Chromium and zinc contamination of parenteral nutrient solution components commonly used in infants and children

Transcription

1 Chromium and zinc contamination of parenteral nutrient solution components commonly used in infants and children EMILY B. HAK, MICHAEL C. STORM, AND RICHARD A. HELMS Abstract: Chromium and zinc contamination of components of parenteral nutrient (PN) solutions used in infants and children was studied. Solutions of amino acids, L- cysteine hydrochloride, dextrose, electrolytes, minerals, vitamins, multiple trace elements, and individual trace elements were obtained. A variety of manufacturers, lots, and expiration dates were represented when possible. The solutions were analyzed for chromium and zinc by flame atomic absorption spectrophotometry. In all amino acid products, chromium concentration was below the limit of detection and zinc concentration ranged from 0.06 to 4.97 mg/ L. In the L-cysteine hydrochloride products, chromium was measurable in only two lots (0.11 and 0.23 mg/l); zinc was measurable in all lots (32 86 mg/l). Sodium and potassium salts of chloride and acetate had chromium concentrations of mg/l and zinc concentrations of mg/ L. Phosphate salts contained chromium mg/l and zinc mg/l. In We recently determined serum and urine levels of chromium and zinc in infants and children receiving parenteral nutrition for longer than three months and measured the amounts of those elements in the parenteral nutrient (PN) solutions. 1,2 We found, as have others, that many PN solutions contained more chromium than was added as traceelement supplementation. 3-5 However, we also unexpectedly found that some dextrose and amino acid solutions, in particular those with added L-cysteine hydrochloride, contained significantly more zinc than was added. 1 The objectives of the study reported here were to measure chromium and zinc in PN solution components, to determine if there is a relationship between the amount of contaminant and the storage time and container size, and to calculate the amount of contaminant chromium and zinc that typically would be provided to an infant by a PN solution. calcium gluconate, zinc concentration was mg/ L. In four lots of multiple trace elements, chromium was % and zinc was % of the labeled amount. A PN solution for a <10-kg infant compounded from the components assayed would provide up to an additional 0.7 µg of chromium per kilogram and 200 µg of zinc per kilogram. Zinc and chromium contaminants were detected in many of the products that are common components of PN solutions for infants and children; the contamination may be sufficient to result in the administration of zinc and chromium in amounts exceeding current recommendations. Index terms: Amino acids; Calcium gluconate; Caloric agents; Chromium; Concentration; Contamination; Cysteine hydrochloride; Dextrose; Elements; Injections; Metals; Minerals; Nutrition; Pediatrics; Phosphates; Potassium chloride; Replacement solutions; Sodium chloride; Toxicity; Vitamins; Zinc Am J Health-Syst Pharm. 1998; 55:150-4 Methods Solutions of amino acids, L-cysteine hydrochloride, dextrose, electrolytes, minerals, vitamins, multiple trace elements, and individual trace elements were obtained from various manufacturers (Table 1). Multiple lots of individual products were included when available. Products that were past their expiration dates were also included. M.V.I. Pediatric was reconstituted in doubledistilled, metal-free water according to the package insert. All other products were sampled directly. Chromium and zinc concentrations were measured with a Perkin-Elmer 3300 flame atomic absorption spectrophotometer equipped with a flow spoiler, an AS 90 autosampler, and an AS 91 controller (Perkin-Elmer, Foster City, CA). Air:acetylene mixtures of 10:3.8 L/min for chromium and 10:2 L/min for zinc were used. Perkin-Elmer Intensitron lamps were used, and settings for initial lamp current, slit, and energy followed the EMILY B. HAK, PHARM.D., BCNSP, is Associate Professor; MICHAEL C. STORM, PH.D., is Associate Professor; and RICHARD A. HELMS, PHARM.D., is Professor, Department of Clinical Pharmacy, The Center for