Tricks for Interpreting and Making a Good Report on Hydrogen and 13C Breath Tests

Home , Hydrogen breath test, Urea breath test

European Review for Medical and Pharmacological Sciences 2013; 17(Suppl 2): -90-98 Tricks for interpreting and making a good report on hydrogen and 13C breath tests

G. D’ANGELO, T.A. DI RIENZO, F. SCALDAFERRI, F. DEL ZOMPO, M. PIZZOFERRATO, L.R. LOPETUSO, L. LATERZA, G. BRUNO, V. PETITO, M.C. CAMPANALE, V. CESARIO, F. FRANCESCHI, G. CAMMAROTA, E. GAETANI, A. GASBARRINI, V. OJETTI

Department of Internal Medicine, Gastroenterology Division, School of Medicine Catholic University of the Sacred Hearth, Rome, Italy

Abstract. Breath tests (BT) represent a valid Introduction and non-invasive diagnostic tool in many gastroenterological disorders. Their wide diffusion Breath tests (BT) represent a valid and non-in- is due to the low cost, simplicity and reproducibility and their common indications include vasive diagnostic tool in many gastroenterological diagnosis of carbohydrate malabsorption, Heli- disorders. Hydrogen-breath tests are based on the cobacter pylori infection, small bowel bacterial fact that gases produced by colonic bacterial fer- overgrowth, gastric emptying time and orocae- mentation diffuses into the blood and are excreted cal transit time. in breath samples, where it can be easily quanti- The review deals with key points on method- fied. There is a growing interest in breath tests but ology, which would influence the correct inter- many differences persist in the methodology used. pretation of the test and on a correct report. While a clear guideline is available for lac- Hydrogen Breath test (HBT) has become an tose and glucose breath tests, no gold stan- important diagnostic tool in many Gastroentero- dard is available for Sorbitol, Fructose or other logical conditions, characterized by abdominal H2 BTs. pain, discomfort, bloating etc. Its wide diffusion Orocaecal transit time (OCTT) defined as is due to the low cost, simplicity and repro- time between assumption of 10 g lactulose and ducibility: it represents a valid and non-invasive a peak > 10 ppm over the baseline value, is a well-defined breath test. The possible value of test to diagnose carbohydrate malabsorption, lactulose as a diagnostic test for the diagnosis small bowel bacterial overgrowth, and for the de- of small bowel bacterial overgrowth is still un- termination of orocaecal transit time. der debate. 13C-breath tests use 13C-labelled substrates in Among 13C breath test, the best and well order to perform a safe non-invasive evaluation of characterized is represented by the urea breath several metabolic pathways. The turnover of the test. Well-defined protocols are available also 13 13 C-substrate throughout the specific explored for other C tests, although a reimbursement 13 for these tests is still not available. pathway is assessed by measuring CO2 in exhaled 13 Critical points in breath testing include the air. The availability of plenty of C-substrates patient preparation for test, type of substrate gives these tests a wide range of potential applica- utilized, reading machines, time between when tions, although the routine clinical use of certain the test is performed and when the test is substrates may be sometimes affected by their high processed. Another crucial point involves clini- costs. The most used 13C breath test is the one for cal conclusions coming from each test. For ex- Helicobacter pylori ample, even if lactulose could be utilized for di- diagnosing infection. agnosing small bowel bacterial overgrowth, In this paper we summarize key element on 13 this indication should be only secondary to assessing and interpreting H2 and major C orocaecal transit time, and added into notes, as breath tests, and we will provide a guide for an clinical guidelines are still uncertain. accurate breath test report.

Key words: Tricks on hydrogen and methane excretion Hydrogen isn’t produced in healthy humans Breath test report, H2 breath test, C13 breath test. fasting and at rest, but it is only generated during

90 Corresponding Author: Veronica Ojetti MD, Ph.D; e-mail: [email protected] Tricks for interpreting and making a good report on hydrogen and 13C breath tests

