Sonographic Evaluation of the Lower Uterine Segment in Patients With Previous Cesarean Delivery

Transcription

1 Article Sonographic Evaluation of the Lower Uterine Segment in Patients With Previous Cesarean Delivery Vincent Y. T. Cheung, MBBS, FRCOG, FRCSC, RDMS, Oana C. Constantinescu, MD, RDMS, Birinder S. Ahluwalia, MBBS, RDMS Objective. To evaluate the appearance of the lower uterine segment (LUS) in pregnant women with previous cesarean delivery and to compare the LUS thickness with that in women with unscarred uteri. Methods. In a prospective study, sonographic examination was performed on 53 pregnant women with previous cesarean delivery (cesarean group), 40 nulliparas (nullip-control), and 40 women who had 1 or more childbirths with unscarred uteri (multip-control) between 36 and 38 weeks gestation to assess the appearance and compare the thickness of the LUS. In the cesarean group, the sonographic findings were correlated with the delivery outcome and the intraoperative LUS appearance. Results. In the cesarean group, 44 patients (83.0%) had a normal-appearing LUS indistinguishable from that of control groups; 2 patients (3.8%) had an LUS defect suggestive of dehiscence; and 7 patients (13.2%) had thickened areas of increased echogenicity with or without myometrial thinning. Although the cesarean group had a thinner LUS (1.9 ± 1.4 mm) when compared with both the nullip-control group (2.3 ± 1.1 mm; P >.05) and the multip-control group (3.4 ± 2.2 mm; P <.001), only the latter difference achieved statistical significance. One of the 2 patients who had a sonographically suspected LUS defect had confirmed uterine dehiscence during surgery. An intraoperatively diagnosed paper-thin LUS, when compared with an LUS of normal thickness, had significantly smaller sonographic LUS measurements (1.1 ± 0.6 versus 2.0 ± 0.8 mm, respectively; P =.004). Conclusions. Prior cesarean delivery is associated with a sonographically thinner LUS when compared with those with prior vaginal delivery. Prenatal sonographic examination is potentially capable of diagnosing a uterine defect and determining the degree of LUS thinning in patients with previous cesarean delivery. Key words: cesarean delivery; lower uterine segment thickness; previous cesarean delivery; sonography; uterine rupture; vaginal birth after cesarean delivery. Abbreviations LUS, lower uterine segment; VBAC, vaginal birth after previous cesarean delivery Received May 10, 2004, from the Department of Obstetrics and Gynecology, North York General Hospital, Toronto, Ontario, Canada (V.Y.T.C.); and BSA Diagnostic Imaging, Toronto, Ontario, Canada (O.C.C., B.S.A.). Revision requested June 10, Revised manuscript accepted for publication July 4, We thank Terence Lao, MBBS, FRCOG (Department of Obstetrics and Gynecology, University of Hong Kong, Hong Kong), for comments and suggestions. This study was partly funded by the Iris and Dalton Bales Research Endowment Fund, North York General Hospital. Address correspondence and reprint requests to Vincent Y. T. Cheung, MBBS, FRCOG, FRCSC, RDMS, Department of Obstetrics and Gynecology, North York General Hospital, 4001 Leslie St, Toronto, Ontario M2K 1E1, Canada. vcheung@nygh.on.ca. Attempted vaginal birth after previous cesarean delivery (VBAC) remains controversial. Although it has been reported as safe and has contributed to a reduced cesarean delivery rate, 1,2 VBAC is associated with a risk of uterine rupture. 3 Because the maternal and fetal consequences of uterine rupture can be serious and potentially life threatening, 4,5 the proper selection of patients would be an important prerequisite. It is generally considered that, among carefully selected patients who have full participation in decision making, most women with 1 previous lower segment transverse cesarean delivery are suitable candidates for VBAC and should be offered a trial of labor. 6,7 Although the efficacy and safety of VBAC have been shown, 1,2 to our knowledge, there are no reliable methods to predict the risks of uterine rupture in this group of patients. Studies have shown 2004 by the American Institute of Ultrasound in Medicine J Ultrasound Med 2004; 23: /04/$3.50

