European Journal of Radiology. 15 (1992) 149- 153
149
0 1992 Elsevier Science Publishers B.V. All rights reserved. 0720-048X/92/$05.00
EURRAD 00290
Sonographic findings in infants with suspected necrotizing enterocolitis T. Biimelburg”, aKinderklinik
H.-J.
von Lengerkeb
and b Institut fiir Klinische Radiologie der Universitdt Miinster, Germany
(Received 4 November
1991; accepted after revision 20 February 1992)
Key words: Colitis, necrotizing; Colitis, radiography;
Colitis, infants; Colitis, US studies; Ultrasound,
intestines
Abstract
During a three-year period, 27 infants with clinical signs of necrotizing enterocolitis (NEC) underwent ultrasonography; 22 of these infants also had abdominal radiographs within a few hours. Sonographically, portal venous gas (PVC) was seen in 10 and intestinal pneumatosis without PVG in 8 infants. Six children underwent laparotomy within 24 hours after ultrasonography and showed evidence of recent NEC. Three other children, laparotomized weeks later were found to have intestinal strictures as signs of previous NEC. None of the patients died from NEC. Nine infants without abnormal gas distribution did not develop overt NEC. With respect to PVG, sonography is able to support the tentative diagnosis of NEC prior to radiography. However, absence of PVG does not exclude NEC.
Introduction In patients with clinical signs of necrotizing enterocolitis (NEC), radiographic evidence of intestinal pneumatosis or the serious finding of portal venous gas (PVG) is generally accepted as confirmation of the diagnosis [l-3]. While radiographic abnormalities are not consistently observed or may be difficult to recognize [ 1,3], sonography has recently been described as a simple method to detect PVG, even in the absence of characteristic radiographic signs [4-61. So far, little is known about the diagnostic value of sonographic findings in suspected NEC [7]. In order to investigate this question, we studied retrospectively the sonographic and radiographic results, as well as the outcome, of 27 infants examined for clinically suspected NEC during a three-year period. Materials and Methods From January 1987 through March 1990, abdominal sonography was performed in 27 children (14 girls, Correspondence
Albert-Schweitzer
to: Dr.
T. Bbmelburg, Universitats-Kinderklinik, Str. 33. 4400 Mnnster, Germany
13 boys, bodyweight 700-4010 g) with clinically suspected NEC. All but 5 infants also underwent abdominal radiography within the same period. Sonography was performed with a real-time computing system using a 5 or 7.5 MHz transducer (Acuson 128). Parenchymal organs and portal venous structures were examined using standard planes. The gastrointestinal tract was screened by transverse and longitudinal planes. For imaging of intestinal walls, intraluminary gas was moved by cautious massage with the transducer. Especially for the assessment of gas bubbles, integration of imaging was lowered for a faster build-up. Gas bubbles within the intestinal walls or portal vein were interpreted as confirmation of the tentative diagnosis of NEC. Gastrointestinal symptoms prompting sonographic examination were reviewed on the basis of clinical records. Abdominal manifestation was graded as mild if distension was the only sign, moderate with additional tenderness or considerably diminished or absent bowel noise, and severe with supervening signs of peritonitis or presumed sepsis. Stool characteristics were noted separately. Finally, outcome in terms of abdominal complications was recorded.
150 TABLE 1 Clinical data of patients with portal venous gas Patient (body weight)
Presenting abdominal findings, and stool characteristics
Evidence of gas (sonographic) Pv
g) g) g) g)
Moderate, fresh blood Mild, fresh blood Mild, fresh blood Severe, guaiac-neg.
5 (2770 g) 6 (2300 g)
Moderate, guaiac-neg. Moderate, fresh blood
1 (3510 2 (2500 3 (1570 4 (1530
7 (3550 8 (3010 9 (1530 10 (1640
g) g) g) g)
Liver
Comments and complications
TPN/AB TPN/AB TPN/AB TPN/AB
Coronavirus Rotavirus Coronavirus Lap., 24 days later: colonic strictures Lap.: NEC Lap., 30 days later: colonic strictures
(radiographic) Mural
Pv
Mural
0
0
X
ni 0 ni
ni 0 ni
X
X
0
X
X
ni
0
X
X X
ni
Severe, guaiac-pos. Mild, guaiac-pos. Severe, fresh blood Moderate, fresh blood
Therapy
X X
ni ni
TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB
Lap.: NEC Lap.: NEC
Pv: portal vein, Liver: microbubbles trapped within the minor branches of the portal vein producing patterns of hyperechogeneity in liver parenchyma, TPN: total parenteral nutrition, AB: antibiotics, ni: not investigated, np: not possible, NEC: necrotizing enterocolitis, Lap: laparotomy.
Results In lo/27 infants examined for suspected NEC, PVG was detected by sonography (Table 1). However, radiographs obtained in 8 infants failed to demonstrate PVG and intestinal pneumatosis was found in only 4/ 8 patients. Sonographically, gas bubbles appeared as high-amplitude echoes streaming through the portal vein in hepatopetal direction (Fig. la). In S/l0 patients, these were trapped within the minor branches of the
portal vein producing varying patterns of focal hyperechogeneity in the hepatic parenchyma (Fig. lb). Intramural gas was always found when the intestinal tract was examined by sonography. Laparotomy served to confirm the diagnosis of NEC in 3 children. The surgical findings showed severe dark-livid segments of dilated intestine, predominantly the colon, but no perforation. Intestinal pneumatosis was present in two patients (No. 9 and No. 10). Two patients (Nos. 5 and 10) underwent segmental and ileocoecal resection, re-
Fig. 1 (a). Intense reflections of gas bubbles within the major branches of the portal vein. (b) Transverse scan of the liver demonstrating tiple echogenic areas of trapped air within the minor branches of the portal vein.
mul-
151 TABLE 2 Clinical data of patients with intramural gas only Patient (body weight)
Presenting abdominal findings and stool character
Evidence of gas
Therapy
(sonographic)
(radiographic)
Pv
Liver
Mural
Pv
Mural
11 (1160 g) 12 (1950 g)
Moderate, guaiac-pos. Severe, guaiac-pos.
0
0
x
0
0
0
0
x
0
0
13 (1230 g) 14 (2950 g)
Moderate, guaiac-neg. Moderate, guaiac-neg.
0
0
x
0
x
0
0
x
ni
ni
15 (1480 16 (3750 17 (1500 18 (1425
Moderate, guaiac-neg. Severe, guaiac-neg. Severe, fresh blood Moderate. fresh blood
0
0
0
0
ni x
np
np
ni 0 x 0
g) g) g) g)
0
x
Comments and complications
;
TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB
Lap.: Perforation, Adhesions, NEC Lap., 18 days later: Colonic strictures
Lap.: Adhesions, NEC Lap.: Early NEC
See Table 1 for notes
spectively. One patient (No. 9) had a hemicolectomy; another two patients were laparotomized for mechanical ileus later on and showed colonic strictures. Eight patients had sonographic evidence of intramural gas, but no PVG (Table 2). In one of these patients, a suspicious pattern was seen in the hepatic parenchyma. Only 3 of 6 children who underwent radiography showed intestinal pneumatosis. In one patient, who presented with additional PVG, sonographic examination of the liver was unsuccessful due to intestinal me-
teorism. Sonographically, air in the intestinal walls was seen as gaseous strips or gas bubbles of high-amplitude echogeneity (Fig. 2 a and b). Laparotomy confirmed the diagnosis of NEC in those two infants who had a suspicious hepatic parenchyma and radiographic evidence of PVG, respectively, and in one child presenting with covered ileocecal perforation. At surgery segmental hyperaemia was present in one patient (No. 18) and dark-livid segmental areas of intestine were present in two other patients (Nos. 12 and 17). In another in-
Fig. 2 (a). Cross-section through thickened intestinal walls showing gaseous strips of intramural air with shadowing. (b). Transverse scan through a pair of intestinal loops demonstrate intraluminal air on the right side and a gaseous strip of intramural air on the left side of the sonogram.
152 TABLE 3 Clinical data of patients without any pathological gas Patient body weight
Presenting abdominal findings and stool character
Therapy
Evidence of gas (radiographic)
(sonographic) Pv
Liver
Mural
Pv
Mural
0
0
AB
0
0
0
0
-iTPN/AB
0
TPN/AB
0
TPN/AB
ni 0 0 0
-ITPN/AB
19 (700 g) 20 (1220 g) 21 (1830 g)
Mild, guaiac-neg. Moderate, guaiac-neg. Mild, guaiac-pos.
0
0
0
0
0
0
ni 0 0
22 (710 g)
Severe, guaiac-neg.
0
0
0
23 24 25 26 27
Moderate, guaiac-pos. Mild, fresh blood Mild, guaiac-pos. Moderate, guaiac-neg. Moderate, guaiac-neg.
(2440 (4010 (2500 (2540 (2050
g) g) g) g) g)
Comments and complications
Lap., 8 days later: Colonic strictures Lap.: Meconiumileus, Perforation H. simplex
-lTPN/AB
See Table 1 for notes.
fant laparotomy was performed for colonic strictures later on. In all 6 children with confirmed NEC laparotomy had been performed within 24 hours after sonography due to rapid clinical deterioration. Nine patients showed no pathological gas formation (Table 3). Two of these underwent laparotomy. One child was admitted from another hospital with clinical evidence of NEC established 4 weeks previously. He presented with intestinal stenosis due to colonic strictures and adhesions. The other child showed intestinal perforation caused by meconium ileus. Discussion NEC is thought to be the final common response to the interaction of varying etiologic factors, including viral infections [ 2,3]. Pathoanatomical examinations reveal progressive necrosis and ulceration of the intestinal layers leading to perforation or pseudomembrane formation [ 81. Intestinal injury permits the passage of air, delivered by bacterial fermentation of formula feedings, into the intestinal wall or portal vein [l-3]. Pertinent radiographic evidence is considered pathognomanic disease. Kosloske et al., who used radiographic criteria for the diagnosis of NEC, found intestinal pneumatosis in 95% and PVG in 23 % of 147 infants. All patients studied by the authors belonged to a group of 242 infants, who were treated for NEC within the same period [9].
Radiography produces static images of dynamic processes and requires certain minimum quantities of air to demonstrate PVG. Indeed, the radiographic finding of PVG is associated with pan-necrosis and is considered a valid indication for operation [ 3, lo]. Infants presenting with clinical symptoms of NEC, but no radiographic abnormalities, are not counted as established cases of NEC. This important group is thought to represent early or developing NEC [ 81. Sonographically, we found intramural gas and/or PVG in 18/27 infants examined for suspected NEC. Radiographs of 22 children obtained within the same period showed intestinal pneumatosis in only seven infants, one of these also had PVG. The sonographic diagnosis was confirmed by laparotomy in nine patients. Intestinal strictures found in three of these infants are well-known late complications of the disease [2,3]. Unlike static radiographs, real-time sonography shows gas bubbles streaming through the portal vein and might therefore serve to demonstrate PVG prior to radiographic evidence. This was confirmed by Merritt et al. [4] and Lindley et al. [7] in 6/S and 5/5 infants, respectively and by our own results in all of 10 infants with sonographically detected PVG. Due to the large acoustic impedance of air, gas bubbles are clearly visible as bright echoes. When transiently trapped within the peripheral branches of the portal vein they produce varying patterns of hyperechogeneity and may be the only suspicious finding supporting the diagnosis of NEC [4]. Although sonography is fascinating for its simplicity
153
in showing PVG and intrahepatic patterns of hyperechogeneity, absence of these findings does not exclude NEC. Intramural gas, the sonographic equivalent preceding radiologically detectable intestinal pneumatosis [lo], may then be the only sign of suspected NEC. The examining physician should be aware that intramural gas, which pathological examinations revealed to be associated with pseudomembrane formation [ 81, should be ruled out by sonography, especially in the absence of PVG. Patients who were examined carefully and showed no pathological gas formation did not develop overt NEC, even without therapy. Among cases with radiographic diagnosis of NEC, mortality was 65 7; in patients with PVG and 86% in those with pneumatosis and PVG combined [ 91. In our series with sonographically detected PVG and/or intestinal pneumatosis none of the infants died from NEC. This might be due to detection at an earlier stage by ultrasonography as compared with radiography. Nevertheless, it remains unclear why patients who presented with intestinal pneumatosis showed no PVG, although abdominal findings were comparable to those obtained in patients with PVG. One possible explanation might be concurrent antibiotic therapy in some patients. This question might be clarified in a prospective study with sonographic examinations prior to the start of therapy.
References 1 Amon RG, Fishbein JF. Portal venous gas in the pediatric age group. J Pediatr 1971; 79: 255-259. 2 Koloske AM, Musemeche CA. Necrotizing enterocolitis of the neonate. Clin Perinatol 1989; 16: 97-111. 3 Walsh MC, Kliegmann RM. Necrotizing enterocolitis: Treatment based on staging criteria. Pediatr Clin North Am 1986; 33: 179201. 4 Merritt CRB, Goldsmith JP, Sharp MJ. Sonographic detection of portal venous gas in infants with necrotizing enterocolitis. AJR 1984; 143: 1059-1062. 5 Robberecht EA, Afschrift M, De Be1 CE, Van Haesebrouck PJ, Van Bever HP, De Wit M, Leroy JG. Sonographic demonstration of portal venous gas in necrotizing enterocolitis. Eur J Pediatr 1988; 147: 192-194. 6 Mahn SW, Bhutani VK, Ritchie WW, Hall ML, Paul D. Echogenic intravascular and hepatic microbubbles associated with necrrotizing enterocolitis. J Pediatr 1983; 103: 637-640. I Lindley S, Mollitt DL, Seibert JJ, Golladay ES. Portal vein ultrasonography in the early diagnosis of necrotizing enterocolitis. J Pediatr Surg 1986; 21: 530-532. 8 Santulli TV, Schullinger JN, Heird WC, Gongaware RD, Wigger J, Barlow B, Blanc WA, Berdon WE. Acute necrotizing enterocolitis in infancy: A review of 64 cases. Pediatr 1975; 55: 376387. 9 Kosloske AM, Musemeche CA, Ball WS, Ablin DS, Bhattacharyya N. Necrotizing enterocolitis: Value of radiographic findings to predict outcome. AJR 1988; 151: 771-774. 10 Vemacchia FS, Jeffrey RB, Laing FC, Wing VW. Sonographic recognition of pneumatosis intestinalis. AJR 1985; 145: 51-52.
149
0 1992 Elsevier Science Publishers B.V. All rights reserved. 0720-048X/92/$05.00
EURRAD 00290
Sonographic findings in infants with suspected necrotizing enterocolitis T. Biimelburg”, aKinderklinik
H.-J.
von Lengerkeb
and b Institut fiir Klinische Radiologie der Universitdt Miinster, Germany
(Received 4 November
1991; accepted after revision 20 February 1992)
Key words: Colitis, necrotizing; Colitis, radiography;
Colitis, infants; Colitis, US studies; Ultrasound,
intestines
Abstract
During a three-year period, 27 infants with clinical signs of necrotizing enterocolitis (NEC) underwent ultrasonography; 22 of these infants also had abdominal radiographs within a few hours. Sonographically, portal venous gas (PVC) was seen in 10 and intestinal pneumatosis without PVG in 8 infants. Six children underwent laparotomy within 24 hours after ultrasonography and showed evidence of recent NEC. Three other children, laparotomized weeks later were found to have intestinal strictures as signs of previous NEC. None of the patients died from NEC. Nine infants without abnormal gas distribution did not develop overt NEC. With respect to PVG, sonography is able to support the tentative diagnosis of NEC prior to radiography. However, absence of PVG does not exclude NEC.
Introduction In patients with clinical signs of necrotizing enterocolitis (NEC), radiographic evidence of intestinal pneumatosis or the serious finding of portal venous gas (PVG) is generally accepted as confirmation of the diagnosis [l-3]. While radiographic abnormalities are not consistently observed or may be difficult to recognize [ 1,3], sonography has recently been described as a simple method to detect PVG, even in the absence of characteristic radiographic signs [4-61. So far, little is known about the diagnostic value of sonographic findings in suspected NEC [7]. In order to investigate this question, we studied retrospectively the sonographic and radiographic results, as well as the outcome, of 27 infants examined for clinically suspected NEC during a three-year period. Materials and Methods From January 1987 through March 1990, abdominal sonography was performed in 27 children (14 girls, Correspondence
Albert-Schweitzer
to: Dr.
T. Bbmelburg, Universitats-Kinderklinik, Str. 33. 4400 Mnnster, Germany
13 boys, bodyweight 700-4010 g) with clinically suspected NEC. All but 5 infants also underwent abdominal radiography within the same period. Sonography was performed with a real-time computing system using a 5 or 7.5 MHz transducer (Acuson 128). Parenchymal organs and portal venous structures were examined using standard planes. The gastrointestinal tract was screened by transverse and longitudinal planes. For imaging of intestinal walls, intraluminary gas was moved by cautious massage with the transducer. Especially for the assessment of gas bubbles, integration of imaging was lowered for a faster build-up. Gas bubbles within the intestinal walls or portal vein were interpreted as confirmation of the tentative diagnosis of NEC. Gastrointestinal symptoms prompting sonographic examination were reviewed on the basis of clinical records. Abdominal manifestation was graded as mild if distension was the only sign, moderate with additional tenderness or considerably diminished or absent bowel noise, and severe with supervening signs of peritonitis or presumed sepsis. Stool characteristics were noted separately. Finally, outcome in terms of abdominal complications was recorded.
150 TABLE 1 Clinical data of patients with portal venous gas Patient (body weight)
Presenting abdominal findings, and stool characteristics
Evidence of gas (sonographic) Pv
g) g) g) g)
Moderate, fresh blood Mild, fresh blood Mild, fresh blood Severe, guaiac-neg.
5 (2770 g) 6 (2300 g)
Moderate, guaiac-neg. Moderate, fresh blood
1 (3510 2 (2500 3 (1570 4 (1530
7 (3550 8 (3010 9 (1530 10 (1640
g) g) g) g)
Liver
Comments and complications
TPN/AB TPN/AB TPN/AB TPN/AB
Coronavirus Rotavirus Coronavirus Lap., 24 days later: colonic strictures Lap.: NEC Lap., 30 days later: colonic strictures
(radiographic) Mural
Pv
Mural
0
0
X
ni 0 ni
ni 0 ni
X
X
0
X
X
ni
0
X
X X
ni
Severe, guaiac-pos. Mild, guaiac-pos. Severe, fresh blood Moderate, fresh blood
Therapy
X X
ni ni
TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB
Lap.: NEC Lap.: NEC
Pv: portal vein, Liver: microbubbles trapped within the minor branches of the portal vein producing patterns of hyperechogeneity in liver parenchyma, TPN: total parenteral nutrition, AB: antibiotics, ni: not investigated, np: not possible, NEC: necrotizing enterocolitis, Lap: laparotomy.
Results In lo/27 infants examined for suspected NEC, PVG was detected by sonography (Table 1). However, radiographs obtained in 8 infants failed to demonstrate PVG and intestinal pneumatosis was found in only 4/ 8 patients. Sonographically, gas bubbles appeared as high-amplitude echoes streaming through the portal vein in hepatopetal direction (Fig. la). In S/l0 patients, these were trapped within the minor branches of the
portal vein producing varying patterns of focal hyperechogeneity in the hepatic parenchyma (Fig. lb). Intramural gas was always found when the intestinal tract was examined by sonography. Laparotomy served to confirm the diagnosis of NEC in 3 children. The surgical findings showed severe dark-livid segments of dilated intestine, predominantly the colon, but no perforation. Intestinal pneumatosis was present in two patients (No. 9 and No. 10). Two patients (Nos. 5 and 10) underwent segmental and ileocoecal resection, re-
Fig. 1 (a). Intense reflections of gas bubbles within the major branches of the portal vein. (b) Transverse scan of the liver demonstrating tiple echogenic areas of trapped air within the minor branches of the portal vein.
mul-
151 TABLE 2 Clinical data of patients with intramural gas only Patient (body weight)
Presenting abdominal findings and stool character
Evidence of gas
Therapy
(sonographic)
(radiographic)
Pv
Liver
Mural
Pv
Mural
11 (1160 g) 12 (1950 g)
Moderate, guaiac-pos. Severe, guaiac-pos.
0
0
x
0
0
0
0
x
0
0
13 (1230 g) 14 (2950 g)
Moderate, guaiac-neg. Moderate, guaiac-neg.
0
0
x
0
x
0
0
x
ni
ni
15 (1480 16 (3750 17 (1500 18 (1425
Moderate, guaiac-neg. Severe, guaiac-neg. Severe, fresh blood Moderate. fresh blood
0
0
0
0
ni x
np
np
ni 0 x 0
g) g) g) g)
0
x
Comments and complications
;
TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB TPN/AB
Lap.: Perforation, Adhesions, NEC Lap., 18 days later: Colonic strictures
Lap.: Adhesions, NEC Lap.: Early NEC
See Table 1 for notes
spectively. One patient (No. 9) had a hemicolectomy; another two patients were laparotomized for mechanical ileus later on and showed colonic strictures. Eight patients had sonographic evidence of intramural gas, but no PVG (Table 2). In one of these patients, a suspicious pattern was seen in the hepatic parenchyma. Only 3 of 6 children who underwent radiography showed intestinal pneumatosis. In one patient, who presented with additional PVG, sonographic examination of the liver was unsuccessful due to intestinal me-
teorism. Sonographically, air in the intestinal walls was seen as gaseous strips or gas bubbles of high-amplitude echogeneity (Fig. 2 a and b). Laparotomy confirmed the diagnosis of NEC in those two infants who had a suspicious hepatic parenchyma and radiographic evidence of PVG, respectively, and in one child presenting with covered ileocecal perforation. At surgery segmental hyperaemia was present in one patient (No. 18) and dark-livid segmental areas of intestine were present in two other patients (Nos. 12 and 17). In another in-
Fig. 2 (a). Cross-section through thickened intestinal walls showing gaseous strips of intramural air with shadowing. (b). Transverse scan through a pair of intestinal loops demonstrate intraluminal air on the right side and a gaseous strip of intramural air on the left side of the sonogram.
152 TABLE 3 Clinical data of patients without any pathological gas Patient body weight
Presenting abdominal findings and stool character
Therapy
Evidence of gas (radiographic)
(sonographic) Pv
Liver
Mural
Pv
Mural
0
0
AB
0
0
0
0
-iTPN/AB
0
TPN/AB
0
TPN/AB
ni 0 0 0
-ITPN/AB
19 (700 g) 20 (1220 g) 21 (1830 g)
Mild, guaiac-neg. Moderate, guaiac-neg. Mild, guaiac-pos.
0
0
0
0
0
0
ni 0 0
22 (710 g)
Severe, guaiac-neg.
0
0
0
23 24 25 26 27
Moderate, guaiac-pos. Mild, fresh blood Mild, guaiac-pos. Moderate, guaiac-neg. Moderate, guaiac-neg.
(2440 (4010 (2500 (2540 (2050
g) g) g) g) g)
Comments and complications
Lap., 8 days later: Colonic strictures Lap.: Meconiumileus, Perforation H. simplex
-lTPN/AB
See Table 1 for notes.
fant laparotomy was performed for colonic strictures later on. In all 6 children with confirmed NEC laparotomy had been performed within 24 hours after sonography due to rapid clinical deterioration. Nine patients showed no pathological gas formation (Table 3). Two of these underwent laparotomy. One child was admitted from another hospital with clinical evidence of NEC established 4 weeks previously. He presented with intestinal stenosis due to colonic strictures and adhesions. The other child showed intestinal perforation caused by meconium ileus. Discussion NEC is thought to be the final common response to the interaction of varying etiologic factors, including viral infections [ 2,3]. Pathoanatomical examinations reveal progressive necrosis and ulceration of the intestinal layers leading to perforation or pseudomembrane formation [ 81. Intestinal injury permits the passage of air, delivered by bacterial fermentation of formula feedings, into the intestinal wall or portal vein [l-3]. Pertinent radiographic evidence is considered pathognomanic disease. Kosloske et al., who used radiographic criteria for the diagnosis of NEC, found intestinal pneumatosis in 95% and PVG in 23 % of 147 infants. All patients studied by the authors belonged to a group of 242 infants, who were treated for NEC within the same period [9].
