|
|
|
Original Articles Indian Pediatrics 2000;37: 837-844 |
|
A
DOUBLE-BLIND PLACEBO-CONTROLLED STUDY ON PROPHYLACTIC USE OF CISAPRIDE
ON FEED INTOLERANCE AND GASTRIC EMPTYING IN PRETERM NEONATES
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
P.S. Reddy, A.K. Deorari, C.S. Bal*, V.K. Paul and Meharban Singh
Preterm neonates have immature gastrointestinal motility, which manifests as gastro-esophageal reflux, delayed gastric emptying and delay in intestinal transit. This results in frequent occurrence of feed intole-rance, which usually manifests as vomiting and or abdominal distension. Feed intolerance has been reported in 2% to 67% of infants weighing less than 1500 g depending on the type of milk feed in different studies(1-3). Intolerance to oral feeding results in prolonged parenteral fluid therapy, increased hospital stay, increased risk of sepsis and consequently increased cost of treatment. To facilitate gastrointestinal transit, various prokinetic drugs have been used(4). These enhance contractile force and accelerate luminal transit time. Cisapride, a piperidinyl benzamide derivative chemically related to metoclopra-mide acts by releasing acetylcholine from postganglionic enteric neurons. It lacks adverse CNS side effects as it does not interact with dopamine receptors. It is at least 100 times more potent than metoclopramide in enhancing gastrointestinal motility(4). Many studies hve shown beneficial effects of cisapride in conditions causing gastroparesis (decreased emptying without mechanical obstruction) by enhancing gastric motility(5,6). There are only few studies regarding the use of cisapride for the management of feed intolerance in neonates(4-9). The present study was designed to test the hypothesis that use of cisapride may result in decreased incidence of gastro-esophageal reflux, increased gastric emptying thus altering the feeding volumes and improving the tolerance of feeds in preterm neonates(10).
A prospective study was conducted in the Neonatal Unit of All India Institute of Medical Sciences from April 97 to August 98. Forty nine preterm neonates between 29-34 weeks of gestational age (GA) were enrolled after obtaining informed consent. They were stratified into two groups (29-32 wks and 33-34 wk). Babies were block randomized in study or placebo group, in a double blind fashion. Entry & Exclusion Criteria Following one day of stable vital signs babies were enrolled into study during the first week of life once they reached 30 ml/kg of enteral feeds, or 25% of total fluid intake was enteral. Gestational age was calculated from the date of last menstrual period. In cases of unsure dates gestational age was assessed according to physical and neurological criteria(11). Babies with gross congenital malformations, sepsis and those on aminophylline were excluded. Feeding Protocols Throughout the study period, neonates were kept preferably on expressed breastmilk (EBM). When EBM was insufficient, the LBW formula feeds were used in a standard dilution. Initiation and advancement of feeding were decided by the treating physician. The usual protocol followed was: if birth weight <1250 g, baby was started on intravenous fluids, if stable for >6 hours, minimal enteral feedings were introduced and then gradually increased. If birth weight was >1250 g, the baby was started on oro gastric (OG) feeds after observing respiratory distress for 2 hours. For babies weighing >1500 g and >32 weeks gestation without respiratory distress, palladay spoon (PS) feeds were offered from the beginning. In babies >1250 g PS feeds were offered per each nursing shift (8 hourly). If babies were unable to take PS feeds properly then they were kept on OG feeding. Babies <32 wks of gestation received multivitamin and calcium supplement once they were on total enteral feeds but these were introduced only after nuclear scan had been done. Therapy/Placebo According to the random numbers, babies were allocated either cisapride or placebo at a dosage of 0.2 mg/kg in a suspension form orally at 8 hourly intervals for 2 weeks. The drug was mixed with milk then flushed down through orogastric tube or given with a spoon. The feed intolerance was managed by the attending physician either by omitting the feeds once or twice or starting intravenous fluids depending on the severity. Outcome Variables Babies were observed for next 14 days of study period for feed intolerance, weight gain pattern and side effects of cisapride. The feeding volumes were noted as well as the time taken to reach the full enteral feeds (defined as 150 ml/kg per day). Gastric emtying and gastro-esophageal reflux were studied by Technetium phytate scan on day 7 ± 1 after enrollment. Feed intolerance was defined as presence of any of the following criteria, prefeed gastric residues of >25% of the previous feed (in those on at least >10 ml OG feeding), increase in the abdominal girth from baseline by more than 2 cm in the region of epigastrium over 24 hours by regular monitoring and vomiting(12-14). Weight gain pattern was observed by daily weight record on an electronic weighing scale with sensitivity of ±1 g. Weight on day 14 and/ or at the time of discharge was recorded. Nuclear Scan Study At the end of 7 ± 1 days, gastric emptying study was done. Two ml of milk mixed with 100 microcurie of Tc99 phytate was flushed down with 10 ml of milk through OG tube. After OG flushing, tube was taken out immediately. The imaging of anterior abdomen was done in a supine position using a large field of view gamma camera connected to a microdelta computer. Acquisition of data was done in a total number of 120 frames with each frame size of 16 seconds. Techenetium scan was considered safe as the radiation exposure was very minimal compared to a single X-ray(15). On the assumption that gastric emptying occurs exponentially like other biological functions, the following formula was used to calculate T1/2 of gastric emptying where a difference between time of start of study (To) from a point A to the end of the study (T1) point B is taken as "t"(15) (Fig. 1). 0.3 ´ t Gastric emptying T1/2 = ––––––– log A/B Sample Size Sample size was calculated with gastric emptying as the main outcome variable. To detect a difference of 30 minutes in both groups with a power of 90%, assuming gastric emptying without drug as 120 minutes with a standard deviation of ±30 minutes, a total number of 20 patients per group had to be recruited for a significance level of 5% (a = 0.05) The mean and standard deviation of gastric emptying time were taken based on the previous studies carried out in children and adults with and without cisapride(16,17). Statistical Analysis Data was analyzed by using two sample "t" test and chi square test wherever applicable. Non parametric tests such as Wilcoxon rank sum test (Mann-Whitney U test) was applied whenever the distribution was skewed. Fisher’s exact test was applied when sum of the components in 2 ´ 2 table was less than 30. Statistix (SX) software version 4.0 was used for calculations.
A total of 49 preterm neonates were enrolled into study. All babies were stable before enrollment into the study in terms of heart rate, blood pressure, oxygen saturation for 24 hours. Three babies were excluded from the study due to subsequent complications of prematurity – one each due to necrotising enterocolitis, respiratory distress due to aspiration and posterior uretheral valve diagnosed later on day-4. All these three cases belonged to placebo group. Two babies in the study group had to be dropped, because they had to be discharged before completion of study period due to strike by the staff. Clinical characteristics of the babies in drug or placebo groups were similar (Table I). The baseline vital signs before entry into the study (24 h observation period) before enrollment did not have any difference when compared between drug and placebo group. Majority of the babies did not have any significant neonatal morbidity before enrollment into the study. Respiratory distress was noted to be significantly higher in drug group (p = 0.03; Fisher’s exact test) Antenatal steroids for preterm labor were used equally in both the groups according to standard protocol. Feed intolerance occurred in 59% of the babies (13 out of 22 babies) in the drug group compared to 41% (9 out of 22) in the placebo group. Though, it was seen in frequent number of patients, it never resulted in the dis-continuation of oral feedings. There was no significant difference between the two groups in number of babies picked up by each criteria as well as in the total number feed intolerance episodes detected by all criteria. Incidence of feed intolerance as judged by increased abdominal girth or percentage of prefeed gastric residuals was noticed to be similar in both groups. It was 21 per 100 babies per day in the the drug group compared to 16 per 100 babies per day in placebo group. Number of feed intolerance as judged by abdominal girth was less. Residue rarely exceeded 10%. As the babies were upgraded to palladay feeding, the residues were mainly calculated in the initial few days. The mean number of vomiting during the study was 1.09 ± 2.4 in the drug group and 0.7 ± 1.2 in placebo