A simple guide to ultrasound screening for placenta accreta spectrum for improving detection and optimizing management in resource limited settings.

When screening for PAS, in a woman with previous cesarean delivery with low‐lying placenta previa, the urinary bladder must be filled to the extent that the entire lower uterine segment and the uterovesical interface can be well visualized. 17 This may be uncomfortable for some women, but it is necessary for proper evaluation. Most women will report the urge to urinate well before they have a sufficiently filled bladder, but they should be encouraged to wait as incontinence is unlikely to occur. The choice of the scanning approach should be based on the experience of the operator, clinical and cultural context. Both transabdominal and transvaginal approaches give excellent results depending on the skill of the operator and quality of the gray‐scale resolution of the ultrasound equipment. In patients with high body mass index (calculated as weight in kilograms divided by the square of height in meters), scanning difficulty with transabdominal approach can be overcome by asking the woman to lift her pannus and scanning beneath it or by using the transvaginal approach. The ideal time for thorough PAS screening is above 28 weeks to reduce false‐positive diagnosis because two or more PAS imaging signs are seen in 98% of normal placentation in the second trimester. 8 Also, most signs of PAS become more prominent in the third trimester. Abnormally invasive placentation (increta and percreta) will usually be accompanied by neovascularity, which occurs at the level of the serosal surface of the uterus (within the utero‐vesical fold of the peritoneum in the case of a low‐lying/previa placenta). This is seen at laparotomy as multiple, newly formed large vessels on the uterine surface at the area of abnormal invasion (Figure  1a ). 9 The underlying pathophysiology remains unclear, however, recent histologic evidence indicates that PAS provokes rapid increased growth of originally much smaller vessels, which is why they have such immature vascular architecture including a poorly formed vessel wall. 18 This neovascularity has been described with a variety of different imaging signs, including uterovesical hypervascularity, sub‐placental hypervascularity, bridging vessels, and bladder wall interruption. 19 The most useful of these is probably bridging vessels (Figure  1a,b ), because these only occur when there is a significant amount of neovascularity at the serosal surface. Demonstrating the same neovascularity seen at laparotomy (a), as bridging vessels with color Doppler ultrasound (b) and as bladder wall interruption (c). (a) Demonstrating the massive neovascularity seen at laparotomy between the anterior aspect of the uterus and the posterior bladder. (b) Demonstrating the same vessels as in (a) seen as ‘bridging vessels’ with color Doppler ultrasound. (c) Demonstrating the same vessels as in (a) seen as bladder wall interruption with gray‐scale ultrasound. Depending on the number, size and course of the new vessels, this sign can manifest on gray‐scale imaging in different ways at the uterine serosal surface (uterovesical interface in an anterior low/previa placenta). It can appear as pairs of hyperechoic lines (sometimes referred to as the “=” sign) running parallel to the uterine serosa and posterior bladder wall. This is generated by the ultrasound being reflected from both walls of the vessel producing an “=” sign or a scalloped appearance. This is the sign often referred to as “bladder wall interruption” (Figure  1b ) and is seen as bridging vessels with color Doppler imaging (Figure  1c ). It must be emphasized that the appearance of “bridging vessels” is an ultrasound artefact caused by the cross‐sectional two‐dimensional imaging of the contorted neovascularity as it curves around a three‐dimensional structure (the front of the uterus). It does not represent blood vessels actually “connecting” the placental bed with the urinary bladder. 20 The placenta is a highly vascular organ, and some signs such as “hypervascularity” are entirely subjective and must therefore be interpreted with caution depending on operator experience. It must be emphasized that the surgically challenging neovascularity associated with PAS occurs at the serosal surface (Figure  1c ) and not within the myometrium. Care must be taken not to confuse conditions, such as adenomyosis, which manifest as thick cystic and vascular myometrium (Figure  2 ) from the neovascularization of PAS, which is associated with an abnormally thin myometrium. Also, the Doppler gain and pulse repetitive frequency must be adjusted to reduce blooming and motion artefacts as well as avoiding exaggerating normal sub‐placental vascularity, which may give a false representation of hypervascularity. A pulse repetitive frequency setting of above 15 cm/s satisfies these requirements in most cases. Demonstrating the thick, cystic vascular myometrium of a woman with proven adenomyosis at subsequent hysterectomy and a normal placenta—this should not be mistaken for sub‐placental hypervascularity. The presence of abnormal lacunae is the most common ultrasound sign of PAS present in literature. 