrial thickness [21]. PAI index used in our patients are presented in Table 1.

All MRI examinations were performed using a 3 Tesla magnet Achieva/Intera Release 2.6.1 Series 12.6 (Philips Medical Systems, Eindhoven, The Netherlands) whole-body scanner device in the Department of Radiology.

The MRI protocol began with the acquisition of a three-plane localizer image of the maternal anatomy, sagittal 5-mm T2-weighted single-shot fast spin-echo, and 5-mm T1-weighted images of the entire gravid uterus were performed to evaluate the placental position. A time-of-flight acquisition was performed in its axial plane to identify placental and maternal vascularity.

Each patient group was also scored according to the MRI findings such as uterine bulging, heterogeneous signal, dark placental bands, focal interruption of myometrium and, tenting of bladder wall. The imaging was done in supine position, and left lateral decubitus position when patient could not tolerate.

For the detection of postpartum haemorrhage we adopted to evaluate the signs and symptoms of hypovolemia as described by Bonnar [23]. As all cases were delivered by caesarean section. Aspirated blood, used sponges and, visual estimation of blood spilled on floor were used to calculate overall bleeding. The amount of aspirated amniotic fluid during delivery of the fetus was marked separately in the collector of aspirator and was not included in the oveall bleeding calculation. The decision for red blood cell transfusion was made if the blood loss was estimated to exceed 1500 mls or haemoglobin concentration was lower than 7 mg/dl.

The collected data were analyzed using the SPSS software version 25.0 (IBM Corp., Armonk, NY, USA). The normality of the demographic data was assessed using the Shapiro-Wilk test. Demographic data are summarized as the median ± inter-quartile range for non-normally distributed data and as the mean ± standard deviation for normally distributed data. Student’s t-test and the chi-square test were used to compare the means and rates of independent groups as indicated. A p-value of < 0.05 was considered to indicate a significant difference. Logistic regression analysis using Forward Wald method was used to predict the dichotomous variables, presence and absence of hemorrhage > 1000 ml, > 2000ml, and need for > 5 units of RBC transfusion separetely after inclusion of demografic data, PAI score parameters and MRI parameters separetely.

Fortythree patients were included in our study. Table 2 shows the comparison of demographic parameters such as age, parity,gestational age, number of previous deliveries,gestational diabetes and, duration of operation and histologic presence of PAS according to each group. The mean age of the patients with severe hemmorrhage was higher (p < 0.001). The mean parity, duration of operation, presence of gestastional diabetes and, histopathologically confirmed PAS rates, were higher in patients with severe haemorrhage and massive transfusion.

Patients were evaluated according to their ultrasound PAI scores in each three classification schemes. PAI scores were found to be statistically higher in both haemorrhage classification groups and also in those who needed more than 5 units of RBC transfusion (Table 3). Other parameters that take part in PAI scoring were also evaluated. Number of 2 or more caesarean section were found to be higher in only bleeders > 2000 ml according to first two classifications. However, number of caesarean sections did not differ according to transfusion groups. Anterior localisation of the placenta of PAI were found to be more frequent in patients that needed more than 5 U of RBC transfusion. On the other hand, this parameter did not show any difference in the other two haemorrhage classification subgroups (Table 3).

The patients were also evaluated according to overall MRI reported PAS, uterine bulging, heterogeneus signal, dark placental bands on t2 weighted MRI, focal interruption of myometrium and, tenting of bladder (table 4). The finding of dark placental bands on T2 weighted images in MRI were found to be higher in patients who bleed more than 1000 ml. and in those of patients who needed transfusion more than 5 Units. Every increase in PAI value and every additional PAS sign increased the risk of bleeding > 2000 ml. The MRI findings of uterine bulging, focal interruption of myometrium and tanting of bladder was significantly more frequent in patients bleeding > 1000 ml, > 2000ml and needed massive transfusion (Table 4).

In the multivariate analysis of data from patients in either group of classification according to set point of 1000 ml., none of MRI and ultrasound findings were found to be predictive. The multivariate analysis done to predict bleeding > 2000ml PAI values (OR:2.3(1.4–3.8)) and overall MRI reported PAS (OR:4.9 (1.8–12.9)) were detected as important. While MRI findings were not discriminative between transfusion groups, grade 3 loculation in ultrasound examination was found to be predictive for need of transfusion more than 5 units (OR: 67,5 (8,2-549,4)).

Labour can be planned in an elective fashion with a multi-disciplinary approach before any life threatening bleeding commenced whenever PAS diagnosis is predicted antenatally. When such an approach is adopted, maternal morbidities such as blood loss and, need for transfusion, or urological injuries are reported to be significantly decreased [22, 24].

