Goetzinger, Katherine R. MD; Cahill, Alison G. MD, MSCI; Macones, George A. MD, MSCE; Odibo, Anthony O. MD, MSCE
The prevalence of echogenic bowel on routine second-trimester ultrasonograph ranges from 0.2% to 1.8%.1 The differential diagnosis for this finding is broad and includes normal variant, primary gastrointestinal pathology, congenital viral infection, cystic fibrosis, aneuploidy, and intra-amniotic bleeding.2–7 Previous reports have also suggested an increased incidence of intrauterine growth restriction (IUGR) and intrauterine fetal demise in fetuses with echogenic bowel with reported incidences ranging from 14.0% to 23.3% for IUGR and from 3.8% to 8.0% for intrauterine fetal demise.7–11 Fetal redistribution of blood flow to vital organs with resultant bowel hypoperfusion and ischemia is the proposed explanation for the hyperechoic appearance of the fetal bowel in these situations.8,10,12
Despite these associations, prior studies have not been able to estimate a risk for each of these adverse outcomes with any precision. This is likely the result of the overall rarity of the primary outcomes in relation to the small number of cases of echogenic bowel observed in these studies. Estimating these risks will be useful in counseling patients and may also provide insight into the appropriate antenatal management strategy for these patients. The aims of this study were to estimate the risk of IUGR and intrauterine fetal demise associated with the finding of echogenic bowel on second-trimester ultrasonography using a robust ultrasound and genetics database and then to further refine those risk estimates by performing subanalyses excluding fetuses with chromosomal abnormalities, congenital infections, and associated anomalies.
MATERIALS AND METHODS
This is a retrospective cohort study of consecutive patients with singleton gestations referred to Washington University Medical Center in St Louis for second-trimester ultrasonography between 1990 and 2008. Study approval was obtained from the institutional review board at Washington University. Demographic information, maternal medical and obstetric history, indication for ultrasonography, ultrasonographic findings, and any genetic screening or diagnostic results are entered into a perinatal database at the time of the ultrasonography for all patients seen at our institution. All pregnancy and neonatal outcome information is also routinely obtained and entered into this database by dedicated outcome coordinators. For the greater than 90% of patients who deliver within our healthcare system, outcome information is obtained from the electronic medical record. The small remainder of patients who do not deliver within our healthcare system are given a standardized questionnaire regarding pregnancy outcome and neonatal follow-up. If a questionnaire is not returned within 4 weeks from the anticipated date of delivery, the patient is then contacted by telephone. In cases in which a patient cannot be reached, the referring healthcare provider is contacted to obtain outcome information. Patients with multiple gestations and those with incomplete outcome information were excluded from the study.
The primary exposure in this study was the presence of fetal echogenic bowel, defined as bowel with a sonographic density equal to or greater than that of surrounding bone at the time of second-trimester ultrasonography. (Fig. 1) The frequency of the ultrasonographic transducer ranged from 3.5 to 5.0 MHz depending on maternal body habitus. In cases in which the bowel was suspected to be echogenic, the ultrasonographic gain was turned down as low as possible. If the bowel continued to meet this criterion, then the diagnosis of echogenic bowel was made. All diagnoses of echogenic bowel were confirmed by a maternal-fetal medicine subspecialist at the time of the ultrasonography. Our primary outcomes of interest included IUGR and intrauterine fetal demise. IUGR was defined as birth weight less than the 10th percentile for gestational age using national standards derived by the Alexander growth curve.13 Intrauterine fetal demise was defined as fetal death at 20 weeks or more of gestation.
