- Email: [email protected]
Does laparoscopy-aided gastrostomy placement improve or worsen gastroesophageal reflux in patients with neurological impairment?
Journal of Pediatric Surgery 49 (2014) 1742–1745
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Does laparoscopy-aided gastrostomy placement improve or worsen gastroesophageal reflux in patients with neurological impairment?☆ Hisayoshi Kawahara a,⁎, Yuko Tazuke b, Hideki Soh b, Akihiro Yoneda b, Masahiro Fukuzawa b a b
Department of Pediatric Surgery, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka, Japan Department of Pediatric Surgery, Osaka Medical Center and Research Institute for Maternal and Child Health, Izumi, Osaka, Japan
a r t i c l e
i n f o
Article history: Received 22 August 2014 Accepted 5 September 2014 Key words: Gastrostomy Gastroesophageal reflux pH-multichannel intraluminal impedance Gastric emptying
a b s t r a c t Background/purpose: The purpose of this study was to retrospectively investigate whether laparoscopy-aided gastrostomy placement (LGP) improved or worsened gastroesophageal reflux (GER) in neurological impairment (NI) patients. Methods: Subjects included 26 NI patients nourished via nasogastric tubes (age, 1–17 years; median, 6 years). They were divided into groups based on the percentage of time with an esophageal pH b 4.0 (reflux index: RI) before LGP: Group 1 (GI, n = 13), RI b 5.0%; Group II (GH, n = 13), RI ≥ 5.0%. Acid/nonacid reflux episodes (RE) were evaluated using combined pH-multichannel intraluminal impedance (pH-MII) monitoring, and gastric emptying was measured with the C breath test before and after LGP. Results: RI and number of RE evaluated with pH analyses and number of total/acid distal and proximal bolus RE with pH-MH increased significantly in GI. RI and acid clearance time with pH analyses and number of total bolus RE with pH-MII decreased significantly in GH. Gastric emptying parameters did not change significantly in GI, whereas the half-gastric emptying time and gastric emptying coefficient improved significantly in GH. Conclusion: LGP reduces GER in NI patients with pathological GER by improving gastric emptying, although it has a paradoxical influence on those without pathological GER. © 2014 Elsevier Inc. All rights reserved.
Gastrostomy feeding has been increasingly used as a method of nutritional support for patients with neurological impairment (NI) [1]. However, the development of gastroesophageal reflux (GER) after gastrostomy placement (GP) is still a matter of debate, and the influence of GP on esophagogastric motility has not yet been clarified [2–4]. Our previous data using 24-h esophageal pH monitoring have shown that esophageal acid exposure increased in patients without pathologic GER after GP, but decreased in those with pathologic GER [5]. Recently, combined esophageal pH-multichannel intraluminal impedance (MII) monitoring has been added to the repertoire of tests available for esophageal physiology [6–9]. Combined pH-MII monitoring has an advantage over traditional pH monitoring, which can detect both acid and nonacid GER. In addition, the 13C-acetate breath test has been newly introduced for the evaluation of gastric emptying in pediatric patients [10,11]. We retrospectively investigated whether laparoscopy-aided gastrostomy placement (LGP) improved or worsened GER in patients with NI.
☆ This study was supported in part by Health and Labour Sciences Research Grants for research on pediatric military disorders from the Ministry of Health, Labour, and Welfare of Japan to HK (26462706). ⁎ Corresponding author at: Department of Pediatric Surgery, Hamamatsu University School of Medicine, 1-20-1 Handayama Higashi-ku, Hamamatsu, Shizuoka, Japan 431–3192. Tel.: +81 53 435 2891; fax: +81 53 435 2312. E-mail address: [email protected] (H. Kawahara). http://dx.doi.org/10.1016/j.jpedsurg.2014.09.008 0022-3468/© 2014 Elsevier Inc. All rights reserved.
1. Methods 1.1. Patients Approval of the institutional review board was obtained from our hospital to conduct a retrospective review of the inhospital and outpatient medical charts of 26 patients with NI, who had undergone LGP and whose motor function of the upper gastrointestinal tract was evaluated with a 24-h combined esophageal pH-MII monitoring and the 13C-acetate breath test. Both examinations had been performed routinely before and 6–13 days (median, 8 days) after LGP. The patients had been nourished via nasogastric tubes and were aged 1–17 years (median, 6 years). All patients had profound NI, such as cerebral palsy (n = 11), chromosomal anomaly (n = 4), West syndrome (n = 3), posttraumatic brain damage (n = 1), advanced mental retardation (n = 1), postherpes encephalitis (n = 1), Menkes disease (n = 1), holoprosencephaly (n = 1), metachromatic leukodystrophy (n = 1), myoclonic encephalopathy (n = 1), and mitochondrial disease (n = 1). Any respiratory and gastrointestinal conditions were controlled adequately with nasogastric tube feeding in the 26 patients. Five patients had undergone tracheostomy, of which three had undergone laryngotracheal separation before LGP. None of the 26 patients had an irreducible hiatal hernia on a preoperative upper GI study. The indication for LGP and the operative procedure of LGP have been described previously in detail [2,6]. No other surgery was conducted concomitantly with LGP.
