Hypogastrinemia By M. Davenport,
and Esophageal
M. Mughal,
Manchester,
Esophageal atresia, plasma gastrin.
E
SOPHAGEAL atresia (EA) and tracheoesophageal fistula (TEF) are complex congenital anomalies affecting 1 in 3,000 to 3,500 live births.’ The embyrological basis for the lesion is believed to be abnormalities in foregut septation,’ although this has been disputed recently.2 Extraesophageal manifestations of the EA anomaly have been shown in infants dying without surgical intervention and include abnormalities of tracheal innervation3 and myenteric innervation of the distal esophagus and gastric fundus. The cephalic phase of gastric acid secretion is mediated by cephalovagal pathways principally in response to the presence of food in the mouth and pharynx. Long vagovagal pathways also mediate a secretory response to the stimulus of gastric distension. It has been estimated that one third of the acid response to a meal in an adult is caused by the cephalic and, therefore, vagally mediated phase.5 The effect of vagal fibers on gastrin release is competitive with both excitatory and inhibitory pathways being demonstrable. In the dog the dominant influence is excitatory, vagal blockade abolishing the normal gastrin response to a meal and further reducing basal gastrin levels.6 From the Sub Department of Paediatric Surgery, St May s Hospital, Manchester, England; the University Department of Surgery Hope Hospital, Salford, England; and the University Department of Surgery, Manchester Royal Infirmary, Manchester, England. Date accepted: November 20, 1990. Address reprint requests to Mark Davenport, FRCS, Lecturer in Surgery, Firm III Office, King’s College Hospital, Denmark Hill, LONDON SE5 9RS, England. Copyright o 1992 by W.B. Saunders Company 0022-346819212705-0008$03.00l0
568
and C.M. Doig
England and Salford, England
l To investigate extraesophageal anomalies in infants with esophageal atresia, preoperative plasma gastrin was measured in 12 infants. The median plasma gastrin was 32 rig/L (interquartile range, 24 to 44). There was significant correlation with birth weight (rs = .73, P < .05) and gestational age (rs = .74, P < .05). Within this group 9 infants of >36 weeks’ gestation were matched to a group of 20 control infants without esophageal atresia. Infants with esophageal atresia had a significantly lower median plasma gastrin (36 rig/L v 55 rig/L,, P < .05). This may indicate preexisting vagal abnormalities in esophageal atresia unrelated to surgical intervention. Copyright o 1992 by W.B. Saunders Company
INDEX WORDS:
R.F. McCloy,
Atresia
If the EA anomaly has associated vagal abnormalities then this may lead to changes in gastrin levels in comparison to the normal range of gastrin levels seen in the neonatal period. This study was designed to investigate infants with EA preoperatively and, therefore, to exclude any surgical cause for any changes seen. MATERIALS
AND
METHODS
Twelve infants with EA were transferred to the regional Neonatal Surgical Unit (NSU) between January 1988 and June 1989. Median gestational age was 38 weeks (range, 32 to 40) with the median birth weight of 2.5 kg (range, 1.5 to 3.2). Eleven of 12 infants had an associated TEF. One infant had the CHARGE (choanal atresia, vertebral anomalies, renal anomalies) association and a another had Down’s syndrome. A OS- to 1-mL heparinized blood sample was obtained by peripheral venepuncture or heelstab during the preoperative period. This varied from 6 to 48 hours of birth. During that period a blood sample was obtained by similar technique from 20 infants of gestation 36 to 41 weeks and aged 0 to 48 hours. These infants either had a diagnosis of anorectal atresia and, therefore, were transferred to the NSU within 24 hours of birth and unfed or had a surgical condition causing early transfer and delay of feeding (eg, sacrococcygeal teratoma, exomphalos minor) or were admitted to the Neonatal Medical Unit for delivery-related problems (eg, asphyxia). All the infants in the control group (median birth weight, 3.1 kg; range, 1.5 to 3.9) were unfed, not vomiting, and not clinically unwell. No infant required respiratory support. If an intravenous infusion was present the solution infused did not contain calcium. This study was approved by the Central Manchester Health Authority Ethical Committee.
