A clinical pilot trial of metoclopramide therapy for gastric residuals in preterm infants

Feeding intolerance characterized by significant gastric residuals is common in preterm infants. This usually prompts caregivers to ‘hold’ or decrease feeds, thereby postponing the establishment of full enteral feeding. The cause of gastric residuals in preterm infants is probably multifactorial. At times, gastric residuals represent the earliest sign of serious pathology such as intestinal obstruction, sepsis, paralytic ileus, lactobezoar (1) or necrotizing enterocolitis (NEC) (2). Most often gastric residuals are attributed to gut immaturity. Gastric emptying is greatly delayed in preterm infants (3); thus, drugs that may enhance gastric emptying, such as metoclopramide, have been tried in a series of cases of infants with gastric residuals, with various degrees of success (4–7). Moreover, metoclopramide is not without side effects, which include irritability, vomiting, dystonic reactions and extrapyramidal symptoms (8). A major practical textbook of neonatal pharmacology (9) states that metoclopramide ‘facilitates gastric emptying and GI motility’ and ‘may improve feeding intolerance’. At this point in time, whether metoclopramide is effective and safe in preterm infants with gastric residuals is unknown. We therefore designed the following pilot study to determine the feasibility of a prospective, double-blinded, placebo-controlled clinical trial of metoclopramide therapy for feeding intolerance in preterm infants. We hypothesized that this pilot study would enable us to determine the sample size necessary to demonstrate the superiority of metoclopramide over placebo. One investigator (DH) enrolled 20 preterm, gavage-fed, very low birth weight infants (500–1500 g birth weight) hospitalized at the neonatal intensive care unit of the Lis Maternity Hospital, Tel Aviv Medical Center between February 2004 and May 2005. At their first episode of significant gastric residuals, which we defined as gastric residuals greater than 20% volume of the previous feeding, infants were randomized to receive metoclopramide therapy (n = 9) or placebo (n = 9). Metoclopramide and placebo were prepared by a pharmacist blinded to patient allocation. The drug was distributed in similar vials marked only A and B, so all personnel treating the study infants were blinded to group assignment. Metoclopramide or a similar volume of placebo was given intravenously at a dose of 0.05 mg/kg body weight every 8 h (9), until the patients received a total feeding volume of >100 mL/kg body weight/day, and thereafter, the medication (or placebo) was administered enterally. It was then continued for 7 days after the infant had tolerated full feeds (160 mL/kg body weight/day). At entry to the study all infants had no electrolyte imbalance and they were free of congenital anomalies or dysmorphism. The major outcome variables of the study were time to reach full enteral feeds (10, 11), the daily number of significant gastric residuals and the incidence of NEC. A feeding protocol was approved by all attending physicians in our institution, and strictly adhered to. The schedule clearly indicated when feeds should be started, the rate of daily increments, feeding composition (pumped human milk whenever available, diluted or non-diluted preterm infant formula), and time to reach full feeds. Bolus feeds were given every 3 h by gravity drainage using a nasogastric tube after priming of the tubing as described by us elsewhere (11). Gastric residuals were assessed every 3 h, prior to each meal. For clinical and research purposes, significant gastric residuals were defined as gastric residuals >20% of the volume of the previous feed with an absolute minimum volume of at least 2 mL. Feeds were initiated at 20 mL/kg/day, with small equal daily increments such that full volume feeds (160 mL/ kg/day) were reached within 8–14 days, depending upon the birth weight of the infant (the lower the birth weight, the lower the incremental progression of feeds). Feeds were started on day 2 at the earliest or 24 h following stabilization of an unstable blood pressure (using normal values of Versmold et al. (12)). Expressed human milk was the nutrition of choice, and was used undiluted. When unavailable, it was replaced by Similac Special Care ready-to-feed formula (Ross