H. Kawahara et al. / Journal of Pediatric Surgery 49 (2014) 1742–1745
The patients were divided into two groups on the basis of the range of the percentage of total preoperative time for which the esophageal pH was less than 4.0 (reflux index, RI) which was determined using the 24-h combined esophageal pH-MII monitoring. The goal was to separately analyze the esophagogastric motility data in those with and without pathologic GER: Group I (n = 13) with an RI b 5.0% and a median age of 6 years (range, 2–17 years) and Group II (n = 13) with an RI ≥5.0% and a median age of 5 years (range, 1–14 years). The postoperative follow-up period ranged from 10 months to 5 years and 6 months (median, 46 months). 1.2. 24-h combined esophageal pH-MII monitoring Any antireflux medications, such as prokinetic agents, histamine-2 receptor antagonists, and proton-pump inhibitors, were discontinued at least 72 h before the 24-h combined esophageal pH-MII monitoring. We used a combined pH-MII flexible probe (outer diameter, 2.1 mm; Sandhill Scientific, Highlands Ranch, CO, USA) that was equipped with 7 or 8 impedance rings for the 6 esophageal impedance channels and 2 antimony pH sensors for the distal esophageal and gastric pH. The pH-MII probe for each patient was chosen from 3 probes types, according to the patient's height. The impedance channels were located 3, 5, 7, 9, 11, 13, and 15 cm from the distal tip of the probe (ZPN-S62C21E) in patients b75 cm in height, at 11, 13, 15, 17, 19, 21, and 23 cm (ZPNS62C07E) in patients 75–130 cm in height, and at 12, 14, 16, 18, 20, 24, 26, and 28 cm (ZAN-BG-44) for those N130 cm in height, respectively. The gastric pH sensor was located at the distal end of the probe, and the distal esophageal pH sensor was located at 6, 12, or 15 cm from the distal tip of the 3 probes, respectively. Before the pH measurements were made, the probe was calibrated using pH 4.0 and 7.0 buffer solution as specified by the manufacturer. The pH-MII probe was inserted nasally under fluoroscopic observation, and the distal esophageal pH sensor was placed at the level of a vertebral body length above the line between the bilateral cranial peak points of the diaphragmatic dome. The probe was connected to a data logger (Sleuth System; Sandhill Scientific) that stored data from the probe. The pH-MII data obtained in each study were downloaded onto a personal computer and analyzed with the Bioview Analysis software, version 5.5.8.0 (Sandhill Scientific). The impedance and pH data were automatically analyzed with the AutoSCAN software. This identified pH drops in the distal esophagus, which indicated traditional chemical reflux episodes (RE), and also identified impedance drops by liquid in the proximal channels, which indicated bolus RE irrespective of pH changes. One of the authors then manually analyzed the 3 min tracings. A bolus (impedance) RE was defined as a retrograde impedance drop N 50% (in Ohms), and N2 s from the pre-episode baseline mean in at least two consecutive channels of the four distal channels. The end of the bolus RE was defined as the time the impedance level returned to 50% of the pre-episode impedance baseline for N 5 s. The bolus RE were divided into two categories on the basis of a nadir pH value during RE and were either labeled either acid or nonacid. An acid RE was defined as RE during which the esophageal pH was lower than 4.0 or that occurred when the distal esophageal pH was already below 4.0. A nonacid RE was defined as RE in which the distal esophageal pH reduced by at least 1 unit and the nadir esophageal pH did not drop below 4.0. An RE that reached the channel two was defined as a proximal extent bolus RE. The clearance dynamics were quantified by both acid and bolus clearance. Acid clearance referred to the chemical neutralization of the remaining hydrogen ions as measured by the pH sensor at the clearance level. The acid clearance time was the elapsed time between the acid entry and acid clearance in a pH channel located in the distal esophagus. If the pH decreased to less than 4.0, the points in time when the pH was 4.0 and when it recovered to above 4.0 were marked, and the resulting acid clearance times were calculated. Bolus clearance referred to the
1743
physical clearance of the bolus as measured by the impedance at the clearance level. A bolus clearance time was defined as the elapsed time between the bolus entry and bolus clearance as measured by the impedance. The mean acid clearance time and median bolus clearance time were calculated in each patient using the Bioview Analysis software. 1.3.
13
C-acetate breath test
The method of 13C-acetate breath has been previously reported in detail [12]. This test measured changes in the 13CO2/ 12CO2 isotope ratio in the patient's exhaled breath, which resulted from the absorption and metabolism of a 13C-labeled test meal marker as it exited in the stomach. Specifically, after a baseline collection of the patient's breath, 100 mg of 13C-labeled sodium acetate dissolved in a liquid meal was administered. The liquid formulas given to patients via a nasogastric tube before LGP and via a gastrostomy tube after LGP were pediatric elemental diet formula (Elental P™, n = 1), low residual formula, including Racol™ (n = 10), Ensure™ (n = 9), and Twinline-NF™ (n = 3), and a hypoallergic formula (New MA-1™, n = 3). The volume of the liquid meal was calculated at 300 mL per 1.73 m 2 of body surface area. Patients were tested for four hours in the supine position after more than six hours of fasting. The gastric emptying was evaluated via a continuous breath test in real time using the BreathID™ system (Oridion BreathID Ltd., Jerusalem, Israel). Patients were connected to the device via a cannula, which was attached to the nose for 21 patients or via a tracheostomy tube for five patients. The test meal was processed and emptied by the stomach, and after absorption and metabolism, 13C was produced and exhaled in the breath as CO2. The device continuously measured the ratio of 13C and 12C in the exhaled CO2 using molecular correlation spectroscopy, which absorbed the radiation emitted by the radioisotope. The BreathID™ device measured the ratio of C isotopes, which was normalized for patient weight, height, and 13C substrate and dose, and provided a percentage-dose recovery and cumulative percentagedose recovery. The excretion of 13CO2 in the breath was mathematically analyzed using two nonlinear regression equations to find the best fitting curve through the measured data points with the affiliated computer software (Oridion Research Software™). The data from the breath test were used to calculate three common parameters for gastric emptying, using formulas based on the analysis described by Ghoos et al. [12]. The first parameter was the half emptying time (T1/2), which was the time required for half of the gastric contents to be emptied. The second parameter was the lag time (Tlag), which was the time that corresponded to the point of inflection of the curve after mathematical integration (i.e., the time of maximal 13CO2 excretion based on the fitted curve). The third parameter was the gastric emptying coefficient (GEC), which was an index of the global gastric emptying rate. In the current study, delayed gastric emptying was defined as a T1/2 value N100 min according to data shown by Lysy et al. [13], who reported using the same technique that the normal T1/2 value was 50–94 min in adult patients. All data are expressed as medians and ranges. The Wilcoxon signedrank test was used to compare data before and after LGP. Statistical significance was accepted at p b .05. 2. Results The data of each pH and impedance parameter before and after LGP are detailed in Table 1. No parameters in the overall 26 patients showed any significant changes. In the analyses of the pH alone, RI (p = 0.01) and the numbers of RE (p = 0.01) increased significantly in Group I. In the analyses of combined impedance and pH, the numbers of acid (p = 0.02)/total (p = 0.01) bolus RE and acid (p = 0.04)/total (p = 0.03) proximal extent bolus RE increased significantly, although an increase in the percent time of acid bolus RE did not reach significance (p = 0.12). Bolus clearance time decreased significantly (p = 0.03).