Technique The samples were centrifuged, separated, and stored within 1 hour of collection. Storage was at -20°C until analysis. A commercial kit (Becton Dickinson Immunodiagnostics, Orangeburg, NY) was used to measure plasma gastrin. This is a radioimmunoassay based upon the original technique of Berson and Yalow,’ which reacts equally with both sulphated and nonsulphated forms of G17 and G34. In our laboratory the median fasting plasma gastrin is 32 rig/L with a normal upper limit of 85 rig/L.. Intraassay variation as assessed by within batch precision was 4.5% at 36 rig/L and 4.2% at 92 rig/L.. Interassay variation as assessed by between batch precision was 8.4% at 38 rig/L..
Statistics The data were analyzed initially using a Kolmogirnov-Smirnov test to assess normality of distribution. This indicated nonnormality of distribution (P < .05) and, therefore, the data are described using medians and interquartile range. Groups were compared using a nonparametric statistical test (Mann-Whitney LI test). A Spearman’s Rho (rs) coefficient was calculated to indicate correlation. Significance was assumed at P < .05.
JournalofPediatric
Surgery,
Vol 27,
No 5
(May), 1992: pp 568-571
ESOPHAGEAL
ATRESIA AND SERUM GASTRIN
569
RESULTS
(nGIL)
6. rtrin
Esophageal Atresia .
The median basal gastrin for 12 infants with EA was 32 rig/L (range, 24 to 44). There was significant correlation of plasma gastrin with gestational age (rs = .74, P < .Ol; Fig 1) and birth weight (rs = .73, P < .Ol; Fig 2).
.
. . .
Comparison With Control Infants
.
In order to reduce the influence of gestational age only infants of > 36 weeks were used in the comparison (Fig 3). Therefore, this includes 9 infants with EA and 20 controls. The groups were similarly distributed for birth weight. There was a significant difference between the median of the OA group (38 rig/L;; interquartile range, 31 to 48) and the median of the control group (5.5 rig/L;; interquartile range, 38 to 90), with U = 46, P = .038. DISCUSSION
.
.
20-
.
lo-
I
I
1 .O
1.5
2.0
Birth
Weight
I
L
2.5
3.0
3.5
(Kg)
Fig 2. Effect of birth weight on plasma gastrin in 12 infants with EA (rs = .73, P < .Ol).
Gastrin 200
Gastrin
.
.
01
AIthough the normal range for basa1 unfed gastrin levels in an adult population is well established, such a situation does not exist for neonates. Rogers et a1 in 1974 were the first authors to measure gastrin in normal healthy full-term neonates.x Subsequent groups have measured basal gastrin in normal fullterm”” and preterm”J” infants. Although there are conflicting values in the “normal” ranges quoted
.
30-
(nG/L)
r
(nG/L)
. 6ol
. .
50 -
150
. .
.
40 -
.
. . 100
m
30
.
I
.
.
. 20 -
.
50
.
lo-
01 30
I
I
32
34
I
I
36
Gestational
38
Age
01
I
40
42
(Wks)
Fig 1. Effect of gestational age on plasma gastrin in 12 infants with EA (rs = .74, P c .Ol).
EA
CONTROL
Fig 3. Plasma gastrin level in g infants with EA (~36 weeks gestation) compared with 20 control infants (a36 weeks gestation). The median in EA group (38 ng/ L) is significantly lower than in control infants (55 rig/L;; P = ,038).
DAVENPORT ET AL
570
Table 1. Quoted Values for Serum and Plasma Gastrin in Neonates Meall Gestational
Day of
Gastrin
NO.