21 , 22 Lacunae are multiple, large, irregular anechoic areas noted within the placenta, which give the placenta a “moth‐eaten” appearance (Figure  3a ). 20 In severe PAS, placental tissue is found more deeply within the myometrium than it should be, often passing the level of the spiral arteries and reaching the radial and arcuate vessels. This causes excessive dilatation of these higher pressure arteries 23 and a massively increased velocity of flow into the delicate intervillous space (Figure  3b ). 24 This powerful flow distorts the architecture of one or more cotyledons and its corresponding interlobar septa, resulting in lacunae formation. 20 Although these lacunae can sometimes be seen throughout the whole placenta because of the underlying pathophysiology, on careful inspection they should be seen to be adjacent to the basal plate of the placenta. If the “placental holes” seen are only on the fetal side of the placenta or just at the edges, it must be carefully considered as to whether they actually are PAS‐related lacunae. Demonstrating the irregular large lacunae continuous with the placental bed and “fed” by large, high‐velocity vessels. (a) Placenta accrete spectrum (PAS) lacunae; (b) PAS lacunae with color Doppler demonstrating high‐velocity blood flow into the lacunae. PAS‐related lacunae (Figure  3a ) must not be confused with placental lakes (Figure  4a ) or echogenic cystic lesions (Figure  4b ). Placental lakes are cystic spaces (greater than 10 mm) usually centrally located within the cotyledon or lobule surrounded by placental tissue of normal echogenicity. 25 Placental lakes are frequently confused with PAS‐related lacunae and the terms “lake” and “lacunae” are often incorrectly used interchangeably. This is evidenced by studies that report the presence of lacunae in normal placentation in low‐risk pregnancies. 8 , 26 On real time gray‐scale imaging, lacunae and lakes appear as hypoechoic areas within the placenta. However, typical PAS‐related lacunae are often irregularly ellipsoid in shape and extend from the placental bed where they receive blood supply from the feeder vessels (deep myometrial vessels, i.e. radial or arcuate arteries). Placental lakes often, but not always, contain slow‐moving blood and are easily compressible with the ultrasound probe whereas typical PAS‐related lacunae are not compressible. It must be noted that lacunae and lakes can be present in the same PAS placenta. However, they should be differentiated based on their size and location. Also, PAS‐related lacunae are often numerous in the region of the abnormal placentation whereas lakes are usually few and widely distributed. (a) Placental lake, (b) echogenic cystic lesion, (c) lake with low‐velocity flow visible on gray‐scale (B mode) imaging. (a) Simple placental lakes – note there is no increased brightness in the surrounding tissue. (b) Echogenic cystic lesion (resulting from a placental infarct). (c) Placental lake with slow flow visible in gray‐scale—note the position away from the placental basal plate. Placental infarcts result from interrupted maternal blood supply to the placenta and often present as hypoechoic regions with hyperechogenic rim (echogenic cystic lesion) or well‐circumscribed lesions with mixed echogenicity and are often associated with pre‐eclampsia and fetal growth restriction. 27 Placental infarcts can be differentiated from lakes and lacunae by the characteristic hyperechoic rim. PAS‐related lacunae can usually be confirmed by demonstrating high‐velocity flow (>10 cm/s) with feeder vessels on Doppler interrogation (Figure  3b ), whereas lakes sometimes show very low flow velocity for which the signals are rarely detected on Doppler but can be seen in gray‐scale imaging (Figure  4c ). Echogenic cystic lesions resulting from placental infarcts show no flow on color Doppler interrogation but occasionally contain static or very slow moving, irregularly shaped contents. A placental bulge describes the outpouching of the uterus containing the placenta due to inadequate residual myometrium to maintain the structural integrity of the uterus (Figure  5 ). 