1744
H. Kawahara et al. / Journal of Pediatric Surgery 49 (2014) 1742–1745
Table 1 Changes in pH and impedance parameters. Group I (n = 13)
pH RI Number of RE Number of RE N5 min Acid clearance time (sec) Impedance + pH %time bolus RE Acid Nonacid Total Number of bolus RE Acid Nonacid Total Number of proximal extent bolus RE Acid Nonacid Total Bolus clearance time (sec)
Group II (n = 13)
Total (n = 26)
Before LGP
After LGP
p value
Before LGP
After LGP
p value
Before LGP
After LGP
p value
1.4(0.0–4.7) 15(0–46) 1(0 ~ 4) 56(0–313)
5.1(0.2–7.9) 37(6–84) 2(0–4) 81(28–203)
0.01 0.01 0.10 0.78
12.9(6.7–35.7) 61(29–129) 7(1–22) 194(48–507)
6.7(0.2–20.4) 68(9–98) 2(0–14) 85(12–218)
0.02 0.26 0.07 0.01
5.7(0.0–35.7) 36(0–129) 4(0–22) 132(0–507)
5.4(0.2–20.4) 41(6–98) 2(0–14) 83(12–218)
0.60 0.52 0.40 0.14
0.5(0–1.6) 0.5(0–3.2) 1.2(0.0–3.9)
0.8(0.1–2.5) 0.6(0.1–2.6) 1.3(0.3–3.6)
0.12 0.91 0.42
1.2(0.3–4.2) 0.5(0–2.7) 1.8(0.9–6.9)
0.8(0.1–1.7) 0.4(0–1.3) 1.7(0.4–2.1)
0.18 0.31 0.13
0.8(0.0–4.2) 0.5(0.0–3.2) 1.5(0.0–6.9)
0.8(0.1–2.5) 0.6(0.0–2.6) 1.5(0.3–3.6)
0.90 0.71 0.78
9(0–34) 20(1–46) 37(2–53)
19(3–55) 21(6–56) 59(12–94)
0.02 0.31 0.01
38(13–104) 25(2–50) 69(37–159)
29(5–70) 17(1–44) 50(20–79)
0.42 0.18 0.02
23(0–104) 22(1–55) 49(2–159)
25(3–70) 21(1–56) 58(12–94)
0.53 0.99 0.52
6(0–19) 5(0–30) 15(0–36) 20(13–45)
14(2–36) 17(0–49) 34(4–70) 19(9–28)
0.04 0.12 0.03 0.03
20(4–55) 10(1–31) 41(14–79) 16(11–24)
10(0–61) 6(0–31) 27(6–61) 15(7–42)
0.23 0.92 0.15 0.28
13(0–55) 10(0–31) 24(0–79) 17(11–45)
14(0–61) 13(0–49) 33(4–70) 17(7–42)
0.80 0.49 0.59 0.15
LGP: laparoscopy-aided gastrostomy placement, reflux index (RI): the percentage of time with an esophageal pH b4.0, RE: reflux episode.
In Group II, RI (p = 0.02) and acid clearance time (p = 0.01) decreased significantly; a decrease in the number of RE N5-min was close to significance (p = 0.07) in the analyses of pH alone. The number of total bolus RE decreased significantly (p = 0.02) in the analyses of the combined impedance and pH, although the percent time of total bolus RE (p = 0.13) and the number of the total proximal extent bolus RE (p = 0.15) did not show a significant decrease. Data of each gastric emptying parameter before and after LGP are detailed in Table 2. The preoperative values of the T1/2 (p = 0.37), Tlag (p = 0.70), and GEC (p = 0.15) did not show significant differences between Groups I and II. No parameters in the overall 26 patients and Group I showed any significant changes. In contrast, T1/2 decreased (p = 0.04) and GEC increased (p = 0.002) significantly in Group II, indicating that gastric emptying improved after LGP. Of the 26 patients, seven in Group I and eight in Group II had a T1/2 of over 100 min, signifying delayed gastric emptying. All gastric emptying parameters, including T1/2 (128 min [102 min–204 min] vs. [112 min [81 min–128 min], p = 0.004], Tlag 63 min [43 min–97 min] vs. 58 min [40 min–76 min], p = 0.008), and GEC (3.4 [2.8–4.3] vs. 3.5 [3.1–3.8], p = 0.003), improved after LGP. In contrast, none of the three parameters, including T1/2 (p = 0.16), Tlag (p = 0.18), and GEC (p = 0.93), showed significant changes in 11 patients with a T1/2 of below 100 min. No patients showed a deterioration of GER requiring fundoplication after LGP in the follow-up period.
3. Discussion The combined esophageal pH-MII monitoring has made it possible to detect all RE and to determine whether they were acid or nonacid, which was not possible with a traditional pH monitoring. In the current
study, it was demonstrated in NI patients with pathologic GER using a combined pH-MII monitoring system that RI, acid clearance time, and the number of total bolus RE decreased significantly after LGP. With a pH monitoring system, we previously demonstrated that RI decreased significantly in NI patients with pathologic GER, while it increased significantly in those without pathologic GER after LGP [5]. Wilson et al. also reported that 19 of 25 symptomatic GER patients (76%) with abnormal pH data improved after endoscopic GP and only 2 (8%) required subsequent antireflux surgery [4]. Scintigraphy has been considered the gold standard for measuring gastric emptying, but this technique is associated with radiation exposure to patients. Recently, breath tests using a substrate labeled with the stable isotope 13C, a naturally occurring nonradioactive isotope, have been used an alternative technique to measure gastric emptying in children. The influence of GP on gastric emptying had not been investigated in pediatric NI patients so far. The T1/2 and GEC improved in NI patients with pathologic GER after LGP, yet they did not change in those without pathologic GER. Considering pH-MII and gastric emptying results in NI patients with pathologic GER, the improvement of delayed gastric emptying caused by the LGP procedure could be a mechanism responsible for the reduction in GER. The finding of a significant improvement of three gastric emptying parameters in 15 patients with delayed gastric emptying may also support the idea that LGP had a favorable influence on delayed gastric emptying. As pediatric NI patients often have the stomach located cranially above the costal arch, it was necessary to pull down the stomach caudally below the arch when we conducted LGP. This action might have corrected the chronic organoaxial gastric volvulus, which was often present in pediatric NI patients, and possibly improved delayed gastric emptying. Vincente et al. reported that esophageal acid clearance was more volume-dependent than motility-dependent in healthy piglets [14]. A
Table 2 Changes in gastric emptying parameters. Group I (n = 13)
T1/2 (min) T lag (min) GEC
Group II (n = 13)
Total (n = 26)
Before LGP
After LGP
p value
Before LGP
After LGP
p value
Before LGP
After LGP
p value
105(76–193) 56(32–97) 3.7(2.9–4.3)
105(70–131) 54(37–76) 3.5(2.9–4.3)
0.70 0.42 0.65
112(75–204) 53(18–96) 3.3(2.8–3.9)
108(78–127) 56(29–71) 3.6(3.1–4.1)
0.04 0.31 0.002
106(75–204) 56(18–97) 3.5(2.8–4.3)
107(70–131) 54(29–76) 3.5(2.9–4.3)
0.61 0.48 0.37
LGP: laparoscopy-aided gastrostomy placement, GEC: gastric emptying coefficient.