Age W4
Life
bxl/L)
Rogers et al*
12
Term
Cord blood
Euler et al9
32
Term
Cord blood
76
Rodgers et aI”
13
Term
Cord blood
130
22
Term
1
155
Lucas et alI2
20
Term
Cord blood
88
Preterm
Cord blood
80
Author
6
89
Moazam et all0
20
?
1
66
Hyman et aI’3
34
~37
7
64
NOTE. Values have been converted where appropriate: 1 pg/mL = 1 rig/L:: 2.1 rig/L = 1 pmol/L.
between studies (Table l), certain facts are established. The source of neonatal gastrin is the infant itself as no transplacental passage has been demonstrated and there is no umbilical arteriovenous difference.9 Furthermore, as the half-life in an adult is of the order of 7 minutes, then any residual maternal gastrin would be quite short-lived. The normal feeding infant exhibits a basal hypergastrinaemic state that lasts up to to 2 to 4 months.‘O This mayI or may not” require a first feed “trigger.” Initially there is no evidence of gastrin release in response to feeding and once again this may last from 2 weeks to 3 months.“J2J4 No study has investigated abnormalities of gastrin levels in relation to pathological states or demonstrated a change in gastrin levels with gestational age or birth weight. l2 This study has demonstrated that the preoperative basal plasma gastrin level is lower in infants with EA in comparison to our control infants. Although our control group is not in any sense normal it is comparable in terms of timing of sampling, lack of enteral stimulation, and gestational age. Furthermore, the quoted normal ranges in the literature are usually much higher. Actual hypergastrinaemia was not seen in any infant with EA (Fig 3).
There are two explanations for such a result. First, that there is a vagal abnormality in EA that is expressed as a defect in innervation of the body and antrum of stomach. The basal gastrin level released depends on a dominant excitatory vagal tone. The alternative hypothesis is that there is a mechanical interruption in fetal swallowing caused by the atresia. The lack of presumed gastric stimulation by distension causes inhibition of fetal gastric growth and, therefore, a lowered postnatal basal gastrin from the smaller G-cell mass. If a TEF is present some amniotic fluid undoubtably passes to stomach via the fistula. The only infant in our series with a pure atresia had a gastrin level of < 10 rig/L,, the lowest in this series. In the adult a gross example of iatrogenic, yet desired, vagal damage is a truncal vagotomy. This causes a marked and chronic postoperative hypergastrinemia. If we accept that vagal stimulation liberates gastrin from G-cells, the explanation for this apparent paradox is that it is compensatory to the hypochlorhydria resulting from the desired effect of the surgery. This study further emphasizes that EA should not be thought of in isolation and that it perhaps has further subtle manifestations expressed as foregut physiological abnormalities. As gastrin is an important trophic hormone to foregut structures (stomach and duodenum), it may be that there is a relative failure to adapt to enteral feeding in the postoperative, postnatal life despite a satisfactory functioning, swallowing esophagus. ACKNOWLEDGMENT The authors thank Drs Sims and Chiswick for allowing infants under their care to be included in this study. Technical help in the performance of the Gastrin assays was by Dr I. Laing, Principle Biochemist at St Mary’s Hospital. All assays were performed by Dr R. Warburton. Senior Biochemist in the Department of Pathology, Wythenshawe Hospital, Manchester.
REFERENCES 1. Cudmore RE: Oeophageal atresia and tracheoesophageal fistula, in Rickham PP, Lister J, Irving IM (eds): Neonatal Surgery (ed 2). London, England, Butterworths, 1978, pp 189-208 2. Kluth D, Steding G, Seidl W: The embryology of foregut malformations. J Pediatr Sure 22389-393. 1987 2
hTnl,,,,,r.
”
T n..A:rm
DU
,X,_ll-
-i-D.
