Introduction

Pancreatic enzyme replacement therapy (PERT) is essential to prevent malnutrition in patients with exocrine pancreatic insufficiency (EPI). Good nutritional status and adequate growth are particularly important in patients with EPI due to cystic fibrosis (CF), as these parameters are associated with better lung function and improved survival in this patient population.[14] PERT is usually administered orally with meals and snacks. However, patients with CF who experience excessive weight loss because of infection or illness, and those who are malnourished and unable to take enough food by mouth to maintain adequate body weight, may require aggressive nutritional therapy such as tube feeding to assist with weight gain.[58] The gastric feeding tube (gastrostomy or G-tube) is more suitable for long-term use (e.g. more than a few weeks) than the nasogastric tube. In clinical practice, the need therefore sometimes arises to administer PERT via G-tube by mixing the pellets contained in PERT capsules with soft food.

CREON® (pancrelipase) Delayed-Release Capsules (Abbott, Marietta, GA, USA; Abbott, Hannover, Germany) is an orally administered PERT that is available worldwide (generic name pancreatin outside the US). CREON® pancrelipase capsules contain enteric-coated pellets (spheres) 0.7–1.6 mm in diameter.[9] The efficacy and safety of pancrelipase have been demonstrated in randomized, double-blind, placebo-controlled trials enrolling patients aged ≥7 years with EPI due to CF[1012] and patients with EPI due to chronic pancreatitis or pancreatic surgery.[13,14] Open-label studies have demonstrated the safety of CREON® pancrelipase in children aged 1 month to 6 years with EPI due to CF, with control of fat malabsorption and clinical symptoms,[1517] and in adults with EPI due to chronic pancreatitis or pancreatic surgery (6-month study).[18]

The objective of this in vitro study was to investigate the feasibility of administering CREON® pancrelipase gastro-resistant pellets via G-tube by mixing them with baby food. Our aim was to identify the appropriate types and sizes of G-tube that allow administration without clogging, sticking or visible damage to the pellets and without loss of pellet integrity (gastric resistance and lipase activity).

Methods

This was an in vitro study carried out in a single research laboratory at Abbott, Hannover, Germany between July and September 2010. The objective was to identify the smallest diameter and longest stoma length of G-tube that would allow administration of the pancrelipase pellets using baby food as the vehicle. Following determination of the appropriate tubes, an additional objective was to determine the integrity of the pancrelipase pellets following G-tube administration. Here we tested pancrelipase 12 000-lipase unit CREON® capsules from a single manufacturing batch. The assessment of multiple batches was not necessary because all batches are identical, being produced using a manufacturing process that has been validated according to current regulatory requirements, with strict specifications that are always met regarding quality, composition, pellet size and gastric resistance.

G-Tube Feasibility Testing

Based on G-tubes predominantly used in clinical practice, feasibility testing was initially planned for the following: Kimberly-Clark MIC Bolus® sizes 14, 16, 18, 20 and 24 French (Fr) and Kimberly-Clark MIC-KEY® low-profile sizes 14, 16, 18, 20 and 24 Fr (stoma length 5.0 cm) [Kimberly-Clark Health Care, Dallas, TX, USA]. Fr denotes the outer tube diameter (1 Fr = 0.33 mm).

A high-viscosity vehicle is required for the administration of pancrelipase pellets by G-tube to ensure that they are suspended properly, thereby guaranteeing passage through the entire G-tube to reach the intended site of action without sedimentation, which may lead to clogging of the tube. Baby food was selected as the vehicle in this study as it meets this specification, is readily available and is commonly used (particularly applesauce) for administering PERT in infants and young children. Four types of baby food available in the US were used: applesauce and bananas manufactured by both Gerber Products Company (Florham Park, NJ, USA) and Beech-Nut Nutrition Corporation (Amsterdam, NY, USA). Prior to administration, pH testing was carried out to ensure that all four baby foods had a pH <4.5 to avoid potential disruption of the enteric coating, early release of enzymes and/or loss of enzyme activity.

CREON® pancrelipase capsules were opened and the pellets sprinkled onto approximately 15 mL of baby food per capsule. The mixture was stirred gently and, after 15 minutes, poured into a 35 mL syringe from which the plunger had been removed and the outlet blocked to prevent spillage. With the syringe tip already in the G-tube, the plunger was inserted and the mixture pushed slowly through the G-tube at a rate of approximately 15 mL every 10–15 seconds until all of the mixture had passed through the tube. Pellets pushed through the tube were collected on a sieve and the tube was flushed with 10–30 mL water to collect any remaining material. G-tubes were inspected visually to assess clogging or sticking to the tube. Recovered pellets were then flushed with 200 mL simulated gastric fluid and visually inspected for damage. Simulated gastric fluid was prepared according to US Pharmacopeia (USP) guidelines: dissolve 2.0 g of sodium chloride in 80 mL of 1 mol/L hydrochloric acid and make up to 1000 mL (volumetric flask).[19] This process was repeated three times for each G-tube and food type.

Integrity Testing of Pancrelipase Pellets

For each tube type, if there was no clogging, sticking or damage with all four foods, pellet integrity (gastric resistance and lipase activity) was assessed using an in vitro dissolution method modified from the current USP monograph for ‘Pancrelipase delayed-release capsules’[19] and the European Pharmacopoeia,[20] with a 2-hour gastric simulation step. The predetermined level of activity required to confirm pellet integrity was ≥80% of actual USP lipase activity per capsule after dissolution within 30 minutes at pH 6.0 (an approved in-house specification). Following the successful feasibility testing with the Kimberly-Clark G-tubes noted above, integrity testing was also carried out using tubes by a different manufacturer with a similar diameter: Bard® Tri-Funnel replacement tube size 18 Fr and the Bard® Button size 18 Fr with stoma length 3.4 cm (Bard Access Systems, Salt Lake City, UT, USA). For comparison, untreated pancrelipase pellets were also subjected to integrity testing. Pellets were removed from the capsule and tested directly using the same in vitro dissolution method.

