- Email: [email protected]
Third trimester ultrasound of fetal pyelectasis: Predictor for postnatal surgery
Journal of Pediatric Urology (2008) 4, 51e54
Third trimester ultrasound of fetal pyelectasis: Predictor for postnatal surgery* Loralei L. Thornburg a,*, Eva K. Pressman a, Satya Chelamkuri a, William Hulbert b, Ronald Rabinowitz b, Robert Mevorach b a
Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University of Rochester, 601 Elmwood Avenue, Box 668, Rochester, NY 14642, USA b Department of Pediatric Urology, University of Rochester, 601 Elmwood Avenue, Box 655, Rochester, NY 14642, USA Received 10 April 2007; accepted 18 April 2007 Available online 8 June 2007
KEYWORDS Pyelectasis; Ultrasound; Hydronephrosis; Surgery
Abstract Objective: The ability to predict surgically relevant fetal renal pyelectasis is limited. We sought to determine whether the intrauterine timing of prenatal pyelectasis can predict the need for postnatal surgery. Methods: We retrospectively reviewed all patients with ultrasound measurements of the fetal renal pelvis during the 2nd and 3rd trimesters and postnatally. Pyelectasis was defined as >7 mm for renal pelvis antero-posterior diameter in the 2nd trimester and >10 mm in the 3rd trimester. Results: Fifty-nine patients out of 2397 candidates met the criteria for inclusion. A total of 116 kidneys were analyzed independently. Second trimester pyelectasis was associated with grade of postnatal hydronephrosis but not the need for surgery. Third trimester pyelectasis was associated with both postnatal hydronephrosis and surgery. The positive and negative predictive values for postnatal surgery were respectively 18% and 95% in the second trimester, and 27% and 100% in the third trimester. Conclusions: Third trimester ultrasound for fetal pyelectasis was better able to predict surgically relevant hydronephrosis than second trimester ultrasound. ª 2007 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
Introduction
*
Poster presentation. ESPU XVIIIth Annual Congress, Brugge, Belgium on 27 April 2007. Poster number S12-8 (PwP). * Corresponding author. Tel.: þ1 585 275 7480; fax: þ1 585 256 1416. E-mail address: [email protected] (L.L. Thornburg).
There is little consistency in the literature regarding the evaluation and impact of prenatally diagnosed renal disease in childhood. One commonly diagnosed prenatal renal condition is mild to moderate pyelectasis, or dilation of the renal pelvis. Although pyelectasis can be easily identified in the second trimester by prenatal ultrasound, the ability of prenatal ultrasound to predict those fetuses that will go on
1477-5131/$30 ª 2007 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jpurol.2007.04.005
52 to demonstrate significant renal or urinary tract pathology is poor, and there is little agreement on the best follow up [1]. There is also no agreement on the degree of anteroposterior pelvis diameter necessary to make the diagnosis of fetal pyelectasis, although several studies have suggested that a previously used cut-off of 4 or 5 mm in the second trimester lacks specificity [2,3]. It has been shown that persistent pyelectasis in the third trimester is correlated with persistent disease in the postnatal period, but several studies have refuted the value of third trimester ultrasound when more stringent cut-offs for second trimester diagnosis are used [1,4].We sought to determine whether repeat ultrasound in the third trimester could more accurately predict the need for postnatal surgery secondary to persistent hydronephrosis and underlying clinically relevant pathology.