Baseline characteristics as well as the incidence of the primary outcomes were compared between patients with and without fetal echogenic bowel. Student's t tests were used to compare continuous variables, and chi-square and Fisher's exact tests were used to compare categorical variables. The incidence of aneuploidy, cytomegalovirus (CMV) infection, cystic fibrosis, and associated anomalies in patients with fetal echogenic bowel was determined. Univariable analysis was used to estimate the relative risk of the primary outcomes. Multivariable logistic regression was then used to further define these risks controlling for potential confounding factors. Factors identified as significant in the univariable analysis as well as those with biologic plausibility or historic significance were considered in the logistic regression models. Nonsignificant variables were removed from the models in a backward, stepwise fashion to obtain adjusted odds ratios (ORs) and 95% confidence intervals (CIs). Differences in the hierarchical models were tested using the likelihood ratio test or Wald test. Only variables that were statistically significant were included in the final model. To further examine the association between echogenic bowel and intrauterine fetal demise, stratified analyses based on the presence or absence of IUGR were also performed.
Given the known risk of IUGR and intrauterine fetal demise in patients with aneuploidy, CMV infection, or both, a subanalysis was performed after excluding these cases. Our institutional standard for the evaluation of echogenic bowel found on second-trimester ultrasonography includes patient counseling toward invasive diagnostic testing for aneuploidy and congenital infection as well as parental cystic fibrosis carrier status determination. Patients who accept amniocentesis have their amniotic fluid sent for karyotype as well as CMV, toxoplasmosis, and parvovirus polymerase chain reaction. Patients who decline amniocentesis are offered maternal serologic testing for congenital infection, recognizing that this is not diagnostic. The finding of echogenic bowel is also relayed to the pediatricians for postnatal evaluation, which includes CMV polymerase chain reaction from a neonatal specimen such as urine or saliva. Postnatal karyotype is also sent if newborn examination findings are suspicious for aneuploidy. Cases without a documented newborn examination were considered to have an unknown chromosomal status and were also excluded in this subanalysis. Cystic fibrosis cases were not excluded from the subanalysis because there is currently no known evidence-based increased risk for IUGR or intrauterine fetal demise in these patients.
Finally, to evaluate the primary outcomes in relationship to cases of isolated echogenic bowel, an additional subanalysis was performed after excluding all major structural anomalies. Given the association between abnormal serum analytes and adverse pregnancy outcomes, particularly elevated maternal serum alpha fetoprotein (AFP) in cases of echogenic bowel, the analysis was also repeated after excluding all patients with abnormal second-trimester serum screening results.8,10 As the approach to serum screening has evolved over the 18-year period of this study, an abnormal second-trimester serum screen was defined as 1) an abnormal triple screen (β-human chorionic gonadotropin, estriol, and maternal serum AFP); 2) an abnormal quadruple screen (β-human chorionic gonadotropin, estriol, maternal serum AFP, and inhibin); or 3) an isolated abnormal maternal serum AFP result.
The approach to the primary and subanalyses is outlined in Figure 2. P<.05 was considered statistically significant. All statistical analyses were performed using STATA 10 Special Edition.
Of 72,368 patients in the cohort, complete pregnancy outcome data were available for 64,048 patients (88.5%). There were 260 cases of echogenic bowel, giving an incidence of 0.4% in our patient population. Of these 260 cases of echogenic bowel, there were 13 cases (5%) of trisomy 21, one case (0.4%) of trisomy 18, two cases (0.8%) of trisomy 13, one case (0.4%) of chromosomal mosaicism, five documented cases (1.9%) of CMV infection, six documented cases (2.3%) of cystic fibrosis, and one case (0.4%) of both documented cystic fibrosis and chromosomal mosaicism. Of the remaining fetuses with echogenic bowel, 43 cases had associated anomalies, and 188 (72.3%) cases had echogenic bowel as an isolated finding. There were 579 cases (0.9%) of intrauterine fetal demise and 8,173 cases (12.8%) of IUGR in the study cohort. Maternal demographics and pregnancy characteristics for our population are shown in Table 1. Compared with the 260 patients with echogenic bowel, the 63,788 patients without echogenic bowel were of similar age, gravidity, parity, gestational age at the time of ultrasonography, and also had similar rates of tobacco use, chronic hypertension, preeclampsia, pre-existing diabetes mellitus, and gestational diabetes. However, patients with a diagnosis of echogenic bowel were statistically more likely to be of white race. Although patients with a diagnosis of echogenic bowel were more likely to have reported vaginal bleeding during the pregnancy, only 15% of patients with echogenic bowel who actually experienced an intrauterine fetal demise reported a positive history of vaginal bleeding. A total of 47.3% of the echogenic bowel cases were diagnosed in the first 9 years of the study, and 52.7% were diagnosed during the subsequent 9 years.