Results

Feasibility of G-Tube Administration

As shown in table I, administration of CREON® pancrelipase pellets was feasible through the following G-tubes, based on the absence of clogging, sticking and visible damage to pellets: Kimberly-Clark MIC Bolus® size 18 Fr and Kimberly-Clark MIC-KEY® size 16 Fr. Feasibility testing of tubes with a greater diameter was not carried out as the successful testing of a particular size permits the use of any tube of the same type and supplier with a larger diameter or shorter stoma length. Based on multiple tests carried out in the integrity testing phase, it was clear that this administration method was also feasible with the Bard® Tri-Funnel size 18 Fr and the Bard® Button size 18 Fr.

Table I
figure Tab1

Summary of G-tube feasibility testing

Pellet Integrity

Pellet integrity was maintained following passage through all four of the above-mentioned G-tubes. As shown in table II, the mean lipase activity of pellets was ≥90% relative to actual lipase activity with all four tubes and all four foods tested, thereby exceeding the predetermined specification (≥80%). There were no differences compared with untreated pellets that had not been mixed with baby food or pushed through a G-tube.

Table II
figure Tab2

Lipase activity of CREON® pancrelipase pellets following G-tube administrationa

These data apply to all CREON® pancrelipase capsule strengths as there are no differences in the composition, size and quality of the gastro-resistant pellets contained in the capsules, regardless of their strength in lipase units. Non-capsule formulations available outside the US require further evaluation.

Discussion

The results of this study indicate that the pellets contained in CREON® pancrelipase capsules can be mixed with baby foods of pH <4.5 and administered via the following G-tubes without clogging, sticking or visible damage to the pellets, and with no loss of gastric resistance or lipase activity: Kimberly-Clark MIC Bolus® size 18 Fr and larger, Kimberly-Clark MIC-KEY® size 16 Fr and larger, Bard® Tri-Funnel size 18 Fr and larger and Bard® Button size 18 Fr and larger.

Patients with CF often experience significant challenges in achieving adequate nutrition, and some require interventions such as tube feeding to maintain appropriate nutritional status and body weight.[58] The method described here of administering PERT via G-tube using baby food is current clinical practice based on clinical experience and anecdotal evidence, although there are very few published data available to support it. Another PERT manufacturer has documented a procedure for administration of pancrelipase (Pancrecarb® MS-4; Digestive Care Inc., Bethlehem, PA, USA) in applesauce through a G-tube.[21] In addition, the results of a prospective, open-label, crossover, single-centre study assessing the safety and efficacy of pancrelipase (Pancrecarb® MS-4) administered by G-tube compared with pancreatic enzyme powder mixed with formula in a feeding bag have been presented in a congress abstract.[22] In six patients completing that study, G-tube administration of PERT did not have any adverse effects and was as effective as higher doses of PERT mixed with formula. Other authors describe the use of thickened fruit juice as a vehicle for administering PERT through G-tubes but do not provide data to confirm that this method does not affect PERT performance.[23] Alternative procedures for administering PERT by G-tube are also sometimes used in clinical practice. However, these procedures have not yet been formally assessed using pancrelipase pellets and we are therefore able to recommend only the procedure described here for CREON® pancrelipase administration by G-tube, using baby food of pH <4.5 as a vehicle. The aim of this method is to deliver the pellets directly to the stomach intact, as they would be if administered orally in a capsule. In order to achieve this, a high-viscosity vehicle (such as baby food) of pH <4.5 is recommended for the reasons described previously. If administration procedures outside these specifications are used, the same performance cannot be guaranteed as they may result in clogging/sticking in the tube, damage to pellets and disruption of the enteric coating. It is important to note that there is variation in the pH of different baby foods; therefore if foods other than those tested in this study are used, the pH should be checked first to guarantee the integrity of the enteric coating and thus the expected lipase activity.

The tube sizes initially tested in this study were selected because of their frequent use in clinical practice; G-tubes of size 14 Fr are commonly used in children with CF, as are larger tubes of size 16 Fr and 18 Fr. The results reported here are therefore applicable to the clinical care of patients with CF, and also support the administration of CREON® pancrelipase pellets through G-tubes of larger diameter, which may be used in adult patients with EPI associated with CF or other diseases such as chronic pancreatitis and pancreatic surgery. Although feasibility testing of larger diameter tubes was not carried out in this study, the successful testing of the sizes mentioned above indicates that the use of any larger diameter tube of the same type and manufacturer is acceptable when using baby food of pH <4.5. Theoretically, vehicles other than baby food could be used in adult patients if they meet the criteria of high viscosity and pH <4.5 (e.g. other acidic soft foods). In addition, PERT dosing is individualized and the number of capsules required per tube feeding will therefore vary greatly between patients. In this study, the contents of one 12 000 lipase unit capsule were mixed with ∼ 15 mL of baby food but this is scaleable where higher doses and more capsules are required.

Conclusion

The results of this study confirm that it is feasible to mix CREON® pancrelipase gastro-resistant pellets with baby food of pH <4.5 and administer them via G-tubes of the specified sizes without clogging, sticking or visible damage to the pellets, and with no loss of gastric resistance or lipase activity.