Patients and methods We performed a retrospective review of all patients from May 1996 to June 2006 undergoing prenatal ultrasound at the University of Rochester Strong Memorial Hospital perinatal ultrasound unit. This study was institutional review board approved as an exempted study. The prenatal ultrasound database was queried for a diagnosis of fetal pyelectasis, any sonographer or physician comments regarding the fetal kidneys, or any mention of the words kidney, pyelectasis or renal in the physician comment of the ultrasound report. Patient were eligible for inclusion if fetal renal pelvis measurements were preformed in both the second and third trimester, and postnatal follow up including renal ultrasound information was available. All postnatal follow up was preformed at a single universitybased pediatric urology practice, which serves the entire metropolitan region. Fetal pyelectasis was defined as a renal pelvis antero-posterior diameter of >7 mm in the second trimester (12e26 weeks), and >10 mm in the third trimester (>26 weeks). Kidneys were divided into four groups based on the presence of pyelectasis in each trimester: Group 1, normal both trimesters (/); Group 2, pyelectasis in the second but not in the third trimester (þ/); Group 3, pyelectasis in the third but not in the second trimester (/þ); Group 4, pyelectasis in both trimesters (þ/þ). Grade of hydronephrosis on postnatal ultrasound was defined as absent (0), minimal (1), mild (2), moderate (3) or severe (4) as reported by treating pediatric urologist or attending radiologist. Postnatal surgery was at the discretion of the treating physician, but reviewed for ‘necessity’ verses parental choice by all three pediatric urologists. Baseline group characteristics and degree of prenatal pyelectasis in relationship to postnatal grade of hydronephrosis was analyzed using analysis of variance and chi-squared testing.
Results Records were reviewed for a total of 2397 pregnancies. The majority of infants were excluded due to lack of both second and third trimester ultrasound data, lack of postnatal follow up, or inability to link maternal and infant medical records. A total of 59 infants with 116 kidneys were identified (one contralateral kidney was absent, and one
L.L. Thornburg et al. dysplastic). The majority of infants were male (66%) and white (52%) which was consistent throughout the groups. Groups also did not significantly differ in laterality of the kidney involved; however, the trend was toward more left kidneys in Group 4 (57%) versus Group 1 (49%). Groups diagnosed with pyelectasis in the second or third trimester (Groups 2, 3 and 4) did not differ in gestational age at diagnostic ultrasound from each other or from those without the diagnosis of pyelectasis in any trimester (Group 1). Those diagnosed with pyelectasis in the second trimester that persisted into the third trimester (Group 4) received significantly more ultrasounds than the other groups. The four groups differed significantly from each other in the maximal degree of renal pelvis dilation in both the second and third trimester, as well as the grade of hydronephrosis on postnatal ultrasound. In addition, the number of subjects with persistence of moderate or severe hydronephrosis (grade 3 or 4) after delivery differed significantly between the groups by chi-squared analysis, indicating an association of postnatal hydronephrosis and the diagnosis and persistence of fetal renal pyelectasis (Table 1). Although the number of patients in these groups was small, the significance was retained after elimination of the smallest group, Group 2, in the chi-squared analysis (p Z 0.0004 with Groups 1, 3, 4). There was thus no evidence that this one small group introduced statistical instability into the four-group chisquared analysis, even though 20% of values were outside of the expected range. When comparing pyelectasis in the second trimester regardless of third trimester ultrasound result, the grade of postnatal hydronephrosis was worse in those with second trimester pyelectasis (mean grade 2.3 1.5) than a normal second trimester ultrasound (mean grade 1.7 1.3) p Z 0.03. Second trimester pyelectasis was also predictive of moderate or severe hydronephrosis (Grade 3 or 4) on postnatal ultrasound (11 of 22), compared to a normal second trimester ultrasound (23 of 94) p Z 0.03. If similar comparisons were done between those with and without pyelectasis in the third trimester regardless of second trimester ultrasound result, the groups also differed significantly in the mean grade of postnatal hydronephrosis: 1.5 1.5 for those with a positive third trimester ultrasound compared to 1.3 1.1 for those without (p Z 0.0002). Moderate or severe hydronephrosis on postnatal ultrasound occurred in 18 of 33 for those with a positive third trimester ultrasound compared to 16 of 94 for those without (p Z 0.0005). Nine postnatal surgeries were required, but an additional three surgeries were performed at parental discretion as an alternative to medical management of VUR. Of the clinically required surgeries, five were done for PUJ obstruction with progression, three for duplicated ureters with ectopia, and one for obstructing persistent uterocele with sepsis. Surgeries occurred up to 5 years postnatally and included pyeloplasty, ureteroureterosyomy, ureteroneocystostomy, nephrectomy, and endoscopic bulking for reflux. There were five surgeries required in Group 3 and four required surgeries in Group 4. The presence of pyelectasis in neither the second nor third trimester was able to predict the need for postnatal surgery, with a positive predictive value of only 18% for the second trimester and 27% for the third trimester. The negative predictive value for