The importance of preparedness in the fight of coping with life threatening bleeding in the case of PP can not be underestimated. The population based studies state that the diagnosis of placenta previa or low-lying placenta can be made in half to two thirds of patients in women with history of previous uterine surgery. According to a study done by Cali et al which proposed an ultrasound staging system for PAS disorders, all cases of placenta previa with evidence of increased vascularity in the inferior part of the lower uterine segment, potentially extending into the parametrial region (PAS3 in the proposed system), had been categorized as Stage 6 according to the FIGO postnatal clinical grading system, with good correlation between this ultrasound finding and surgical outcome [25].

Although not routinely utilized, MRI may help to increase diagnostic accuracy, or discriminate suspicious findings obtained from an ultrasound study [26]. In our study we wanted to augment the efforts of obstetric teams in the field of potential life threatening bleeding events experienced secondary to PP and/or PAS situations by creating an accurate estimation model of surgical outcome via combining ultrasound and MRI findings together with demographic patient data obtained from medical records. For this purpose, we stratified patients into groups in 3 classification schemes. The different set points were used in order to test their predictive competency for catastrophic surgical events. According to our results PAI score is capable of predicting excessive haemorrhage (> 1000ml or > 2000ml ) or need for transfusion of more than 5 U of RBCs ( Table 3) which is inconsistent with the study of Martha W et al which found that the smallest anterior myometrial thickness measured in the sagittal plane was the only prognostic criterion to successfully predict massively adherent placentas [27]. Up to our knowledge either ultrasound or MRI was utilized to predict surgical outcomes in the literature. We found only the study of Budorick et al. integrating both modalities at the same time. In their study comparison of these two modalities were made without mentioning clinically significant end points, while we analysed their predictive capabilities for different set off points and, applied multivariate analyses for each parameter statistically [28]. For predicting bleeding more than 1000 mls neither USG nor MRI findings were found to be predictive. On the other hand scoring with both modalities were found to be predictive for haemorrhage more than 2000 mls. Curiously, only grade 3 lacunae detected by USG were found to be successful in predicting need for transfusion more than 5 units. But to be marked again, each marker of invasion might be correlated with clinically significant bleeding and, should not be overlooked.

Once it is predicted, it is always a step forward in team efforts to challenge unwanted scenarios. According to our data, high PAI and overall MRI reported PAS scores can predict that the patient might bleed more than 2000 ml. Additionally, if loculation in ultrasound study was noted, it is also highly likely that the patient may need a blood product transfusion of more than 5 units.

Acknowledgements

The authors acknowledge their gratitude to all of the participants in this work.

Ethics

Ethics Committee Approval: The approval of Alaaddin Keykubat University Clinical Research Ethics Committee (decision dated 02.11.2022 and numbered 11/09).

Informed Consent: Participation in the study was voluntary and all participants read and approved the informed consent form.

Peer-review: Internally peer-reviewed.

Authorship Contributions

Concept: E.Ç, Design: E.Ç, B.A, Data Collection or Processing: B.A, E.D, Y.A, Analysis or Interpretation: E.Ç, M.Ç.K, Literature Search: E.K, E.Ç, Writing: E.K, E.Ç

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.

