PAPER Esophageal pH Monitoring Abnormalities and Gastroesophageal Reflux Disease in With

Stephen G. Jolley, MD; Mary L. Lorenz, RN, BSN; Margo Hendrickson, MD; J. Parker Kurlinski, MD

Hypothesis: Infants with rotational abnormalities of the GERD was reported as a combination of clinical outcome are at high risk for gastroesophageal and subsequent extended esophageal pH monitoring. reflux disease (GERD) and for sudden death (SID) from GERD. Results: Of the 91 infants studied, 80 (88%) had GERD and 26 (29%) had the risk factor for SID from GERD. Of Design: A survey of the prevalence of GERD and the risk 55 infants with intestinal malrotation alone, 52 (95%) factor for SID from GERD in a case series of infants treated had GERD, and 20 (36%) had the risk factor for SID from for congenital anomalies that include intestinal malro- GERD. Although GERD was found in 19 (83%) of 23 pa- tation. Eighty-one (89%) of the infants studied for GERD tients with repaired abdominal wall defects, the preva- had a mean follow-up of 23.2 months (median, 12 lence of the risk factor for SID from GERD was signifi- months). cantly lower (13% [3 patients]; P = .03) than in patients with intestinal malrotation alone. The prevalence of GERD Setting: Patients treated in 2 tertiary care centers con- in infants with repaired congenital diaphragmatic her- sisting of a children’s hospital and a university medical nia was significantly lower (69% [9/13]; P = .02) than in center. infants with intestinal malrotation alone but not for the prevalence of the risk factor for SID from GERD (23% Patients: Two hundred eighty-six consecutive infants [3/13]; P = .19). Both symptomatic and asymptomatic pa- were treated for congenital anomalies from September 1, tients had similar prevalences of GERD (91% [52/57] vs 1985, through May 31, 1998. The patients selected for study 82% [28/34]; P = .17) and for the risk factor for SID from were 91 infants who had 18- to 24-hour esophageal pH GERD (31% [18/57] vs 24% [8/34]; P = .28). On follow- monitoring performed and no prior operation on the stom- up, the prognosis for GERD in infants with intestinal mal- ach or esophagus. The studied infants had intestinal mal- rotation was better in the infants who were asymptom- rotation either alone (n = 55) or associated with a re- atic than in those who were symptomatic at the initial paired abdominal wall defect (n = 23) or congenital extended esophageal pH monitoring. diaphragmatic (n = 13). Of the 91 infants, 34 were asymptomatic at the time of esophageal pH monitoring. Conclusions: The prevalence of GERD in infants with intestinal malrotation is high, and the prevalence of the Interventions: Eighteen- to 24-hour esophageal pH risk factor for SID from GERD is a significant concern. monitoring was used to determine the presence of GERD The prevalence of GERD is lower in infants with con- (abnormal pH score Ͼ2 hours postcibal) and the risk fac- genital diaphragmatic hernia. Infants with repaired ab- tor for SID from GERD (type I or III reflux pattern in com- dominal wall defects have a lower prevalence of the risk bination with a prolonged mean duration of sleep re- factor for SID from GERD. We recommend careful evalu- flux). ation and follow-up of infants with intestinal malrota- tion for problems, such as SID, from GERD. Main Outcome Measures: The prevalence of GERD and the risk factor for SID from GERD. The follow-up of Arch Surg. 1999;134:747-753

OTATIONAL abnormalities of toms from intermittent intestinal obstruc- the midgut mesentery are a tion as a consequence of midgut From the Departments of ubiquitous congenital or internal herniation. These symptoms re- Surgery (Drs Jolley and anomaly. Intestinal malro- ferable to obstruction are usually re- Hendrickson and Ms Lorenz) tation is found either alone lieved successfully by the Ladd operative and Pediatrics (Dr Kurlinski), or in combination with other congenital procedure with or without suture fixa- University of Nevada School of R anomalies such as abdominal wall de- tion of the midgut mesentery. The Ladd Medicine, Sunrise Children’s Hospital, and the University fects or congenital diaphragmatic hernia. operation appears to be less commonly Medical Center of Southern The lack of normal mesenteric fixation needed in infants with abdominal wall de- Nevada, Las Vegas. leads most often to acute or chronic symp- fects or congenital diaphragmatic her-