Patients with echogenic bowel were at an increased risk for intrauterine fetal demise compared with those without echogenic bowel (relative risk 8.6, 95% CI 5.6–13.2). This association remained statistically significant even after adjusting for African American race (adjusted OR 10.3, 95% CI 6.5–16.5). Similarly, patients with echogenic bowel were also at an increased risk for IUGR compared with those without echogenic bowel (relative risk 1.6, 95% CI 1.3–2.1). After adjusting for African American race, tobacco use, preeclampsia, and pre-existing diabetes mellitus, echogenic bowel remained associated with an increased risk for IUGR (adjusted OR 2.1, 95% CI 1.5–3.0).
To further delineate this relationship, all patients with documented cases of CMV infection, aneuploidy, or unknown chromosomal status were excluded, leaving 50,171 patients for analysis. Results from this subanalysis are shown in Table 2. The association between echogenic bowel and both intrauterine fetal demise (adjusted OR 9.6, 95% CI 5.8–15.9) and IUGR (adjusted OR 2.1, 95% CI 1.5–2.9) remained statistically significant after controlling for potential confounders. The median gestational age at the time of intrauterine fetal demise in the patients with echogenic bowel was 24.1 weeks (range, 20.1–30.0 weeks). Given the well-established association between IUGR and intrauterine fetal demise, the relationship between echogenic bowel and intrauterine fetal demise was also evaluated after stratifying patients based on the presence or absence of IUGR. Although the risk for intrauterine fetal demise was fourfold higher in the growth-restricted fetuses, there still remained a statistically significant risk of intrauterine fetal demise (relative risk 6.2, 95% CI 2.0–19.2) in those fetuses whose growth was appropriate for gestational age. Given the relatively small number of fetal demise cases in each stratum, an adjusted analysis could not reliably be performed.
Finally, to estimate the risk of intrauterine fetal demise and IUGR in patients with isolated echogenic bowel, all patients with major anomalies (n=1,180) were excluded from the cohort. There were 188 cases of isolated echogenic bowel in the remaining patient population. Patients with isolated echogenic bowel were at increased risk for IUGR compared with those with no major fetal anomalies (19.7% compared with 12.9%; adjusted OR 2.1, 95% CI 1.4–3.0) after adjusting for African American race, tobacco use, pre-existing diabetes mellitus, and preeclampsia. Patients with isolated echogenic bowel were also at increased risk for intrauterine fetal demise compared with those with no major fetal anomalies (5.3% compared with 0.8%; adjusted OR 7.4, 95% CI 3.9–14.1). After stratification, this risk association remained significant in both the growth-restricted and appropriate for gestational age groups. In the final subanalysis, all patients with abnormal second-trimester serum screening results (n=5,148) were excluded. The positive risk association between isolated echogenic bowel and both IUGR and intrauterine fetal demise remained significant in patients with normal serum screening results (Table 3).
The median gestational age at the time of intrauterine fetal demise in patients with isolated echogenic bowel was 23.7 weeks (range, 20.1–27.0 weeks) compared with 23.6 weeks (range, 20.1–40.6 weeks) in patients without isolated echogenic bowel (P=.80). Although the median gestational age at the time of intrauterine fetal demise was not significantly different between these groups, the range of gestational ages at which intrauterine fetal demise occurred was much wider in those without isolated echogenic bowel. Of the 149 patients with isolated echogenic bowel who did not have either IUGR or an intrauterine fetal demise, 93 (62.4%) had resolution of the echogenic bowel on follow-up ultrasonography, 18 (12.1%) had no resolution, and 38 (25.5%) did not have a follow-up ultrasonogram at our institution.