2 Lower Uterine Segment in Patients With Previous Cesarean Delivery that the risk of uterine rupture in the presence of a defective scar is related directly to the degree of thinning of the lower uterine segment (LUS). 8,9 Although LUS thickness as measured by sonography at or near term is being used by 16% of obstetricians in Canada to determine which women are good candidates for VBAC, 10 the value of applying sonographic LUS thickness measurement in the management of VBAC remains unclear. Sonographically, the LUS appears as a 2-layered structure that consists, from the urinary bladder inward, of the echogenic visceral-parietal reflection, including the muscularis and mucosa of the urinary bladder (the outer layer), and the relatively hypoechoic myometrial layer (Figure 1). 8,11 Usually at late gestation, the chorioamniotic membrane and the decidualized endometrial layer cannot be seen as layers separate from the myometrium. If the fetus is vertex presenting, the presenting part may be sitting against the LUS, and no amniotic fluid can be seen in between these 2 structures. However, very little has been published on sonographic LUS measurement, and the technique for measuring the LUS thickness has not been standardized. The purpose of this study was to evaluate the appearance of the LUS in patients with previous cesarean delivery and to compare their LUS thickness with that in subjects with unscarred uteri. Materials and Methods This study was approved by the Research Ethics Committee at North York General Hospital, and written consent was obtained from each study patient. During the study period from January 2003 to October 2003, pregnant women attending the antenatal clinic or the fetal assessment unit were invited to participate in the study. A study group consisting of 53 women with 1 or more previous cesarean deliveries (cesarean group) and 2 control groups consisting of nulliparas (nullip-control) and women with 1 or more previous childbirths (multip-control) were recruited. The enrollment of the 2 control groups was completed when 40 subjects were included in each group. All sonographic examinations were performed at 36 to 38 weeks gestation. All patients were not in labor at the time of scanning. Because the uterine thickness might be affected by abnormal intrauterine volume, women with multiple gestations and abnormal amniotic fluid volumes were excluded from the study. Also, patients with placenta previa in whom the LUS might not be clearly identifiable were not candidates for LUS measurement. For patients who participated in the study with no obstetric indications, the sonographic examination included, in addition to the evaluation of the LUS, identification of fetal presentation and exclusion of unexpected abnormal amniotic fluid volume and placenta previa. Figure 1. Longitudinal sonogram of the LUS from a control patient showing the urinary bladder wall myometrium interface (arrows) and the myometrium/chorioamniotic membrane amniotic fluid interface (arrowheads). B indicates urinary bladder; and H, fetal head. Figure 2. Longitudinal sonogram showing the LUS appearance from a cesarean group patient. The arrows indicate where the LUS thickness is measured (2.2 mm). The LUS was reported to be of normal thickness during another cesarean delivery J Ultrasound Med 2004; 23:

3 Cheung et al Transabdominal sonographic examination was carried out with a full urinary bladder (to the extent that the patient had the urge to void) to allow good imaging of the LUS. The LUS was examined longitudinally and transversely to identify any areas of obvious dehiscence or rupture. Any balloon effect, as described by Michaels et al, 11 consisting of any abnormal bulging of the outer layer associated with fetal movement or changes in amniotic fluid pressure against the urinary bladder base, was noted. If the LUS appeared intact, an attempt would be made to identify the previous uterine scar, and the appearance was noted. The thinnest zone of the lower segment was identified visually at the midsagittal plane along the cervical canal. This area was magnified to the extent that any slight movement of the caliper would produce a change in measurement by only 0.1 mm. The measurement was taken with the cursors at the urinary bladder wall myometrium interface and the myometrium/chorioamniotic membrane amniotic fluid interface (Figure 2). At least 2 measurements were made, and the lowest value was taken as the