ARCH SURG/ VOL 134, JULY 1999 747

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 PATIENTS AND METHODS to a method reported previously.5 The esophageal pH score more than 2 hours after a meal was used to determine the PATIENT POPULATION presence or absence of GERD. A score greater than 64 was thought to be abnormal and indicative of GERD. In pa- From September 1, 1985, to May 31, 1998, a total of 286 tients with GERD, the type I, II, or III pattern of reflux was consecutive infants with rotational abnormalities of the mid- also determined from the esophageal pH recording. These gut mesentery were treated at Sunrise Children’s Hospital reflux patterns occur following feedings of apple juice (300 and the University Medical Center of Southern Nevada, Las mL/m2 of body surface area) and appear as continuous re- Vegas. Ninety-one of these infants were selected for this re- flux episodes (type I), discontinuous episodes of reflux port because they had no prior operation on the or (type II), or a mixture of continuous and discontinuous oc- esophagus and underwent extended esophageal pH moni- currences of reflux episodes (type III).6 The mean duration toring. Informed written consent for the extended esopha- of acid reflux episodes during sleep more than 2 hours after geal pH studies was obtained from the patients’ parents or a feeding (ZMD) was determined. A ZMD was either pro- legal guardian. Of the patients studied, 55 had intestinal mal- longed (Ͼ3.8 minutes) or normal (Յ3.8 minutes). The risk rotation without an abdominal wall defect or congenital dia- factor in the esophageal pH recording that has been associ- phragmatic hernia, 23 had a repaired abdominal wall defect ated with SID from GERD is the coexistence of a type I or ( [n = 5] and [n = 18]), and 13 were III reflux pattern and a prolonged ZMD.9 survivors of a repaired congenital Bochdalek diaphrag- matic hernia. Fifty-nine infants (65%) had additional con- PATIENT TREATMENT AND FOLLOW-UP ditions, shown in Table 1. The ages at the time of initial extended esophageal pH monitoring ranged from 2 weeks All patients with GERD documented by extended esoph- to 9 months (mean, 2.3 months; median, 2 months). There ageal pH monitoring were offered treatment of their GERD. were 59 boys and 32 girls. Fifty-seven patients (63%) had Patients with symptoms were offered aggressive treat- symptoms suggestive of GERD. These included repeated non- ment with medical antireflux therapy (postural therapy, bilious emesis (n = 36), growth retardation (n = 15), recur- thickened feedings, and a combination of prokinetic and rent respiratory tract symptoms (n = 25), and irritability histamine 2-receptor–blockade ). An antire- (n = 2). Eighteen infants had more than 1 symptom. Thirty- flux operation was offered if medical therapy for 6 weeks four infants had no symptoms to suggest GERD and were did not control symptoms or if symptoms were life- studied because of a perceived increased risk for SID from threatening. In patients with a prolonged ZMD, medical GERD.2 Of the 42 symptomatic infants studied with intes- therapy also included home cardiac and apnea monitor- tinal malrotation alone, 19 were found to have unsuspected ing. Infants with no symptoms of GERD were offered pos- and untreated intestinal malrotation at the time of an anti- tural therapy as long as they remained asymptomatic. Home reflux operation after GERD had been identified by ex- cardiac and apnea monitoring was recommended in the tended esophageal pH monitoring. Although all symptom- treatment of asymptomatic infants with a prolonged ZMD. atic infants had a barium esophagram and upper Follow-up information was available in 64 (89%) of the 72 series before esophageal pH monitor- patients in whom intestinal malrotation was diagnosed be- ing, intestinal malrotation was not detected in the 19 infants fore extended esophageal pH monitoring, and in 17 (89%) with intestinal malrotation found at the time of an antire- of the 19 patients in whom intestinal malrotation was di- flux operation. In the remaining 23 symptomatic infants, the agnosed at an antireflux operation (Table 2). Ladd operation without intestinal suture fixation was per- formed before extended esophageal pH monitoring. STATISTICAL ANALYSIS

EXTENDED ESOPHAGEAL pH MONITORING The prevalence of GERD and the risk factor for SID from GERD were determined. Patient groups were compared us- Intraesophageal pH monitoring was performed in the distal ing the Fisher exact test. Differences between patient groups esophagus for 18 to 24 hours and was evaluated according were considered to be significant if PϽ.05.