anaerobic metabolism, which occurs, for example, Tricks on sampling H2 from alveolar during bacterial fermentation. Anaerobic bacteria expirates for H2 breath testing metabolize sugar molecules, producing CO2, We can measure hydrogen concentrations in short-chain fatty acid (SCFA) and Hydrogen1. Un- end-expiratory breath samples using gas chro- absorbed carbohydrates are so fermented by matography or electrochemical cells. Stationary colonic bacteria and the hydrogen, like other gases dedicated gas chromatographs represent the gold produced within the intestine, is absorbed through standard for hydrogen determinations in breath, as the intestinal wall, passes in the blood circulation, they were previously validated in comparison with reaches the lungs, and finally it is released and ex- non-dedicated instruments in terms of linearity and haled in the breath samples. The amount of hydro- reproducibility of results10-12. Stationary dedicated gen produced by carbohydrate fermentation varies gas-chromatographs could be made with solid state according to the individual’s microbiota, as hydro- sensors, and they represent the gold standard for gen is also used by bacteria for other metabolic hydrogen determination, or with electrochemical processes, including conversion to methane. sensor, which also show a good accuracy, although In fasting state, normal hydrogen baseline val- they will undergo to a natural “drift” over time. ues are reported to be 7 ± 5 ppm2, if it results too Gas measurement must be performed in alveolar high (>10-16 ppm), breath test shouldn’t be per- air, avoiding the interference represented by respira- formed3. The ingestion of rice flour and meat is tory dead space air. The three systems are available not associated with a detectable increase in breath to collect alveolar air: the modified Haldan-Priestley hydrogen excretion, so they are eaten the evening tube, the Y-Piece device and the two bag-system13. before the test4,5. In response to the test, some peo- The most efficient method to remove dead space air ple produce huge amount of hydrogen in response and guarantee complete respiratory exchanges is to to sugars, some produce lower amounts, and about inhale maximally, hold the inhalation for 15 s and 15% produce apparently none (designated ‘non- expire into the bag14. It could be also considered as a hydrogen producers’), as the expired hydrogen marker of correct sampling the concentration of throughout the test dose not reach the limit detec- CO2, which levels in alveolar air are stable around tion rate. In some case of apparent hydrogen non 5-5.5%:, it is possible to reduce variability due to the production, methane determination could repre- death space, measuring CO2 in breath sample and 15 sent a valid alternative, although insufficient data correcting for the ideal value of CO2 . are presented in this topic. In patients that do not Breath samples are currently stored in plastic produce methane neither hydrogen (a condition syringes, an inexpensive method allowing the present in about 5% of the population), there is no analysis of gases with no further handling. Be- indication of performing breath tests6 and no indi- cause of the high diffusion capacity of hydrogen, cation exists on whether increase test time would samples should be analyzed as soon as possible, exert any role. This condition is mainly due to a avoiding storage longer than 12 h; however, sim- different intestinal flora composition, which in ple refrigeration of plastic syringes is sufficient turn is influenced by many factors including age, to ensure the stability of hydrogen concentrations susceptibility to infections, dietary habits, im- for a long time. At room temperature, after 5 munological factors, intraluminal pH, interaction days, the hydrogen concentration is reduced up among components of intestinal flora and fer- to 30%, while at –20 °C the reduction is equal to mentable substrates, recent antibiotic therapy, use 5% and only 7% after 15 days. Moreover, at –20 of laxatives (preparation for colonoscopy), or for °C no hydrogen loss is detectable for 2 days13. the predominance of methane producing bacteria Hydrogen concentration in breath samples must, in the colon7-9. Some groups try to overcome this therefore, be determined within 6 h of sampling. problem by testing initially lactulose: this synthet- ic disaccharide cannot be absorbed in humans and Tricks on Lactose breath test is, therefore, always fermented producing hydro- interpretation gen, except for the “non-H2-producer”. It is, there- Lactose breath test shows good sensitivity fore, conceivable that a slow gastrointestinal tran- (mean value of 77.5%) and excellent specificity sit of the substrate may be responsible for false (mean value of 97.6%) for lactose intolerance in negative results. It was shown that the prolonga- patients which display or not symptoms16, 17. False tion of measurement from 2 to 4 hours induce a negative tests, however, can be due to “hydrogen significant reduction of the prevalence of “H2 non non-producers”; false positive breath tests are less producer”8. frequent and are mainly secondary to of small