Our study demonstrates that the finding of echogenic bowel on second-trimester ultrasonography, even when isolated, is associated with an increased risk for both IUGR and intrauterine fetal demise. Using a rigorous stepwise approach to the analysis, we observed a twofold increase in IUGR in patients with both isolated echogenic bowel as well as in patients with associated anomalies and an even more substantial risk of intrauterine fetal demise in these patients.
Using a large retrospective cohort of patients from our institution's 18-year experience, our study was able to evaluate adverse outcomes and quantify risks for IUGR and intrauterine fetal demise in pregnancies with fetal echogenic bowel diagnosed on a second-trimester ultrasonogram. In 1993, Nyberg et al7 reported a 6.5-fold increased risk for adverse outcomes in 95 fetuses with echogenic bowel compared with 110 control fetuses. This relative risk decreased to 4.9 but remained statistically significant when evaluating only isolated cases of echogenic bowel. Although this study did produce quantitative estimates, their definition of adverse outcome was a composite that included aneuploidy and cystic fibrosis in addition to IUGR and intrauterine fetal demise. Our study demonstrates that the association between echogenic bowel and IUGR and intrauterine fetal demise persists even when patients with other risk factors are excluded from the analysis. This suggests an association that is independent of aneuploidy, congenital infection, and associated anomalies.
In a case–control study of 156 fetuses diagnosed with echogenic bowel before 24 weeks of gestation, IUGR, oligohydramnios, and an elevated maternal serum AFP level were observed more frequently in cases of fetal demise compared with liveborn neonates. Despite these trends, only oligohydramnios and elevated maternal serum AFP levels were found to be independently associated with intrauterine fetal demise in their logistic regression model.10 In contrast, IUGR was present in two of four (50%) cases of fetal demise in a prospective study by Ghose et al.11 Findings from our stratified analysis demonstrate a significant risk association between echogenic bowel, both isolated and nonisolated, and intrauterine fetal demise regardless of the presence or absence of fetal growth restriction. These findings suggest that the presence of IUGR does not completely account for the increased intrauterine fetal demise risk in patients with echogenic bowel, although causal inferences should only be made with caution in observational studies such as ours. A similar phenomenon also has been observed in cases of gastroschisis in which perinatal death is postulated to occur through a pathway that is independent of fetal growth restriction.14,15
One of the most clinically important findings from our study is the early gestational age at which the fetal demises occurred in patients with echogenic bowel. The median gestational age of demise in the nonisolated echogenic bowel group was 24.1 weeks compared with 23.7 weeks in the isolated echogenic bowel group. Of note, all fetal demises occurred before 30 weeks of gestation in those with echogenic bowel in our population. Al-Koutaly et al10 observed a similar median gestational age of fetal demise (22.0 weeks), although the upper limit of the gestational age range of fetal demise in that study was 39 weeks. Our finding certainly has clinical implications for the antenatal monitoring of these fetuses. Given that most antenatal surveillance protocols do not begin until 28–32 weeks of gestation, this may preclude the use of antenatal surveillance as a means to identify and intervene on those fetuses at risk for imminent demise.