20 This sign can be seen with both ultrasound and MRI and is highly predictive of increta or percreta when used in combination with other imaging signs of PAS. 28 This sign represents the absence of sufficient myometrial tissue to support the placenta and hence is extremely useful in both the prenatal diagnosis of PAS and the subsequent management plan. If there is a bulge, the clinician can be confident that there is not enough residual muscle to contract and provide the living ligature required to stop bleeding from the placental bed therefore, forced removal of the placenta should not be attempted because it will result in bleeding. Examples of a placental bulge on ultrasound scan and after delivery—the “bulge” is caused by a loss of structural integrity in the muscle of the lower segment causing it to bulge outwards, it becomes more pronounced after delivery as the upper segment contracts. A placental bulge can also occur with a normal placenta as a result of progressive dehiscence of a uterine scar resulting in the underlying placenta bulging through it. 16 However, in these cases, the bulge is usually smaller, the placenta is fairly homogeneous with no evidence of placental lacunae, neovascularization, or other signs of PAS (Figure  6 ). Example of a uterine dehiscence—the bladder is full at ultrasound but empty at laparotomy, hence it has collapsed down revealing the bulge of placenta, note the completely normal uterine tissue around the defect and the lack of signs of placenta accrete spectrum on the ultrasound. The retroplacental hypoechoic zone is the echolucent space between the placenta and endometrium. A clear understanding of what the retroplacental hypoechoic zone represents is yet to be determined; however, it has been linked to the presence of decidua glands and vascular plexus involving basal arteries and terminal branches of the spiral arteries. 20 The absence of this sign indicates a loss or deficiency of Nitabuch's layer and the subsequent “fusion” of placenta and myometrium. This is currently the only known direct marker for PAS. However, the specificity of this sign is an issue of controversy because of its susceptibility to false‐positive results. The retroplacental hypoechoic space is influenced by external compressive force, usually from pressing hard with the ultrasound probe. Due to the susceptibility of the retroplacental hypoechoic zone to compressive effect, care must be taken to prevent/minimize the compressive effect of the probe when assessing the uteroplacental bed for the presence of the clear zone. 17 This sign becomes more prominent in advancing gestation because of myometrial thinning and prominent dilatation of the uteroplacental circulation. 29 To have a clear assessment on ultrasound, the dynamic range, chroma(tint) and focus must be used to improve contrast resolution and zoom feature, to enlarge the image for thorough assessment. Myometrial thickness less than 1 mm, or an area of imperceptible myometrium behind the placenta, has been reported as a sign of PAS. 30 Myometrial thinning has been attributed to the progressive migration/invasion of the extravillous trophoblast through the abnormally healed myometrium resulting in minimal/absent myometrium to support the placental bed. In this case, any attempt to separate the placenta, could result in torrential hemorrhage. However, data obtained do not establish a clear association between myometrial thickness and the severity of PAS. Both myometrial thicknesses of more than 2 mm 31 , 32 , 33 and abnormally thin myometrium 29 have been reported in PAS. To be clear, thinning of the lower uterine segment can be a normal finding in the third trimester and may be related to fetal presentation. This may be further emphasized after a previous cesarean delivery due to significant scar thinning resulting from poor myometrial healing. 34 Occasionally, the placenta may overlie an area of simple uterine scar dehiscence without any abnormal placentation, this phenomenon is known as a “uterine window” because the placenta can be seen through the lower segment at delivery 4 (Figure  6 ). Even though this is not PAS, there is a significant risk of uterine rupture so it must be approached with care. The presence or absence of additional sonographic and clinical markers of PAS will differentiate between the two pathologies. 16 Hence, when screening for PAS, myometrial thickness should not be used as the sole parameter for diagnosis but must be used in conjunction with other imaging signs. In rare cases, the placenta can involve the cervix (probably as a result of the previous cesarean scar being on the cervix as can occur with a fully dilated cesarean delivery). In such cases, the whole lower segment appears to be filled with bulging placenta and the cervix shows massive hypervascularity. Often the cervix itself is virtually invisible transabdominally. It is important to know about this from the imaging because it should guide subsequent management. Any attempt at focal resection or sub‐total hysterectomy in these situations risks massive hemorrhage.

TAB and SC conceptualized this review and drafted the initial manuscript. MJ, AJN‐C, VS, RAA, and KF reviewed the initial draft and contributed to finalizing the manuscript.