nia.1 Some chronic symptoms in infants following the Ladd (prolonged ZMD) or absence (normal ZMD) of reflux- operation for intestinal malrotation have been attrib- induced respiratory symptoms.8 A prolonged ZMD in uted to gastric emptying abnormalities,2 but others have combination with a type I or III reflux pattern is associ- been associated with gastroesophageal reflux disease ated with the highest risk for SID from untreated GERD, (GERD).3,4 Infants with intestinal malrotation have been even in asymptomatic infants.9 thought to be at risk for GERD, including sudden infant In the present study, we describe a large group of death (SID) from GERD.2 infants with intestinal malrotation who were studied with When performed properly, extended (18- to 24- extended esophageal pH monitoring in an attempt to de- hour) esophageal pH monitoring has become a useful tool termine the prevalence of GERD and the risk for SID from in the identification of GERD in children. An abnormal GERD. pH score more than 2 hours after a meal is used to iden- 5 tify GERD. Three patterns of GERD—types I, II, and III— RESULTS also can be detected with esophageal pH monitoring and correlate with the course and prognosis of treated child- Of the 91 infants studied with extended esophageal pH hood GERD.6,7 The mean duration of acid reflux epi- monitoring, 80 (88%) had GERD and 26 (29%) had the sodes during sleep (ZMD) correlates with the presence risk factor for SID from GERD. The prevalence of GERD

ARCH SURG/ VOL 134, JULY 1999 748

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 Table 1. Ages and Associated Conditions in 59 Infants Table 2. Length of Follow-up in the Patient Groups With Intestinal Malrotation Studied for Gastroesophageal With Intestinal Malrotation Reflux Disease With Extended Esophageal pH Monitoring* Follow-up, mo No. of Infants With Intestinal Malrotation Group (No. of Patients) Range Mean Median With With Repaired Intestinal malrotation alone (48) Repaired Congenital Diagnosis before pH study (31) 1-132 25.5 12 Associated Abdominal Diaphragmatic Diagnosis after pH study (17) 1-101 18.8 10 Condition Alone Wall Defect Hernia Total Repaired abdominal wall defect (21) 2-72 21.6 19 Age, range, mo 0.8-9.0 0.2-8.0 0.5-2.0 0.2-9.0 Repaired diaphragmatic hernia (12) 1-73 26.4 9 Mean (median) 2.6 (2) 2.2 (2) 1.1 (1) 2.3 (2) All patients (81) 1-132 23.2 12 Associated conditions CNS disease 8 1 2 11 Congenital 51 1 7 disease testinal malrotation diagnosed at the time of antireflux Prematurity 14 18 1 33 operation and after GERD had been documented by ex- Jejunoileal atresia 3 1 0 4 Anorectal atresia 2 0 0 2 tended esophageal pH monitoring (type I, 63% [12/19]; Cystic fibrosis 1 0 0 1 type II, 16% [3/19]; and type III, 21% [14/19]). 1 0 0 1 Laryngomalacia 1 0 0 1 REPAIRED ABDOMINAL WALL DEFECTS Cleft palate 1 1 0 2 Bronchopulmonary 00 1 1 Nineteen (83%) of the 23 infants with a repaired abdomi- dysplasia alone DiGeorge syndrome 1 0 0 1 nal wall defect were found to have GERD, and only 3 in- Meconium aspiration 1 0 0 1 fants (13%) were found to have the risk factor for SID from Pulmonary 00 1 1 GERD. The prevalence of the risk factor for SID from GERD sequestration was significantly lower (P = .03) in these infants than in Ͼ1 Condition 5 2 1 8 infants with isolated intestinal malrotation. Even though the risk factor for SID from GERD was less prevalent, the *CNS indicates central nervous system. prevalence of a prolonged ZMD was 39% (9/23). For symp- tomatic and asymptomatic patients with repaired abdomi- nal wall defects, the prevalences of GERD (75% [6/8] vs and the risk factor for SID from GERD were not higher 87% [13/15]; P = .90) and the risk factor for SID from GERD in patients with additional conditions (88% [52/59] and (0% [0/8] vs 20% [3/15]; P = .25) were not significantly 29% [17/59], respectively) than in those without addi- different. In the patients with a repaired abdominal wall tional conditions (88% [28/32] and 28% [9/32], respec- defect and GERD, the prevalence of each reflux pattern tively). Infants with symptoms were not more likely than was 47% (9/19) for type I, 42% (8/19) for type II, and 11% asymptomatic infants to have a significantly higher preva- (2/19) for type III. lence of GERD (91% [52/57] vs 82% [28/34]; P = .17) or the risk factor for SID from GERD (31% [18/57] vs 24% REPAIRED CONGENITAL [8/34]; P = .28). DIAPHRAGMATIC HERNIA

