Cohen, Jacqueline R*; Schall, Joan I*; Ittenbach, Richard F†; Zemel, Babette S*; Stallings, Virginia A*
Cystic fibrosis (CF) is a disease that affects multiple organ systems, including the lungs and pancreas. Approximately 85% of patients with CF develop exocrine pancreatic insufficiency (PI) (1). If not adequately treated, PI leads to malabsorption, malnutrition and growth failure in children with CF. According to the 2002 CF Patient Registry, 92% of CF patients take exogenous pancreatic enzymes (2).
Direct pancreatic function evaluation by endoscopic sampling of pancreatic enzyme activity is both invasive and costly. Fecal chymotrypsin activity is influenced by exogenous enzymes (1), so enzyme regimens must be interrupted to assess function accurately in children after initial diagnosis. The 72-hour collection of dietary intake and stool for determination of fecal fat content, often used in clinical care and research, is inconvenient and prone to collection errors in the non-research setting. The pancreatic enzyme elastase 1 is stable during intestinal transit, is measurable as fecal elastase 1 (FE) and is not affected by enzyme medication. The commercially available assay is relatively inexpensive and easy to perform. Many studies (3-13) have compared the specificity and sensitivity of the FE assay to other methods of assessment of exocrine pancreatic status in CF. Others have followed the progression of pancreatic status of subjects with CF from pancreatic sufficiency to insufficiency (13). Very few studies have explored the association between FE status and nutritional status in CF, and there are no longitudinal studies of growth, nutritional and pulmonary status related to FE in children with CF. The aim of this study was to evaluate FE status in prepubertal children with CF with a clinical or laboratory diagnosis of PI and to prospectively explore the relationship between FE and growth, nutritional and pulmonary status and fat absorption over 24 months.
Children with mild-to-moderate CF lung disease and PI, aged 6.0-8.9 years, were enrolled from 13 CF centers in the United States to participate in a 24-month longitudinal study of growth, nutrition and pulmonary status. Children were excluded from the study if they had forced expiratory volume in 1 second (FEV1) <40% of predicted, significant liver disease, insulin-dependent diabetes mellitus or sputum colonized with Burkholderia cepacia. The diagnosis of CF was determined by the home CF center based on clinical symptoms and laboratory values, including duplicate quantitative pilocarpine iontophoresis sweat sodium and chloride concentrations >60 mEq/L. The diagnosis of PI was determined at the home CF center based on clinical or laboratory values including 72-hour fecal fat analyses with <93% absorption or stool trypsin concentration <80 μg/g. The protocol was approved by the Committee for the Protection of Human Subjects of the Institutional Review Board at Children's Hospital of Philadelphia and at the subject's home institution. Informed, written consent was obtained from the parent or guardian of each child and assent was obtained from each subject.
Random stool samples were obtained at the time of one of the protocol hospital visits, stored at −20°C and analyzed with an enzyme-linked immunosorbent assay FE kit (ScheBo Biotech, Giessen, Germany). Additional samples for subjects with FE concentration ≥15 μg/g were obtained for a repeat evaluation. Additional samples were also obtained randomly from four children with FE concentration <15 μg/g for retesting. For purposes of this study, subjects with FE ≥15 μg/g stool were classified as the group with residual pancreatic activity (R-FE) and those with FE <15 μg/g stool were classified as the group with no pancreatic activity (NO-FE). Genotype was obtained from review of subjects' medical records when available. When unknown, a blood sample was submitted for genotype analysis (Genzyme Genetics, Pittsburgh, PA). For the purposes of this study, subjects were categorized as ΔF508/ΔF508 homozygous, ΔF508/other mutation heterozygous and other/other genotypes. Other mutations included both known non-ΔF508 mutations and unknown mutations.
Subjects were seen annually (at baseline and 12 and 24 months) at the Children's Hospital of Philadelphia General Clinical Research Center while in their usual state of good health. The 6-month and 18-month protocol visits were conducted at the subject's home CF center for growth, clinical status and dietary intake assessment. Evaluations included pulmonary function, anthropometric assessment, blood, urine and fecal collections. Dietary assessments were based on 7-day weighed food records. After each annual visit, the family collected 72-hour stool collections. Skeletal age was determined from hand/wrist radiographs and assessed using the TW3 method at baseline (14). Relative delay in skeletal age was calculated by subtracting chronological age from skeletal age. Pulmonary function was evaluated at the time of each annual visit using standard methods (15) for spirometry, lung volumes, and conductance after albuterol and chest physiotherapy. FEV1 was measured and compared with appropriate reference values (16,17).