Third trimester ultrasound of fetal pyelectasis Table 1
53
Study population characteristics
Characteristic
Group 1 (/) n Z 75
20.3 2.2 GAa first measurement (weeks) Number of USc (2nd and 3rd 3 1.2 trimester) Maximal degree 3.8 1.9 pyelectasis 2nd trimester (mm) Maximal degree 3.5 3.4 pyelectasis 3rd trimester (mm) Grade postnatal 1.2 1.1 hydronephrosis Grade 3 or 4 (moderate/ 13 of 75 severe) on postnatal US
Group 2 (þ/) n Z 8
Group 3 (/þ) n Z 19
Group 4 (þ/þ) n Z 14
p-value
19.7 3
19.7 2.1
20.1 1.7
NSb
3 1.3
3.3 1.3
4.1 1.5
0.02b
7.6 1.1
5.2 0.5
9.9 3.4
<0.0001b
5 3.4
14.6 6.4
16.4 5.7
<0.0001b
2.5 1.3
2.3 1.7
<0.0001b
2.3 1 3 of 8
10 of 19
8 of 14
0.001d
Group 1, normal both trimesters (/); Group 2, pyelectasis in the second, but not in the third trimester (þ/); Group 3, pyelectasis in the third, but not in the second trimester (/þ); Group 4, pyelectasis in both trimesters (þ/þ). Values are means and standard deviation, except for ‘Grade 3 or 4 (moderate/severe) on postnatal ultrasound’ which is presented as number of patients out of total patients in each group. a Gestational age. b Analysis of variance. c Ultrasound. d Chi-squared analysis. NS, not significant.
both the second and third trimesters was excellent at 95% and 100%, respectively (Table 2).
Discussion Although the grade of postnatal hydronephrosis appears to be associated with the degree of prenatal pyelectasis, the real value in these data lies in the ability of a normal ultrasound in the second and third trimesters to rule out the need for postnatal surgery, and possibly eliminate the need for postnatal evaluation beyond a renal bladder ultrasound. Although the negative predictive value of a second trimester ultrasound is high, combining this ultrasound with a third trimester ultrasound further improves the negative predictive value, as well as the sensitivity and specificity of prenatal ultrasound to predict the need for postnatal surgery. Therefore, third trimester ultrasound for fetal pyelectasis is a better predictor of postnatal surgery than second trimester ultrasound. This means that parents can be reassured that it is unlikely that their infant will require surgery if there is no
Table 2 Predictive value (%) of prenatal ultrasound for postnatal surgery
Specificity Sensitivity Positive predictive value Negative predictive value
Second trimester pyelectasis
Third trimester pyelectasis
83 44 18 95
78 100 27 100
pyelectasis by the third trimester, even in the case of previously diagnosed pyelectasis. The strengths of our study include that all patients were seen at a single ultrasound unit and were followed up at a single pediatric urology practice, both at a single university, which gives consistency to the care and definitions used within each practice. In addition, our study includes both prenatal and postnatal renal evaluations for many patients with borderline or normal renal pelvis measurements in either the first trimester, second trimester or both. These normal patients allow us to compare our results to a concurrent control group, biologically identical to the study group, since the normal kidney was often the contralateral kidney in the same patient. In addition, as the current cut-offs used for diagnosis of prenatal pyelectasis vary from 5 to 7 mm in the second trimester and from 10 to 14 mm in the third trimester, the excellent negative predictive value of the 7-mm and 10-mm cut-offs used in our study lends them further strength for use in defining pyelectasis. Increasing the cut-offs would probably improve the positive predictive value, but when assuring normalcy is the goal, a high negative predictive value is more important. The main weakness of the study is the retrospective nature, such that the patients most likely to receive complete follow up were those with the worst appearance on prenatal ultrasound, or with postnatal complications. In addition, although every attempt was made to collect complete information, there may be variation in the care that patients received, or an inherent selection bias towards those that were referred for further evaluation by pediatric urology. The standard of care in this community is to refer significant postnatal hydronephrosis for pediatric urological evaluation, limiting this bias.
54 Additionally, as this is the only pediatric urology practice within the city of Rochester and the region, it is unlikely that patients received evaluation and treatment outside of this group. All patients with pyelectasis diagnosed in any trimester had an increase in grade on postnatal ultrasound compared to those with normal ultrasounds in both trimesters. Although this is not unexpected, it gives further weight to the need for close postnatal follow up of patients with prenatally diagnosed pyelectasis in any trimester even with resolution in the third trimester. Regardless, third trimester ultrasound is a better predictor of the need for postnatal surgical intervention, which will allow improved counseling of patients. Given the predictive nature of third trimester ultrasound, we can better reassure our patients that if fetal pyelectasis has normalized by the third trimester, the likelihood of postnatal surgery is small. Further studies on the value of third trimester ultrasound in a prospective manner are ongoing, as not all prior studies have noted that prenatal ultrasound is associated with the degree of postnatal hydronephrosis, or have confirmed that prenatal
L.L. Thornburg et al. ultrasound can predict the need for postnatal surgery. From our data, however, third trimester ultrasound for fetal pyelectasis is a better predictor of postnatal surgery than second trimester ultrasound.
References [1] Ahmad G, Green P. Outcome of fetal pyelectasis diagnosed antenatally. J Obstet Gynaecol 2005;25:119e22. [2] Twining P. Genitourinary malformations. In: Nyberg DA, McGahan JP, Pretorius DH, Pilu G, editors. Diagnostic imaging of fetal anomalies. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 603e60. [3] Cohen-Overbeek TE, Wijngaard-Boom P, Ursem NT, Hop WC, Wladimiroff JW, Wolffenbuttel KP. Mild renal pyelectasis in the second trimester: determination of cut-off levels for postnatal referral. Ultrasound Obstet Gynecol 2005;25:378e83. [4] Gramellini D, Fieni S, Caforio E, Benassi G, Bedocchi L, Beseghi U, et al. Diagnostic accuracy of fetal renal pelvis anteroposterior diameter as a predictor of significant postnatal nephrouropathy: second versus third trimester of pregnancy. Am J Obstet Gynecol 2006;194:167e73.
Third trimester ultrasound of fetal pyelectasis: Predictor for postnatal surgery* Loralei L. Thornburg a,*, Eva K. Pressman a, Satya Chelamkuri a, William Hulbert b, Ronald Rabinowitz b, Robert Mevorach b a
Division of Maternal Fetal Medicine, Department of Obstetrics and Gynecology, University of Rochester, 601 Elmwood Avenue, Box 668, Rochester, NY 14642, USA b Department of Pediatric Urology, University of Rochester, 601 Elmwood Avenue, Box 655, Rochester, NY 14642, USA Received 10 April 2007; accepted 18 April 2007 Available online 8 June 2007
KEYWORDS Pyelectasis; Ultrasound; Hydronephrosis; Surgery
Abstract Objective: The ability to predict surgically relevant fetal renal pyelectasis is limited. We sought to determine whether the intrauterine timing of prenatal pyelectasis can predict the need for postnatal surgery. Methods: We retrospectively reviewed all patients with ultrasound measurements of the fetal renal pelvis during the 2nd and 3rd trimesters and postnatally. Pyelectasis was defined as >7 mm for renal pelvis antero-posterior diameter in the 2nd trimester and >10 mm in the 3rd trimester. Results: Fifty-nine patients out of 2397 candidates met the criteria for inclusion. A total of 116 kidneys were analyzed independently. Second trimester pyelectasis was associated with grade of postnatal hydronephrosis but not the need for surgery. Third trimester pyelectasis was associated with both postnatal hydronephrosis and surgery. The positive and negative predictive values for postnatal surgery were respectively 18% and 95% in the second trimester, and 27% and 100% in the third trimester. Conclusions: Third trimester ultrasound for fetal pyelectasis was better able to predict surgically relevant hydronephrosis than second trimester ultrasound. ª 2007 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
Introduction
*
Poster presentation. ESPU XVIIIth Annual Congress, Brugge, Belgium on 27 April 2007. Poster number S12-8 (PwP). * Corresponding author. Tel.: þ1 585 275 7480; fax: þ1 585 256 1416. E-mail address: [email protected] (L.L. Thornburg).
There is little consistency in the literature regarding the evaluation and impact of prenatally diagnosed renal disease in childhood. One commonly diagnosed prenatal renal condition is mild to moderate pyelectasis, or dilation of the renal pelvis. Although pyelectasis can be easily identified in the second trimester by prenatal ultrasound, the ability of prenatal ultrasound to predict those fetuses that will go on
1477-5131/$30 ª 2007 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jpurol.2007.04.005
52 to demonstrate significant renal or urinary tract pathology is poor, and there is little agreement on the best follow up [1]. There is also no agreement on the degree of anteroposterior pelvis diameter necessary to make the diagnosis of fetal pyelectasis, although several studies have suggested that a previously used cut-off of 4 or 5 mm in the second trimester lacks specificity [2,3]. It has been shown that persistent pyelectasis in the third trimester is correlated with persistent disease in the postnatal period, but several studies have refuted the value of third trimester ultrasound when more stringent cut-offs for second trimester diagnosis are used [1,4].We sought to determine whether repeat ultrasound in the third trimester could more accurately predict the need for postnatal surgery secondary to persistent hydronephrosis and underlying clinically relevant pathology.
Patients and methods We performed a retrospective review of all patients from May 1996 to June 2006 undergoing prenatal ultrasound at the University of Rochester Strong Memorial Hospital perinatal ultrasound unit. This study was institutional review board approved as an exempted study. The prenatal ultrasound database was queried for a diagnosis of fetal pyelectasis, any sonographer or physician comments regarding the fetal kidneys, or any mention of the words kidney, pyelectasis or renal in the physician comment of the ultrasound report. Patient were eligible for inclusion if fetal renal pelvis measurements were preformed in both the second and third trimester, and postnatal follow up including renal ultrasound information was available. All postnatal follow up was preformed at a single universitybased pediatric urology practice, which serves the entire metropolitan region. Fetal pyelectasis was defined as a renal pelvis antero-posterior diameter of >7 mm in the second trimester (12e26 weeks), and >10 mm in the third trimester (>26 weeks). Kidneys were divided into four groups based on the presence of pyelectasis in each trimester: Group 1, normal both trimesters (/); Group 2, pyelectasis in the second but not in the third trimester (þ/); Group 3, pyelectasis in the third but not in the second trimester (/þ); Group 4, pyelectasis in both trimesters (þ/þ). Grade of hydronephrosis on postnatal ultrasound was defined as absent (0), minimal (1), mild (2), moderate (3) or severe (4) as reported by treating pediatric urologist or attending radiologist. Postnatal surgery was at the discretion of the treating physician, but reviewed for ‘necessity’ verses parental choice by all three pediatric urologists. Baseline group characteristics and degree of prenatal pyelectasis in relationship to postnatal grade of hydronephrosis was analyzed using analysis of variance and chi-squared testing.
Results Records were reviewed for a total of 2397 pregnancies. The majority of infants were excluded due to lack of both second and third trimester ultrasound data, lack of postnatal follow up, or inability to link maternal and infant medical records. A total of 59 infants with 116 kidneys were identified (one contralateral kidney was absent, and one
L.L. Thornburg et al. dysplastic). The majority of infants were male (66%) and white (52%) which was consistent throughout the groups. Groups also did not significantly differ in laterality of the kidney involved; however, the trend was toward more left kidneys in Group 4 (57%) versus Group 1 (49%). Groups diagnosed with pyelectasis in the second or third trimester (Groups 2, 3 and 4) did not differ in gestational age at diagnostic ultrasound from each other or from those without the diagnosis of pyelectasis in any trimester (Group 1). Those diagnosed with pyelectasis in the second trimester that persisted into the third trimester (Group 4) received significantly more ultrasounds than the other groups. The four groups differed significantly from each other in the maximal degree of renal pelvis dilation in both the second and third trimester, as well as the grade of hydronephrosis on postnatal ultrasound. In addition, the number of subjects with persistence of moderate or severe hydronephrosis (grade 3 or 4) after delivery differed significantly between the groups by chi-squared analysis, indicating an association of postnatal hydronephrosis and the diagnosis and persistence of fetal renal pyelectasis (Table 1). Although the number of patients in these groups was small, the significance was retained after elimination of the smallest group, Group 2, in the chi-squared analysis (p Z 0.0004 with Groups 1, 3, 4). There was thus no evidence that this one small group introduced statistical instability into the four-group chisquared analysis, even though 20% of values were outside of the expected range. When comparing pyelectasis in the second trimester regardless of third trimester ultrasound result, the grade of postnatal hydronephrosis was worse in those with second trimester pyelectasis (mean grade 2.3 1.5) than a normal second trimester ultrasound (mean grade 1.7 1.3) p Z 0.03. Second trimester pyelectasis was also predictive of moderate or severe hydronephrosis (Grade 3 or 4) on postnatal ultrasound (11 of 22), compared to a normal second trimester ultrasound (23 of 94) p Z 0.03. If similar comparisons were done between those with and without pyelectasis in the third trimester regardless of second trimester ultrasound result, the groups also differed significantly in the mean grade of postnatal hydronephrosis: 1.5 1.5 for those with a positive third trimester ultrasound compared to 1.3 1.1 for those without (p Z 0.0002). Moderate or severe hydronephrosis on postnatal ultrasound occurred in 18 of 33 for those with a positive third trimester ultrasound compared to 16 of 94 for those without (p Z 0.0005). Nine postnatal surgeries were required, but an additional three surgeries were performed at parental discretion as an alternative to medical management of VUR. Of the clinically required surgeries, five were done for PUJ obstruction with progression, three for duplicated ureters with ectopia, and one for obstructing persistent uterocele with sepsis. Surgeries occurred up to 5 years postnatally and included pyeloplasty, ureteroureterosyomy, ureteroneocystostomy, nephrectomy, and endoscopic bulking for reflux. There were five surgeries required in Group 3 and four required surgeries in Group 4. The presence of pyelectasis in neither the second nor third trimester was able to predict the need for postnatal surgery, with a positive predictive value of only 18% for the second trimester and 27% for the third trimester. The negative predictive value for
Third trimester ultrasound of fetal pyelectasis Table 1
53
Study population characteristics
Characteristic
Group 1 (/) n Z 75
20.3 2.2 GAa first measurement (weeks) Number of USc (2nd and 3rd 3 1.2 trimester) Maximal degree 3.8 1.9 pyelectasis 2nd trimester (mm) Maximal degree 3.5 3.4 pyelectasis 3rd trimester (mm) Grade postnatal 1.2 1.1 hydronephrosis Grade 3 or 4 (moderate/ 13 of 75 severe) on postnatal US
Group 2 (þ/) n Z 8
Group 3 (/þ) n Z 19
Group 4 (þ/þ) n Z 14
p-value
19.7 3
19.7 2.1
20.1 1.7
NSb
3 1.3
3.3 1.3
4.1 1.5
0.02b
7.6 1.1
5.2 0.5
9.9 3.4
<0.0001b
5 3.4
14.6 6.4
16.4 5.7
<0.0001b
2.5 1.3
2.3 1.7
<0.0001b
2.3 1 3 of 8
10 of 19
8 of 14
0.001d
Group 1, normal both trimesters (/); Group 2, pyelectasis in the second, but not in the third trimester (þ/); Group 3, pyelectasis in the third, but not in the second trimester (/þ); Group 4, pyelectasis in both trimesters (þ/þ). Values are means and standard deviation, except for ‘Grade 3 or 4 (moderate/severe) on postnatal ultrasound’ which is presented as number of patients out of total patients in each group. a Gestational age. b Analysis of variance. c Ultrasound. d Chi-squared analysis. NS, not significant.
both the second and third trimesters was excellent at 95% and 100%, respectively (Table 2).
