Stable isotope breath tests

RESEARCH TECHNIQUES APPLICABLE TO PEDIATRICS

Nutrition Vol. 14, No. 10, 1998

Stable Isotope Breath Tests LAWRENCE T. WEAVER, MA, MD, DCH, FRCP From the Department of Child Health, University of Glasgow, Yorkhill Hospitals, Glasgow, Scotland, UK ABSTRACT

There is a need for non-invasive tests of gastrointestinal and nutritional function. Clinical problems peculiar to infancy and childhood require prompt diagnosis, and methods that are invasive or involve the use of radioisotopes are often impractical or ethically unacceptable. What the pediatrician and clinical scientist seek are tests that are simple, repeatable, and unequivocal in their result for diagnosis, to assess the effects of treatment, and to measure the development of gastrointestinal function during early life. Stable isotope breath tests offer a ready and attractive answer to these needs. They involve the ingestion of substrates labeled with the non-radioactive isotope of carbon (13C), followed by the collection of serial breath samples for analysis of the enrichment of 13CO2, the end product of substrate metabolism. Their non-invasive nature recommends them for use in infancy and childhood, and they can be performed in the ward, clinic, laboratory, and home. In this article I discuss to what degree stable isotope breath tests fulfill the pediatrician’s and scientist’s needs. I have chosen two examples from the work of myself and my colleagues to illustrate the principles and use of 13C breath tests to detect Helicobacter pylori infection and to measure fat digestion in infancy and childhood. Nutrition 1998;14:826 – 829. ©Elsevier Science Inc. 1998 Key words: isotope, nutrition, gastrointestinal tract, breath test, children

PRINCIPLES AND METHODOLOGY OF 13

13

C BREATH TESTS

At present the principal uses of C breath tests in clinical practice are for the measurement of digestion, the detection of enteric infection, assessment of the integrity of the small intestine, the measurement of gastrointestinal motility, and the quantitation of hepatic microsomal activity.1 13 C breath tests involve the measurement of the 13C:12C ratio present in breath carbon dioxide after ingestion of a nutrient, meal, or other substrate containing 13C. 13C is a non-radioactive stable isotope that occurs naturally at an abundance of 1.1%. When a substrate that is relatively 13C rich is ingested, the 13C contained within it, after digestion and absorption, enters oxidative metabolic pathways leading to enrichment of a variety of organic metabolites, including bicarbonate, protein, fat, and carbohydrate, within the body. The end product of oxidative metabolism is 13 CO2, which is expired in the breath. This is separated by cryogenic or chromatographic means from other components of expired air and its 13C enrichment can be measured by isotope ratio mass spectrometry.2 Measurement of 13C enrichment in expired CO2 may be all that is required in the simplest form of these breath tests (see proceeding section MEASUREMENT OF 12 C ENRICHMENT). If the amount of 13C tracer ingested is sufficient to cause a detectable increase in breath 13CO2 enrichment, the measurement of the ratio of 13C:12C can be used to assess the digestion, absorption, and oxidation of the nutrient, food, or other substrate rich in the heavier isotope. The test procedure is relatively similar in all cases. The choice of substrate, dose, enrichment, method of

administration, test meal, exercise, duration of breath sample collection, and expression of results can vary according to the test performed.2 In general, after a fast (overnight in older children; for 3 or 4 h in infants) the subject ingests the 13C-labeled substrate. Collection of Breath Samples Breath samples are obtained before, and at intervals after, administration of the substrate. The choice of breath collection method is determined largely by the age of the subject. For infants and toddlers a soft plastic face mask connected via two one-way valves to a reusable oxygen reservoir bag that has been fitted with a three-way tap can be used. Samples of expired air are aspirated from the bag and stored in evacuated tubes. In young children and adolescents a modified party-blower is an alternative. This contains a low-pressure valve and expired air is aspirated when the inflated blow-out portion is full.1 Older children can exhale directly, via a straw, into an open tube. Unlike the H2 breath test it is not necessary to obtain end-expiratory breath as recovery of 13C is measured as a 13C:12C ratio. Measurement of 13C Enrichment and Expression of Results The 13C enrichment of each sample is measured by isotope ratio mass spectrometry and the results are usually quoted as the relative difference (delta per mL %o) between the sample and an international limestone standard, PeeDee Belemnite (PDB), using a formula described in detail elsewhere,3 where: delta 13C 5 [(Rs/RPDB) 2 1] 3 103, Rs 5 13C/12C in the sample and RPDB 5 13C/12C in PDB: 0.0112372.