group per week (the median values in both the groups was zero, signifying that many babies did not have feed intolerance in the form of vomiting). There was no statistical difference between drug and placebo groups in terms of feed intolerance when they were compared after stratification (Table II). Full enteral feeding (150 ml/kg/d) was reached on 6.5 ± 3 days of life in drug group compared to the 6.2 ± 3 days of placebo. Number of days to reach full enteral feeding was 3.3 ± 2 days in drug group compared to 3.3 ± 2 days in placebo group (p >0.05). No significant net gain or loss of weight was observed during first week of study whereas during the second week babies receiving cisapride gained 8.8 ± 2.9% of birth weight in comparison to 9.1 ± 5.7% in placebo group (p >0.05). Cisapride use did not result in any increased stool frequency. Incidence of jaundice was the same in both groups. During continuous ECG monitoring while in intensive care unit, no abnormal ECG patterns were observed. Gastric emptying time (mean (SD) and median) in the study group (58.1 (32.2) 48.8 min; 95% CI being 44.6-71.6) was not signi-ficantly different from that of the placebo group (53.8 (34.6) 43.4 min 95% CI being 38-74). Asymptomatic gastro-esophageal reflux was seen in nearly 50% of babies in each group. Table I - Clinical Characteristics
* Expressed as No course/complete/incomplete. Table II - Feed Intolerance Based on Clinical Criteria (Mean ± SD)
Feed intolerance is commonly reported in preterm neonates due to various factors related to their gut immaturity. This manifests as increasing gastric aspirates, regurgitation and or abdominal distension. Increase in abdominal girth by 2 cm from the baseline or prefeed gastric residue of >25% are good clinical indicators of feed intolerance(12-14). Various studies have shown that feed intolerance ranges from 2 to 67%(1-3). In the present study this was noticed in 59% of babies receiving cisapride and in 41% of babies on placebo. The babies in two groups were comparable in all aspects in terms of stability before inclusion, feeding, use of antenatal steroids, and incidence of birth asphyxia. All these are known to affect gastrointestinal function in preterm neonates(18,19). Cisapride use was found to be safe in babies <34 wk as has been reported earlier(8). Prolongation of QT interval on electrocordio-gram and cholestasis have been described as possible side-effects of cisapride use(8,20). The mean GET (half life) was 58 ± 32 min in the study group compared to 53 ± 34 min in the placebo group. As GET may be altered due to type of feeds (EBM or formula), it was ensured that two feeds prior to the nuclear scan, the baby always received EBM. The gastric emptying time was not altered by cisapride significantly (p >0.05) in this study. In another study gastric emptying was around 87 ± 29 min by scintigraphy(21). These differences may be due to different populations, types of feeds and postnatal age of assessment. Incidence of GER as well as relationship to symptomatology in preterm infants has not been well studied till now. In this study GER was noticed in nearly one half of study population which is comparable to other studies(22). It was asymptomatic in all except one which required medical treatment. The use of cisapride did not reduce the incidence of GER in study group. The clinical parameters as judged by feed intolerance, weight gain pattern and gastric motility as indicated by GET showed that cisapride did not have any beneficial effect on the premature gut. As the sample size was not large enough to detect improvement in GET, future multicentric studies are warranted. A recent study using ultrasonography did not show improvement in gastric emptying in preterm neonates receiving cisapride(23). A few studies done previously with a small sample size have shown equivocal results with cisapride in the babies with feed intolerance(7,8). In conclusion, cisapride given prophy-lactically may not have any beneficial effect on feed intolerance and gastric emptying of preterm babies. Tailoring of volume and type of feeds may play an important role to improve enteral tolerance of feeds. Whether cisapride will be of use in preterm neonates with feed intolerance needs further evaluation by a well designed randomized placebo controlled clinical trial.
Contributors: AKD, VKP, MS coordinated the study (design and interpretation); PSR helped in data collection; AKD, PSR drafted the paper; and CSB conducted nuclear scan on babies. AKD will act as guarantor for the paper. Funding:
None.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