In patients with cystic fibrosis, aneuploidy, or both, the characteristic ultrasound appearance of echogenic bowel is thought to be the result of thickened, viscous meconium caused by decreased levels of microvillar enzymes in the gut and hypoperistalsis.7,16,17 Ewer et al12 proposed that echogenic bowel may also be a reflection of intrauterine gut ischemia, an effect that may be observed in growth-restricted fetuses who are exhibiting a “brain-sparing” circulatory response with resultant hypoperfusion of the splanchnic vessels. More recently, Doppler studies of the splanchnic circulation in growth-restricted fetuses with echogenic bowel have demonstrated the opposite effect, a relative hyperperfusion of the fetal gut circulation represented by a low pulsatility index in both the celiac trunk and superior mesenteric artery.18 Although the precise mechanism is still unclear, it appears that abnormal vascular supply to the fetal gut contributes to the echogenic appearance of fetal bowel, thereby providing biologic plausibility to the association with IUGR observed in our study. In fetuses that experience intrauterine fetal demise in the absence of growth restriction, the mechanism is less clear. Prior studies have demonstrated that elevated maternal serum AFP levels are associated with a worse prognosis in fetuses with echogenic bowel.8,10 Elevated maternal serum AFP levels can be seen in cases of intra-amniotic bleeding and may represent an alteration in the maternal–placental barrier. Al-Kouatly et al10 speculated that echogenic bowel may appear before fetal demise in such cases, although only one of nine patients (11.1%) with fetal demise in that study reported a history of vaginal bleeding during the pregnancy. This is similar to the findings from our study in which only 15% of patients with echogenic bowel who had a subsequent intrauterine fetal demise reported a history of vaginal bleeding.
Strengths of our study include our robust ultrasonography and genetics database from which obstetric history, pregnancy outcome, and neonatal outcome data are extracted. Our large sample size allowed us to estimate risks for our primary outcomes, and our staged approach to the analysis further allowed us to determine whether these associations were independent of karyotype, congenital infection, other associated malformations, and abnormal second-trimester serum screening results. In addition, the incidence of aneuploidy, infection, and cystic fibrosis observed in our study is within similar ranges as those reported in prior studies, thereby validating our cohort.5,10,11,19 Despite our large numbers, the analysis was still somewhat limited by the rarity of both the exposure and the outcome of intrauterine fetal demise, resulting in wide CIs and preventing adjusted subgroup analysis.
Limitations of our study include its retrospective design in which data collection is limited to chart review and is contingent on the correct coding of ultrasonographic findings and karyotype data. In addition, all patients in the cohort did not undergo invasive aneuploidy screening or testing for congenital infection or cystic fibrosis; however, our institution uses coordinators who are dedicated to obtaining neonatal outcomes; therefore, the possibility of missing an affected child diagnosed after birth is minimal. Furthermore, if there were any undiagnosed cases of CMV and aneuploidy in the nonechogenic bowel group, who did not undergo a standardized testing approach, our results would be biased toward the null hypothesis, producing an underestimation of association. If this misclassification bias does exist in our cohort, then our reported estimates of risk may actually be conservative. To further ensure the validity of our cohort, all patients with incomplete outcome information and unknown chromosomal status were excluded in the subanalysis. Finally, echogenic bowel may be perceived as a somewhat subjective diagnosis. All ultrasonograms in our unit are interpreted by a maternal-fetal medicine subspecialist, and all ultrasonographic data are entered into our database prospectively at the time of the examination. Again, if misclassification of echogenic bowel diagnoses did occur in our cohort, this would also bias our results toward the null hypothesis. When we divided our 18-year study cohort in half, we noticed a comparable proportion of echogenic bowel cases being diagnosed in the first and second halves of our study, indicating that the diagnosis has been stable over time as ultrasonographic technology and practice have advanced.
In conclusion, our large retrospective study confirms the association between echogenic bowel diagnosed on second-trimester ultrasonogram and an increased risk for IUGR and intrauterine fetal demise. Our study also demonstrates that this risk association is independent of karyotypic abnormalities and congenital infections and even persists when echogenic bowel is observed as an isolated finding. Serial growth ultrasonograms may be warranted in these patients, and although antenatal surveillance may be considered, its value is inconclusive given the early gestational age at which fetal demise occurs in these patients.
1.Bronshtein M, Blazer S, Zimmer EZ. The gastrointestinal tract and abdominal wall. In: Callen PW, editor. Ultrasonography in obstetrics and gynecology. 5th ed. Philadelphia (PA): Saunders Elsevier; 2008. p. 587–639.