At.-..---1
Auerbac,,
PUXUJ LUUK eaupnagus ano SLUU~~FII or parrenrs wror esophageal atresia and tracheoesophageal fistula. J Pediatr Surg 21:831-837, 10%
l,“”
4. Nakazato Y, Wells TR, Landing BH: Abnormal tracheal innervation in patients with esophageal atresia and tracheoesophageal fistula: Study of the intrinsic tracheal nerve plexuses by a microdissection technique. J Pediatr Surg 21:838-844, 1986 5. Richardson CT, Walsh JH, Cooper KA, et al: Studies on the
role of cephalic-vagal stimulation in the acid secretory response to eating in normal human subjects. J Clin Invest 60:435-441, 1977 6. Greenberg GR: Influence of vagal integrity on gastrin and somatostatin release in dogs. Gastroenterology 93:994-1001,1987 7. Berson SA, Yalow RS: Radioimmunoassay in gastroenterology. Gastroenterology 62:1061-1084, 1972 8. Rogers IM, Davidson DC, Lawrence J, et al: Neonatal secretion of gastrin and glucagon. Arch Dis Child 49:796-801, 1974 9. Euler AR, Byrne WJ, Cousins LM, et al: Increased serum gastrin concentrations and gastric acid hyposecretion in the immediate newborn period. Gastroenterology 72:1271-1273, 1977 10. Moazam F, Kirby WJ, Rodgers BM, et al: Physiology of serum gastrin production in neonates and infants. Ann Surg 199:389-392,1984 11. Rodgers BM, Dix PM, Talbert JL, et al: Fasting and
ESOPHAGEAL
ATRESIA AND SERUM GASTRIN
postprandial serum gastrin in normal human neonates. J Pediatr Surg 13:13-16. 1978 12. Lucas A, Adrian TE, Christofides N, et al: Plasma motilin, gastrin, and enteroglucagon and feeding in the human newborn. Arch Dis Child .55:673-677,198O
571
13. Hyman PE, Clarke DD, Everett SL, et al: Gastric acid secretory function in preterm infants. J Pediatr 106:467-471, 1985 14. Marchini G, Winberg J, Uvnas-Moberg K: Plasma concentrations of gastrin and somatostatin after breast feeding in 4 day old infants. Arch Dis Child 63:1218-1221. 1988
and Esophageal
M. Mughal,
Manchester,
Esophageal atresia, plasma gastrin.
E
SOPHAGEAL atresia (EA) and tracheoesophageal fistula (TEF) are complex congenital anomalies affecting 1 in 3,000 to 3,500 live births.’ The embyrological basis for the lesion is believed to be abnormalities in foregut septation,’ although this has been disputed recently.2 Extraesophageal manifestations of the EA anomaly have been shown in infants dying without surgical intervention and include abnormalities of tracheal innervation3 and myenteric innervation of the distal esophagus and gastric fundus. The cephalic phase of gastric acid secretion is mediated by cephalovagal pathways principally in response to the presence of food in the mouth and pharynx. Long vagovagal pathways also mediate a secretory response to the stimulus of gastric distension. It has been estimated that one third of the acid response to a meal in an adult is caused by the cephalic and, therefore, vagally mediated phase.5 The effect of vagal fibers on gastrin release is competitive with both excitatory and inhibitory pathways being demonstrable. In the dog the dominant influence is excitatory, vagal blockade abolishing the normal gastrin response to a meal and further reducing basal gastrin levels.6 From the Sub Department of Paediatric Surgery, St May s Hospital, Manchester, England; the University Department of Surgery Hope Hospital, Salford, England; and the University Department of Surgery, Manchester Royal Infirmary, Manchester, England. Date accepted: November 20, 1990. Address reprint requests to Mark Davenport, FRCS, Lecturer in Surgery, Firm III Office, King’s College Hospital, Denmark Hill, LONDON SE5 9RS, England. Copyright o 1992 by W.B. Saunders Company 0022-346819212705-0008$03.00l0
568
and C.M. Doig
England and Salford, England