91 G. D’Angelo, T.A. Di Rienzo, F. Scaldaferri, F. Del Zompo et al bowel bacterial overgrowth. LBT was considered colon, so it is a suitable substrate to detect proxi- positive when the peak of hydrogen was 20 ppm mal small bowel overgrowth. A rise in H2, after over the baseline. In children some laboratories the assumption of the substrate, means that glu- have used > 10 ppm as cutoff value, but a rise of > cose meets bacteria in the small bowel, before its 20 ppm over baseline correlates better with symp- absorption. Because of its early absorption, GBT tom development18. may not able to diagnose SIBO of the distal If the value rises more than 10 ppm but less small intestine (ileum). Sensitivity and specifici- than 20 ppm above the base value, the test could ty are 62.5% and 77.8 %, respectively compared be considered as “borderline”19. In this condition, to the gold standard (jejunal culture)27. it must be verified when this peak occurs: if the Lactulose has also been used for diagnosis SI- hydrogen value seems to increase at the end of BO. Many criteria have been established to try to the test, should be considered the possibility of a detect SIBO using Lactulose: some authors sug- prolonged test, to augment the sensibility, reduc- gested that a rise in breath H2 of 20 ppm above ing “false negative” due to a possible slow in- base levels within 90 minutes after ingestion of testinal transit time. lactulose should be considered as diagnostic of SI- BO26, but this criterion has not been validated. Ac- Tricks on other carbohydrates tually the most diffuse criterion is the presence of breath test interpretation a hydrogen “double peak”, the earlier representing Sorbitol is a sugar alcohol widespread in presence of bacteria fermenting in the small bow- plants, particularly in fruits and juices, which is el, the second and bigger one linked with the flora only partially absorbed. Sorbitol BT is effective in usually colonizing the colon. However, it is diffi- detecting small bowel damage, especially in celiac cult to distinguish between an early H2 peak disease20. Tursi et al published many papers on caused by SIBO and a fast transit; thus, the non- Sorbitol Breath test (5 g) for the diagnosis of sub- standardized criteria, and lower diagnostic accura- clinical or silent celiac disease, and for the asses cy (55 versus 71 of Glucose14) makes lactulose histological recovery after gluten-free diet21,22. breath test not indicated to diagnostic for SIBO. Fructose, widely used as a sweetener in different foods, is a monosaccharide, which is ab- Tricks on orocaecal transit time sorbed by a specific receptor for fructose (GlUT5 interpretation fructose transporter mediated facilitative diffu- Lactulose is a manufactured disaccharide, sion), which occurs along the entire small bowel consisting of galactose and fructose, which is not and it is slow; and by a second receptor which is metabolized by the human organism, producing rapid and efficient, but occurs only together with always fermentation in the colon. Thus, the time glucose6. The mean daily fructose intake is about from ingestion of lactulose to the hydrogen in- 16 g per day, but eating enriched food and drinks crease in breath samples reflects the orocaecal can achieve up to 100 g per day23,24. transit time. An increase >20 ppm compared to The clinical importance of sorbitol and fruc- baseline is considered significant in diagnostic tose malabsorption remains to be evaluated, so routine28. The orocaecal transit time in healthy BT should not be recommended in clinical prac- subjects ranges between 40 and 170 min for a tice in both adults and children, according to cur- lactulose meal (192-232 min for a solid meal). A rent guidelines14. delayed orocaecal transit has been reported in constipation, alcoholism, depression, diabetes Tricks on small intestinal bacterial mellitus, IBS, pregnancy, obesity, cholecystecto- overgrowth interpretation my, cirrhosis, scleroderma, acromegaly and dys- Small intestinal bacterial overgrowth (SIBO) pepsia. Meanwhile, a fast transit in partial gas- is usually defined as a growth of >105 colony trectomy, post vagotomy diarrhea, hyperthy- forming units (CFU)/ml in fluid aspirated from roidism, IBS, alcoholics. LHBT is also useful to the jejunum14, which could be associated to ab- demonstrate drug effects on orocaecal transit. dominal pain, bloating, diarrhea and/or signs of malabsorption, similar to those observed in irrita- Tricks for sampling and ble bowel syndrome (IBS)14,25,26. Glucose Breath measuring 13C/12C ratio test (GBT) is currently used for the diagnosis of 13C stable isotope is used in breath testing to SIBO. Glucose is rapidly absorbed in the proxi- label specific substrates. The 13C-labelled tracer mal small bowel and usually does not reach the probe, once ingested, is cleaved in the gastroin-