H. Kawahara et al. / Journal of Pediatric Surgery 49 (2014) 1742–1745
significant decrease in acid clearance time demonstrated in NI patients with pathologic GER may be related to the volume reduction of the refluxates, which possibly occurred through the improvement of gastric emptying. These findings are consistent with the finding that rikkunshito, a traditional Japanese medicine, improves gastric emptying and esophageal acid clearance [15]. Whether GP increases reflux and worsens reflux symptoms in NI patients without pathologic GER also has been a subject of controversy. The finding in the this study suggesting the increase of reflux in NI patients without pathologic GER after LGP was consistent with data reported with combined esophageal pH-MII monitoring by Thomson et al., which were that the number of acid bolus RE and the percent time of total bolus RE increased significantly after percutaneous endoscopic GP in pediatric NI patients without pathologic GER [16]. Possibly, the fixation of the stomach to the anterior abdominal wall exerted a silent adverse effect on GER, although gastric emptying parameters did not change significantly in this study. Further study using different motility tests, which could evaluate gastric motor function in detail, would clarify the mechanisms responsible for the increase of reflux in patients without pathologic GER prior to LGP. Such changes rarely precipitate deterioration of reflux symptoms, but the possibility that GP increases GER in NI patients without pathologic GER should be recognized by pediatric surgeons. The weakness of this study was the heterogeneity of profound NI in the subjects. Motility changes were not possible to be analyzed separately in subgroups of NI owing to limitations of patient numbers. As a different NI potentially had a different influence on esophagogastric motility especially after general anesthesia and surgical interventions, the changes seen, or lack thereof, in this study might have been skewed toward or away from a subgroup of NI. Further separate analyses with more patients in each NI would clarify whether LGP is equally impactful regardless of the type of NI. The authors declare that they have no conflicts of interest. The authors would like to thank Enago (www.enago.jp) for the English language review.
1745
References [1] Kawahara H, Kubota A, Okuyama H, et al. One-trocar laparoscopy-aided gastrostomy in handicapped children. J Pediatr Surg 2006;41:2076–80. [2] Jolley SG, Smith EI, Tunell WP. Protective antireflux operation with feeding gastrostomy. Experience with children. Ann Surg 1985;201:736–40. [3] Plantin I, Arnbjörnsson E, Larsson LT. No increase in gastroesophageal reflux after laparoscopic gastrostomy in children. Pediatr Surg Int 2006;22:581–4. [4] Wilson GJ, van der Zee DC, Bax NM. Endoscopic gastrostomy placement in the child with gastroesophageal reflux: is concomitant antireflux surgery indicated? J Pediatr Surg 2006;41:1441–5. [5] Kawahara H, Mitani Y, Nose K, et al. Should fundoplication be added at the time of gastrostomy placement in patients who are neurologically impaired? J Pediatr Surg 2010;45:2373–6. [6] van Wijk MP, Benninga MA, Omari TI. Role of the multichannel intraluminal impedance technique in infants and children. J Pediatr Gastroenterol Nutr 2009;48: 2–12. [7] Salvatore S, Arrigo S, Luini C, et al. Esophageal impedance in children: symptombased results. J Pediatr 2010;157:949–54. [8] Francavilla R, Magistà AM, Bucci N, et al. Comparison of esophageal pH and multichannel intraluminal impedance testing in pediatric patients with suspected gastroesophageal reflux. J Pediatr Gastroenterol Nutr 2010;50:154–60. [9] Sifrim D, Castell D, Dent J, et al. Gastro-oesophageal reflux monitoring: review and consensus report on detection and definitions of acid, non-acid, and gas reflux. Gut 2004;53:1024–31. [10] Gatti C, di Abriola FF, Dall’Oglio L, et al. Is the 13C-acetate breath test a valid procedure to analyse gastric emptying in children? J Pediatr Surg 2000;35:62–5. [11] Kawahara H, Mitani Y, Nomura M, et al. Impact of rikkunshito, an herbal medicine, on delayed gastric emptying in profoundly handicapped patients. Pediatr Surg Int 2009;25:987–90. [12] Ghoos YF, Maes BD, Geypens BJ, et al. Measurement of gastric emptying of solids by means of a carbon-labeled octanoic acid breath test. Gastroenterology 1993;104: 1640–7. [13] Lysy J, Israeli E, Strauss-Liviatan N, et al. Relationships between hypoglycaemia and gastric emptying abnormalities in insulin-treated diabetic patients. Neurogastroenterol Motil 2006;18:433–40. [14] Vicente Y, da Rocha C, Hernandez-Peredo G, et al. Esophageal acid clearance: more volume-dependent than motility-dependent in healthy piglets. J Pediatr Gastroenterol Nutr 2002;35:173–9. [15] Kawahara H, Kubota A, Hasegawa T, et al. Effects of rikkunshito on the clinical symptoms and esophageal acid exposure in children with symptomatic gastroesophageal reflux. Pediatr Surg Int 2007;23:1001–5. [16] Thomson M, Rao P, Rawat D, et al. Percutaneous endoscopic gastrostomy and gastrooesophageal reflux in neurologically impaired children. World J Gastroenterol 2011; 17:191–6.