l To investigate extraesophageal anomalies in infants with esophageal atresia, preoperative plasma gastrin was measured in 12 infants. The median plasma gastrin was 32 rig/L (interquartile range, 24 to 44). There was significant correlation with birth weight (rs = .73, P < .05) and gestational age (rs = .74, P < .05). Within this group 9 infants of >36 weeks’ gestation were matched to a group of 20 control infants without esophageal atresia. Infants with esophageal atresia had a significantly lower median plasma gastrin (36 rig/L v 55 rig/L,, P < .05). This may indicate preexisting vagal abnormalities in esophageal atresia unrelated to surgical intervention. Copyright o 1992 by W.B. Saunders Company
INDEX WORDS:
R.F. McCloy,
Atresia
If the EA anomaly has associated vagal abnormalities then this may lead to changes in gastrin levels in comparison to the normal range of gastrin levels seen in the neonatal period. This study was designed to investigate infants with EA preoperatively and, therefore, to exclude any surgical cause for any changes seen. MATERIALS
AND
METHODS
Twelve infants with EA were transferred to the regional Neonatal Surgical Unit (NSU) between January 1988 and June 1989. Median gestational age was 38 weeks (range, 32 to 40) with the median birth weight of 2.5 kg (range, 1.5 to 3.2). Eleven of 12 infants had an associated TEF. One infant had the CHARGE (choanal atresia, vertebral anomalies, renal anomalies) association and a another had Down’s syndrome. A OS- to 1-mL heparinized blood sample was obtained by peripheral venepuncture or heelstab during the preoperative period. This varied from 6 to 48 hours of birth. During that period a blood sample was obtained by similar technique from 20 infants of gestation 36 to 41 weeks and aged 0 to 48 hours. These infants either had a diagnosis of anorectal atresia and, therefore, were transferred to the NSU within 24 hours of birth and unfed or had a surgical condition causing early transfer and delay of feeding (eg, sacrococcygeal teratoma, exomphalos minor) or were admitted to the Neonatal Medical Unit for delivery-related problems (eg, asphyxia). All the infants in the control group (median birth weight, 3.1 kg; range, 1.5 to 3.9) were unfed, not vomiting, and not clinically unwell. No infant required respiratory support. If an intravenous infusion was present the solution infused did not contain calcium. This study was approved by the Central Manchester Health Authority Ethical Committee.
Technique The samples were centrifuged, separated, and stored within 1 hour of collection. Storage was at -20°C until analysis. A commercial kit (Becton Dickinson Immunodiagnostics, Orangeburg, NY) was used to measure plasma gastrin. This is a radioimmunoassay based upon the original technique of Berson and Yalow,’ which reacts equally with both sulphated and nonsulphated forms of G17 and G34. In our laboratory the median fasting plasma gastrin is 32 rig/L with a normal upper limit of 85 rig/L.. Intraassay variation as assessed by within batch precision was 4.5% at 36 rig/L and 4.2% at 92 rig/L.. Interassay variation as assessed by between batch precision was 8.4% at 38 rig/L..
Statistics The data were analyzed initially using a Kolmogirnov-Smirnov test to assess normality of distribution. This indicated nonnormality of distribution (P < .05) and, therefore, the data are described using medians and interquartile range. Groups were compared using a nonparametric statistical test (Mann-Whitney LI test). A Spearman’s Rho (rs) coefficient was calculated to indicate correlation. Significance was assumed at P < .05.
JournalofPediatric
Surgery,
Vol 27,
No 5
(May), 1992: pp 568-571
ESOPHAGEAL
ATRESIA AND SERUM GASTRIN
569
RESULTS
(nGIL)
6. rtrin
Esophageal Atresia .
The median basal gastrin for 12 infants with EA was 32 rig/L (range, 24 to 44). There was significant correlation of plasma gastrin with gestational age (rs = .74, P < .Ol; Fig 1) and birth weight (rs = .73, P < .Ol; Fig 2).
.
. . .
Comparison With Control Infants
.