Height and weight were measured using standard techniques (18) with a stadiometer accurate to 0.1 cm (Holtain, Crymych, UK) and a digital scale accurate to 0.1 kg (Scaletronix, White Plains, NY). Body mass index was calculated as weight (in kg)/height (in m2). Measured or estimated height by recall was collected for both biologic parents and used to obtain a mid-parent height. Using this mid-parent height and the subject's measured height, a parent-specific height was generated. This method of obtaining a parent-adjusted height is applicable for children aged birth to 18 years (19). Z scores for parental-adjusted height (AHAZ), weight (WAZ) and body mass index (BMIZ) were computed using the Centers for Disease Control 2000 growth charts (20). Upper arm circumference was measured using a flexible plastic measuring tape (Ross Laboratories, Columbus, OH), and a skinfold caliper (Holtain, Crymych, UK) was used to measure triceps, biceps, subscapular and suprailiac skin folds on the right side. Total upper arm muscle and fat areas were calculated (21) and Z scores for upper arm muscle area (UAMAZ) and upper arm fat area (UAFAZ) were calculated (22).
Subjects and parents were provided with measuring cups, spoons and digital food scale to use at home for a 7-day weighed food record. Detailed verbal and written instructions were given to ensure the recording procedures were understood. Completed diet records were reviewed and analyzed by a research dietitian (Nutrition Data System NDS, Minneapolis, MN). Adequacy of energy intake was assessed using the recent Dietary Reference Intakes for estimated energy requirement (23), which adjusts for height, weight, and physical activity level. In a previous study of preadolescent children with CF (24), we demonstrated that the ratio of total energy expenditure to resting energy expenditure was 1.68, which corresponds to the Dietary Reference Intakes physical activity level of “active” (23). Therefore, energy intake was expressed as a percentage of the estimated energy requirement (% EER) calculated for active children.
A 72-hour stool collection was performed at home after annual visits to Children's Hospital of Philadelphia while subjects were on their usual pancreatic enzyme regimen. Subjects were given a home collection kit and detailed instructions. Specimens were stored frozen until analysis of total fat content by a gravimetric method (25) (Mayo Medical Laboratories, Rochester, MN). The coefficient of fat absorption (%CoA) was calculated from 7-day weighed food records and 72-hour stool by subtracting the average daily fecal fat excretion (g) from the average daily dietary fat intake (g) and then dividing by the average daily dietary fat intake (g) and multiplying by 100 to yield the percentage of fat absorbed.
Bomb calorimetry for total caloric content of stool was performed (Covance, Madison, WI). Routine enzyme supplementation regimens were maintained for all the collections. Fasting blood samples were obtained and total serum cholesterol was assessed using standard methods at the 24-month visit only (Clinical Laboratories, Children's Hospital of Philadelphia).
Data analysis occurred in two phases. Phase I consisted of descriptive statistics, initially for the group as a whole and then for the R-FE and NO-FE groups specifically. To estimate the comparability of groups at the time of entry into the study, baseline comparisons between both groups of children were conducted using Student's t-test for all growth and nutritional status, energy intake and pulmonary status variables. In addition, the relationship between FE group status and genotype was estimated using the χ2 test of association.
In Phase II, the inferential phase, longitudinal mixed-effects analysis (26) was used to determine differences in longitudinal data between FE groups in growth and nutritional status, pulmonary function, fat absorption and energy intake over the 24-month study period. For growth and nutritional status and for energy intake all five time points (baseline, 6, 12, 18 and 24 months) were included in the longitudinal mixed-effects analysis.
Initial age was entered into the model to account for baseline differences in children even over the narrow age range. Time (in years) was entered into the model because all children are expected to grow over a 24-month time period. For measures of absolute body size such as height, the coefficient for time was expected to be positive and significant. For age-adjusted measures such as HAZ, the coefficient for time is expected to be nonsignificant if children maintain their usual pattern of growth over time. Gender differences were also examined in the model. The interaction term (group × time) was the term of interest because it indicates the difference in the rate of growth between the FE groups over time. A total of nine separate models was tested for statistical significance. The experiment-wise error rate was held constant at the α = 0.05 level across all nine models, with a hypothesis-wise error rate of α = 0.017 for each of the nine related models, separately, using Tukey, Ciminera and Heyse's adjustment for multiple, moderately related comparisons (27,28). All statistical analyses were performed with STATA 7.0 (STATA, College Station, TX).
Of the 91 children enrolled in the study, FE levels were obtained for 85 (41 male, 44 female). The six children without FE samples were not different in growth, pulmonary function, %CoA or energy intake from the 85 with FE samples. Seventy-five of the 85 children with FE levels were in the NO-FE group, with FE levels <15 μg/g stool, and 10 children were in the R-FE group, with FE levels ≥15 μg/g stool. One of the 10 children in the R-FE group did not complete the 24-month study and was excluded from analysis. The final sample was 84 subjects, with 75 (89%) in the NO-FE group and nine (11%) in the R-FE group. FE levels were reassessed for eight subjects in the R-FE group (≥15 μg/g stool) and the status was confirmed (Table 1). A second stool sample of one subject was not available. First assessment FE levels ranged from 17 to 667 μg/g stool and the second ranged from 24 to >500 μg/g stool. Samples from four subjects with initial FE status <15 μg/g stool were chosen for status reassessment, and all remained below <15 μg/g stool.