Faiz AS, Ananth CV. Etiology and risk factors for placenta previa: an overview and meta-analysis of observational studies. J Matern Fetal Neonatal Med 2003; 13:175.
Cresswell JA, Ronsmans C, Calvert C, Filippi V. Prevalence of placenta praevia by world region: a systematic review and meta-analysis. Trop Med Int Health 2013; 18:712.
King LJ, Dhanya Mackeen A, Nordberg C, Paglia MJ. Maternal risk factors associated with persistent placenta previa. Placenta 2020; 99:189
Long SY, Yang Q, Chi R, et al. Maternal and Neonatal Outcomes Resulting from Antepartum Hemorrhage in Women with Placenta Previa and Its Associated Risk Factors: A Single-Center Retrospective Study. Ther Clin Risk Manag 2021; 17:31.
Lavery JP. Placenta previa. Clin Obstet Gynecol 1990; 33:414.
Weis MA, Harper LM, Roehl KA, et al. Natural history of placenta previa in twins. Obstet Gynecol 2012; 120:753
Wu S, Kocherginsky M, Hibbard JU. Abnormal placentation: twenty-year analysis. Am J Obstet Gynecol 2005;192:1458-61
Tantbirojn P, Crum CP, Parast MM. Pathophysiology of placenta creta: the role of decidua and extravillous trophoblast. Placenta 2008; 29:639
Silver RM, Landon MB, Rouse DJ, et al. Maternal morbidity associated with multiple repeat cesarean deliveries. Obstet Gynecol 2006; 107:1226
Levine D, Hulka CA, Ludmir J, et al. Placenta accreta: evaluation with color Doppler US, power Doppler US, and MR imaging. Radiology 1997; 205:773
Comstock CH. Antenatal diagnosis of placenta accreta: a review. Ultrasound Obstet Gynecol 2005; 26:89.
Finberg HJ, Williams JW. Placenta accreta: prospective sonographic diagnosis in patients with placenta previa and prior cesarean section. J Ultrasound Med 1992; 11:333
Pagani G, Cali G, Acharya G, et al. Diagnostic accuracy of ultrasound in detecting the severity of abnormally invasive placentation: a systematic review and meta-analysis. Acta Obstet Gynecol Scand 2018; 97:25.
Kirkinen P, Helin-Martikainen HL, Vanninen R, Partanen K. Placenta accreta: imaging by gray-scale and contrast-enhanced color Doppler sonography and magnetic resonance imaging. J Clin Ultrasound 1998; 26:90.
De Oliveira Carniello M, Oliveira Brito LG, Sarian LO, Bennini JR. Diagnosis of placenta accreta spectrum in high-risk women using ultrasonography or magnetic resonance imaging: systematic review and meta-analysis. Ultrasound Obstet Gynecol 2022; 59:428.
Jha P, Pōder L, Bourgioti C, 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.
Bourgioti C, Zafeiropoulou K, Fotopoulos S, 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.
International Federation of Gynecology and Obstetrics (FIGO): Recommendations on the management of postpartum haemorrhage, 2022
RCOG: Green-top guideline for the prevention and management of postpartum haemorrhage, 2nd edition, 2016
Stephen J. McCall, Catherine Deneux-Tharaux, Loïc Sentilhes, Rema Ramakrishnan, Sally L. Collins, Aurélien Seco, Jennifer J. Kurinczuk, Marian Knight, Gilles Kayem. Placenta accreta spectrum – variations in clinical practice and maternal morbidity between the UK and France: a populationbased comparative study. BJOG. 2022;129:1676–1685
Abuhamad A. Morbidly adherent placenta. Semin Perinatol 2013;37:359-64
Warshak CR, Ramos GA, Eskander R, et al. Effect of predelivery diagnosis in 99 consecutive cases of placenta accreta. Obstet Gynecol 2010;115:65-9.
Bonnar J. Massive obstetric haemorrhage. Baillieres Best Pract Res Clin Obstet Gynaecol 2000; 14:1.
Rac MWF, Dashe JS, Wells CE, et al. Ultrasound predictors of placental invasion: the Placenta Accreta Index. Am J Obstet Gynecol 2015;212:343.e1-7.
Cali G, Forlani F, Lees C, Timor-Trisch I, Palacios-Jaraquemada J, Dall’Asta A, Bhide A, Flacco ME, Manzoli L, Labate F, Perino A, Scambia G, D’Antonio F. Prenatal ultrasound staging system for placenta accreta spectrum disorders. Ultrasound Obstet Gynecol 2019; 53: 752–760.
Ting Chen, Xiao-Quan Xu, Hai-Bin Shi, Zheng-Qiang Yang, Xin Zhou, Yi Pan. Conventional MRI features for predicting the clinical outcome of patients with invasive placenta. Diagn Interv Radiol. 2017 May-Jun;23(3):173-179.
Martha W. F. Rac MD, Elysia Moschos MD, C. Edward Wells MD, Donald D. McIntire PhD, Jodi S. Dashe MD, Diane M. Twickler MD. Sonographic Findings of Morbidly Adherent Placenta in the First Trimester. J Ultrasound Med 2016; 35:e11–e17
Budorick NE, Figueroa R, Vizcarra M, Shin J. Another Look at Ultrasound and Magnetic Resonance Imaging for Diagnosis of Placenta Accreta. J Matern Fetal Neonatal Med. 2017 Oct;30(20):2422-2427