In order to reduce the influence of gestational age only infants of > 36 weeks were used in the comparison (Fig 3). Therefore, this includes 9 infants with EA and 20 controls. The groups were similarly distributed for birth weight. There was a significant difference between the median of the OA group (38 rig/L;; interquartile range, 31 to 48) and the median of the control group (5.5 rig/L;; interquartile range, 38 to 90), with U = 46, P = .038. DISCUSSION
.
.
20-
.
lo-
I
I
1 .O
1.5
2.0
Birth
Weight
I
L
2.5
3.0
3.5
(Kg)
Fig 2. Effect of birth weight on plasma gastrin in 12 infants with EA (rs = .73, P < .Ol).
Gastrin 200
Gastrin
.
.
01
AIthough the normal range for basa1 unfed gastrin levels in an adult population is well established, such a situation does not exist for neonates. Rogers et a1 in 1974 were the first authors to measure gastrin in normal healthy full-term neonates.x Subsequent groups have measured basal gastrin in normal fullterm”” and preterm”J” infants. Although there are conflicting values in the “normal” ranges quoted
.
30-
(nG/L)
r
(nG/L)
. 6ol
. .
50 -
150
. .
.
40 -
.
. . 100
m
30
.
I
.
.
. 20 -
.
50
.
lo-
01 30
I
I
32
34
I
I
36
Gestational
38
Age
01
I
40
42
(Wks)
Fig 1. Effect of gestational age on plasma gastrin in 12 infants with EA (rs = .74, P c .Ol).
EA
CONTROL
Fig 3. Plasma gastrin level in g infants with EA (~36 weeks gestation) compared with 20 control infants (a36 weeks gestation). The median in EA group (38 ng/ L) is significantly lower than in control infants (55 rig/L;; P = ,038).
DAVENPORT ET AL
570
Table 1. Quoted Values for Serum and Plasma Gastrin in Neonates Meall Gestational
Day of
Gastrin
NO.
Age W4
Life
bxl/L)
Rogers et al*
12
Term
Cord blood
Euler et al9
32
Term
Cord blood
76
Rodgers et aI”
13
Term
Cord blood
130
22
Term
1
155
Lucas et alI2
20
Term
Cord blood
88
Preterm
Cord blood
80
Author
6
89
Moazam et all0
20
?
1
66
Hyman et aI’3
34
~37
7
64
NOTE. Values have been converted where appropriate: 1 pg/mL = 1 rig/L:: 2.1 rig/L = 1 pmol/L.
between studies (Table l), certain facts are established. The source of neonatal gastrin is the infant itself as no transplacental passage has been demonstrated and there is no umbilical arteriovenous difference.9 Furthermore, as the half-life in an adult is of the order of 7 minutes, then any residual maternal gastrin would be quite short-lived. The normal feeding infant exhibits a basal hypergastrinaemic state that lasts up to to 2 to 4 months.‘O This mayI or may not” require a first feed “trigger.” Initially there is no evidence of gastrin release in response to feeding and once again this may last from 2 weeks to 3 months.“J2J4 No study has investigated abnormalities of gastrin levels in relation to pathological states or demonstrated a change in gastrin levels with gestational age or birth weight. l2 This study has demonstrated that the preoperative basal plasma gastrin level is lower in infants with EA in comparison to our control infants. Although our control group is not in any sense normal it is comparable in terms of timing of sampling, lack of enteral stimulation, and gestational age. Furthermore, the quoted normal ranges in the literature are usually much higher. Actual hypergastrinaemia was not seen in any infant with EA (Fig 3).
There are two explanations for such a result. First, that there is a vagal abnormality in EA that is expressed as a defect in innervation of the body and antrum of stomach. The basal gastrin level released depends on a dominant excitatory vagal tone. The alternative hypothesis is that there is a mechanical interruption in fetal swallowing caused by the atresia. The lack of presumed gastric stimulation by distension causes inhibition of fetal gastric growth and, therefore, a lowered postnatal basal gastrin from the smaller G-cell mass. If a TEF is present some amniotic fluid undoubtably passes to stomach via the fistula. The only infant in our series with a pure atresia had a gastrin level of < 10 rig/L,, the lowest in this series. In the adult a gross example of iatrogenic, yet desired, vagal damage is a truncal vagotomy. This causes a marked and chronic postoperative hypergastrinemia. If we accept that vagal stimulation liberates gastrin from G-cells, the explanation for this apparent paradox is that it is compensatory to the hypochlorhydria resulting from the desired effect of the surgery. This study further emphasizes that EA should not be thought of in isolation and that it perhaps has further subtle manifestations expressed as foregut physiological abnormalities. As gastrin is an important trophic hormone to foregut structures (stomach and duodenum), it may be that there is a relative failure to adapt to enteral feeding in the postoperative, postnatal life despite a satisfactory functioning, swallowing esophagus. ACKNOWLEDGMENT The authors thank Drs Sims and Chiswick for allowing infants under their care to be included in this study. Technical help in the performance of the Gastrin assays was by Dr I. Laing, Principle Biochemist at St Mary’s Hospital. All assays were performed by Dr R. Warburton. Senior Biochemist in the Department of Pathology, Wythenshawe Hospital, Manchester.
REFERENCES 1. Cudmore RE: Oeophageal atresia and tracheoesophageal fistula, in Rickham PP, Lister J, Irving IM (eds): Neonatal Surgery (ed 2). London, England, Butterworths, 1978, pp 189-208 2. Kluth D, Steding G, Seidl W: The embryology of foregut malformations. J Pediatr Sure 22389-393. 1987 2
hTnl,,,,,r.
”
T n..A:rm
DU
,X,_ll-
-i-D.
At.-..---1
Auerbac,,
PUXUJ LUUK eaupnagus ano SLUU~~FII or parrenrs wror esophageal atresia and tracheoesophageal fistula. J Pediatr Surg 21:831-837, 10%
l,“”
4. Nakazato Y, Wells TR, Landing BH: Abnormal tracheal innervation in patients with esophageal atresia and tracheoesophageal fistula: Study of the intrinsic tracheal nerve plexuses by a microdissection technique. J Pediatr Surg 21:838-844, 1986 5. Richardson CT, Walsh JH, Cooper KA, et al: Studies on the
role of cephalic-vagal stimulation in the acid secretory response to eating in normal human subjects. J Clin Invest 60:435-441, 1977 6. Greenberg GR: Influence of vagal integrity on gastrin and somatostatin release in dogs. Gastroenterology 93:994-1001,1987 7. Berson SA, Yalow RS: Radioimmunoassay in gastroenterology. Gastroenterology 62:1061-1084, 1972 8. Rogers IM, Davidson DC, Lawrence J, et al: Neonatal secretion of gastrin and glucagon. Arch Dis Child 49:796-801, 1974 9. Euler AR, Byrne WJ, Cousins LM, et al: Increased serum gastrin concentrations and gastric acid hyposecretion in the immediate newborn period. Gastroenterology 72:1271-1273, 1977 10. Moazam F, Kirby WJ, Rodgers BM, et al: Physiology of serum gastrin production in neonates and infants. Ann Surg 199:389-392,1984 11. Rodgers BM, Dix PM, Talbert JL, et al: Fasting and
ESOPHAGEAL
ATRESIA AND SERUM GASTRIN
postprandial serum gastrin in normal human neonates. J Pediatr Surg 13:13-16. 1978 12. Lucas A, Adrian TE, Christofides N, et al: Plasma motilin, gastrin, and enteroglucagon and feeding in the human newborn. Arch Dis Child .55:673-677,198O
571
13. Hyman PE, Clarke DD, Everett SL, et al: Gastric acid secretory function in preterm infants. J Pediatr 106:467-471, 1985 14. Marchini G, Winberg J, Uvnas-Moberg K: Plasma concentrations of gastrin and somatostatin after breast feeding in 4 day old infants. Arch Dis Child 63:1218-1221. 1988