92 Tricks for interpreting and making a good report on hydrogen and 13C breath tests testinal lumen, transported and degraded follow- performance, but also the diffusion of 13C- ing the same metabolic pathways of its native breath tests in primary care settings. Generally, 13 isoform. The final result is the excretion of CO2 there is a good agreement between mass spec- through the alveolocapillary membrane into the troscopy and infrared based technology, for 13 13 12 exhaled air, in which CO2 amount is measured. C/ C ratio estimate, particularly for Urea Since 13C stable isotope is commonly found in breath test30,31. The agreement of infrared tech- nature – accounting for 1.11% of all carbon nologies on dynamic 13C breath tests is not well atoms – and its amount may vary among individ- addressed and some concerns remains on repro- uals, it is necessary to refer results to a baseline ducibility and accuracy. value in order to get accurate results. Thus, total

CO2 production at baseline must be measured or Tricks on Helicobacter pylori breath test estimated in each patient. Usually it is calculated Helicobacter pylori (H. pylori) is a common 2 29 with 300 mmol CO2/m body surface area/h . Gram-negative pathogen bacterium colonizing Moreover, factors influencing the endogenous the gastric mucosal environment in about 50% 13 CO2 production (e.g. food ingestion, physical ac- individuals. C-urea breath test [UBT] is consid- tivity, respiratory diseases, thyroid dysfunctions, ered the gold standard for non-invasive diagnosis fever) should be considered. of H. pylori infection, because of its accuracy Breath samples are collected at baseline and and its easy execution32. Urea hydrolysis can be for a variable number of times and at different generally entirely accounted to the presence of time intervals depending on the type of 13C H. pylori, since it is the most common gastric breath test performed. pathogen expressing urease33. Several studies re- The tracer could be detected and expressed as ported the high diagnostic reliability of UBT, that a static variable, e.g. the delta over baseline reaches values >95% both for sensitivity and δ δ 34-36 [DOB: sample – basal value before tracer application] in urea specificity prior and after eradication therapy . breath test, or as a dynamic one, in tests where 13C-UBT protocol underwent multiple modi- 13C/12C is expressed as recovery rate per hour or fications through years. Nowadays, 75 mg 13C- percent dose recovery and more than 2 dosages urea in 200 ml water with the adjunction of 1-4 are necessary and special algorithms are used. g citric acid37 represents the most commonly Moreover, in order to get accurate results, mea- accepted protocol. Adding citric acid enhances surement or estimation of basal 13C value is rec- H. pylori urea hydrolisation by both increasing ommended. 13C recovery is never complete as a intra-citoplasmic urea availability and enhanc- substantial amount of tracer is retained in the car- ing urease activity38. Hence it allows to spare bon pool of the body. Thus, 13C breath test analy- urea and to contain costs. Breath samples are sis is a semi-quantitative diagnostic tool. collected through a straw in a glass phial prior Breath samples are analyzed by high-resolu- to the substrate administration and after 20 or tion mass spectrometers, which are able to mea- 30 minutes. sure the slight mass difference of one neutron The coexistence of some factors might alter between 13C-labelled carbon dioxide and the UBT results affecting its accuracy. In particular, naturally most common carbon dioxide with the the assumption of proton pump inhibitors re- carbon isotope 12C, on the base of different light duces UBT sensitivity in 31%39 since it reduces absorbance between the two isotopes. The pre- the bacterial load especially in the antrum32,40,41. cision of high-resolution mass spectrometers is False-positive tests may be provided by ure- very high, so very low isotopic quantity can be ase expressed by oropharyngeal flora. Mouth detected, permitting the use of small samples of washing or administering the substrate with a exhaled air. On the other hand, mass spectrome- straw or in pills may be helpful strategies to ters are highly expensive, they have relatively avoid this disadvantage. long analysis time and they require well-trained Although performing the test in a non-fasting personnel for their use. The introduction of in- condition has been associated with a higher rate frared technologies (non-dispersive isotope se- of both false-negative and false-positive results42, lected infrared spectrometers, NDIRS), based several other studies found only small differ- on the use of photoacustic detectors, made this ences between DOB values in fasting and non- goal cheaper with adequate accuracy. Operating fasting status43,44 or even no differences45. How- and handling of NDIRS is easy, even for non- ever, it seems prudent to perform the test in a experienced users. This enables not only the fasting status until this issue is better clarified.