2.Fakhary J, Reiser M, Shapiro LR, Scheuchter A, Pait LP, Glennon A. Increased echogenicity in the lower fetal abdomen: a common normal variant in the second trimester. J Ultrasound Med 1986;5:489–92.
3.Caspi B, Elchalal U, Lancet M, Chemke J. Prenatal diagnosis of cystic fibrosis: ultrasonographic appearance of meconium ileus in the fetus. Prenat Diagn 1998;8:379–82.
4.Forouzan I. Fetal abdominal echogenic mass: an early sign of intrauterine cytomegalovirus infection. Obstet Gynecol 1992;80:535–7.
5.Strocker AM, Snijders RJ, Carlson DE, Greene N, Gregory KD, Walla CA, et al. Fetal echogenic bowel: parameters to be considered in the differential diagnosis. Ultrasound Obstet Gynecol 2000;16:519–23.
6.Sepulveda W, Reid R, Nicolaidis P, Prendiville O, Chapman RS, Fisk NM. Second-trimester echogenic bowel and intraamniotic bleeding: association between fetal bowel echogenicity and amniotic fluid spectrophotometry at 410 nm. Am J Obstet Gynecol 1996;174:839–42.
7.Nyberg DA, Dubinsky T, Resta RG, Mahony BS, Hickok DE, Luthy DA. Echogenic fetal bowel during the second trimester: clinical importance. Radiology 1993;188:527–31.
8.Achiron R, Seidman D, Horowitz A, Mashiach S, Goldman B, Lipitz S. Hyperechogenic fetal bowel and elevated serum alpha-fetoprotein: a poor fetal prognosis. Obstet Gynecol 1996;88:368–71.
9.MacGregor SN, Tamura R, Sabbagha R, Brenhofer JK, Kambich MP, Pergament E. Isolated hyperechoic fetal bowel: significance and implications for management. Am J Obstet Gynecol 1995;173:1254–8.
10.Al-Kouatly HB, Chasen ST, Karam AK, Ahner R, Chervenak FA. Factors associated with fetal demise in fetal echogenic bowel. Am J Obset Gynecol 2001;185:1039–43.
11.Ghose I, Mason GC, Martinez D, Harrison KL, Evans JA, Ferriman EL, et al. Hyperechogenic fetal bowel: a prospective analysis of sixty consecutive cases. BJOG 2000;107:426–9.
12.Ewer AK, McHugo J, Chapman S, Newwell SJ. Fetal echogenic gut: a marker of intrauterine gut ischemia? Arch Dis Child 1993;69:510–3.
13.Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan M. A United States national reference for fetal growth. Obstet Gynecol 1996;87:163–8.
14.Towers CV, Carr MH. Antenatal fetal surveillance in pregnancies complicated by gastroschisis. Am J Obstet Gynecol 2008;198:686.e1–5.
15.Adair CD, Rosnes J, Frye AH, Burrus DR, Nelson LH, Veille JC. The role of antepartum surveillance in the management of gastroschisis. Int J Gynaecol Obstet 1996;52:141–4.
16.Scioscia AL, Pretorius DH, Burdorick NE, Cahill TC, Axelrod FT, Leopold GR. Second-trimester echogenic bowel and chromosomal abnormalities. Am J Obstet Gynecol 1992;167:889–94.
17.Brock DJ. A comparative study of microvillar enzyme activities in the prenatal diagnosis of cystic fibrosis. Prenat Diagn 1985;5:129–34.
18.Achiron R, Mazkereth R, Orvieto R, Kuint J, Lipitz S, Rotstein Z. Echogenic bowel in intrauterine growth restriction fetuses: does this jeopardize the gut? Obstet Gynecol 2002;100:120–4.
19.Scotet V, Dugueperoux I, Audrezet MP, Audebert-Bellanger S, Muller M, Blayau M, et al. Focus on cystic fibrosis and other disorders evidenced in fetuses with sonographic finding of echogenic bowel: 16-year report from Brittany, France. Am J Obstet Gynecol 2010;203:592.e1–6.
Figure. No caption available.