INTESTINAL MALROTATION The prevalence of GERD in infants with repaired con- WITHOUT ABDOMINAL WALL genital diaphragmatic hernia was 69% (9/13). This preva- OR DIAPHRAGMATIC DEFECT lence was significantly lower (P = .02) than that in in- fants with isolated intestinal malrotation. The prevalence Fifty-two (94%) of the 55 infants with isolated intesti- of the risk factor for SID from GERD in these patients nal malrotation had GERD. Twenty (36%) had the risk was 23% (3/13). Symptomatic and asymptomatic pa- factor for SID from GERD. The prevalence in patients with tients with repaired congenital diaphragmatic hernia had extended esophageal pH monitoring performed after in- similar prevalences for GERD (71% [5/7] vs 67% [4/6]; testinal malrotation was corrected was 92% (33/36) for P = .78) and the risk factor for SID from GERD (29% [2/7] GERD and 33% (12/36) for the risk factor for SID from vs 17% [1/6]; P = .56). In patients with repaired diaphrag- GERD. The presence or absence of symptoms did not sig- matic hernia and GERD, the prevalence of the type I re- nificantly affect the prevalence of GERD (96% [22 of 23 flux pattern was 89% (8/9) and of the type II reflux pat- patients] vs 85% [11 of 13 patients]; P = .29) or the preva- tern was 11% (1/9). No patients were seen with the type lence of the risk factor for SID from GERD (38% [8 of III pattern of GERD. 23 patients] vs 31% [4 of 13 patients]; P = .55). Among the infants with GERD and isolated intestinal malrota- FOLLOW-UP IN ALL PATIENTS WITH tion treated before extended esophageal pH monitor- INTESTINAL MALROTATION ing, the prevalence of each reflux pattern was 70% (23/ 33) for type I, 15% (5/33) for type II, and 15% (5/33) for Even though the presence or absence of symptoms did type III. This did not differ significantly from the preva- not have a significant bearing on the prevalence of GERD lence of reflux patterns in the 19 patients who had in- or the risk factor for SID from GERD, such was not the

ARCH SURG/ VOL 134, JULY 1999 749

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 Normal pH Study (n = 4) Normal pH Study (n = 4) Antireflux Operation (n = 0) Antireflux Operation (n = 0) No Antireflux Operation (n = 4) No Antireflux Operation (n = 4)

Asymptomatic (n = 2) Died of Bronchopulmonary Dysplasia (n = 2) Asymptomatic (n = 4)

Type I Reflux Pattern (n = 21) Type I Reflux Pattern (n = 17) Antireflux Operation (n = 11) Antireflux Operation (n = 0) No Antireflux Operation (n = 10) No Antireflux Operation (n = 17)

Persistent Symptoms (n = 2) Asymptomatic (n = 8) GERD Symptoms (n = 2) Asymptomatic (n = 15)

Died of Candida Medical Therapy pH Study, No (n = 4) pH Study, Yes (n = 4) pH Study, Yes (n = 1) pH Study, No (n = 1) pH Study, No (n = 8) pH Study, Yes (n = 7) (n = 1) (n = 1) Normal GERD GERD Medical Therapy Normal GERD (n = 0) (n = 4) (n = 1) (n = 1) (n = 3) (n = 4)

Type II Reflux Pattern (n = 6) Type II Reflux Pattern (n = 5) Antireflux Operation (n = 0) Antireflux Operation (n = 0) No Antireflux Operation (n = 6) No Antireflux Operation (n = 5)

Persistent Symptoms (n = 1) Asymptomatic (n = 5) GERD Symptoms (n = 0) Asymptomatic (n = 5)

Medical Therapy (n = 1)pH Study, No (n = 3) pH Study, Yes (n = 2) pH Study, No (n = 4) pH Study, Yes (n = 1)

Normal GERD Normal (n = 1) (n = 1) (n = 1) Type III Reflux Pattern (n = 3) Type III Reflux Pattern (n = 4) Antireflux Operation (n = 0) Antireflux Operation (n = 2) No Antireflux Operation (n = 3) No Antireflux Operation (n = 2) GERD Symptoms (n = 0) Asymptomatic (n = 3) Persistent Symptoms (n = 0) Asymptomatic (n = 2) pH Study, No (n = 1) pH Study, Yes (n = 2) pH Study, No (n = 2) Normal GERD Figure 1. Follow-up information in 35 infants with intestinal malrotation (n = 1) (n = 1) diagnosed before extended esophageal pH monitoring (pH study) and with symptoms suggestive of gastroesophageal reflux disease (GERD) at the time Figure 2. Follow-up information in 29 infants with intestinal malrotation of initial esophageal pH monitoring. The subsequent clinical course and diagnosed before extended esophageal pH monitoring (pH study) who had follow-up extended esophageal pH monitoring are presented according to no symptoms suggestive of gastroesophageal reflux disease (GERD) at the the findings on initial monitoring (normal or type I, type II, or type III reflux time of initial esophageal pH monitoring. The subsequent clinical course and pattern). All patients with GERD received medical antireflux therapy until the follow-up extended esophageal pH monitoring are presented according to GERD resolved or an antireflux operation was performed. the findings on initial monitoring (normal or type I, type II, or type III reflux pattern).

case for the subsequent clinical course in the 81 pa- lution of GERD with time was more likely in patients who tients observed. Two of the 4 patients with symptoms sug- were asymptomatic at the time of initial extended esoph- gestive of GERD and normal results on extended esoph- ageal pH monitoring. ageal pH monitoring died. Symptoms and the cause of None of the 25 patients observed with the risk fac- death were directly attributable to cor pulmonale from tor for SID from GERD and treated for GERD died of severe bronchopulmonary dysplasia. The remaining 2 GERD or of the sudden infant death syndrome (SIDS). symptomatic patients with normal results of extended There were 5 unrelated deaths in infants with GERD. Four esophageal pH monitoring subsequently became asymp- infants died of congenital heart disease beyond the post- tomatic without treatment of GERD. In contrast, all 4 operative period for their antireflux operation, and 1 in- asymptomatic patients with normal results of extended fant died of Candida albicans sepsis during medical treat- esophageal pH monitoring remained asymptomatic with- ment of both the sepsis and GERD. out any treatment of GERD. Thirteen (42%) of the 31 symptomatic infants observed with GERD detected af- COMMENT ter the diagnosis of intestinal malrotation required an an- tireflux operation to control GERD (7 Nissen fundopli- The prevalences of GERD and the risk factor for SID from cations and 6 Thal fundoplications). None of the 25 GERD in infants with intestinal malrotation are much asymptomatic infants with GERD required an antire- higher than expected. Because GERD has been associ- flux operation, however, even though in some, symp- ated with SIDS,9 an estimate of the prevalence of GERD toms from GERD subsequently developed (PϽ.001). The and the risk factor for SID from GERD in the general in- subsequent course of patients with GERD treated non- fant population can be made with certain assumptions. operatively is summarized in Figure 1 for symptom- The risk of sudden death from GERD was 9.1% in un- atic patients and in Figure 2 for asymptomatic pa- treated infants with the risk factor,9 and the risk of SIDS tients. Seventeen of the patients with GERD treated before the avoidance of the prone-flat position was 1.3 nonoperatively had extended esophageal pH monitor- per 1000 live births.10 By assuming that all SIDS deaths ing repeated after 1 year of age. The spontaneous reso- occur from having the risk factor for SID from GERD,

ARCH SURG/ VOL 134, JULY 1999 750

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 Table 3. Course of Treated Gastroesophageal Reflux Disease (GERD) by Age 1 Year According to Reflux Pattern, Symptoms, and Intestinal Malrotation*

Infants With Intestinal Malrotation Symptomatic Infants Without Intestinal Malrotation GERD Symptoms Asymptomatic Reflux Pattern and With GERD† (n = 35) (n = 29) Type I Antireflux operation needed 50 52 0 Fate of GERD in nonoperated-on infants‡ GERD present 40 48 63 GERD absent 10 0 37 Type II Antireflux operation needed 10 0 0 Fate of GERD in nonoperated-on infants‡ GERD present 20 50 50 GERD absent 70 50 100 Type III Antireflux operation needed 50 50 0 Fate of GERD in nonoperated-on infants‡ GERD present 40 Unknown 50 GERD absent 10 Unknown 50

*Data are given as percentage of patients. †From Jolley.15 ‡Results of extended esophageal pH monitoring.

then the prevalence of the risk factor for SID from GERD lence of the risk factor for SID from GERD. A shortened in the infant population would be 1.4%. Because the preva- intra-abdominal esophagus from increased intra- lence of the risk factor for SID from GERD in infants with abdominal pressure following repair of the abdominal wall GERD is 37%,11 the prevalence of GERD in the infant defect probably does not contribute to the development population would be 3.9%. With these assumptions re- of GERD because a higher prevalence of the type I or III garding the infant population, the prevalence of GERD reflux pattern would be expected because of their asso- would be at least 24 times higher and the prevalence of ciation with hiatal hernia.6 The type II reflux pattern is the risk factor for SID from GERD would be at least 25 not associated with a hiatal hernia or a decreased lower times higher in infants with intestinal malrotation alone. esophageal sphincter pressure.6 The prevalence of GERD associated with intestinal The high prevalence of GERD in infants with intes- malrotation is lower in infants with repaired congenital tinal malrotation suggests that these infants usually have diaphragmatic hernia. This lower prevalence of GERD a defective gastroesophageal junction that may result from suggests that factors other than intestinal malrotation are the same conditions that lead to the interference of nor- responsible for the observed prevalence. One possible fac- mal intestinal rotation and fixation during fetal devel- tor may be the high mortality associated with repaired opment. Such simultaneous defects in the function of the congenital diaphragmatic hernia. Patients with GERD and foregut and anatomy of the midgut do not necessarily re- the risk factor for SID from GERD may be less likely to sult in an anatomic defect of the foregut. This concept is survive congenital diaphragmatic hernia repair and further supported by the gastric emptying abnormali- thereby give the impression that survivors have a lower ties often seen in patients with intestinal malrotation.2 prevalence of GERD. Although the prevalence of the risk Mechanical gastric-outlet obstruction does not appear to factor for SID from GERD was not significantly lower with be a factor because a high prevalence of GERD persists repaired congenital diaphragmatic hernia, statistical sig- despite the relief of gastric-outlet obstruction with a suc- nificance might be achieved with a larger series of in- cessful Ladd operation. The GERD in these patients can fants. A higher prevalence for GERD was expected in in- also be seen with rapid and slow gastric emptying.2 The fants with repaired congenital diaphragmatic hernia factors responsible for the development of the risk for because of the deficient diaphragmatic attachments and SID from GERD are unknown, but problems with swal- function near the esophageal hiatus.12,13 lowing, esophageal emptying, and gastric acid secretion Infants with repaired abdominal wall defects have are likely major components. Unfortunately, the mea- a lower prevalence of the risk factor for SID from GERD. surement of these factors was not part of this study. This finding also points to factors other than intestinal The course of treated infants with GERD and intes- malrotation as a cause for the GERD in these infants. Pre- tinal malrotation depends on the presence or absence of maturity can be a factor in the ability to detect a pro- symptoms to suggest GERD. The course in symptom- longed ZMD.14 It is unlikely that this was a factor be- atic infants is similar to that in infants with GERD with- cause 39% (9/23) of infants with a repaired abdominal out intestinal malrotation. Their outcome depends on the wall defect had a prolonged ZMD. The disproportion- reflux pattern type (Table 3).15 In contrast, asymptom- ately high prevalence of the type II reflux pattern with a atic patients with GERD have a better prognosis than prolonged ZMD was responsible for the lower preva- symptomatic patients. Although these infants have a bet-

ARCH SURG/ VOL 134, JULY 1999 751

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 ter prognosis, a significant number of infants with the DISCUSSION type I and III reflux patterns continue to have GERD on follow-up. Because Barrett esophagus in children is seen Thomas R. Weber, MD, St Louis, Mo: Steve Jolley and his col- primarily in association with the type I reflux pattern,16 leagues in Las Vegas have been leaders in bringing science to the infants with persistent GERD on follow-up remain at risk confusing and frequently controversial topic of reflux in chil- for this disorder. The observed course of GERD in symp- dren. Through the extensive use of pH monitoring, they have tomatic and asymptomatic patients included some treat- been able to identify groups of high-risk patients who benefit from early and aggressive antireflux therapy, including surgical therapy, ment of the GERD. Without any treatment of the GERD involving a variety of scenarios. The present study extends these in our infant population with intestinal malrotation, we observations to the group of children with intestinal rotational would have expected at least 2 sudden deaths. disorders and severe reflux, and they have identified a sub- group who they feel is at great risk for SID from aspiration. The CONCLUSIONS data in this study, although convincing, generate several ques- tions. Why do so many infants and children with intestinal ro- Gastroesophageal reflux disease and the risk factor for tational anomalies have such severe reflux? Is it due in part to duodenal obstruction from congenital bands? Or is there abnor- SID from GERD are highly prevalent in infants with in- mal duodenal or gastric peristalsis? Do you have any gastric emp- testinal malrotation. This high prevalence of GERD in- tying studies to explain the high incidence of reflux in these pa- dicates that infants with intestinal malrotation require tients? Second, you elected to treat nonoperatively several infants careful evaluation and follow-up for problems from GERD. whom you have identified as being at high risk for SID. Many pediatric surgeons would consider near SID as an absolute in- Presented at the 106th Scientific Session of the Western Sur- dication for immediate fundoplication. Aren’t you somewhat gical Association, Indianapolis, Ind, November 18, 1998. nervous about treating these high-risk patients medically? This Reprints: Stephen G. Jolley, MD, 3201 S Maryland is particularly worrisome in the group of asymptomatic pa- tients in whom you have found continued reflux by pH stud- Pkwy, Suite 414, Las Vegas, NV 89109. ies. These are precisely the children who seem to be at greatest risk for near or complete SID episodes. REFERENCES You obviously used pH monitoring as the gold standard for the detection of reflux in children and have great experience in its application and interpretation. Do you use other diagnostic 1. Rescorla FJ, Shedd FJ, Grosfeld JL, Vane DW, West KW. Anomalies of intestinal studies in these patients? How do they correlate with your pH rotation in childhood: analysis of 447 cases. Surgery. 1990;108:710-716. studies? Several patients in your series were found to have intes- 2. Jolley SG, Tunell WP, Thomas S, Young J, Smith EI. The significance of gastric tinal rotation abnormalities at the time of fundoplication. More emptying in children with intestinal malrotation. J Pediatr Surg. 1985;20:627- liberal use of upper gastrointestinal (GI) contrast studies, for ex- 631. 3. Kumar D, Brereton RJ, Spitz L, Hall CM. Gastro-oesophageal reflux and intesti- ample, would probably have identified these anomalies preop- nal malrotation in children. Br J Surg. 1988;75:533-535. eratively, thus eliminating this surprise factor intraoperatively. 4. Chung C, Bautista N, Rowe V, Krishnamoorthy M, Applebaum H. Simultaneous Jay S. Grosfeld, MD, Indianapolis, Ind: I get concerned correction of malrotation and gastroesophageal reflux in infants. Am Surg. 1996; about correlating malrotation and gastroesophageal reflux with 62:800-802. the SID syndrome (SIDS). If you wanted to say these conditions 5. Jolley SG, Johnson DG, Herbst JJ, Pena A, Garnier R. An assessment of gas- are associated with aspiration or the threat of aspiration and death, troesophageal reflux in children by extended pH monitoring of the distal esopha- that is one thing, but the SIDS is different. Especially since the gus. Surgery. 1978;84:16-22. publication of the book by Firstman and Talen, The Death of In- 6. Jolley SG, Herbst JJ, Johnson DG, Book LS, Matlak ME, Condon VR. Patterns of nocence, which documents that, unfortunately, many of the in- postcibal gastroesophageal reflux in symptomatic infants. Am J Surg. 1979;138: 946-950. fants who were considered to have SIDs were subsequently shown 7. Jolley SG, Johnson DG, Herbst JJ, Matlak ME. The significance of gastroesopha- to be suffocated by their parents. The relationship of malrota- geal reflux patterns in children. J Pediatr Surg. 1981;16:859-865. tion to SIDS leaves me cold. I would rather you consider this as 8. Halpern LM, Jolley SG, Tunell WP, Johnson DG, Sterling CE. The mean duration a threat for aspiration than suggest that it is a cause of SIDS. Many of gastroesophageal reflux during sleep as an indicator of respiratory symp- pediatric specialists do not believe gastroesophageal reflux has toms from gastroesophageal reflux in children. J Pediatr Surg. 1991;26:686- anything to do with SIDS. Would you please comment on this? 690. Dr Jolley, does the motility problem in the foregut associated with 9. Jolley SG, Halpern LM, Tunell WP, Johnson DG, Sterling CE. The risk of sudden gastroesophageal reflux extend to the in patients who infant death from gastroesophageal reflux. J Pediatr Surg. 1991;26:691-696. have intestinal malrotation? Did some of the patients who were 10. Wegman ME. Annual summary of vital statistics—1991. Pediatrics. 1992;90: previously operated on for malrotation before the diagnosis of 835-845. 11. Jolley SG. Esophagus: gastroesophageal reflux disease. In: Ziegler MM, Aziz- gastroesophageal reflux have a midgut volvulus? Did they re- khan RG, Gauderer MWL, Weber TR, eds. Operative . Stam- quire bowel resection? Did they have delayed return of gastric ford, Conn: Appleton & Lange. In press. emptying after their last procedure? 12. Kieffer J, Sapin E, Berg A, Beaudoin S, Bargy F, Helardot PG. Gastroesophageal Dr Jolley: I do not understand why children with malro- reflux after repair of congenital diaphragmatic hernia. J Pediatr Surg. 1995;30: tation have a high prevalence of reflux. We do use other mo- 1330-1333. dalities to study these patients. All the patients who were symp- 13. Qi B, Soto C, Diez-Pardo JA, Tovar JA. An experimental study on the pathogen- tomatic had an upper GI series and an esophagram, including esis of gastroesophageal reflux after repair of diaphragmatic hernia. J Pediatr a small bowel follow-through. In answer to Dr Grosfeld’s ques- Surg. 1997;32:1310-1313. tion, an x-ray was done to be sure they did not have an ob- 14. Jolley SG, Halpern CT, Sterling CE, Feldman BH. The relationship of respiratory struction postoperatively. None had the resection of bowel, even complications from gastroesophageal reflux to prematurity in infants. J Pediatr Surg. 1990;25:755-757. though they have had a volvulus, because the bowel was vi- 15. Jolley SG. Gastroesophageal reflux disease as a cause for emesis in infants. Semin able and did not need to be resected. Pediatr Surg. 1995;4:176-189. Regarding gastric emptying studies in these patients, we 16. Jolley SG. Indications for surgery in childhood Barrett’s esophagus. Acta Endosc. have published this previously and have shown that these pa- 1993;23:119-123. tients can have slow gastric emptying, normal gastric empty-