Radiography produces static images of dynamic processes and requires certain minimum quantities of air to demonstrate PVG. Indeed, the radiographic finding of PVG is associated with pan-necrosis and is considered a valid indication for operation [ 3, lo]. Infants presenting with clinical symptoms of NEC, but no radiographic abnormalities, are not counted as established cases of NEC. This important group is thought to represent early or developing NEC [ 81. Sonographically, we found intramural gas and/or PVG in 18/27 infants examined for suspected NEC. Radiographs of 22 children obtained within the same period showed intestinal pneumatosis in only seven infants, one of these also had PVG. The sonographic diagnosis was confirmed by laparotomy in nine patients. Intestinal strictures found in three of these infants are well-known late complications of the disease [2,3]. Unlike static radiographs, real-time sonography shows gas bubbles streaming through the portal vein and might therefore serve to demonstrate PVG prior to radiographic evidence. This was confirmed by Merritt et al. [4] and Lindley et al. [7] in 6/S and 5/5 infants, respectively and by our own results in all of 10 infants with sonographically detected PVG. Due to the large acoustic impedance of air, gas bubbles are clearly visible as bright echoes. When transiently trapped within the peripheral branches of the portal vein they produce varying patterns of hyperechogeneity and may be the only suspicious finding supporting the diagnosis of NEC [4]. Although sonography is fascinating for its simplicity
153
in showing PVG and intrahepatic patterns of hyperechogeneity, absence of these findings does not exclude NEC. Intramural gas, the sonographic equivalent preceding radiologically detectable intestinal pneumatosis [lo], may then be the only sign of suspected NEC. The examining physician should be aware that intramural gas, which pathological examinations revealed to be associated with pseudomembrane formation [ 81, should be ruled out by sonography, especially in the absence of PVG. Patients who were examined carefully and showed no pathological gas formation did not develop overt NEC, even without therapy. Among cases with radiographic diagnosis of NEC, mortality was 65 7; in patients with PVG and 86% in those with pneumatosis and PVG combined [ 91. In our series with sonographically detected PVG and/or intestinal pneumatosis none of the infants died from NEC. This might be due to detection at an earlier stage by ultrasonography as compared with radiography. Nevertheless, it remains unclear why patients who presented with intestinal pneumatosis showed no PVG, although abdominal findings were comparable to those obtained in patients with PVG. One possible explanation might be concurrent antibiotic therapy in some patients. This question might be clarified in a prospective study with sonographic examinations prior to the start of therapy.
References 1 Amon RG, Fishbein JF. Portal venous gas in the pediatric age group. J Pediatr 1971; 79: 255-259. 2 Koloske AM, Musemeche CA. Necrotizing enterocolitis of the neonate. Clin Perinatol 1989; 16: 97-111. 3 Walsh MC, Kliegmann RM. Necrotizing enterocolitis: Treatment based on staging criteria. Pediatr Clin North Am 1986; 33: 179201. 4 Merritt CRB, Goldsmith JP, Sharp MJ. Sonographic detection of portal venous gas in infants with necrotizing enterocolitis. AJR 1984; 143: 1059-1062. 5 Robberecht EA, Afschrift M, De Be1 CE, Van Haesebrouck PJ, Van Bever HP, De Wit M, Leroy JG. Sonographic demonstration of portal venous gas in necrotizing enterocolitis. Eur J Pediatr 1988; 147: 192-194. 6 Mahn SW, Bhutani VK, Ritchie WW, Hall ML, Paul D. Echogenic intravascular and hepatic microbubbles associated with necrrotizing enterocolitis. J Pediatr 1983; 103: 637-640. I Lindley S, Mollitt DL, Seibert JJ, Golladay ES. Portal vein ultrasonography in the early diagnosis of necrotizing enterocolitis. J Pediatr Surg 1986; 21: 530-532. 8 Santulli TV, Schullinger JN, Heird WC, Gongaware RD, Wigger J, Barlow B, Blanc WA, Berdon WE. Acute necrotizing enterocolitis in infancy: A review of 64 cases. Pediatr 1975; 55: 376387. 9 Kosloske AM, Musemeche CA, Ball WS, Ablin DS, Bhattacharyya N. Necrotizing enterocolitis: Value of radiographic findings to predict outcome. AJR 1988; 151: 771-774. 10 Vemacchia FS, Jeffrey RB, Laing FC, Wing VW. Sonographic recognition of pneumatosis intestinalis. AJR 1985; 145: 51-52.