The signs described so far demonstrate the uterine/placental morphologic changes when there is abnormal placentation. However, lack of these signs does not definitely rule out all cases of PAS, as abnormal adherence has virtually no associated signs. However, to have clinically significant invasion there should be some myometrial compromise, so a thick myometrium with no placental bulge is very unlikely to be seen in cases of increta or percreta. Currently, the only direct markers described in literature are the loss of the retroplacental hypoechoic zone. This marker is valuable for ruling out PAS but not for ruling it in. In simple words, the presence of the retroplacental hypoechoic zone excludes the presence of PAS in that area of placental bed. However, care must be taken to develop sufficient experience to assess these signs and to examine the whole placenta bed before PAS is ruled out. PAS is not a binary condition; it is a spectrum from abnormal adherence to severe percreta; hence the ultrasound appearance will significantly differ from case to case. The sonographer must therefore consider the underlying pathophysiology of the known ultrasound markers of PAS when screening high‐risk patients and report on the clinical implications of all the signs seen. The steps have been summarized in Table  1 . Simplified steps in PAS screening Assess risk factors for PAS ( pre‐test probability ) Anterior low‐lying (<2 cm from internal os) or placenta previa + previous cesarean delivery/ies History of uterine surgery or myometrial/endometrial damage Anterior low‐lying (<2 cm from internal os) or placenta previa + previous cesarean delivery/ies History of uterine surgery or myometrial/endometrial damage Ultrasound signs represent different anatomical features, e.g. neovascularity Severity and intra‐operative findings vary significantly Each PAS will have different ultrasound signs representing anatomical features unique to that case Utero‐placental bed Placental bulge = defect in uterine muscle (PAS or dehiscence) definitely insufficient muscle to contract Myometrial thinning (<1 mm or undetectable) = probably insufficient muscle to contract Loss of “clear zone” = loss of smooth placental surface with probable “fusion” of placenta to uterus Placental bulge = defect in uterine muscle (PAS or dehiscence) definitely insufficient muscle to contract Myometrial thinning (<1 mm or undetectable) = probably insufficient muscle to contract Loss of “clear zone” = loss of smooth placental surface with probable “fusion” of placenta to uterus Abnormal lacunae Large, irregular, anechoic areas connecting with the myometrium = destruction of the placental tissue by high pressure “feeder” blood vessels from deep within the uterus (radial/arcuate arteries) If Doppler is available the feeder vessels can be seen (>10 cm/s) Large, irregular, anechoic areas connecting with the myometrium = destruction of the placental tissue by high pressure “feeder” blood vessels from deep within the uterus (radial/arcuate arteries) If Doppler is available the feeder vessels can be seen (>10 cm/s) Neovascularization signs Bladder wall interruption = presence of tangled mat of new blood vessels between the anterior uterine wall and posterior bladder wall (ultrasound artifacts from vessel walls cause “=” appearance or “scalloping”) If Doppler is available this is seen as “bridging vessels” Bladder wall interruption = presence of tangled mat of new blood vessels between the anterior uterine wall and posterior bladder wall (ultrasound artifacts from vessel walls cause “=” appearance or “scalloping”) If Doppler is available this is seen as “bridging vessels” Abbreviation: PAS, placenta accreta spectrum.

In a high pre‐test probability for PAS, we recommend thorough screening of the entire placental bed by an expert in the prenatal diagnosis of PAS. Modern technology and the use of telemedicine have performed well in some low‐resource settings as a novel way to bridge the gap of distance, for example when the closest referral center is hours away or in a neighboring country. 35 When available, use of telemedicine has the potential to obtain an expert second opinion without adding undue financial burdens or disruption to the daily life of the woman and her family, simply to repeat imaging elsewhere. Placentation process is completed in the middle of the second trimester. 36 Beyond this point, further trophoblastic invasion does not take place. However, the need for follow up in diagnosed cases of PAS remains uncertain and is a topic of debate. Currently, only two published studies longitudinally assessed the progression of PAS from the first trimester till delivery. 37 , 38 Both studies revealed no significant changes in the ultrasound signs between the second and third trimesters. Changes seen within the placenta in the third trimester are only related to ageing of the placenta, making the signs easier for the sonographer to see, not the progression of invasion. 38 Improving outcomes in pregnancies complicated by PAS should not end with prenatal diagnosis, adequate preparation and appropriate management play a crucial role in improving the outcome. Our next article will focus on how these two very crucial factors can be achieved in a limited‐resource setting.