LUS thickness. All examinations were performed on an Aloka SSD-5000 machine (Aloka Co, Ltd, Tokyo, Japan) or a Toshiba SSH-140A machine (Toshiba Medical Systems Co, Ltd, Tokyo, Japan) with a 3.5- or 5.0-MHz convex transducer by a single sonographer (V.Y.T.C.). The examiner bias with respect to the knowledge of prior cesarean delivery was difficult to eliminate because the presence or absence of an abdominal scar was obvious. The referring physicians were aware of the sonographic findings. The patients labor and delivery outcomes were reviewed. After a vaginal delivery, routine manual palpation of the LUS was not performed. If any patient in the cesarean group had repeat cesarean deliveries, the obstetrician who performed the surgery was asked to comment on the appearance of the LUS under the following categories (modified from those of Michaels et al 11 and Fukuda et al 12 ): (1) normal thickness, similar to that seen with primary cesarean delivery; (2) evidence of rupture (represented by complete separation of the uterine scar of any length, resulting in communication between the uterine and peritoneal cavities) or dehiscence (represented by subperitoneal separation of the uterine scar, with the chorioamniotic membrane visible through the peritoneum of the LUS); and (3) paper-thin but not thin enough to visualize the uterine contents. Statistical analysis was performed with the Student t test, χ 2 test, and Fisher exact test when appropriate. P <.05 was taken as significant. Because most related studies used different measuring techniques and had no comparison with control subjects or information on population size at 36 to 38 weeks gestation, we were unable to use these data to guide us in determining the sample size. Hence, for a preliminary study with no idea of the likely difference, we decided to choose 40 or more subjects in each group as the sample size, which was larger than those in several similar reports. 11,13,14 Results The study population consisted of 53 women in the cesarean group, 40 women in the nullip-control group, and 40 women in the multip-control group. In the cesarean group, 2 patients had 2 previous cesarean deliveries, and 4 patients had 1 or more vaginal births in addition to the cesarean delivery. Tables 1 and 2 show the patient characteristics and the indications for sonographic examination, respectively. In the cesarean group, the previous surgical reports were available for 40 patients (75.5%), and all indicated a previous Table 1. Patient Characteristics Cesarean 1 Nullip-Control 2 Multip-Control 3 Characteristic (n = 53) (n = 40) P 1&2 (n = 40) P 1&3 Maternal age, y 34.6 ± ± 5.2 < ± 5.1 NS Parity 2 7 (13.2) 0 < (15.0) NS Maternal weight, kg 66.5 ± ± 9.0 NS 69.2 ± 10.1 NS Obstetric history 1 spontaneous abortion 13 (24.5) 5 (12.5) NS 6 (15.0) NS 1 therapeutic abortion 24 (45.3) 17 (42.5) NS 18 (45.0) NS 1 preterm labor 1 (1.9) 0 (0) NS 1 (2.5) NS Data are presented as mean ± SD and n (%). NS indicates not significant (P >.05) J Ultrasound Med 2004; 23:

4 Lower Uterine Segment in Patients With Previous Cesarean Delivery Table 2. Indications for Sonographic Examination Indication Cesarean (n = 53) Nullip-Control (n = 40) Multip-Control (n = 40) Fetal weight assessment 1 (1.9) 8 (20.0) 9 (22.5) Suspected malpresentation 1 (1.9) 7 (17.5) 9 (22.5) Gestational diabetes mellitus 4 (7.5) 6 (15.0) 3 (7.5) Hypertensive disorders 1 (1.9) 1 (2.5) 0 Decreased fetal movement 1 (1.9) 3 (7.5) 2 (5.0) Suspected fetal hydronephrosis 0 2 (5.0) 3 (7.5) Fibroids/adnexal cysts 0 3 (7.5) 2 (5.0) Previous stillbirth (2.5) No indication 45 (84.9) 10 (25.0) 11 (27.5) Data are presented as n (%). lower segment transverse uterine scar. Of these 40 patients, 34 (85.0%) had a 2-layered uterine closure. In the remaining 6 patients, the type of closure was not reported. Sonographic Appearances of the LUS in the Cesarean Group Normal-Appearing LUS In 44 patients (83.0%), the previous uterine scar could not be identified. A normal-appearing LUS indistinguishable from that in the control groups, consisting of an echogenic outer layer on top of the intact myometrial layer, was shown (Figure 2). LUS Defect Suggestive of Dehiscence This appearance was defined as a defective area in the LUS with no myometrial layer evident in this defect (Figure 3). Only the outer