Correspondence to: L. T. Weaver, MA, MD, DCH, FRCP, Department of Child Health, University of Glasgow, Yorkhill Hospitals, Glasgow G38SJ, Scotland, UK.

Nutrition 14:826 – 829, 1998 ©Elsevier Science Inc. 1998 Printed in the USA. All rights reserved.

0899-9007/98/$19.00 PII S0899-9007(98)00094-X

STABLE ISOTOPE BREATH TESTS

FIG. 1. Results of positive and negative urea breath test. PDB, PeeDee Belemnite.

827 relative to PDB above baseline at 15 and 30 min after urea ingestion is sufficient to distinguish children with H. pylori from those without, and that it is unnecessary to calculate PDR.5 Such is the discriminatory power of the test that a fixed dose of 50 mg of 13C-urea and a single breath sample taken 30 min after its ingestion can be used to diagnose gastric colonization with H. pylori.6 The 13C-urea breath test is a safe, convenient, easily repeatable method, acceptable to mothers and children, for the rapid detection of H. pylori infection. It is a model of the ideal stable isotope breath test. A simple isotopically labeled substrate (urea) is given by mouth. A single factor controls its digestion (urease) that is indicative of a single infection (H. pylori). Discrimination between the presence and absence of infection is unequivocal (Fig. 1) and the result can be expressed as a clear positive or negative value. However, when stable isotope breath tests are used to measure the digestion of macronutrients, such as fat or complex carbohydrates, several factors that regulate physiologic handling between ingestion of the labeled substrate and recovery of expired CO2 reduce the possibility of achieving such an unequivocal result.1,2 13

C BREATH TESTS TO MEASURE FAT DIGESTION

Results are usually expressed as the cumulative percentage of dose (in this case, 13C) recovered (PDR) over time calculated for each time interval.2 This requires calculation of the product of 13C enrichment in breath CO2 with time and CO2 production rate. The latter can be measured using a metabolic monitor (e.g., Deltatrac, Datex, Helsinki, Finland), and it is important to do so when altered metabolic or respiratory rates are suspected.3 When CO2 production is not measured, a value of 5 mmol z m2 z min21 is generally used in calculations.3 The precision of duplicate analyses of breath samples is generally , 0.1%o SD. Background variability, due to the natural variations in 13C abundance of food stuffs eaten prior to the test, is , 1%o. The coefficient of variation of PDR over 5 h in duplicate tests of fat digestion using the methods that will be outlined in this article can be , 3%. 13

C-UREA BREATH TEST

Helicobacter pylori is probably the most common bacterial infection of mankind. In the developed world it is the principal cause of chronic gastritis and duodenal ulcer disease in adults, but in the developing world it may be associated with the syndrome of malnutrition, diarrhea, and growth failure in childhood.4 In the Gambia, H. pylori appears to be acquired in infancy,5 and we have used the 13C-urea breath test to measure its prevalence in early life. The use of the 13C-urea breath test to detect H. pylori infection is based on the unique capacity of the organism to secrete urease in large quantities. 13C-urea is given by mouth, where it is hydrolyzed in the stomach by urease undergoing the reaction: 13

Urea 1 Water 2. Ammonia 1 Carbon Dioxide 13 H2O CO(NH2)2 CO2 2NH3

In our studies Gambian children received, after an overnight fast, 10 mg/kg 13C-urea (99 atom % excess) in a solution of 10% polycose (to delay gastric emptying) of volume 100 mL. Breath samples were collected, in duplicate, at 15-min intervals to 1 h and at 90 min. The results of a typical positive and negative test are shown in Figure 1. Using cluster analysis it is possible to define a cut-off value to distinguish between children that are H. pylori-positive and -negative.6 We have studied several hundred children using this method, and have found that measurement of the delta 13C

Exocrine Pancreatic Development and Insufficiency Exocrine pancreatic function is far from mature at birth. Lipase activity is detected first at around 30 wk of fetal life, and in the neonate, levels of lipolytic activity are only 10% of those seen in adulthood. Postnatally there is gradual maturation so that by around 2 y of age adult levels of activity have been achieved.7 During infancy, this apparent functional insufficiency is compensated for by other sources of lipase: lingual (from glands in the posterior third of the tongue), gastric (sometimes together with lingual called preduodenal lipase), and human milk.8 In children with exocrine pancreatic insufficiency, such as in cystic fibrosis, steatorrhea is marked, owing to endogenous lipase deficiency, and enzyme supplementation is usually required to achieve normal fat balance. Lipid Substrates Fat absorption may be divided into three phases: lipolysis (lipase dependent), solubilization (bile-salt dependent), and mucosal uptake (dependent on functional integrity of the epithelium). Selection of appropriate substrates help to distinguish between defects of each of these. The specificity of lipolytic enzymes secreted at different sites is such that different substrates may also be chosen to measure the contribution of each, or a combination of each, to fat digestion. In practice, except in infancy, when the contribution of breast milk lipase (BSSL) to digestion is the subject of study, a lipid substrate that is digested by pancreatic and lingual lipase (gastric lipolysis is probably a product of lingual lipase) is chosen.9 Lipid substrates that have been used in 13C breath tests include trioctanoin,10 triolein,10 tripalmitin,10 cholesteryl-octanoate,11 and 1,3-distearyl, 2-octanyl glycerol.12 Trioctanoin absorption is regulated largely by lipolysis and it may be used to distinguish between pancreatic insufficiency and bile-salt deficiency and mucosal defects.10 Triolein absorption depends on all phases listed here, and it may be used to distinguish between subjects with steatorrhea (of any cause) and those with no defect of digestion or absorption. It discriminates poorly between pancreatic and nonpancreatic causes of fat malabsorption.10 Cholesteryl octanoate does not require micellar solubilization for absorption and octanoate is rapidly taken up and oxidized. Because it is not hydrolyzed by lingual and gastric lipases, cholesteryl octanoate may be used to measure the contribution of pancreatic enzymes alone (exclud-

828

STABLE ISOTOPE BREATH TESTS

ing preduodenal lipases) to fat digestion.11 1,3 distearyl 2-carboxyl-octanoyl glycerol (mixed triacylglycerol) measures pancreatic lipase activity.12 The rate limiting step in its digestion is hydrolysis of the 1 and 3 stearyl groups. The 13C label is on the medium-chain fatty acid (octanoic), which is rapidly absorbed and completely oxidized.13 Normal Infants Hoshi et al.14 studied five term and five preterm infants at 3 d of age using 13C-trioctanoin as a substrate. They reported mean PDRs of 53% in term neonates and 46% in the preterm group. McClean et al.,15 using 13C-trioactanoin and 13C-cholesteryl octanoate as substrates to measure “total” intraluminal lipolysis and the contribution of BSSL, reported PDRs of 35% and 26%, respectively, with each substrates, results that were unaffected by preheating the breast milk, suggesting that BSSL did not contribute to the rate of fat digestion. Intravenous infusion of 1-13C-trioctanoin in preterm infants revealed fatty acid oxidation rates of between 29 and 89%, varying in response to the quantity of carbohydrate administered concurrently.16 When octanoate is infused intravenously it appears to be completely oxidized, and in healthy adults comparable rates of oxidation are also achieved when it is administered orally, suggesting complete absorption in later life.13 After oral ingestion of 1-13C-potassium octanoate (free fatty acid rather than triacylglycerol), Sulkers et al.17 reported oxidation of 47% of the substrate in preterm infants. Using the mixed triacylglycerol (MTG) breath test to measure lipase activity in preterm infants of gestational age 27–35 wk and ages 14 –55 d, Van Aalst et al.18 reported a mean PDR of 25%. When we performed MTG breath tests on neonates ages 1–3 d, we recorded a mean PDR of 16% with a range of 0 –32%, which increased to 23% at 7–21 d, and to 29% in late infancy.19 Octanoic acid has been a favored labeled fatty acid to use in substrates to measure dietary fat digestion because it is thought to be efficiently absorbed and completely oxidized,13 although it may be necessary to validate these assumptions in different physiologic and pathologic conditions. These properties have led to its use as a substrate with which to measure gastric emptying rate in infants.20 It appears, however, that although it may be an appropriate substrate to incorporate in the MTG to measure intraluminal lipolysis in adults,12 in early life not only may absorption be incomplete, but also oxidation rates may be less than 100%.17 Children With Cystic Fibrosis Children with cystic fibrosis (CF) have variable degrees of exocrine pancreatic insufficiency that, if untreated, is the main cause of fat malabsorption. Assessment of enzyme supplement requirements can be difficult: 3-d fecal fat collection is cumbersome and lacks sensitivity, the “steatocrit” is not quantitative, and clinical measures (stool frequency and consistency) are subjective. We measured the impact of pancreatic enzyme supplementation on fat digestion using a non-invasive test of intraluminal lipolysis in 41 children with CF, 11 healthy controls, and 5 children with mucosal disease.21 Children ingested the 13C-labeled mixed triacylglycerol in a long-chain triacylglycerol emulsion, and the results were expressed as 13C cumulative PDR over 6 h. The children with CF without pancreatic supplements had a median (range) PDR of 3.1% (0 –31.7), the controls 31% (21.8 – 41.1), and the subjects with mucosal disease 27.8% (19.7–32.5). In 16 children with CF, ingestion of their usual dose of pancreatic enzyme supplements increased PDR to a median of 23.9% (0 – 45.6), and twice the usual dose of enteric-coated microspheres to PDR of 31.1% (11.1– 47.8). There was no significant difference between the median PDR of normal controls and children with mucosal disease, but there was a highly significant difference

FIG 2. Cumulative percentage recovery curves following ingestion of 13 C-MTG in a child with CF without pancreatic enzyme supplementation (f), with regular dose of supplements (M), and with double regular dose (r).

between these groups and children with untreated CF (P , 0.0001). Only 6 children (15%) with CF had a PDR within the normal range. All but 1 of the 5 children with mucosal disease were in the normal range. Thirteen children with CF had no 13C recovery in their breath without enzymes; 11 showed marked rises with regular enzymes. In 8 subjects, doubling the dose of enzymes caused no or minimal improvement. There was considerable intersubject variation in PDR, reflecting a wide degree of pancreatic insufficiency. These findings are comparable to those reported using 13C-trioctanoin,22,23 which also showed overlap in results between healthy children and those with CF. The failure to distinguish unequivocally between normal children and those with depressed pancreatic function should not be seen as a deficiency of the test, but a reflection of the interplay of physiologic maturation and pathologic damage of the pancreas. Patients with CF may be undernourished with altered body composition, energy stores, and metabolism, may hyperventilate, and have bile-salt deficiency contributing to fat malabsorption. All these factors add to the physiologic variables that affect the recovery of 13C.2,24 The MTG breath test offers a simple, non-invasive way of assessing the need for pancreatic enzyme supplementation in children with CF, and could be used to measure optimization of therapy (Fig. 2). CONCLUSIONS

Stable isotope breath tests have great potential as methods for the repeated measurement of fat digestion in infants and children. They are safe25 and do not involve the use of radioactive isotopes, they are non-invasive, and may be used in the clinic, home, laboratory, or ward. However, they require the availability of a mass spectrometer, isotopically labeled substrates may be expensive, and the results obtained from some breath tests are semiquantitative. Nevertheless, the 13C urea breath test has become established as the preferred non-invasive method for the detection of H. pylori infection. To measure fat digestion 13C-lipid breath tests have not yet found a secure place in clinical practice, but remain useful tools for clinical scientists. The selection of appropriate substrates, the standardization of conditions in which tests are undertaken,25 agreement amongst investigators of protocols used (including dose, enrichment, and mode of administration of substrate), methods of analysis, and expression of results will improve their potential for diagnosis and for monitoring therapy. Work remains to be done to define the full

STABLE ISOTOPE BREATH TESTS

829

rate of ingested lipid substrates: what proportion of the label is retained in the bowel26 and/or within slowly turning over body pools. Collaborative studies between research groups27 aimed at answering these questions and defining the sensitivity and specificity of 13C breath tests will lead to their increasing use for physiologic investigation and clinical diagnosis. It is hoped that before long the use of stable isotope breath tests in some areas of clinical medicine will be as common, and even replace that of more direct and invasive procedures and radionuclides.

ACKNOWLEDGMENTS

I am grateful for the help, collaboration, and advice of Julian Thomas, Sergio Amarri, Andrew Coward, Marilyn Harding, Patricia McClean, William Manson, Thomas Preston, and Simon Ling. Some of the work described here was supported by the Thrasher Research Fund, Medical Research Council, Bristol Myers Research Foundation, EC (Biomed Programme) and British Digestive Foundation.