The incidence of PAS may be rare, but it is definitely rising with the increasing rate of cesarean deliveries worldwide. This article is intended to provide a simplified guide to aid the thought process of ultrasound operators when they need to undertake prenatal screening for PAS. It is vital that they do not try to just identify the signs but consider what each sign represents anatomically and how this correlates with subsequent surgical challenges. Then report this to their surgical colleagues in a way that they understand.

Placenta accreta spectrum (PAS) is a rare pregnancy complication where the placenta fails to separate spontaneously after delivery and cannot be forcibly separated without causing catastrophic obstetric hemorrhage. 1 It is caused by abnormal placental implantation over a myometrial scar, and results in extrusion of placental tissue beyond the usual confines of the intrauterine cavity with fibrinoid deposition, and massive neovascularity. 2 PAS is typically graded according to the extent of placenta involvement ranging from abnormal adherence to the myometrium (accreta), to deep myometrial implantation (increta), and percreta, which breaches the serosal surface and may involve other surrounding structures. The degree of morbidity is highly dependent on the degree of extension, amount of neovascularity and involvement of other pelvic structures. In placenta accreta, the placenta can sometimes be detached because it is only adherent, as long as there is sufficient myometrium underlying the placenta to enable adequate uterine contraction to prevent catastrophic hemorrhage. Any attempt to manually remove the placenta in the more serious phenotypes (increta and percreta) can cause a uterine rupture and significant bleeding. Even if it does not, there is insufficient myometrium to contract and provide the “living ligature” of the terminal arteries. The normal physiologic changes of pregnancy result in placental perfusion approaching 800 mL/min at term. 3 Hence, rupture of the uterus or inadequate myometrial contraction results in torrential hemorrhage. The mainstay of management for PAS is not to disturb the placental bed. 4 The involvement of other pelvic structures in percreta requires a multidisciplinary team with experience of PAS to ensure safe dissection of surrounding structures at hysterectomy. 5 PAS is associated with a very high risk of maternal mortality, especially if the surgeon is caught unaware. In high‐income countries with an abundance of experienced surgeons and readily available life‐saving resources the maternal mortality rate has been reported to be 7% at the severe end of the spectrum. 6 Although the exact comparative figure remains unknown, it is likely that women with PAS in resource‐limited settings have a much greater risk of death due to technical, logistic, and resourcing inadequacies. In 1994, the three‐delay model was developed to offer health programs options to prevent maternal death with an emphasis on strategies to mobilize and adapt existing resources. The focus was to understand barriers in the interval between onset of obstetric complications and their outcome, and identify the point at which healthcare delivery could be optimized 7 : (1) delay in the decision to seek care; (2) delay in arrival at a health facility; and (3) delay in the provision of adequate care. We have used the philosophy of the three‐delay model to propose an “IS‐PAS 4‐A strategy” to ensure optimal outcomes of PAS namely; (1) A wareness of the risk factors of the condition, (2) A ccurate prenatal diagnosis, (3) A dequate preparation, and (4) A ppropriate management. In this paper we aim to concentrate on the first two parts of the IS‐PAS 4‐A strategy. Our second article will address preparation and management in low‐ and middle‐income countries. Although the incidence of PAS is still relatively rare, the numbers are rising as the result of mounting cesarean section rates. An increased awareness leading to appropriate screening is important in improving maternal outcomes. Accurate prenatal diagnosis is often challenging because the published literature on the sonographic signs of PAS can be contradictory and potentially confusing to sonographers with limited experience with the condition. Even though the diagnosis can be readily made with B‐mode (gray‐scale) ultrasound, recent reports focus on Doppler technology, which is often not present on ultrasound equipment used in limited‐resource settings. The role of ultrasound in PAS screening also needs to move beyond making a binary diagnosis of presence or absence of PAS. Rather, there is the need to consider the anatomical findings that each sign represents and how that could influence planning for the surgery and the potential need for referral to a more experienced team, which can be extremely challenging in low‐resource settings. This article aims to provide a simplified three‐step approach for fetal imaging personnel to improve their understanding of the process for the basic screening for PAS, with and without Doppler ultrasound.

The authors have declared that there are no conflicts of interest.