93 G. D’Angelo, T.A. Di Rienzo, F. Scaldaferri, F. Del Zompo et al

The choice of a reliable cut-off to define the rect technique for the study of gastric emptying. positivity of the test is another incompletely ad- Half-emptying gastric time [t½] indicates the time dressed question. A unique DOB cut-off is not required to dismiss half of the gastric content; definable because it has to be adapted to different gastric emptying coefficient [GEC] is an index of factors, such as the test meal, doses and types of the gastric emptying rate; lag phase [tlag] is the urea, or the pre-/post-treatment setting in which time occurring for the solid meal to be triturated the test is employed40. Although, because posi- into <2 mm3 fragments46. tive and negative UBT results tend to cluster out- Test protocols vary depending on the test side of the DOB range between 1.8 and 5‰, a meal used. The main variations regard the formu- change in cut-off value within this range would lation of test meals, caloric burden, time intervals be expected to have little effect on clinical accu- between sample collections and the total duration racy of the test41. of sampling. There is a reasonable agreement on a 4-6 hours-long sampling at time intervals of 15 Tricks on 13C octanoic acid breath or 30 minutes47-50. Regarding the test meal, the for studying gastric emptying time standardisation of the test meals is required but Gastric emptying is a complex process requir- to date it is still lacking. To this purpose, Perri et ing variable durations, depending on quantity and al. proposed a new standardised test meal in form quality of food ingested, and on the presence of of a gluten-, lactose-, glucose-free muffin51. certain pathologies. In a clinical setting, a gastric 13C-acetate is used as marker in liquid phase emptying study might be particularly useful to as- studies because of the hydrophilic character of the sess how gastric physiology varies depending on molecule52 and has showed an excellent correlation age and diet, to study functional dyspepsia, dia- toward scintigraphy53. Liquid phase studies are betic gastroparesis and to evaluate the effect of particularly useful for research purposes, whilst in prokynetic therapy. Radioscintigraphy represents a clinical setting only certain neurological patholo- the gold standard for gastric emptying studies. gies affect gastric emptying for the liquids. Breath testing for gastric emptying assessing Breath tests can be easily performed by chil- is justified by the assumption that the time limit- dren, pregnant women and critically ill patients. 13 ing step for CO2 to appear in the exhalation is Their analysis does not require expert personnel represented by the passage of the test meal into operating in the same place where they are per- duodenum, whilst the subsequent processes oc- formed, thanks to the easy transportability of the cur negligibly fast or at a constant rate29. samples. Although, breath-test and scintigraphy 13C-octanoate and 13C-acetate are the most results are not identical46,54. Different results used substrate in breath testing, respectively used originate from different approaches in describing to label solid and liquid components of test the gastric emptying. Radioactive tests study the meals. progressive tracer disappearing from the stomach In the original description by Ghoos et al, 13C- and are affected by tissue attenuation, angle of octanoic acid is mixed together with egg yolk detection, radiation decay. 13C-Breath-tests study and baked. The frying process anchors the mole- absorption of 13C-labelled molecules in the small cule in the solid phase of the egg yolk46. After its bowel and are delayed pulmonary excretion and administration, when the meal has passed partial retention of the tracer amount in the body through the pylorus, the 13C-octanoic acid is and is potentially influenced by a delayed pul- rapidly absorbed in the duodenum and jejunum monary excretion and partial retention of the and is oxidised to 13C-labelled carbon dioxide, tracer amount in the body29. then eliminated by lungs. Collecting breath sam- On the other hand, simultaneous labelling of 13 ples for more than 4 h gives reproducible CO2- solid and liquid phases, although could be done kinetics in breath, which reflects gastric empty- in nuclear medicine, is possible in breath-testing 13 14 55 ing; thus the appearance of CO2 in breathed air only with the addition of C , whose clinical use reflects the transit of the test meal from the stom- is obsolete because of its radioactivity. ach through the duodenum, where this is digested and absorbed with the final production of 13C-la- Tricks on 13C mixed triglyceride acid belled carbon dioxide. breath exploring exocrine pancreatic Several parameters are calculated to evaluate function gastric emptying and this issue, requiring mathe- Except for secretin-pancreozymin test and matical formulae, makes breath testing an indi- secretin-enhanced magnetic resonance pancre-

94 Tricks for interpreting and making a good report on hydrogen and 13C breath tests atography [s-MRP] that are both able to reach strates, the high time expenditure and the lack of good levels of sensitivity and specificity, func- standardization limit the utilization and the diffu- tional studying for the diagnosis of chronic pan- sion in common clinical settings. creatitis plays a minor role today56. Neverthe- less, their high costs and invasiveness, makes Drawing a good and complete report useful to develop cheaper tests with similar ac- The medical report is necessary to reduce curacy. variability between tests performed in different Several substrates have been studied for as- centers and to give the opportunity to experi- sessing exocrine pancreatic disorders, e.g. 13C- enced gastroenterologist to fully interpret results, starch57, 13C-egg white58-60 and 13C-peptides61 particularly in the case of tests not fully standard- demonstrated low sensitivity because of the ized. non-exclusivity of exocrine pancreas to diges- It must include all the methodologies and tion processes of carbohydrates and proteins. clinical characteristics to let the test interpretable Thus, the most suitable breath tests for the also by non-specialists. study of exocrine pancreatic disorders seem to In addition to patient data, it should be speci- be those exploring fat maldigestion. This kind fied the substrate administrated, in term of of studies reach sufficiently good accuracy dosage and concentration, the methods of sample even in the study of early or mild disease main- collecting (intervals and duration of the test) and ly because lipase is the first enzyme reduced machine used for the measurement. during pancreatic insufficiency; gastric lipase It is good to indicate some clinical notes, like participates only for 15% to intestinal lipolysis; for instance the clinical indications for requiring moreover, gastric lipase is not able to adequate- the test (ie suspect lactose intolerance, diarrhea, ly compensate the reduced pancreatic lipase ac- constipation, bloating, etc.). tivity62. If necessary, on the basis of the results, gas- The best investigated triglyceride breath test troenterologist or family doctors could require uses 1,3-distearyl, 2-[13C-carboxyl] octanoyl further exams, to complete the diagnostic glycerol and is known as 13C-mixed triglyceride process. breath test63,64 administered at the doses of 200- In Figure 1 we summarized in a virtual format 300 mg. Breath samples are collected before all informations, which should be clearly written the test meal and for the subsequent 5 hours at in a good report. 65 time intervals of 30 minutes . Labeled triglyc- We think that, dealing with H2 breath testing, erides are metabolized in the duodenum under the key points for the validity of the test would physiological conditions; they are easily ab- be the distance between the day of the breath sorbed by mucosal wall and metabolized in the sampling and the day of breath assessment. Type liver cells with production of 13C labeled car- of machines utilized, status of the machines, be- bon dioxide, excreted by lungs. Low 13C/12C ra- ing the “age” of the machine, a limiting step par- tio in expired air indicate a not completed di- ticularly for gas-chromatography with electro- gestion of marked substrates due to impaired chemical sensor. pancreatic secretion. Duration of the test, contemporary measure-

Sensitivity and specificity of the mixed ment of CO2 and/or methane is also another im- triglyceride breath test compared to the stimulat- portant parameter. Even for 13C breath test, a ed lipase output in the pancreocymin test are clear indication of protocol used for the test is 89% and 81%, respectively64-66. crucially important, particularly for non conven-

It is important that the patient follows certain tional breath tests. Despite H2 breath testing, indications prior to the test: food naturally rich in samples for 13C breath testing, when kept in 13C (i.e. corn, pineapple, cane sugar) should be proper tubes, are more stable, being 13C a stable avoided at least from 2 days before the test; a 12 isotope. The indication of the machine utilized hours fasting is recommended, with the excep- could be of interest, as mass spectroscopy has the tion of natural water. Moreover, smoking, sleep- highest accuracy and, usually, highest precision. ing or doing physical exercise should be avoided from 30 minutes before the test through the en- –––––––––––––––––––-–– tire duration of the test. Conflict of interest Even if breath tests are valid methods for The Authors declare that they have no conflict of inter- evaluate pancreatic function, the costs of the sub- ests.

95 G. D’Angelo, T.A. Di Rienzo, F. Scaldaferri, F. Del Zompo et al

Figure 1. Schematic representation of a complete breath test report

References 6) BARRETT JS, GIBSON PR. Fructose and lactose testing. Aust Fam Physician 2012; 5: 293-296. 1) LEVITT MD. Volume and composition of human intesti- 7) Hao WL, Lee YK. Microflora of the gastrointestinal nal gas determined by means of an intestinal tract: a review. Methods Mol Biol 2004; 268: 491-502. washout technic. N Engl J Med 1971; 25: 1394-1398. 8) CORAZZA G, STROCCHI A, SORGE M, BENTAI G, GASBAR- 2) PERMAN JA, MODLER S, BARR RG, ROSENTHAL P. Fast- RINI G. Prevalence and consistency of low breath ing breath hydrogen concentration: normal values H2 excretion following lactulose ingestion. Possi- and clinical application. Gastroenterology 1984; ble implications for the clinical use of the H2 6: 1358-1363. breath test. Dig Dis Sci 1993; 11: 2010-2016. 3) GHOSHAL UC. How to interpret hydrogen breath 9) VERNIA P, C AMILLO MD, MARINARO V, C APRILLI R. Effect of tests. J Neurogastroenterol Motil 2011; 3: 312-317. predominant methanogenic flora on the outcome of 4) ANDERSON IH, LEVINE AS, LEVITT MD. Incomplete ab- lactose breath test in irritable bowel syndrome pa- sorption of the carbohydrate in all-purpose wheat tients. Eur J Clin Nutr 2003; 9: 1116-1119. flour. N Engl J Med 1981; 15: 891-892. 10) SOLOMONS NW, VITERI FE, HAMILTON LH. Application 5) LEVITT MD, HIRSH P, F ETZER CA, SHEAHAN M, LEVINE of a simple gas chromatographic technique for AS. H2 excretion after ingestion of complex car- measuring breath hydrogen. J Lab Clin Med bohydrates. Gastroenterology 1987; 2: 383-389. 1977; 5: 856-862.