layer was visualized, and abnormal ballooning or bulging of this layer when movement of the fetus or amniotic fluid pushed against the base of the urinary bladder, as described by Michaels et al, 11 was occasionally seen on real-time scanning. Two patients (3.8%) had this appearance on sonographic examination. One patient, whose sonographic LUS appearance is shown in Figure 3, had another cesarean delivery by a different obstetrician not involved in this study, and during the surgery a defect was found in the LUS, confirming the sonographic finding. The other patient had another cesarean delivery by the principal author (V.Y.T.C), and during the surgery plenty of adhesions were noted, which limited exposure of the LUS. The uterine incision was therefore made higher than usual, but the LUS was thought to be paper-thin on palpation. Areas of Increased Echogenicity in the Outer Layer With or Without Myometrial Thinning In 7 patients (13.2%), the previous scar was identified as an isolated thickened echogenic area in the outer layer, whereas the myometrial layer underneath this area appeared asymmetrically thinned out when compared with the adjacent myometrium (Figure 4). LUS Thickness Figures 2 and 5 are sonograms obtained from cesarean group patients showing different LUS thicknesses. The mean ± SD (range) LUS thicknesses in the cesarean, nullip-control, and multip-control groups were 1.9 ± 1.4 (0 9.0), 2.3 ± 1.1 ( ), and 3.4 ± 2.2 ( ) mm, respectively (Table 3). Although the difference between the cesarean and nullip-control LUS thickness failed to reach statistical significance, it achieved significance when those patients with more than 1 prior delivery (both vaginal and cesarean) in the Figure 3. Longitudinal sonogram showing the uterine defect (between arrows). The myometrial layer is clearly shown adjacent to the defect (arrowheads) J Ultrasound Med 2004; 23:

5 Cheung et al Figure 4. Longitudinal sonogram showing an area of increased echogenicity in the outer layer (arrow) with myometrial thinning (arrowhead). Figure 5. Longitudinal sonogram from a cesarean group patient showing an extremely thin LUS (between arrows, 0.7 mm). The LUS was confirmed to be paper-thin during another cesarean delivery. cesarean group were excluded from the analysis (1.8 ± 0.9 versus 2.3 ± 1.1 mm; P =.042). Only 2 patients had multiple (2) cesarean deliveries, and no difference was noted in the LUS thickness in these 2 patients when compared with those with a single cesarean delivery (1.9 ± 0.1 versus 1.8 ± 0.9 mm; P >.05). Table 3 shows the labor and delivery outcome. As expected, more patients in the cesarean group had elective cesarean delivery. The difference in the induction rate and the oxytocin infusion rate simply reflected the usual practice in VBAC management. The principal author (V.Y.T.C.) performed the deliveries in 35.8% and 22.5% of the patients in the cesarean and control group patients, respectively. All other patients had delivery by the oncall or their own attending obstetricians. For all patients who had cesarean delivery, the intraoperative findings were compared with the sonographic description and the measurement of the LUS. These comparisons were not totally blinded because some elective repeat cesarean deliveries were performed by the obstetricians who were aware of the sonographic findings. Of the 35 cesarean group patients who had repeat cesarean deliveries, 1 was confirmed to have uterine dehiscence, 9 were reported to have a paper-thin LUS, and 19 had normal LUS thickness, with the preoperative sonographic mean (range) LUS thicknesses of 0 ± 0, 1.1 ± 0.6* (0 1.9), and 2.0 ± 0.8* ( ) mm, respectively (*P =.004). In 6 patients (17.1%), the intraoperative LUS appearance was not described, but neither dehiscence nor rupture was reported. All 6 patients had prenatal sonographic LUS thickness of greater than 1 mm, and, in our opinion, they likely had normal uterine thickness because we thought that most obstetricians would report a uterine defect if noted during surgery. Table 4 summarizes the relationship between various LUS measurements with respect to the delivery outcome and the intraoperative LUS appearance. The sonographic mean (range) LUS thicknesses in patients who had successful and failed VBAC were 2.2 ± 1.8 ( ) and 1.9 ± 0.9 ( ) mm (P >.05). Labor did not seem to affect the intraoperative LUS appearance (Table 5). Discussion The main objective of this study was to determine the appearance and thickness of the LUS in normal and scarred uteri and to relate our findings to those in the literature. Rozenberg et al, 8 using transabdominal sonographic examination of the LUS at 36 to 38 weeks gestation, included the bladder mucosa and the peritoneal layer in the measurement. With a cutoff value of less than 3.5 mm, the negative predictive value was high (99.3%) for predicting uterine defects, but the positive predictive value was low (11.8%). 