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Does laparoscopy-aided gastrostomy placement improve or worsen gastroesophageal reflux in patients with neurological impairment?☆ Hisayoshi Kawahara a,⁎, Yuko Tazuke b, Hideki Soh b, Akihiro Yoneda b, Masahiro Fukuzawa b a b
Department of Pediatric Surgery, Hamamatsu University School of Medicine, Hamamatsu, Shizuoka, Japan Department of Pediatric Surgery, Osaka Medical Center and Research Institute for Maternal and Child Health, Izumi, Osaka, Japan
a r t i c l e
i n f o
Article history: Received 22 August 2014 Accepted 5 September 2014 Key words: Gastrostomy Gastroesophageal reflux pH-multichannel intraluminal impedance Gastric emptying
a b s t r a c t Background/purpose: The purpose of this study was to retrospectively investigate whether laparoscopy-aided gastrostomy placement (LGP) improved or worsened gastroesophageal reflux (GER) in neurological impairment (NI) patients. Methods: Subjects included 26 NI patients nourished via nasogastric tubes (age, 1–17 years; median, 6 years). They were divided into groups based on the percentage of time with an esophageal pH b 4.0 (reflux index: RI) before LGP: Group 1 (GI, n = 13), RI b 5.0%; Group II (GH, n = 13), RI ≥ 5.0%. Acid/nonacid reflux episodes (RE) were evaluated using combined pH-multichannel intraluminal impedance (pH-MII) monitoring, and gastric emptying was measured with the C breath test before and after LGP. Results: RI and number of RE evaluated with pH analyses and number of total/acid distal and proximal bolus RE with pH-MH increased significantly in GI. RI and acid clearance time with pH analyses and number of total bolus RE with pH-MII decreased significantly in GH. Gastric emptying parameters did not change significantly in GI, whereas the half-gastric emptying time and gastric emptying coefficient improved significantly in GH. Conclusion: LGP reduces GER in NI patients with pathological GER by improving gastric emptying, although it has a paradoxical influence on those without pathological GER. © 2014 Elsevier Inc. All rights reserved.
Gastrostomy feeding has been increasingly used as a method of nutritional support for patients with neurological impairment (NI) [1]. However, the development of gastroesophageal reflux (GER) after gastrostomy placement (GP) is still a matter of debate, and the influence of GP on esophagogastric motility has not yet been clarified [2–4]. Our previous data using 24-h esophageal pH monitoring have shown that esophageal acid exposure increased in patients without pathologic GER after GP, but decreased in those with pathologic GER [5]. Recently, combined esophageal pH-multichannel intraluminal impedance (MII) monitoring has been added to the repertoire of tests available for esophageal physiology [6–9]. Combined pH-MII monitoring has an advantage over traditional pH monitoring, which can detect both acid and nonacid GER. In addition, the 13C-acetate breath test has been newly introduced for the evaluation of gastric emptying in pediatric patients [10,11]. We retrospectively investigated whether laparoscopy-aided gastrostomy placement (LGP) improved or worsened GER in patients with NI.
☆ This study was supported in part by Health and Labour Sciences Research Grants for research on pediatric military disorders from the Ministry of Health, Labour, and Welfare of Japan to HK (26462706). ⁎ Corresponding author at: Department of Pediatric Surgery, Hamamatsu University School of Medicine, 1-20-1 Handayama Higashi-ku, Hamamatsu, Shizuoka, Japan 431–3192. Tel.: +81 53 435 2891; fax: +81 53 435 2312. E-mail address: [email protected] (H. Kawahara). http://dx.doi.org/10.1016/j.jpedsurg.2014.09.008 0022-3468/© 2014 Elsevier Inc. All rights reserved.
1. Methods 1.1. Patients Approval of the institutional review board was obtained from our hospital to conduct a retrospective review of the inhospital and outpatient medical charts of 26 patients with NI, who had undergone LGP and whose motor function of the upper gastrointestinal tract was evaluated with a 24-h combined esophageal pH-MII monitoring and the 13C-acetate breath test. Both examinations had been performed routinely before and 6–13 days (median, 8 days) after LGP. The patients had been nourished via nasogastric tubes and were aged 1–17 years (median, 6 years). All patients had profound NI, such as cerebral palsy (n = 11), chromosomal anomaly (n = 4), West syndrome (n = 3), posttraumatic brain damage (n = 1), advanced mental retardation (n = 1), postherpes encephalitis (n = 1), Menkes disease (n = 1), holoprosencephaly (n = 1), metachromatic leukodystrophy (n = 1), myoclonic encephalopathy (n = 1), and mitochondrial disease (n = 1). Any respiratory and gastrointestinal conditions were controlled adequately with nasogastric tube feeding in the 26 patients. Five patients had undergone tracheostomy, of which three had undergone laryngotracheal separation before LGP. None of the 26 patients had an irreducible hiatal hernia on a preoperative upper GI study. The indication for LGP and the operative procedure of LGP have been described previously in detail [2,6]. No other surgery was conducted concomitantly with LGP.
H. Kawahara et al. / Journal of Pediatric Surgery 49 (2014) 1742–1745
The patients were divided into two groups on the basis of the range of the percentage of total preoperative time for which the esophageal pH was less than 4.0 (reflux index, RI) which was determined using the 24-h combined esophageal pH-MII monitoring. The goal was to separately analyze the esophagogastric motility data in those with and without pathologic GER: Group I (n = 13) with an RI b 5.0% and a median age of 6 years (range, 2–17 years) and Group II (n = 13) with an RI ≥5.0% and a median age of 5 years (range, 1–14 years). The postoperative follow-up period ranged from 10 months to 5 years and 6 months (median, 46 months). 1.2. 24-h combined esophageal pH-MII monitoring Any antireflux medications, such as prokinetic agents, histamine-2 receptor antagonists, and proton-pump inhibitors, were discontinued at least 72 h before the 24-h combined esophageal pH-MII monitoring. We used a combined pH-MII flexible probe (outer diameter, 2.1 mm; Sandhill Scientific, Highlands Ranch, CO, USA) that was equipped with 7 or 8 impedance rings for the 6 esophageal impedance channels and 2 antimony pH sensors for the distal esophageal and gastric pH. The pH-MII probe for each patient was chosen from 3 probes types, according to the patient's height. The impedance channels were located 3, 5, 7, 9, 11, 13, and 15 cm from the distal tip of the probe (ZPN-S62C21E) in patients b75 cm in height, at 11, 13, 15, 17, 19, 21, and 23 cm (ZPNS62C07E) in patients 75–130 cm in height, and at 12, 14, 16, 18, 20, 24, 26, and 28 cm (ZAN-BG-44) for those N130 cm in height, respectively. The gastric pH sensor was located at the distal end of the probe, and the distal esophageal pH sensor was located at 6, 12, or 15 cm from the distal tip of the 3 probes, respectively. Before the pH measurements were made, the probe was calibrated using pH 4.0 and 7.0 buffer solution as specified by the manufacturer. The pH-MII probe was inserted nasally under fluoroscopic observation, and the distal esophageal pH sensor was placed at the level of a vertebral body length above the line between the bilateral cranial peak points of the diaphragmatic dome. The probe was connected to a data logger (Sleuth System; Sandhill Scientific) that stored data from the probe. The pH-MII data obtained in each study were downloaded onto a personal computer and analyzed with the Bioview Analysis software, version 5.5.8.0 (Sandhill Scientific). The impedance and pH data were automatically analyzed with the AutoSCAN software. This identified pH drops in the distal esophagus, which indicated traditional chemical reflux episodes (RE), and also identified impedance drops by liquid in the proximal channels, which indicated bolus RE irrespective of pH changes. One of the authors then manually analyzed the 3 min tracings. A bolus (impedance) RE was defined as a retrograde impedance drop N 50% (in Ohms), and N2 s from the pre-episode baseline mean in at least two consecutive channels of the four distal channels. The end of the bolus RE was defined as the time the impedance level returned to 50% of the pre-episode impedance baseline for N 5 s. The bolus RE were divided into two categories on the basis of a nadir pH value during RE and were either labeled either acid or nonacid. An acid RE was defined as RE during which the esophageal pH was lower than 4.0 or that occurred when the distal esophageal pH was already below 4.0. A nonacid RE was defined as RE in which the distal esophageal pH reduced by at least 1 unit and the nadir esophageal pH did not drop below 4.0. An RE that reached the channel two was defined as a proximal extent bolus RE. The clearance dynamics were quantified by both acid and bolus clearance. Acid clearance referred to the chemical neutralization of the remaining hydrogen ions as measured by the pH sensor at the clearance level. The acid clearance time was the elapsed time between the acid entry and acid clearance in a pH channel located in the distal esophagus. If the pH decreased to less than 4.0, the points in time when the pH was 4.0 and when it recovered to above 4.0 were marked, and the resulting acid clearance times were calculated. Bolus clearance referred to the
1743
physical clearance of the bolus as measured by the impedance at the clearance level. A bolus clearance time was defined as the elapsed time between the bolus entry and bolus clearance as measured by the impedance. The mean acid clearance time and median bolus clearance time were calculated in each patient using the Bioview Analysis software. 1.3.
13
C-acetate breath test
The method of 13C-acetate breath has been previously reported in detail [12]. This test measured changes in the 13CO2/ 12CO2 isotope ratio in the patient's exhaled breath, which resulted from the absorption and metabolism of a 13C-labeled test meal marker as it exited in the stomach. Specifically, after a baseline collection of the patient's breath, 100 mg of 13C-labeled sodium acetate dissolved in a liquid meal was administered. The liquid formulas given to patients via a nasogastric tube before LGP and via a gastrostomy tube after LGP were pediatric elemental diet formula (Elental P™, n = 1), low residual formula, including Racol™ (n = 10), Ensure™ (n = 9), and Twinline-NF™ (n = 3), and a hypoallergic formula (New MA-1™, n = 3). The volume of the liquid meal was calculated at 300 mL per 1.73 m 2 of body surface area. Patients were tested for four hours in the supine position after more than six hours of fasting. The gastric emptying was evaluated via a continuous breath test in real time using the BreathID™ system (Oridion BreathID Ltd., Jerusalem, Israel). Patients were connected to the device via a cannula, which was attached to the nose for 21 patients or via a tracheostomy tube for five patients. The test meal was processed and emptied by the stomach, and after absorption and metabolism, 13C was produced and exhaled in the breath as CO2. The device continuously measured the ratio of 13C and 12C in the exhaled CO2 using molecular correlation spectroscopy, which absorbed the radiation emitted by the radioisotope. The BreathID™ device measured the ratio of C isotopes, which was normalized for patient weight, height, and 13C substrate and dose, and provided a percentage-dose recovery and cumulative percentagedose recovery. The excretion of 13CO2 in the breath was mathematically analyzed using two nonlinear regression equations to find the best fitting curve through the measured data points with the affiliated computer software (Oridion Research Software™). The data from the breath test were used to calculate three common parameters for gastric emptying, using formulas based on the analysis described by Ghoos et al. [12]. The first parameter was the half emptying time (T1/2), which was the time required for half of the gastric contents to be emptied. The second parameter was the lag time (Tlag), which was the time that corresponded to the point of inflection of the curve after mathematical integration (i.e., the time of maximal 13CO2 excretion based on the fitted curve). The third parameter was the gastric emptying coefficient (GEC), which was an index of the global gastric emptying rate. In the current study, delayed gastric emptying was defined as a T1/2 value N100 min according to data shown by Lysy et al. [13], who reported using the same technique that the normal T1/2 value was 50–94 min in adult patients. All data are expressed as medians and ranges. The Wilcoxon signedrank test was used to compare data before and after LGP. Statistical significance was accepted at p b .05. 2. Results The data of each pH and impedance parameter before and after LGP are detailed in Table 1. No parameters in the overall 26 patients showed any significant changes. In the analyses of the pH alone, RI (p = 0.01) and the numbers of RE (p = 0.01) increased significantly in Group I. In the analyses of combined impedance and pH, the numbers of acid (p = 0.02)/total (p = 0.01) bolus RE and acid (p = 0.04)/total (p = 0.03) proximal extent bolus RE increased significantly, although an increase in the percent time of acid bolus RE did not reach significance (p = 0.12). Bolus clearance time decreased significantly (p = 0.03).
1744
H. Kawahara et al. / Journal of Pediatric Surgery 49 (2014) 1742–1745
Table 1 Changes in pH and impedance parameters. Group I (n = 13)
pH RI Number of RE Number of RE N5 min Acid clearance time (sec) Impedance + pH %time bolus RE Acid Nonacid Total Number of bolus RE Acid Nonacid Total Number of proximal extent bolus RE Acid Nonacid Total Bolus clearance time (sec)
Group II (n = 13)
Total (n = 26)
Before LGP
After LGP
p value
Before LGP
After LGP
p value
Before LGP
After LGP
p value
1.4(0.0–4.7) 15(0–46) 1(0 ~ 4) 56(0–313)
5.1(0.2–7.9) 37(6–84) 2(0–4) 81(28–203)
0.01 0.01 0.10 0.78
12.9(6.7–35.7) 61(29–129) 7(1–22) 194(48–507)
6.7(0.2–20.4) 68(9–98) 2(0–14) 85(12–218)
0.02 0.26 0.07 0.01
5.7(0.0–35.7) 36(0–129) 4(0–22) 132(0–507)
5.4(0.2–20.4) 41(6–98) 2(0–14) 83(12–218)
0.60 0.52 0.40 0.14
0.5(0–1.6) 0.5(0–3.2) 1.2(0.0–3.9)
0.8(0.1–2.5) 0.6(0.1–2.6) 1.3(0.3–3.6)
0.12 0.91 0.42
1.2(0.3–4.2) 0.5(0–2.7) 1.8(0.9–6.9)
0.8(0.1–1.7) 0.4(0–1.3) 1.7(0.4–2.1)
0.18 0.31 0.13
0.8(0.0–4.2) 0.5(0.0–3.2) 1.5(0.0–6.9)
0.8(0.1–2.5) 0.6(0.0–2.6) 1.5(0.3–3.6)
0.90 0.71 0.78
9(0–34) 20(1–46) 37(2–53)
19(3–55) 21(6–56) 59(12–94)
0.02 0.31 0.01
38(13–104) 25(2–50) 69(37–159)
29(5–70) 17(1–44) 50(20–79)
0.42 0.18 0.02
23(0–104) 22(1–55) 49(2–159)
25(3–70) 21(1–56) 58(12–94)
0.53 0.99 0.52
6(0–19) 5(0–30) 15(0–36) 20(13–45)
14(2–36) 17(0–49) 34(4–70) 19(9–28)
0.04 0.12 0.03 0.03
20(4–55) 10(1–31) 41(14–79) 16(11–24)
10(0–61) 6(0–31) 27(6–61) 15(7–42)
0.23 0.92 0.15 0.28
13(0–55) 10(0–31) 24(0–79) 17(11–45)
14(0–61) 13(0–49) 33(4–70) 17(7–42)
0.80 0.49 0.59 0.15
LGP: laparoscopy-aided gastrostomy placement, reflux index (RI): the percentage of time with an esophageal pH b4.0, RE: reflux episode.
In Group II, RI (p = 0.02) and acid clearance time (p = 0.01) decreased significantly; a decrease in the number of RE N5-min was close to significance (p = 0.07) in the analyses of pH alone. The number of total bolus RE decreased significantly (p = 0.02) in the analyses of the combined impedance and pH, although the percent time of total bolus RE (p = 0.13) and the number of the total proximal extent bolus RE (p = 0.15) did not show a significant decrease. Data of each gastric emptying parameter before and after LGP are detailed in Table 2. The preoperative values of the T1/2 (p = 0.37), Tlag (p = 0.70), and GEC (p = 0.15) did not show significant differences between Groups I and II. No parameters in the overall 26 patients and Group I showed any significant changes. In contrast, T1/2 decreased (p = 0.04) and GEC increased (p = 0.002) significantly in Group II, indicating that gastric emptying improved after LGP. Of the 26 patients, seven in Group I and eight in Group II had a T1/2 of over 100 min, signifying delayed gastric emptying. All gastric emptying parameters, including T1/2 (128 min [102 min–204 min] vs. [112 min [81 min–128 min], p = 0.004], Tlag 63 min [43 min–97 min] vs. 58 min [40 min–76 min], p = 0.008), and GEC (3.4 [2.8–4.3] vs. 3.5 [3.1–3.8], p = 0.003), improved after LGP. In contrast, none of the three parameters, including T1/2 (p = 0.16), Tlag (p = 0.18), and GEC (p = 0.93), showed significant changes in 11 patients with a T1/2 of below 100 min. No patients showed a deterioration of GER requiring fundoplication after LGP in the follow-up period.
3. Discussion The combined esophageal pH-MII monitoring has made it possible to detect all RE and to determine whether they were acid or nonacid, which was not possible with a traditional pH monitoring. In the current
study, it was demonstrated in NI patients with pathologic GER using a combined pH-MII monitoring system that RI, acid clearance time, and the number of total bolus RE decreased significantly after LGP. With a pH monitoring system, we previously demonstrated that RI decreased significantly in NI patients with pathologic GER, while it increased significantly in those without pathologic GER after LGP [5]. Wilson et al. also reported that 19 of 25 symptomatic GER patients (76%) with abnormal pH data improved after endoscopic GP and only 2 (8%) required subsequent antireflux surgery [4]. Scintigraphy has been considered the gold standard for measuring gastric emptying, but this technique is associated with radiation exposure to patients. Recently, breath tests using a substrate labeled with the stable isotope 13C, a naturally occurring nonradioactive isotope, have been used an alternative technique to measure gastric emptying in children. The influence of GP on gastric emptying had not been investigated in pediatric NI patients so far. The T1/2 and GEC improved in NI patients with pathologic GER after LGP, yet they did not change in those without pathologic GER. Considering pH-MII and gastric emptying results in NI patients with pathologic GER, the improvement of delayed gastric emptying caused by the LGP procedure could be a mechanism responsible for the reduction in GER. The finding of a significant improvement of three gastric emptying parameters in 15 patients with delayed gastric emptying may also support the idea that LGP had a favorable influence on delayed gastric emptying. As pediatric NI patients often have the stomach located cranially above the costal arch, it was necessary to pull down the stomach caudally below the arch when we conducted LGP. This action might have corrected the chronic organoaxial gastric volvulus, which was often present in pediatric NI patients, and possibly improved delayed gastric emptying. Vincente et al. reported that esophageal acid clearance was more volume-dependent than motility-dependent in healthy piglets [14]. A
Table 2 Changes in gastric emptying parameters. Group I (n = 13)
T1/2 (min) T lag (min) GEC
Group II (n = 13)
Total (n = 26)
Before LGP
After LGP
p value
Before LGP
After LGP
p value
Before LGP
After LGP
p value
105(76–193) 56(32–97) 3.7(2.9–4.3)
105(70–131) 54(37–76) 3.5(2.9–4.3)
0.70 0.42 0.65
112(75–204) 53(18–96) 3.3(2.8–3.9)
108(78–127) 56(29–71) 3.6(3.1–4.1)
0.04 0.31 0.002
106(75–204) 56(18–97) 3.5(2.8–4.3)
107(70–131) 54(29–76) 3.5(2.9–4.3)
0.61 0.48 0.37
LGP: laparoscopy-aided gastrostomy placement, GEC: gastric emptying coefficient.