Pediatric Pharmacokinetics and Therapeutics, The University of Tennessee, Memphis. Address reprint requests to Dr. Hak at the Department of Clinical Pharmacy, 26 S. Dunlap Street, Memphis, TN 38163, or to ehak@utmem1.utmem.edu. Supported in part by a State of Tennessee Center of Excellence grant and the Pediatric Pharmacology Research Unit at The University of Tennessee, Memphis. Presented at the Clinical Congress of the American Society for Parenteral and Enteral Nutrition, San Francisco, CA, January 28, Copyright 1998, American Society of Health-System Pharmacists, Inc. All rights reserved /98/ $ Am J Health-Syst Pharm Vol 55 Jan

2 Chromium and zinc contamination Reports Table 1. Chromium and Zinc Concentrations in Parenteral Nutrient Solution Components Tested Solution Source manufacturer s recommendations. Atomic absorption spectrophotometry (AAS)-grade chromium and zinc standards (VWR Scientific, South Plainfield, NJ) and samples were diluted with high-purity, metal-free 0.1 N nitric acid (prepared from ultra-pure nitric acid [Ultrex II, J. T. Baker, Inc., Phillipsburg, NJ, lot J25545]). Standards and samples were prepared in duplicate. The autosampler was set so that measurement of each standard was replicated three times and each sample two times. All analyses for each metal was performed on the same day in February Each sample was diluted to 1:10 with AAS-grade 0.1 N nitric acid and analyzed. Samples that were outside the linear range of the working standards at a 1:10 dilution ( mg/l for chromium and mg/l for zinc) were further diluted to 1:100 and 1:1000 with 0.1 N nitric acid if necessary. The analytical range of the standard curve for each element was linear within the accepted range for both elements (correlation coefficients, for chromium and for zinc). Measurements that were less than 0.02 mg/l were considered below the assay s limit of detection. Results are reported as the mean ± S.D. for duplicate measurements of single samples for the individual lots. Expiration Date Chromium Concentration (mg/l) b Zinc Concentration (mg/l) b FreAmine III 8.5% McGaw J1E /83 < ± 0.05 Aminosyn 10% Abbott DM /95 < ± 0.03 TrophAmine 10% McGaw J5H /96 < ± 0.00 FreAmine III 10% McGaw J5K /97 < ± 0.01 Dextrose 70% McGaw J5K096A /97 < ± 0.01 L-cysteine hydrochloride 50 mg/ml McGaw P0B /92 < ± 3.9 McGaw P1P / ± ± 2.2 American Regent / ± ± 3.2 Abbott R1 10 5/95 < ± 6.4 McGaw P3L /95 < ± 9.9 Abbott R1 10 2/96 < ± 4.2 McGaw P5D /97 < ± 0.9 Potassium chloride 2 meq/ml Abbott DM / ± ± 0.14 Abbott DK 20 7/ ± ± 0.03 Potassium acetate 4 meq/ml Lyphomed / ± ± 0.21 Potassium phosphate 3 mmol P/mL Lyphomed / ± ± 0.12 Abbott DK 50 6/ ± ± 0.14 Sodium acetate 2 meq/ml Abbott DK 100 3/ ± ± 0.07 Sodium chloride 14.6% Abbott DM / ± ± 0.09 Sodium phosphate 3 mmol P/mL Abbott DK 50 12/ ± ± 0.06 Calcium gluconate 10% Lyphomed / ± ± 0.53 Int Med System BB053A / ± ± 0.06 Magnesium sulfate 4 meq/ml Abbott DK 50 11/ ± ± 0.04 Selenium 40 µg/ml SoloPak /96 < ± 0.50 Manganese sulfate 0.1 mg/ml Lyphomed /88 < ± 0.11 M.V.I. Pediatric Rorer F /91 < ± 0.07 Astra N /96 < ± 0.09 P.T.E.-4 Lyphomed / ± ± 24 Multitrace-4 American Regent / ± ± 41 American Regent / ± ± 12 American Regent / ± ± 10 a All containers were made of plastic, except for lot DK of potassium chloride, for which containers were made of glass. b Reported as mean ± S.D. for duplicate determinations of single samples. Lot Volume (ml) a The results were examined for correlations between metal content and product volume (in milliliters) and expiration date by using Cricket Graph III, version (Computer Associates Int., Inc., Islandia, NY). The amounts of chromium and zinc that would be provided as contaminants were calculated for each product. These concentrations were then used to calculate the microgram-per-kilogram doses of contaminant chromium and zinc for the individual components of an infant PN solution providing pediatric amino acids 2.5 g/kg, glucose 85 kcal/kg, sodium 38.5 meq/l, potassium chloride 20 meq/l, phosphorus 1 mmol/kg, calcium gluconate 2.5 meq/kg, magnesium sulfate 0.4 meq/ kg, M.V.I. Pediatric 5 ml, selenium 3 µg/kg, and 40 mg of L-cysteine hydrochloride per gram of amino acids on a daily basis. Results Table 1 lists the concentrations of chromium and zinc that were found in the various PN solution components. Chromium concentration was below the limit of detection in all amino acid products, and zinc was measurable in all amino acid products (range, mg/l). In the L-cysteine hydrochloride products, chro- Vol 55 Jan Am J Health-Syst Pharm 151

3 Table 2. Amounts of Contaminant Chromium and Zinc Provided to Infant (<10 kg) in Parenteral Nutrient Solution Compounded from Components Assayed Component Amount Chromium (µg/kg/day) Zinc (µg/kg/day) Amino acids 2.5 g/kg < (TrophAmine) Glucose 85 kcal/kg < Sodium 2.5 meq/kg chloride Potassium 2 meq/kg chloride Sodium 1 mmol P/kg phosphate Calcium 2.5 meq/kg gluconate Magnesium 0.4 meq/kg sulfate M.V.I. Pediatric 5 ml < Selenium 3 µg/kg < L-cysteine 40 mg/g < hydrochloride amino acids Total mium was measurable in only two lots from different manufacturers; zinc was measurable in all lots. A PN solution for a <10-kg infant compounded from the components that were assayed, including pediatric amino acids 2.5 g/kg with 40 mg of L-cysteine hydrochloride per gram of amino acids, would provide up to an additional 0.7 µg of chromium per kilogram and 200 µg of zinc per kilogram from contaminants (Table 2). The labeled amount of chromium in P.T.E.-4 was 1 µg/ml and in Multitrace-4 was 4 µg/ml. Both multipletrace-element products were labeled to contain 1 mg of zinc per milliliter. Variation from the labeled amounts was less than 10%, except for P.T.E.-4, whose zinc concentration was 13.5% greater than the labeled amount. Chromium and zinc contamination did not have any obvious correlation with expiration date. For amino acids (expiration dates, May 1983 to August 1997), L-cysteine hydrochloride (February 1992 to April 1997), and multiple-trace-element products (July 1990 to July 1997), there appeared to be no relationship between the expiration date and the chromium or zinc content. The oldest amino acid product evaluated, one lot of Fre- Amine III 8.5%, did contain substantial amounts of zinc. Potassium chloride was the only product evaluated that was available in both glass and plastic containers. The chromium and zinc contents did not differ between the two container types (for glass, chromium 0.02 mg/ml and zinc 0.45 mg/ml; for plastic, chromium 0.02 mg/ml and zinc 0.43 mg/ml). A wide range of glass product volumes ( ml) were examined to look for a correlation of container size with chromium and zinc contamination. Those products that were formulated with chromium or zinc (Multitrace-4 and P.T.E.-4), the very old unit of Fre- Amine III 8.5% (expiration date, May 1983), and all the L-cysteine hydrochloride products (very low ph) were excluded from this correlation analysis. Chromium and zinc concentrations in the remaining products were inversely correlated with glass product size. For chromium, r 2 was 0.58 and the power relationship was chromium concentration = 6.76(size in milliliters) For zinc, r 2 was 0.5 and the power relationship was zinc concentration = 2.12(size in milliliters) Discussion The problem of trace-element contamination has been described for a wide variety of PN products. 2-9 Trace elements may be inherent in raw materials, be introduced into a product during manufacturing or compounding, or result from product interaction with container components over time. Historically, the most significant problem involving trace-element contamination of PN products was aluminum contamination of dialysis solutions. Aluminum deposition in brain and bone tissue of patients undergoing dialysis caused dialysis dementia 6 and osteodystrophy. 7 The occurrence of these conditions has been greatly decreased by improved manufacturing processes that result in less aluminum in PN products. Aluminum toxicity continues to be a concern in children, and the American Academy of Pediatrics recommends that efforts to reduce the amount of contaminant aluminum in PN products be continued. 10 Chromium contamination of PN products has been recognized for several years and probably results from chromium present in the raw materials. Several investigators have measured the amount of chromium in a variety of PN products, components of PN solutions, and in amino acid and dextrose admixtures and fat emulsion. 