8 Fukuda et al 12 and Suzuki et al, 15 using transabdominal sonography without clearly defining the site of measurement, concluded that a wall thickness of 2 mm or less was a potential sign of a uterine defect. Gotoh et al, 9 using transvaginal sonography, concluded that 74% of women with a LUS of less than 2.0 mm had an incomplete J Ultrasound Med 2004; 23:

6 Lower Uterine Segment in Patients With Previous Cesarean Delivery Table 3. Lower Uterine Segment Thickness, Labor, and Delivery Outcomes Characteristic Cesarean 1 (n = 53) Nullip-Control 2 (n = 40) P 1&2 Multip-Control 3 (n = 40) P 1&3 LUS thickness, mm 1.9 ± ± 1.1* NS 3.4 ± 2.2* <.001 Gestation at scanning, wk 36.9 ± ± 0.6 NS 36.9 ± 0.8 NS Cephalic presentation at scanning 48 (90.6) 38 (95.0) NS 35 (87.5) NS Labor No 26 (49.1) 3 (7.5) < (5.0) <.001 Spontaneous 26 (49.1) 28 (70.0) NS 30 (75).018 Induced 1 (1.8) 9 (22.5) (20).005 Oxytocin infusion 8 (15.1) 24 (60.0) < (45.0).002 Delivery Vaginal 18 (34.0) 29 (72.5) < (92.5) <.001 Emergency cesarean 10 (18.8) 9 (22.5) NS 1 (2.5).021 Elective cesarean 25 (47.2) 2 (5.0) < (5.0) <.001 Gestation at delivery, wk 39.1 ± ± 1.3 NS 39.5 ± Scanning-delivery interval, wk 2.2 ± ± 1.2 NS 2.6 ± 1.2 NS Birth weight, g ± ± NS ± NS Apgar score <7 at 5 min 0 1 (2.5) NS 0 NS Data are presented as mean ± SD and n (%). NS indicates not significant (P >.05). *P =.008, Nullip-Control versus Multip-Control. uterine rupture noted at cesarean delivery. Michaels et al, 11 instead of just measuring the thickness, described the sonographic appearances of the LUS by its symmetry, thickness, movement, ballooning, and the presence of a wedge defect and divided these findings into 3 Table 4. Relationship Between Sonographic LUS Measurement, Delivery Outcome, and Intraoperative LUS Appearance Sonographic LUS Measurement, mm Parameter Delivery outcome (n = 4) (n = 32) (n = 11) (n = 6) Vaginal Elective cesarean Unsuccessful VBAC LUS appearance* (n = 3) (n = 21) (n = 8) (n = 3) Normal Thin Defect Not described *Cesarean deliveries only. Table 5. Intraoperative LUS Appearance in Cesarean Group Patients With and Without Labor Unsuccessful VBAC Elective Cesarean LUS Appearance (n = 10) (n = 25) P Normal 6 (60.0) 13 (52.0) NS Thin 1 (10.0) 8 (32.0) NS Defect 0 1 (4.0) NS Not described 3 (30.0) 3 (12.0) NS Data are presented as n (%). NS indicates not significant (P >.05). classes to assist in identifying uterine defects. It is therefore obvious that the techniques used for measuring the LUS thickness and identifying uterine defects have not been consistent among different studies. Although some studies seem to give good results with different measurement techniques, 11,13,15 our experience suggests that measurement of only the myometrial layer should be more representative of the LUS thickness because 13.2% of our patients showed thickening of the LUS scar at the outer layer but thinning of the underlying myometrial layer. In addition, measurements that include the bladder mucosa and the underlying tissues are more likely influenced by the variation of the bladder size. Our finding of a difference in LUS thickness between the cesarean and control groups is similar to the findings of Gotoh et al 9 and Qureshi et al, 13 which showed that the LUS was thinner in women with previous cesarean delivery than in the control groups. However, the parity factor had not been examined in those studies. Our findings indicate that parity appears to be an important variable that affects the LUS thickness (Table 3). Also, our study suggests that sonographic LUS evaluation is potentially capable of identifying those patients with a thin or defective LUS, which could carry a higher risk of subsequent rupture when a trial of VBAC is attempted. However, because of the limited sample size, further studies are necessary to determine this relationship J Ultrasound Med 2004; 23:

7 Cheung et al The risk of uterine rupture in the presence of a LUS defect, whether identifiable sonographically, is unknown. Also, the cutoff value at which an extremely thin LUS on sonographic examination can predispose to a high risk of uterine rupture has yet to be identified. This obviously could not be determined in this study and, in our opinion, was unable to be determined in previously reported studies because of the inconsistency in the measuring technique and the method used to diagnose uterine rupture. Conversely, most studies suggest a high negative predictive value of LUS measurement with respect to the risk of rupture. 8 This may encourage obstetricians to offer patients a trial of labor, when the LUS thickness is shown to be normal sonographically. When using normal nulliparas as the reference, our study suggests that a normal LUS thickness should be at least 1 mm. It should be useful, when managing a planned trial of VBAC, to have an accurate method for diagnosing uterine defects. It remains to be established whether sonographic LUS assessment can be used to identify patients at high risk for uterine rupture should they undergo VBAC, thus reducing the prevalence of this complication. Although our study suggests that sonographic surveillance for a defective LUS is potentially possible, our study population was small, and the observers were not blinded. These preclude accurate estimates and limit the ability to establish clinically useful relationships. Therefore, further studies with larger numbers of subjects and blinded observers are needed to correlate the degree of LUS thinning with regard to the risk of subsequent uterine rupture. References 1. Flamm BL, Newman LA, Thomas SJ, Fallon D, Yoshida MM. Vaginal birth after cesarean section delivery: results of a 5-year multicenter collaborative study. Obstet Gynecol 1990; 76: Miller DA, Diaz FG, Paul RH. Vaginal birth after cesarean: a 10-year experience. Obstet Gynecol 1994; 84: Lydon-Rochelle M, Holt VL, Easterling TR, Martin DP. Risk of uterine rupture during labor among women with a prior cesarean delivery. N Engl J Med 2001; 345: Jones RO, Nagashima AW, Hatnett-Goodman MM, Goodlin RC. Rupture of low transverse cesarean scars during trial of labor. Obstet Gynecol 1991; 77: Leung AS, Leung EK, Paul RH. Uterine rupture after previous cesarean delivery: maternal and fetal consequences. Am J Obstet Gynecol 1993; 169: Society of Obstetricians and Gynaecologists of Canada. SOGC Clinical Practice Guidelines. Guidelines for vaginal birth after previous caesarean birth. J Obstet Gynaecol Can 2004; 26: American College of Obstetricians and Gynecologists. ACOG Practice Bulletin 54. Vaginal birth after previous cesarean delivery. Obstet Gynecol 2004; 104: Rozenberg P, Goffinet F, Philippe HJ, Nisand I. Ultrasonographic measurement of lower uterine segment to assess risk of defects of scarred uterus. Lancet 1996; 347: Gotoh H, Masuzaki H, Yoshida A, Yoshimura S, Miyamura T, Ishimaru T. Predicting incomplete uterine rupture with vaginal sonography during the late second trimester in women with prior cesarean. Obstet Gynecol 2000; 95: Brill Y, Kingdom J, Thomas J, et al. The management of VBAC at term: a survey of Canadian obstetricians. J Obstet Gynaecol Can 2003; 25: Michaels WH, Thompson HO, Boutt A, Schreiber FR, Michaels SL, Karo J. Ultrasound diagnosis of defects in the scarred lower uterine segment during pregnancy. Obstet Gynecol 1988; 71: Fukuda M, Fukuda K, Mochizuki M. Examination of previous caesarean section scars by ultrasound. Arch Gynecol Obstet 1988; 243: Qureshi B, Inafuku K, Oshima K, Masamoto H, Kanazawa K. Ultrasonographic evaluation of lower uterine segment to predict the integrity and quality of cesarean scar during pregnancy: a prospective study. Tohoku J Exp Med 1997; 183: Lonky NM, Worthen N, Ross MG. Prediction of cesarean section scars with ultrasound imaging during pregnancy. J Ultrasound Med 1989; 8: Suzuki S, Sawa R, Yoneyama Y, Asakura H, Araki T. Preoperative diagnosis of dehiscence of the lower uterine segment in patients with a single previous Caesarean section. Aust NZ J Obstet Gynaecol 2000; 40: J Ultrasound Med 2004; 23: