3312jum_online_Layout 1 11/24/14 10:19 AM Page 2181
ORIGINAL RESEARCH
Association Between Sonographic Cervical Appearance and Preterm Delivery After a History-Indicated Cerclage Emily S. Miller, MD, MPH, Susan E. Gerber, MD, MPH Objectives—To determine which transvaginal cervical sonographic characteristics are associated with preterm delivery after placement of a history-indicated cerclage. Methods—We conducted a case-control study of all women with a singleton gestation and a history-indicated cerclage placed at a single center between January 1, 2008, and October 1, 2012. Cerclages were placed between 12 and 16 weeks’ gestation. Cases were defined as women who delivered before 34 weeks’ gestation; controls delivered at or after 34 weeks. Transvaginal cervical characteristics, such as cervical funneling, proximal cervical length, distal cervical length, and total cervical length, obtained between 18 and 24 weeks were compared between the two groups. Bivariable, multivariable, and log rank analyses were performed. Results—A total of 124 women met inclusion criteria, with 17 (14%) delivering before 34 weeks. Cervical funneling, a proximal cervical length of less than 1.5 cm, and a total cervical length of less than 2.5 cm were associated with an increased odds of preterm birth before 34 weeks in the bivariable analysis [odds ratio (OR), 10.9 (95% confidence interval (CI), 2.3–62.8), 2.9 (95% CI, 1.0–8.5), and 4.5 (95% CI, 1.3–16.4), respectively], whereas distal cervical length of less than 1 cm was not significantly associated with delivery before 34 weeks [OR, 3.0 (95% CI, 0.8–11.1)]. In multivariable analysis, only cervical funneling remained associated with preterm delivery before 34 weeks [adjusted OR, 10.6 (95% CI, 2.2–51.5)]. Conclusions—In women with a history-indicated cerclage, cervical funneling is the only independently significant sonographic finding associated with an increased risk of preterm birth before 34 weeks.
Received December 13, 2013, from the Department of Obstetrics and Gynecology, Division of MaternalFetal Medicine, Northwestern University, Feinberg School of Medicine, Chicago, Illinois USA. Revision requested January 21, 2014. Revised manuscript accepted for publication March 26, 2014. Address correspondence to Emily S. Miller, MD, MPH, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Northwestern University, Feinberg School of Medicine, 250 E Superior St, Suite 05-2185, Chicago, IL 60611 USA. E-mail: [email protected] Abbreviations
CI, confidence interval; OR, odds ratio doi:10.7863/ultra.33.12.2181
Key Words—cerclage; cervical funneling; cervical insufficiency; cervical length; obstetric ultrasound; preterm birth
T
ransvaginal sonographic assessment of cervical length is used to assess the risk of preterm delivery in high-risk women.1 One subgroup of these high-risk women includes those with a history of cervical insufficiency. In these women, often based on a suggestive history, a vaginal cerclage is placed in the early second trimester.2,3 However, even after this intervention, these women remain at risk of preterm delivery. There is limited evidence in the published literature to inform surveillance after cerclage placement. Some authors have assessed the appearance of the cervix proximal to the cerclage as a window into the continuing pathophysiologic process.4–7 However, many of these prior studies incorporated heterogeneous cohorts of
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:2181–2186 | 0278-4297 | www.aium.org
3312jum_online_Layout 1 11/24/14 10:19 AM Page 2182
Miller and Gerber—Cervical Funneling After Cerclage
women, including combinations of women with historyindicated, sonographically indicated, and physical examination–indicated cerclages.4–7 Women in the latter groups are likely at the highest risk of both continuing cervical changes after cerclage as well as preterm delivery. Whether cervical shortening and funneling proximal to the stitch continue to pose an increased risk of preterm delivery in women with a history-indicated cerclage has been studied by one group, who did not find a statistically significant association between proximal cervical length and preterm birth; however, the study was limited by a small sample size.8 The purpose of this study was to assess the relationship between the transvaginal sonographic characteristics of the cervix in the second trimester and the risk of preterm birth before 34 weeks. Specifically, the objectives were to quantify the relationship between cervical length (total, proximal to the cerclage, and distal to the cerclage) and the risk of preterm birth and assess whether funneling is associated with preterm birth. We hypothesized that the presence of cervical funneling, but not cervical length, would be associated with preterm birth before 34 weeks.
Materials and Methods We conducted a retrospective case-control study of women with a singleton pregnancy who had a history-indicated cerclage placed between 12 and 16 weeks’ gestation at Northwestern Memorial Hospital between January 1, 2008, and October 1, 2012. These women were identified through a query of the operating room records during the study time frame. Clinical practice at the time of this study was to offer a history-indicated cerclage to patients who had 1 or more second-trimester losses suggestive of cervical insufficiency. Notably, this study did not include women who underwent cerclage placement on the basis of a short cervix on sonography. Women were included if they were 18 years or older, underwent history-indicated transvaginal cerclage placement, and subsequently had a transvaginal cervical length sonographic examination between 18 and 24 weeks’ gestation. As transvaginal cervical length measurement became universally performed at Northwestern in March 2008, data collection was limited to women who had their cerclage placed after this time. Women with multiple gestations, women who did not complete a transvaginal cervical length sonographic examination, and women with a stillbirth or termination of pregnancy were excluded from the analysis. This study was approved by the Northwestern Institutional Review Board before its initiation. The medical charts of women who met inclusion criteria were reviewed, and baseline demographic factors
2182
as well as the pertinent obstetric and gynecologic histories were abstracted. Specifically, we ascertained maternal age, race/ethnicity, insurance status, marital status, tobacco use, body mass index at the time of cerclage placement, obstetric history (the number and gestational age of prior preterm births and the number of any prior term deliveries), and any prior cervical excisional procedures. Details about the clinical care that were abstracted included the use of adjuvant intramuscular progesterone and the surgical technique employed (ie, McDonald versus Shirodkar technique and suture type). All women had a cervical length assessment as a part of their routine obstetric anatomic survey. The cervical length images were obtained according to the method of Iams et al.1 All of these second-trimester transvaginal cervical length sonograms were reviewed by 1 of 2 maternalfetal medicine subspecialists who were blinded to the gestational age at delivery. The location of the cerclage was identified as an echo-dense structure in the cervical stroma (Figure 1). Variables measured included total cervical length, length proximal to the cerclage, and length distal to the cerclage. Cervical length cutoffs were chosen by using those most highly associated with preterm birth before 34 weeks, examined in intervals of 0.5 cm. The presence of cervical funneling, defined as protrusion of membranes into the cervical canal, was recorded. An example of cervical funneling is shown in Figure 2. Clinical care did not change on the basis of the cervical characteristics observed on sonography. Notably, no women in this study had a cerclage revision performed.
Figure 1. Sonographic appearance of a cervix with a cerclage in situ. Asterisk indicates the location of the cerclage; solid line, proximal cervical length; and dashed line, distal cervical length.
J Ultrasound Med 2014; 33:2181–2186
3312jum_online_Layout 1 11/24/14 10:19 AM Page 2183
Miller and Gerber—Cervical Funneling After Cerclage
Delivery data obtained from the chart included gestational age at delivery and indication for delivery (ie, induced or spontaneous). For the primary analysis, women were dichotomized into those who delivered before 34 weeks’ gestation and those that delivered at or after 34 weeks. A secondary analysis was performed only for those who had spontaneous labor before 34 weeks. Bivariable analyses were also performed, using gestational age cutoffs of 28 and 24 weeks to ascertain whether the associations persisted at various cut points. Bivariable comparisons between groups were performed with the Mann-Whitney U, χ2, or Fisher exact test, as appropriate (Stata 11.1 IC; StataCorp, College Station, TX), and results are reported as odds ratios (ORs) and 95% confidence intervals (CIs). A Kaplan-Meier curve was developed for each sonographic characteristic, and a logrank was used to test the statistical significance of each. Finally, the independence of the association between cervical characteristics and preterm birth before 34 weeks was evaluated by multivariable logistic regression, and results are reported as adjusted ORs and 95% CIs. Variables included in this assessment were those with a significant (P < .05) association with delivery before 34 weeks in the bivariable analysis. A separate regression was performed with potential clinical confounders (included on the basis of clinical plausibility) to assess their effect on the outcome. Finally, interaction terms were created between the cervical length and gestational age at cervical length measurement and the presence of cervical funneling as well as the gestational age at cervical length measurement. These were included in the regression to assess whether the gestational age at ascertainment of the measurement led to any effect modification.
Figure 2. Example of cervical funneling with a cerclage in situ.
J Ultrasound Med 2014; 33:2181–2186
Results During the study period, 140 women underwent a historyindicated cerclage and had a subsequent transvaginal sonogram obtained for cervical length between 18 and 24 weeks. Of these women, 16 were multiple gestations, leaving 124 women for analysis. During the study period, 111 (90%) history-indicated cerclages were performed by the McDonald technique, and 122 (98%) used a 5-mm Mersilene polyester fiber suture (Ethicon Endo-Surgery, Inc, Blue Ash, OH). The mean gestational age ± SD at delivery for the cases was 27.9 ± 4.8 weeks, compared to 38.5 ± 1.4 weeks for the controls. Of the women meeting inclusion criteria, 17 (13.7%) delivered before 34 weeks’ gestation. Of these, 16 (94%) had spontaneous preterm labor, leading to their preterm delivery, and 1 (6%) was induced for severe preeclampsia. The patient characteristics of the study population are shown in Table 1. Those who delivered at or after 34 weeks were more likely to be white and were more likely to be married. Intramuscular progesterone was used by some women due to their history of a preterm birth after 20 weeks. No women received progesterone in response to the sonographic changes. The bivariable analysis of the association between transvaginal cervical characteristics and preterm delivery is shown in Table 2. Characteristics that were significantly associated with a delivery before 34 weeks included the presence of cervical funneling, a proximal cervical length of less than 1.5 cm, and a total cervical length of less than 2.5 cm. Forty-seven percent of women with a secondtrimester cervical funnel present delivered before 34 weeks, compared to 8% of women without a funnel. Similarly, 47% of women with a short proximal cervical length delivered before 34 weeks, compared to 23% of women with a proximal cervical length of at least 1.5 cm. In addition, 36% of women with a short cervix (cervical length
ORIGINAL RESEARCH
Association Between Sonographic Cervical Appearance and Preterm Delivery After a History-Indicated Cerclage Emily S. Miller, MD, MPH, Susan E. Gerber, MD, MPH Objectives—To determine which transvaginal cervical sonographic characteristics are associated with preterm delivery after placement of a history-indicated cerclage. Methods—We conducted a case-control study of all women with a singleton gestation and a history-indicated cerclage placed at a single center between January 1, 2008, and October 1, 2012. Cerclages were placed between 12 and 16 weeks’ gestation. Cases were defined as women who delivered before 34 weeks’ gestation; controls delivered at or after 34 weeks. Transvaginal cervical characteristics, such as cervical funneling, proximal cervical length, distal cervical length, and total cervical length, obtained between 18 and 24 weeks were compared between the two groups. Bivariable, multivariable, and log rank analyses were performed. Results—A total of 124 women met inclusion criteria, with 17 (14%) delivering before 34 weeks. Cervical funneling, a proximal cervical length of less than 1.5 cm, and a total cervical length of less than 2.5 cm were associated with an increased odds of preterm birth before 34 weeks in the bivariable analysis [odds ratio (OR), 10.9 (95% confidence interval (CI), 2.3–62.8), 2.9 (95% CI, 1.0–8.5), and 4.5 (95% CI, 1.3–16.4), respectively], whereas distal cervical length of less than 1 cm was not significantly associated with delivery before 34 weeks [OR, 3.0 (95% CI, 0.8–11.1)]. In multivariable analysis, only cervical funneling remained associated with preterm delivery before 34 weeks [adjusted OR, 10.6 (95% CI, 2.2–51.5)]. Conclusions—In women with a history-indicated cerclage, cervical funneling is the only independently significant sonographic finding associated with an increased risk of preterm birth before 34 weeks.
Received December 13, 2013, from the Department of Obstetrics and Gynecology, Division of MaternalFetal Medicine, Northwestern University, Feinberg School of Medicine, Chicago, Illinois USA. Revision requested January 21, 2014. Revised manuscript accepted for publication March 26, 2014. Address correspondence to Emily S. Miller, MD, MPH, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Northwestern University, Feinberg School of Medicine, 250 E Superior St, Suite 05-2185, Chicago, IL 60611 USA. E-mail: [email protected] Abbreviations
CI, confidence interval; OR, odds ratio doi:10.7863/ultra.33.12.2181
Key Words—cerclage; cervical funneling; cervical insufficiency; cervical length; obstetric ultrasound; preterm birth
T
ransvaginal sonographic assessment of cervical length is used to assess the risk of preterm delivery in high-risk women.1 One subgroup of these high-risk women includes those with a history of cervical insufficiency. In these women, often based on a suggestive history, a vaginal cerclage is placed in the early second trimester.2,3 However, even after this intervention, these women remain at risk of preterm delivery. There is limited evidence in the published literature to inform surveillance after cerclage placement. Some authors have assessed the appearance of the cervix proximal to the cerclage as a window into the continuing pathophysiologic process.4–7 However, many of these prior studies incorporated heterogeneous cohorts of
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:2181–2186 | 0278-4297 | www.aium.org
3312jum_online_Layout 1 11/24/14 10:19 AM Page 2182
Miller and Gerber—Cervical Funneling After Cerclage
women, including combinations of women with historyindicated, sonographically indicated, and physical examination–indicated cerclages.4–7 Women in the latter groups are likely at the highest risk of both continuing cervical changes after cerclage as well as preterm delivery. Whether cervical shortening and funneling proximal to the stitch continue to pose an increased risk of preterm delivery in women with a history-indicated cerclage has been studied by one group, who did not find a statistically significant association between proximal cervical length and preterm birth; however, the study was limited by a small sample size.8 The purpose of this study was to assess the relationship between the transvaginal sonographic characteristics of the cervix in the second trimester and the risk of preterm birth before 34 weeks. Specifically, the objectives were to quantify the relationship between cervical length (total, proximal to the cerclage, and distal to the cerclage) and the risk of preterm birth and assess whether funneling is associated with preterm birth. We hypothesized that the presence of cervical funneling, but not cervical length, would be associated with preterm birth before 34 weeks.
Materials and Methods We conducted a retrospective case-control study of women with a singleton pregnancy who had a history-indicated cerclage placed between 12 and 16 weeks’ gestation at Northwestern Memorial Hospital between January 1, 2008, and October 1, 2012. These women were identified through a query of the operating room records during the study time frame. Clinical practice at the time of this study was to offer a history-indicated cerclage to patients who had 1 or more second-trimester losses suggestive of cervical insufficiency. Notably, this study did not include women who underwent cerclage placement on the basis of a short cervix on sonography. Women were included if they were 18 years or older, underwent history-indicated transvaginal cerclage placement, and subsequently had a transvaginal cervical length sonographic examination between 18 and 24 weeks’ gestation. As transvaginal cervical length measurement became universally performed at Northwestern in March 2008, data collection was limited to women who had their cerclage placed after this time. Women with multiple gestations, women who did not complete a transvaginal cervical length sonographic examination, and women with a stillbirth or termination of pregnancy were excluded from the analysis. This study was approved by the Northwestern Institutional Review Board before its initiation. The medical charts of women who met inclusion criteria were reviewed, and baseline demographic factors
2182
as well as the pertinent obstetric and gynecologic histories were abstracted. Specifically, we ascertained maternal age, race/ethnicity, insurance status, marital status, tobacco use, body mass index at the time of cerclage placement, obstetric history (the number and gestational age of prior preterm births and the number of any prior term deliveries), and any prior cervical excisional procedures. Details about the clinical care that were abstracted included the use of adjuvant intramuscular progesterone and the surgical technique employed (ie, McDonald versus Shirodkar technique and suture type). All women had a cervical length assessment as a part of their routine obstetric anatomic survey. The cervical length images were obtained according to the method of Iams et al.1 All of these second-trimester transvaginal cervical length sonograms were reviewed by 1 of 2 maternalfetal medicine subspecialists who were blinded to the gestational age at delivery. The location of the cerclage was identified as an echo-dense structure in the cervical stroma (Figure 1). Variables measured included total cervical length, length proximal to the cerclage, and length distal to the cerclage. Cervical length cutoffs were chosen by using those most highly associated with preterm birth before 34 weeks, examined in intervals of 0.5 cm. The presence of cervical funneling, defined as protrusion of membranes into the cervical canal, was recorded. An example of cervical funneling is shown in Figure 2. Clinical care did not change on the basis of the cervical characteristics observed on sonography. Notably, no women in this study had a cerclage revision performed.