H. Kawahara et al. / Journal of Pediatric Surgery 49 (2014) 1742–1745
significant decrease in acid clearance time demonstrated in NI patients with pathologic GER may be related to the volume reduction of the refluxates, which possibly occurred through the improvement of gastric emptying. These findings are consistent with the finding that rikkunshito, a traditional Japanese medicine, improves gastric emptying and esophageal acid clearance [15]. Whether GP increases reflux and worsens reflux symptoms in NI patients without pathologic GER also has been a subject of controversy. The finding in the this study suggesting the increase of reflux in NI patients without pathologic GER after LGP was consistent with data reported with combined esophageal pH-MII monitoring by Thomson et al., which were that the number of acid bolus RE and the percent time of total bolus RE increased significantly after percutaneous endoscopic GP in pediatric NI patients without pathologic GER [16]. Possibly, the fixation of the stomach to the anterior abdominal wall exerted a silent adverse effect on GER, although gastric emptying parameters did not change significantly in this study. Further study using different motility tests, which could evaluate gastric motor function in detail, would clarify the mechanisms responsible for the increase of reflux in patients without pathologic GER prior to LGP. Such changes rarely precipitate deterioration of reflux symptoms, but the possibility that GP increases GER in NI patients without pathologic GER should be recognized by pediatric surgeons. The weakness of this study was the heterogeneity of profound NI in the subjects. Motility changes were not possible to be analyzed separately in subgroups of NI owing to limitations of patient numbers. As a different NI potentially had a different influence on esophagogastric motility especially after general anesthesia and surgical interventions, the changes seen, or lack thereof, in this study might have been skewed toward or away from a subgroup of NI. Further separate analyses with more patients in each NI would clarify whether LGP is equally impactful regardless of the type of NI. The authors declare that they have no conflicts of interest. The authors would like to thank Enago (www.enago.jp) for the English language review.
1745
References [1] Kawahara H, Kubota A, Okuyama H, et al. One-trocar laparoscopy-aided gastrostomy in handicapped children. J Pediatr Surg 2006;41:2076–80. [2] Jolley SG, Smith EI, Tunell WP. Protective antireflux operation with feeding gastrostomy. Experience with children. Ann Surg 1985;201:736–40. [3] Plantin I, Arnbjörnsson E, Larsson LT. No increase in gastroesophageal reflux after laparoscopic gastrostomy in children. Pediatr Surg Int 2006;22:581–4. [4] Wilson GJ, van der Zee DC, Bax NM. Endoscopic gastrostomy placement in the child with gastroesophageal reflux: is concomitant antireflux surgery indicated? J Pediatr Surg 2006;41:1441–5. [5] Kawahara H, Mitani Y, Nose K, et al. Should fundoplication be added at the time of gastrostomy placement in patients who are neurologically impaired? J Pediatr Surg 2010;45:2373–6. [6] van Wijk MP, Benninga MA, Omari TI. Role of the multichannel intraluminal impedance technique in infants and children. J Pediatr Gastroenterol Nutr 2009;48: 2–12. [7] Salvatore S, Arrigo S, Luini C, et al. Esophageal impedance in children: symptombased results. J Pediatr 2010;157:949–54. [8] Francavilla R, Magistà AM, Bucci N, et al. Comparison of esophageal pH and multichannel intraluminal impedance testing in pediatric patients with suspected gastroesophageal reflux. J Pediatr Gastroenterol Nutr 2010;50:154–60. [9] Sifrim D, Castell D, Dent J, et al. Gastro-oesophageal reflux monitoring: review and consensus report on detection and definitions of acid, non-acid, and gas reflux. Gut 2004;53:1024–31. [10] Gatti C, di Abriola FF, Dall’Oglio L, et al. Is the 13C-acetate breath test a valid procedure to analyse gastric emptying in children? J Pediatr Surg 2000;35:62–5. [11] Kawahara H, Mitani Y, Nomura M, et al. Impact of rikkunshito, an herbal medicine, on delayed gastric emptying in profoundly handicapped patients. Pediatr Surg Int 2009;25:987–90. [12] Ghoos YF, Maes BD, Geypens BJ, et al. Measurement of gastric emptying of solids by means of a carbon-labeled octanoic acid breath test. Gastroenterology 1993;104: 1640–7. [13] Lysy J, Israeli E, Strauss-Liviatan N, et al. Relationships between hypoglycaemia and gastric emptying abnormalities in insulin-treated diabetic patients. Neurogastroenterol Motil 2006;18:433–40. [14] Vicente Y, da Rocha C, Hernandez-Peredo G, et al. Esophageal acid clearance: more volume-dependent than motility-dependent in healthy piglets. J Pediatr Gastroenterol Nutr 2002;35:173–9. [15] Kawahara H, Kubota A, Hasegawa T, et al. Effects of rikkunshito on the clinical symptoms and esophageal acid exposure in children with symptomatic gastroesophageal reflux. Pediatr Surg Int 2007;23:1001–5. [16] Thomson M, Rao P, Rawat D, et al. Percutaneous endoscopic gastrostomy and gastrooesophageal reflux in neurologically impaired children. World J Gastroenterol 2011; 17:191–6.