3-5 Unlike Ito et al., 4 who found measurable amounts of chromium in amino acid products, we found none. However, our assay was not sensitive enough to detect the small amounts of chromium ( mg/l) previously reported. The amounts of chromium in the different dextrose, electrolyte, and mineral products that were evaluated in our study were greater than previously reported. This suggests that concentrations of contaminant chromium vary widely among lots from a single manufacturer, among manufacturers, and with container size. Contaminant zinc was found to be present in an early crystalline amino acid product (FreAmine II 8.5%) in amounts ranging from to 1.88 mg/l 11 ; however, zinc contamination in components of PN solutions had not been evaluated further until now. We had previously noted that PN solutions that contained L-cysteine hydrochloride had more zinc than was added as trace-element supplementation, 1 and therefore we expected that some L-cysteine products would have measurable amounts of zinc. However, we did not 152 Am J Health-Syst Pharm Vol 55 Jan

4 Chromium and zinc contamination Reports expect that zinc would be present in all L-cysteine products, and we did not expect as much as we found. During manufacturing of rubber, zinc oxide is added to accelerate vulcanization; thus, zinc is present in many rubber stoppers. 12 Zinc is also in the materials used to produce glass bottles. We suspect that contaminant zinc in PN products comes from the leaching of zinc from materials in direct contact with the solution, such as rubber stoppers or glass, rather than from raw material contamination. The amount of zinc in the L-cysteine hydrochloride products was variable and was not correlated with expiration date. Thus, the variability in zinc concentrations in different products likely relates, in part, to storage conditions. Storing L-cysteine hydrochloride, a product with a ph of 1.3, in a container on its side or upside down would result in greater contact between the solution and the stopper and would increase the amount of zinc leached from the stopper. Even though the 1983 amino acid product is not very acidic (ph, ~6.8), the greatest amount of zinc found in an amino acid formulation was in this product; this finding may be related both to the position in which the product was stored and to how long it was stored. The two lots of L-cysteine hydrochloride with the smaller amounts of zinc were from the Abbott universal additive syringe, which has a stopper material different from those of the other L-cysteine containers. The presence of zinc in this product is most likely due to the constant contact of the L-cysteine solution with the stopper material. We postulate that the high zinc concentration in the L-cysteine products is due to the low solution ph, which may facilitate zinc extraction from the stopper. Recommended dosages for trace elements in infants and children receiving PN solutions have been extrapolated from requirements for healthy infants and children, with consideration given to bioavailability. With parenteral nutrition, absorptive mechanisms regulating the amount of trace elements that reach the systemic circulation are bypassed, so it is important to give the appropriate dosages. Current dosage recommendations for zinc are 250 µg/kg/day for term infants less than three months of age and 100 µg/kg/day for term infants more than three months of age. 13 The maximum daily dose of contaminant zinc that would be provided from PN solutions compounded from the components we evaluated is twice the recommended daily dose for a term infant older than three months. The current dosage recommendation for chromium is 0.2 µg/kg/day. 13 The maximum daily dose of contaminant chromium that would be provided from the dextrose and amino acid solutions we studied is three times the recommended daily dose. Small amounts of chromium would also be provided via fat emulsion. 4,5 It is possible that many trace elements are being given in excessive amounts because of contamination of components of PN products. 2,4,5,14 To complicate matters, the amount and type of contamination vary with the product, the lot, and the manufacturer. Patients receiving long-term parenteral nutrition may be exposed to excessive trace elements for extended periods and thus are at increased risk for toxicity. Moukarzel et al. 5 reported that children receiving PN solutions for longer than one year had a decreased glomerular filtration rate that appeared to be related to excessive chromium intake. Others found an inverse relationship between plasma concentrations of chromium and iron that was thought to be due to competition between chromium and iron for transferrin. 15 Thus, routine monitoring of certain trace elements, in particular chromium and perhaps zinc, may be important in patients, notably infants and children, who receive long-term parenteral nutrition. Clinicians should consider instituting practices that minimize trace-element contamination of PN solutions. These include storing containers upright to minimize PN solution contact with stoppers, preferentially using newly manufactured products, and using products from as large a container as is possible and practical. An important potential issue could be trace-element deficiency, should there be changes in manufacturing processes or in containers that essentially eliminate trace-element contamination of products. Conclusion Zinc and chromium contaminants were detected in many of the products that are common components of PN solutions for infants and children; the contamination may be sufficient to result in the administration of zinc and chromium in amounts that exceed current recommendations. References 1. Cochran EB, Mouser J, Helms RA et al. Effects of L-cysteine HCl dose on serum concentration and urinary elimination of zinc in children on home TPN. Pediatr Res. 1994; 35:125A. Abstract. 2. Mouser J, Cochran EB, Helms RA et al. Chromium concentrations in children on home TPN and the relationship to intake. JPEN J Parenter Enteral Nutr. 1994; 18(suppl):34S. 3. Anderson RA. Chromium and parenteral nutrition. Nutrition. 1995; 11(suppl): Ito Y, Alcock NW, Shils ME. Chromium content of total parenteral nutrition solutions. JPEN J Parenter Enteral Nutr. 1990; 14: Moukarzel AA, Song MK, Buchman AL et al. Excessive chromium intake in children receiving total parenteral nutrition. Lancet. 1992; 339: Sedman AB, Wilkening GN, Warady BA et al. Encephalopathy in childhood secondary to aluminum toxicity. J Pediatr. 1984; 105: Ott SM, Maloney NA, Klein GL et al. Aluminum is associated with low bone formation in patients receiving chronic parenteral nutrition. Ann Intern Med. 1983; 98: Milliner DS, Shinaberger JH, Shuman P et al. Inadvertent aluminum administration during plasma exchange due to aluminum contamination of albumin-replacement solutions. N Engl J Med. 1985; 312: Koo WWK. Aluminum in parenteral nutrition solution Vol 55 Jan Am J Health-Syst Pharm 153

5 sources and possible alternatives. JPEN J Parenter Enteral Nutr. 1986; 10: Committee on Nutrition. Aluminum toxicity in infants and children. Pediatrics. 1996; 97: Shearer CA, Bozian RC. The availability of trace elements in intravenous hyperalimentation solutions. Drug Intell Clin Pharm. 1977; 11: Smith EJ, Nash RJ. Elastomeric closures for parenterals. In: Avis KE, Lachman L, Lieberman HA, eds. Pharmaceutical dosage forms. Vol. 2. Parenteral medications. New York: Dekker; 1986: Greene HL, Hambidge KM, Schanler R et al. Guidelines for the use of vitamins, trace elements, calcium, magnesium, and phosphorus in infants and children receiving total parenteral nutrition: report of the Subcommittee on Pediatric Parenteral Nutrient Requirements from the Committee on Clinical Practice Issues of the American Society for Clinical Nutrition. Am J Clin Nutr. 1988; 48: Hambidge KM, Sokol RJ, Fidanza SJ et al. Plasma manganese concentrations in infants and children receiving parenteral nutrition. JPEN J Parenter Enteral Nutr. 1989; 13: Bougle D, Bureau F, Deschrevel G et al. Chromium and parenteral nutrition in children. J Pediatr Gastroenterol Nutr. 1993; 17: Am J Health-Syst Pharm Vol 55 Jan