ARCH SURG/ VOL 134, JULY 1999 752

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 ing, or rapid gastric emptying. There is no pattern of the gas- connotation, primarily because it is a catch-all category, and tric emptying to suggest that they may have persistent chronic there is no question that there are infants who have been as- obstruction. It is clear that they do not always have com- cribed as SIDS victims who have been the victims of trauma or pletely normal motility of their stomach. suffocation at home. We also see patients who are victims of Why do we treat patients who are at risk for SID medi- Munchausen by proxy. However, the patients who have reflux cally? This applies only to patients who are asymptomatic. The and who have died at home have been legitimate SIDS victims. main reason for treating those patients medically is that we do They are victims who have not been the subject of these other not know that we should treat them surgically because the lim- factors. The only reason that I can stand up here and say that ited data that we have on those patients was coupled with the is that I have been doing esophageal pH monitoring in chil- fact that their families discontinued the medical antireflux dren for 20 years. We have done close to 3000 esophageal pH therapy that they were on. At this point in time, we do not have studies. In those 20 years and in all those studies, there has not any evidence that asymptomatic patients at risk for SID will die been a single child who has had an autopsy by a pediatric pa- despite medical antireflux therapy. thologist (who is well trained in the SIDS autopsy) and who The upper GI findings on these patients are typical for non- was called a SIDS victim who has had any findings on their rotation postoperatively. I can tell you that of the 19 children in esophageal pH study other than type I or type III reflux and a whom we were surprised to find intestinal malrotation, all had prolonged ZMD. This now totals 5 SIDS victims and 4 infants an upper GI series performed before the antireflux operation, and with death from GERD (total of 9 infants). This is probably the these were read as normal. Intestinal malrotation was not detected most compelling evidence that reflux is a cause of SID in these on the upper GI series by either the radiologist or surgeon. patients, and it is certainly a large portion of what has been called With respect to labeling these patients as SIDS victims, I SIDS victims, with other portions being those alluded to by Dr agree with Dr Grosfeld in that the term “SIDS” has a negative Grosfeld.

IN OTHER AMA JOURNALS

ARCHIVES OF INTERNAL MEDICINE Do Symptoms Predict Cardiac Arrhythmias and Mortality in Patients With Syncope? Jeong H. Oh, MD; Barbara H. Hanusa, PhD; Wishwa N. Kapoor, MD, MPH Background: Patients with syncope frequently present with multitude of other symptoms but their significance in predicting morbidity or mortality has not been previously studied. Objective: To determine if certain symptoms can be used to identify syncope patients at risk for cardiac arrhythmias, mor- tality, or recurrence of syncope. Patients and Methods: From August 1987 to February 1991, we prospectively evaluated patients with syncope from out- patient, inpatient, and emergency department services of a university medical center. These patients were interviewed, charts were reviewed, and detailed information on 19 symptoms and comorbidities was obtained. A cause of syncope was assigned using standardized diagnostic criteria. All patients were followed up at 3-month intervals for at least 1 year for recurrence of syncope and mortality. Patients in whom the cause of syncope was determined by medical history and physical examination alone were not included in our analysis. Results: History and physical examination led to the cause of syncope in 222 of 497 patients enrolled. In the remaining 275 patients, the absence of nausea and before syncope (odds ratio, 7.1) and electrocardiographic abnormalities (odds ratio, 23.5) were predictors of arrhythmic syncope. Underlying cardiac disease was the only predictor of 1-year mortality. No symptom remained as independent predictor for 1-year mortality or syncope recurrence. Conclusions: Symptoms, although important in assigning many noncardiac causes, are not useful in risk-stratifying patients whose cause of syncope cannot be identified by other history and physical examination. Triage decisions and management plans should be based on preexisting cardiac disease or electrocardiographic abnormalities, which are important predictors of arrhythmic syncope and mortality. (1999;159:375-380) Reprints: Wishwa N. Kapoor, MD, MPH, Montefiore University Hospital, Suite W933, 200 Lothrop St, University of Pittsburgh Medical Center, Pittsburgh, PA 15213-2582 (e-mail: [email protected]).

ARCH SURG/ VOL 134, JULY 1999 753

©1999 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021