Discussion Although the grade of postnatal hydronephrosis appears to be associated with the degree of prenatal pyelectasis, the real value in these data lies in the ability of a normal ultrasound in the second and third trimesters to rule out the need for postnatal surgery, and possibly eliminate the need for postnatal evaluation beyond a renal bladder ultrasound. Although the negative predictive value of a second trimester ultrasound is high, combining this ultrasound with a third trimester ultrasound further improves the negative predictive value, as well as the sensitivity and specificity of prenatal ultrasound to predict the need for postnatal surgery. Therefore, third trimester ultrasound for fetal pyelectasis is a better predictor of postnatal surgery than second trimester ultrasound. This means that parents can be reassured that it is unlikely that their infant will require surgery if there is no
Table 2 Predictive value (%) of prenatal ultrasound for postnatal surgery
Specificity Sensitivity Positive predictive value Negative predictive value
Second trimester pyelectasis
Third trimester pyelectasis
83 44 18 95
78 100 27 100
pyelectasis by the third trimester, even in the case of previously diagnosed pyelectasis. The strengths of our study include that all patients were seen at a single ultrasound unit and were followed up at a single pediatric urology practice, both at a single university, which gives consistency to the care and definitions used within each practice. In addition, our study includes both prenatal and postnatal renal evaluations for many patients with borderline or normal renal pelvis measurements in either the first trimester, second trimester or both. These normal patients allow us to compare our results to a concurrent control group, biologically identical to the study group, since the normal kidney was often the contralateral kidney in the same patient. In addition, as the current cut-offs used for diagnosis of prenatal pyelectasis vary from 5 to 7 mm in the second trimester and from 10 to 14 mm in the third trimester, the excellent negative predictive value of the 7-mm and 10-mm cut-offs used in our study lends them further strength for use in defining pyelectasis. Increasing the cut-offs would probably improve the positive predictive value, but when assuring normalcy is the goal, a high negative predictive value is more important. The main weakness of the study is the retrospective nature, such that the patients most likely to receive complete follow up were those with the worst appearance on prenatal ultrasound, or with postnatal complications. In addition, although every attempt was made to collect complete information, there may be variation in the care that patients received, or an inherent selection bias towards those that were referred for further evaluation by pediatric urology. The standard of care in this community is to refer significant postnatal hydronephrosis for pediatric urological evaluation, limiting this bias.
54 Additionally, as this is the only pediatric urology practice within the city of Rochester and the region, it is unlikely that patients received evaluation and treatment outside of this group. All patients with pyelectasis diagnosed in any trimester had an increase in grade on postnatal ultrasound compared to those with normal ultrasounds in both trimesters. Although this is not unexpected, it gives further weight to the need for close postnatal follow up of patients with prenatally diagnosed pyelectasis in any trimester even with resolution in the third trimester. Regardless, third trimester ultrasound is a better predictor of the need for postnatal surgical intervention, which will allow improved counseling of patients. Given the predictive nature of third trimester ultrasound, we can better reassure our patients that if fetal pyelectasis has normalized by the third trimester, the likelihood of postnatal surgery is small. Further studies on the value of third trimester ultrasound in a prospective manner are ongoing, as not all prior studies have noted that prenatal ultrasound is associated with the degree of postnatal hydronephrosis, or have confirmed that prenatal
L.L. Thornburg et al. ultrasound can predict the need for postnatal surgery. From our data, however, third trimester ultrasound for fetal pyelectasis is a better predictor of postnatal surgery than second trimester ultrasound.
References [1] Ahmad G, Green P. Outcome of fetal pyelectasis diagnosed antenatally. J Obstet Gynaecol 2005;25:119e22. [2] Twining P. Genitourinary malformations. In: Nyberg DA, McGahan JP, Pretorius DH, Pilu G, editors. Diagnostic imaging of fetal anomalies. Philadelphia: Lippincott Williams & Wilkins; 2003. p. 603e60. [3] Cohen-Overbeek TE, Wijngaard-Boom P, Ursem NT, Hop WC, Wladimiroff JW, Wolffenbuttel KP. Mild renal pyelectasis in the second trimester: determination of cut-off levels for postnatal referral. Ultrasound Obstet Gynecol 2005;25:378e83. [4] Gramellini D, Fieni S, Caforio E, Benassi G, Bedocchi L, Beseghi U, et al. Diagnostic accuracy of fetal renal pelvis anteroposterior diameter as a predictor of significant postnatal nephrouropathy: second versus third trimester of pregnancy. Am J Obstet Gynecol 2006;194:167e73.