96 Tricks for interpreting and making a good report on hydrogen and 13C breath tests

11) CHRISTMAN NT, HAMILTON LH. A new chromato- 25) SINGH VV, TOSKES PP. Small bowel bacterial over- graphic instrument for measuring trace concen- growth: presentation, diagnosis, and treatment. trations of breath-hydrogen. J Chromatogr 1982; Curr Treat Options Gastroenterol 2004; 1: 19-28. 2: 259-265. 26) PIMENTEL M, CHOW EJ, LIN HC. Eradication of small 12) SOLOMONS NW, HAMILTON LH, CHRISTMAN NT, ROTH- intestinal bacterial overgrowth reduces symptoms MAN D. Evaluation of a rapid breath hydrogen an- of irritable bowel syndrome. Am J Gastroenterol alyzer for clinical studies of carbohydrate absorp- 2000; 12: 3503-3506. tion. Dig Dis Sci 1983; 5: 397-404. 27) CORAZZA GR, MENOZZI MG, STROCCHI A, RASCITI L, 13) CORAZZA GR, SORGE M, MAURINO E, STROCCHI A, LAT- VAIRA D, LECCHINI R, AVANZINI P, C HEZZI C, GASBARRINI TANZI MC, GASBARRINI G. Methodology of the H2 G. The diagnosis of small bowel bacterial over- breath test. I. Collection and storage for gas mea- growth. Reliability of jejunal culture and inade- surement. Ital J Gastroenterol 1990; 4: 200-204. quacy of breath hydrogen testing. Gastroenterol- 14) GASBARRINI A, CORAZZA GR, GASBARRINI G, MONTALTO ogy 1990; 2: 302-309. M, DI STEFANO M, BASILISCO G, PARODI A, USAI-SATTA P, 28) WILBERG S, PIERAMICO O, MALFERTHEINER P. [The H2-lac- VERNIA P, A NANIA C, ASTEGIANO M, BARBARA G, BENINI tulose breath test in the diagnosis of intestinal tran- L, BONAZZI P, C APURSO G, CERTO M, COLECCHIA A, sit time]. Leber Magen Darm 1990; 3: 129-137. CUOCO L, DI SARIO A, FESTI D, LAURITANO C, MICELI E, 29) BRADEN B. Methods and functions: Breath tests. Best NARDONE G, PERRI F, P ORTINCASA P, R ISICATO R, SORGE Pract Res Clin Gastroenterol 2009; 3: 337-352. M, TURSI A, ST ROME HBTCCWG. Methodology and 30) BRADEN B, HAISCH M, DUAN LP, LEMBCKE B, CASPARY indications of H2-breath testing in gastrointestinal WF, HERING P. Clinically feasible stable isotope diseases: the Rome Consensus Conference. Ali- technique at a reasonable price: analysis of ment Pharmacol Ther 2009; 29: 1-49. 13CO2/12CO2-abundance in breath samples with 15) NIU HC, SCHOELLER DA, KLEIN PD. Improved gas a new isotope selective-nondispersive infrared chromatographic quantitation of breath hydrogen spectrometer. Z Gastroenterol 1994; 12: 675-678. by normalization to respiratory carbon dioxide. J 31) SAVARINO V, M ELA GS, ZENTILIN P, B ISSO G, PIVARI M, Lab Clin Med 1979; 5: 755-763. MANSI C, MELE MR, BILARDI C, VIGNERI S, CELLE G. 16) OJETTI V, G ABRIELLI M, MIGNECO A, LAURITANO C, ZOC- Comparison of isotope ratio mass spectrometry CO MA, SCARPELLINI E, NISTA EC, GASBARRINI G, GAS- and nondispersive isotope-selective infrared BARRINI A. Regression of lactose malabsorption in spectroscopy for 13C-urea breath test. Am J Gas- coeliac patients after receiving a gluten-free diet. troenterol 1999; 5: 1203-1208. Scand J Gastroenterol 2008; 2: 174-177. 32) MALFERTHEINER P, M EGRAUD F, O ’ M ORAIN CA, ATHER- 17) METZ G, JENKINS DJ, PETERS TJ, NEWMAN A, BLENDIS TON J, AXON AT, BAZZOLI F, G ENSINI GF, GISBERT JP, LM. Breath hydrogen as a diagnostic method for GRAHAM DY, ROKKAS T, E L-OMAR EM, KUIPERS EJ, EU- hypolactasia. Lancet 1975; 7917: 1155-1157. ROPEAN HELICOBACTER STUDY G. Management of Heli- 18) VELIGATI LN, TREEM WR, SULLIVAN B, BURKE G, HYAMS cobacter pylori infection—the Maastricht IV/ Flo- JS. Delta 10 ppm versus delta 20 ppm: a reap- rence Consensus Report. Gut 2012; 5: 646-664. praisal of diagnostic criteria for breath hydrogen 33) PERRI F, F ESTA V, C LEMENTE R, QUITADAMO M, ANDRIULLI testing in children. Am J Gastroenterol 1994; 5: A. Methodological problems and pitfalls of urea 758-761. breath test. Ital J Gastroenterol Hepatol 1998; 19) EISENMANN A, AMANN A, SAID M, DATTA B, LEDO- S315-319. CHOWSKI M. Implementation and interpretation of 34) ROBERTS AP, CHILDS SM, RUBIN G, DE WIT NJ. Tests for hydrogen breath tests. J Breath Res 2008; 4: Helicobacter pylori infection: a critical appraisal 046002. from primary care. Fam Pract 2000; S12-20. 20) TURSI A, BRANDIMARTE G, GIORGETTI GM, INCHINGOLO 35) OHARA S, KATO M, SAITO M, FUKUDA S, KATO C, HAMA- CD. Effectiveness of the sorbitol H2 breath test in DA S, NAGASHIMA R, OBARA K, SUZUKI M, HONDA H, detecting histological damage among relatives of ASAKA M, TOYOTA T. Comparison between a new coeliacs. Scand J Gastroenterol 2003; 7: 727-731. 13C-urea breath test, using a film-coated tablet, 21) TURSI A, BRANDIMARTE G, GIORGETTI GM. Sorbitol H2- and the conventional 13C-urea breath test for the breath test versus anti-endomysium antibodies for detection of Helicobacter pylori infection. J Gas- the diagnosis of subclinical/silent coeliac disease. troenterol 2004; 7: 621-628. Scand J Gastroenterol 2001; 11: 1170-1172. 36) GATTA L, RICCI C, TAMPIERI A, OSBORN J, PERNA F, 22) TURSI A, BRANDIMARTE G, GIORGETTI GM. Sorbitol H2- BERNABUCCI V, V AIRA D. Accuracy of breath tests us- breath test versus anti-endomysium antibodies to ing low doses of 13C-urea to diagnose Helicobac- assess histological recovery after gluten-free diet ter pylori infection: a randomised controlled trial. in coeliac disease. Dig Liver Dis 2002; 12: 846- Gut 2006; 4: 457-462. 850. 37) GRAHAM DY, RUNKE D, ANDERSON SY, MALATY HM, 23) GLINSMANN WH, BOWMAN BA. The public health KLEIN PD. Citric acid as the test meal for the 13C- significance of dietary fructose. Am J Clin Nutr urea breath test. Am J Gastroenterol 1999; 5: 1993; 5 Suppl: 820S-823S. 1214-1217. 24) GIBSON PR, NEWNHAM E, BARRETT JS, SHEPHERD SJ, 38) AGHA A, OPEKUN AR, ABUDAYYEH S, GRAHAM DY. Ef- MUIR JG. Review article: fructose malabsorption fect of different organic acids (citric, malic and and the bigger picture. Aliment Pharmacol Ther ascorbic) on intragastric urease activity. Aliment 2007; 4: 349-363. Pharmacol Ther 2005; 9: 1145-1148.

97 G. D’Angelo, T.A. Di Rienzo, F. Scaldaferri, F. Del Zompo et al

39) KLEIN PD, GRAHAM DY. Minimum analysis require- 52) UCHIDA M, SHIMIZU K. 13C-acetic acid is more sensitive ments for the detection of Helicobacter pylori in- than 13C-octanoic acid for evaluating gastric empty- fection by the 13C-urea breath test. Am J Gas- ing of liquid enteral nutrient formula by breath test in troenterol 1993; 11: 1865-1869. conscious rats. Biol Pharm Bull 2007; 3: 487-489. 40) GISBERT JP, PAJARES JM. Review article: 13C-urea 53) BRADEN B, ADAMS S, DUAN LP, ORTH KH, MAUL FD, breath test in the diagnosis of Helicobacter pylori LEMBCKE B, HOR G, CASPARY WF. The [13C]acetate infection — a critical review. Aliment Pharmacol breath test accurately reflects gastric emptying of Ther 2004; 10: 1001-1017. liquids in both liquid and semisolid test meals. 41) GRAHAM DY, OPEKUN AR, JOGI M, YAMAOKA Y, L U H, Gastroenterology 1995; 4: 1048-1055. REDDY R, EL-ZIMAITY HM. False negative urea 54) BRADEN B, PETERKNECHT A, PIEPHO T, S CHNEIDER A, breath tests with H2-receptor antagonists: inter- CASPARY WF, HAMSCHO N, AHRENS P. Measuring gas- actions between Helicobacter pylori density and tric emptying of semisolids in children using the pH. Helicobacter 2004; 1: 17-27. 13C-acetate breath test: a validation study. Dig 42) EPPLE HJ, KIRSTEIN FW, BOJARSKI C, FREGE J, FROMM Liver Dis 2004; 4: 260-264. M, RIECKEN EO, SCHULZKE JD. 13C-urea breath test 55) MAES BD, GHOOS YF, GEYPENS BJ, MYS G, HIELE MI, in Helicobacter pylori diagnosis and eradication. RUTGEERTS PJ, VANTRAPPEN G. Combined carbon-13- Correlation to histology, origin of ʻfalseʼ results, glycine/carbon-14-octanoic acid breath test to and influence of food intake. Scand J Gastroen- monitor gastric emptying rates of liquids and terol 1997; 4: 308-314. solids. J Nucl Med 1994; 5: 824-831. 43) BRADEN B, DUAN LP, CASPARY WF, LEMBCKE B. More 56) DOMINGUEZ MUNOZ JE. Diagnosis of chronic pan- convenient 13C-urea breath test modifications still creatitis: Functional testing. Best Pract Res Clin meet the criteria for valid diagnosis of Helicobacter Gastroenterol 2010; 3: 233-241. pylori infection. Z Gastroenterol 1994; 4: 198-202. 57) LOSER C, MOLLGAARD A, AYGEN S, HENNEMANN O, 44) MANA F, F RANKEN PR, HAM HR, REYNAERT H, URBAIN FOLSCH UR. 13C-starch breath test—comparative D. 13C urea breath test with nondispersive iso- clinical evaluation of an indirect pancreatic func- tope-selective infrared spectrometry: reproducibil- tion test. Z Gastroenterol 1997; 3: 187-194. ity and importance of the fasting status. Heli- 58) EVENEPOEL P, H IELE M, GEYPENS B, GEBOES KP, RUT- cobacter 2000; 2: 104-108. GEERTS P, G HOOS Y. 13C-egg white breath test: a 45) NG FH, LAI KC, WONG BC, WONG WM, WONG SY, CHOW non-invasive test of pancreatic trypsin activity in KC, YUEN ST, LEUNG SY, LAM SK. [13C]-urea breath test the small intestine. Gut 2000; 1: 52-57. without prior fasting and without test meal is accurate 59) GEBOES KP, BAMMENS B, LUYPAERTS A, MALHEIROS R, for the detection of Helicobacter pylori infection in BUYSE J, EVENEPOEL P, R UTGEERTS P, V ERBEKE K. Valida- Chinese. J Gastroenterol Hepatol 2002; 8: 834-838. tion of a new test meal for a protein digestion 46) GHOOS YF, MAES BD, GEYPENS BJ, MYS G, HIELE MI, breath test in humans. J Nutr 2004; 4: 806-810. RUTGEERTS PJ, VANTRAPPEN G. Measurement of gas- 60) HIELE M, GHOOS Y, R UTGEERTS P, V ANTRAPPEN G. tric emptying rate of solids by means of a carbon- Starch digestion in normal subjects and patients labeled octanoic acid breath test. Gastroenterolo- with pancreatic disease, using a 13CO2 breath gy 1993; 6: 1640-1647. test. Gastroenterology 1989; 2 Pt 1: 503-509. 47) ZAHN A, LANGHANS CD, HOFFNER S, HABERKORN U, 61) ISHII Y, K OHNO T, I TO A, SUZUKI S, KOHNO T, TAKAYAMA T, RATING D, HAASS M, ENCK P, S TREMMEL W, R UHL A. ASAI S. Evaluation of pancreatic exocrine secretion Measurement of gastric emptying by 13C-octanoic using 13C-dipeptide (benzoyl-L-tyrosyl-[1- acid breath test versus scintigraphy in diabetics. (13)C]alanine) breath test: focusing on pancreato- Z Gastroenterol 2003; 5: 383-390. duodenectomy cases. Pancreas 2007; 4: 313-319. 48) PFAFFENBACH B, WEGENER M, ADAMEK RJ, WISSUWA H, 62) BRADEN B. (13)C breath tests for the assessment SCHAFFSTEIN J, AYGEN S, HENNEMANN O. [Non-inva- of exocrine pancreatic function. Pancreas 2010; sive 13C octanoic acid breath test for measuring 7: 955-959. stomach emptying of a solid test meal—correla- 63) VANTRAPPEN GR, RUTGEERTS PJ, GHOOS YF, HIELE MI. tion with scintigraphy in diabetic patients and re- Mixed triglyceride breath test: a noninvasive test producibility in healthy probands]. Z Gastroen- of pancreatic lipase activity in the duodenum. terol 1995; 3: 141-145. Gastroenterology 1989; 4: 1126-1134. 49) DUAN LP, BRADEN B, CASPARY WF, LEMBCKE B. Influ- 64) Kalivianakis M, Verkade HJ, Stellaard F, van der ence of cisapride on gastric emptying of solids Were M, Elzinga H, Vonk RJ. The 13C-mixed and liquids monitored by 13C breath tests. Dig Dis triglyceride breath test in healthy adults: determi- Sci 1995; 10: 2200-2206. nants of the 13CO2 response. Eur J Clin Invest 50) SCHADEWALDT P, S CHOMMARTZ B, WIENRICH G, BROSICKE 1997; 5: 434-442. H, PIOLOT R, ZIEGLER D. Application of isotope-selec- 65) LOSER C, BRAUER C, AYGEN S, HENNEMANN O, FOLSCH tive nondispersive infrared spectrometry (IRIS) for UR. Comparative clinical evaluation of the 13C- evaluation of [13C]octanoic acid gastric-emptying mixed triglyceride breath test as an indirect pan- breath tests: comparison with isotope ratio-mass creatic function test. Scand J Gastroenterol 1998; spectrometry (IRMS). Clin Chem 1997; 3: 518-522. 3: 327-334. 51) PERRI F, B ELLINI M, PORTINCASA P, P ARODI A, BONAZZI P, 66) ADAMEK RJ, BODEKER C, SZYMANSKI C, HAGEMANN D, MARZIO L, GALEAZZI F, U SAI P, C ITRINO A, USAI-SATTA P. PFAFFENBACH B. [13C-mixed triglyceride CO2 exha- (13)C-octanoic acid breath test (OBT) with a new lation test. Investigation with an isotope selective, test meal (EXPIROGer): Toward standardization non dispersive infrared spectrophotometer of indi- for testing gastric emptying of solids. Dig Liver rect function of the exocrine pancreas]. Dtsch Dis 2010; 8: 549-553. Med Wochenschr 1999; 5: 103-108.