Table-1 PAİ index ²²

Variable		Value
> 2 ceserean deliveries		3.0
Lacunae	Grade 3	3.5
	Grade 2	1.0
Sagittal smallest myometrial thickness	< 1 mm	1.0
	<1mm but > 3mm	0.5
	<3mm but <5mm	0.25
Anterior placenta previa		1.0
Bridging vessel		0.5

a-If parameter is not present, then value is 0

b-Measured in sagittal plane

c- If any portion of placenta is anterior

Tablo-2 Demographic data of study population

Variable	>1000 ml haemorrhage			>2000 ml severe haemorrhage			>5 unite transfusion
Variable	Yes n=25	No n=18	P	Yes n=12	No n=31	P	Yes n=11	No n=32	p
Age	34.2±6.8	28.1±3.5	0.001	37.1±6.6	29.5±4.9	<0.001	34.6±9.2	30.6±4.9	P=0.19
Parity	1.9±0.7	1±0.5	<0.001	2.3±0.6	1.2±0.6	<0.001	2.1±0.8	1.3±0.6	P=0.001
Gestational Age (week)	34.7±9.7	35.7±1.8	P=0.14	35.2±0.36	35±2.5	P=0.76	35.2±0.9	35±2.5	P=0.78
No Previous Delivery	6 (24%)	9(50%)	P=0.02	2(16.7 %)	13(41.9%)	<0.01	1(9.1%)	14(43.8%)	<0.01
Gestational Diabetes Mellitus	9 (36%)	0	P=0.004	6(50%)	3(9.7%)	P=0.04	6(54.5%)	3(9.4%)	P=0.001
Operation Time	88.6±39.7	36.6±4.8	<0.001	106.6±42.8	51.4±25.8	<0.001	119±35	48±21	<0.001
Histologic PAS	20(80%)	3(16.7%)	<0.001	12(100%)	11(35%)	<0.001	11(100%)	12(37.5%)	<0.001

Table-3 Ultrasound finding with respect to haemorrhage outcome

Variable	>1000 ml haemorrhage			>2000 ml severe haemorrhage			>5 unite transfusion
Variable	Yes n=25	No n=18	P	Yes n=12	No n=31	P	Yes n=11	No n=32	p
PAİ Score	6.6±2.1	3.9±1.4	<0.001	7.7±1.8	4.6±1.8	<0.001	8±1.1	4.6±1.9	<0.001
>2 c/s	15(60%)	3(16.7%)	P=0.004	9(75%)	9(29%)	0.006	7(63%)	11(34%)	0.09
Lacunae	11(44%)	0	P=0.001	7(58.3%)	4(12.9%)	P=0.002	9(81.8%)	2(6.3%)	<0.001
Myometrial Thickness	1.8±1.7	3.7±0.7	<0.001	1.2±2	3.1±1.2	<0.001	0.5±0.8	3.3±1.2	<0.001
Anterior Placenta	14(56%)	5(27.8%)	P=0.06	8(66.7%)	11(35.5%)	0.06	10(90.9%)	9(28.1%)	<0.001
Bridging Vessels	16(64%)	0	<0.001	9(75%)	7(22.6%)	0.001	10(90.9%)	6(18.8%)	<0.001

Table-4 MRI finding with respect to clinical outcome

Variable	>1000 ml hemorrhage			>2000 ml severe hemorrhage			>5 unit transfusion
Variable	Yes n=25	No n=18	P	Yes n=12	No n=31	P	Yes n=11	No n=32	p
Overall MRI Reported PAS	21(84%)	4(22.2%)	<0.001	12(100%)	13(41.9%)	P=0.001	11(100%)	14(43.8%)	P=0.001
Uterine Bulging	15(60%)	1(5.6%)	<0.001	10(83.3%)	6(19.4%)	<0.001	10(90.9%)	6(32.8%)	<0.001
Heterogeneous Signal	8(32%)	2(11%)	0.1	5(41.7%)	5(16.1%)	P=0.07	5(45.5%)	5(15.6%)	P=0.05
Dark Placental bands on T2 Weighted MRI	7(28%)	1(5.6%)	P=0.06	5(41%)	3(9.7%)	P=0.01	4(36.4%)	4(12.5%)	0.08
Focal ınterruption of myometrium	1(5.6%)	18(72%)	<0.001	10(83.3%)	9(29%)	P=0.001	10(90.9%)	3(9.3%)	<0.001
Tantıng of bladder	6(24%)	0	0.002	6(50%)	0	<0.001	6(54.5%)	0	<0.001

Download PDF

Version 1

posted

You are reading this latest preprint version

Is it possible to predict severe postpartum hemorrhage and need for massive transfusion in placenta previa cases?

Status:

Version 1

Abstract

Purpose

Materials and Methods

Results

Conclusions

Introduction

Materials and methods

Results

Discussion

Declarations

References

Tables

Status:

Version 1