REFERENCES 1. Weaver LT, Thomas JE, McLean P, Harding M, Coward WA. Stable isotope breath tests: their use in paediatric practice. In: Janssens J, ed. Progress in understanding and management of gastrointestinal motility disorders. Leuven: University of Leuven Department of Gastroenterology, 1993;155 2. Amarri S, Weaver LT. 13C breath tests to measure fat and carbohydrate digestion in clinical practice. Clin Nutr 1995;14:149 3. Amarri S, Coward WA, Harding M, Weaver LT. Importance of measuring CO2 production rate in 13C breath tests. Proc Nutr Soc 1995;54:111A 4. Sullivan PB, Thomas JE, Wight DGW, et al. Helicobacter pylori infection in Gambian children with chronic diarrhoea and malnutrition. Arch Dis Child 1990;65:189 5. Weaver LT. Helicobacter pylori infection, nutrition and growth of West African infants. Trans Roy Soc Trop Med Hyg 1995;89:347 6. Mion F, Rosner G, Rousseau M, Minaire Y. The 13C urea breath test for Helicobacter pylori: cut-off point determination by cluster analysis. Clin Sci 1997;93:3 7. McClean P, Weaver LT. Ontogeny of human pancreatic exocrine function. Arch Dis Child 1993;68:62 8. Manson WG, Weaver LT. Fat digestion in the neonate. Arch Dis Child 1997;76:F206 9. Weaver LT, Manson WG, Amarri S. Measuring fat digestion in early life using stable isotope breath tests. Prenat Neonat Med 1997;2:116 10. Watkins JB, Klein PD, Schoeller D, Kirschner BS, Park R, Perman JA. Diagnosis and differentiation of fat malabsorption in children using 13C-labeled lipids: trioctanoin, triolein, and palmitic acid breath tests. Gastroenterology 1982;82:911 11. Cole SG, Rossi S, Stern A, Hofmann AF. Cholesteryl octanoate breath test. Preliminary studies of a new non-invasive test of human pancreatic exocrine function. Gastroenterology 1987;93:1372 12. Vantrappen GR, Rutgeerts PJ, Ghoos YF, Hiele MI. Mixed triglyceride breath test: a noninvasive test of pancreatic lipase activity in the duodenum. Gastroenterology 1989;96:1126 13. Schwarbe AD, Bennett LR, Bowman LP. Octanoic acid absorption and oxidation in humans. J Appl Physiol 1964;19:335 14. Hoshi J, Nishida H, Yasui M, Ohishi M, Takahashi M. (13C) breath test of medium-chain triglycerides and oligosaccharides in neonates. Acta Paediatr Jpn 1992;34:674

15. McClean P, Harding M, Coward WA, Prentice A, Austin S, Weaver LT. Bile salt-stimulated lipase and digestion of non-breast milk fat. J Pediatr Gastroenterol Nutr 1998;26:39 16. Paust H, Keles T, Park W, Knoblach G. Fatty acid metabolism in infants. In: Chapman TE, Berger R, Reyngoud DJ, Okken A, eds. Stable isotopes in paediatric nutritional and metabolic research. Andover: Intercept Ltd, 1991:1 17. Sulkers EJ, Lafeber HN, Sauer PJJ. Quantitation of oxidation of medium-chain triglycerides in preterm infants. Pediatr Res 1989;26: 294 18. Van Aalst K, Veerman-Wauters G, van der Schoor S, Ghoos YF, Derlieger H, Eggermont E. The 13C mixed triglyceride breath test for assessment of lipase activity in preterm infants. J Pediatr Gastroenterol Nutr 1995;20:459 19. Manson WG, Dale E, Harding M, Coward WA, Weaver LT. Functional capacity of the newborn to digest dietary fats. J Pediatr Gastroenterol Nutr 1996;22:427 20. Ghoos YF, Maes BD, Geypens BJ, et al. Measurement of gastric emptying rate of solids by means of a carbon labelled octanoic acid breath test. Gastroenterology 1993;104:164 21. Amarri S, Harding M, Coward WA, Evans TJ, Weaver LT. 13C mixed triglyceride breath test and pancreatic enzyme supplementation in children with cystic fibrosis. Arch Dis Child 1997;76:349 22. McClean P, Harding M, Coward WA, Green MR, Weaver LT. Measurement of fat digestion in early life using a stable isotope breath test. Arch Dis Child 1993;69:366 23. Murphy MS, Eastham EJ, Nelson R, Aynsley-Green A. Non-invasive assessment of intraluminal lipolysis using a 13CO2 breath test. Arch Dis Child 1990;65:574 24. Kalivianakis M, Verkade HJ, Stellaard F, Van der Werf M, Elzinga H, Vonk RJ. The 13C-mixed triglyceride breath test in healthy adults: determinants of the 13CO2 response. Europ J Clin Invest 1997;27:434 25. Jones PJH, Leatherdale ST. Stable isotopes in clinical research: safety reaffirmed. Clin Sci 1991;80:277 26. Ling SC, Weaver LT. The fate of fat in the infant’s colon. Q J Med 1997;90:553 27. Weaver LT, Amarri S, Swart GR. 13C mixed triglyceride breath test. Gut 1998; in press