Figure 1. Sonographic appearance of a cervix with a cerclage in situ. Asterisk indicates the location of the cerclage; solid line, proximal cervical length; and dashed line, distal cervical length.
J Ultrasound Med 2014; 33:2181–2186
3312jum_online_Layout 1 11/24/14 10:19 AM Page 2183
Miller and Gerber—Cervical Funneling After Cerclage
Delivery data obtained from the chart included gestational age at delivery and indication for delivery (ie, induced or spontaneous). For the primary analysis, women were dichotomized into those who delivered before 34 weeks’ gestation and those that delivered at or after 34 weeks. A secondary analysis was performed only for those who had spontaneous labor before 34 weeks. Bivariable analyses were also performed, using gestational age cutoffs of 28 and 24 weeks to ascertain whether the associations persisted at various cut points. Bivariable comparisons between groups were performed with the Mann-Whitney U, χ2, or Fisher exact test, as appropriate (Stata 11.1 IC; StataCorp, College Station, TX), and results are reported as odds ratios (ORs) and 95% confidence intervals (CIs). A Kaplan-Meier curve was developed for each sonographic characteristic, and a logrank was used to test the statistical significance of each. Finally, the independence of the association between cervical characteristics and preterm birth before 34 weeks was evaluated by multivariable logistic regression, and results are reported as adjusted ORs and 95% CIs. Variables included in this assessment were those with a significant (P < .05) association with delivery before 34 weeks in the bivariable analysis. A separate regression was performed with potential clinical confounders (included on the basis of clinical plausibility) to assess their effect on the outcome. Finally, interaction terms were created between the cervical length and gestational age at cervical length measurement and the presence of cervical funneling as well as the gestational age at cervical length measurement. These were included in the regression to assess whether the gestational age at ascertainment of the measurement led to any effect modification.
Figure 2. Example of cervical funneling with a cerclage in situ.
J Ultrasound Med 2014; 33:2181–2186
Results During the study period, 140 women underwent a historyindicated cerclage and had a subsequent transvaginal sonogram obtained for cervical length between 18 and 24 weeks. Of these women, 16 were multiple gestations, leaving 124 women for analysis. During the study period, 111 (90%) history-indicated cerclages were performed by the McDonald technique, and 122 (98%) used a 5-mm Mersilene polyester fiber suture (Ethicon Endo-Surgery, Inc, Blue Ash, OH). The mean gestational age ± SD at delivery for the cases was 27.9 ± 4.8 weeks, compared to 38.5 ± 1.4 weeks for the controls. Of the women meeting inclusion criteria, 17 (13.7%) delivered before 34 weeks’ gestation. Of these, 16 (94%) had spontaneous preterm labor, leading to their preterm delivery, and 1 (6%) was induced for severe preeclampsia. The patient characteristics of the study population are shown in Table 1. Those who delivered at or after 34 weeks were more likely to be white and were more likely to be married. Intramuscular progesterone was used by some women due to their history of a preterm birth after 20 weeks. No women received progesterone in response to the sonographic changes. The bivariable analysis of the association between transvaginal cervical characteristics and preterm delivery is shown in Table 2. Characteristics that were significantly associated with a delivery before 34 weeks included the presence of cervical funneling, a proximal cervical length of less than 1.5 cm, and a total cervical length of less than 2.5 cm. Forty-seven percent of women with a secondtrimester cervical funnel present delivered before 34 weeks, compared to 8% of women without a funnel. Similarly, 47% of women with a short proximal cervical length delivered before 34 weeks, compared to 23% of women with a proximal cervical length of at least 1.5 cm. In addition, 36% of women with a short cervix (cervical length