Characteristics of the sample by group at baseline for growth and nutritional status, skeletal age, energy and fat intake, % CoA, bomb calorimetry and pulmonary function are shown in Table 2. This group of young children with CF had suboptimal growth status (mean AHAZ = −0.6 ± 1.1, WAZ = −0.5 ± 1.1, muscle stores as UAMAZ = −0.2 ± 1.1 and fat stores as UAFAZ = −0.2 ± 1.1) compared with healthy children (19). Mean pulmonary function was normal (FEV1 = 100% ± 18%) and the mean energy intake (percentage of the estimated energy requirement with active physical activity level) was increased as expected in children with CF, 110% ± 20% of the estimated energy requirement. Children in the R-FE group had significantly better adjusted height status (AHAZ). No significant differences were noted between the two FE groups at baseline in other growth or nutritional status variables, dietary intake of energy, stool fat, bomb calorimetry or pulmonary function (Table 2). The relative skeletal age differed significantly (P = 0.009), with the R-FE group as 0.7 ± 1.5 years (advanced) and the NO-FE group as −0.4 ± 1.1 years (delayed) compared with chronological age. As expected, children in the R-FE group had significantly higher (P < 0.01) mean %CoA derived from the 72-hour stool and dietary assessment than did those in the NO-FE group (94% ± 3% versus 81% ± 14%, respectively). The range of %CoA values was narrower (87% to 97%) in the R-FE group than in the NO-FE group (30% to 98%) while children were on their usual enzyme regimen (Fig. 1). For this reason, Student's t-test with unequal variances was deemed most appropriate. The significant difference between FE groups in %CoA was similar at the 24-month evaluation. At 24 months, the subjects in the R-FE group had significantly higher mean serum cholesterol levels than those in the NO-FE group (167 ± 22 versus 139 ± 25 mg/dL, respectively, P = 0.002). Genotype differed significantly (P = 0.007) between the two FE groups (Table 2). Table 1 provides genotype information in addition to the FE levels for children in the R-FE group and genotype information for the NO-FE group in the table footnote. Only one child of nine in the R-FE group was ΔF508/ΔF508 homozygous, whereas 46 of the 75 (61%) children in the NO-FE group were homozygous. Five of the nine children in the R-FE group were ΔF508/other mutation compared with 23 of the 75 (31%) of those in the NO-FE group, and three were other/other mutations compared to six of the 75 (8%) in the NO-FE group.
The longitudinal mixed-effects models presented in Table 3 tested for differences between the R-FE group compared to NO-FE, the reference group, in growth and nutritional status (WAZ, UAMAZ, UAFAZ), pulmonary function (FEV1), fecal fat absorption (%CoA) and energy intake (kcal) over the 24-month study period (Table 3). In this analysis, each unit of time represents 1 year, so that the coefficients for time and for the group × time interaction represent the annual change for each variable rather than the change over the entire 24-month observation period. Age at enrollment, across a small range, and gender were not significant predictors of growth or nutritional status variables. Age at enrollment was a significant predictor of energy intake, with older children having higher intake, as expected. Male gender was significantly associated with higher FEV1 (+8.5%), higher energy intake (+340 kcal) and lower %CoA (−5.4%). Mean weight status remained stable over time; however, UAMAZ (muscle) improved significantly whereas UAFAZ (fat) did not. Pulmonary function declined significantly over time, with an annual rate of decline of FEV1 of 4.5%, whereas %CoA increased (1.4% per year) and energy intake increased (163 kcal per year). The interaction term (group × time) indicated a significantly greater improvement over time in weight status (+0.15 WAZ per year), arm muscle stores (+0.26 UAMAZ per year) and arm fat stores (+0.23 UAFAZ per year) for subjects in the R-FE groups. In absolute terms, this amounted to a 2.6-kg greater increase in weight, 2.1-cm2 greater increase in upper arm muscle area and a 3.2-cm2 greater increase in fat area over the 24 months for subjects in the R-FE group. Although subjects in the R-FE group tended to have a slower decline in FEV1 (2.8% per year less), this did not reach significance (P = 0.20). Although all six models presented in (Table 3) would have achieved statistical significance using a traditional (unadjusted) level of significance such as α = 0.05, only four of the models yielded significant results using the more stringent criterion of α = 0.017. Therefore, the models for UAMAZ, FEV1, %CoA and energy intake are significant at this adjusted level. We have presented the models for WAZ and UAFAZ that appear significant in a traditional sense (P < 0.04 and 0.02, respectively), although they do not meet the adjusted criteria for significance. Models were also tested for AHAZ and BMIZ; however, they failed to reach significance using either adjusted or unadjusted criteria and are not shown.
This study describes the pattern of concordance between the clinical diagnosis of PI and the stool FE in a group of preadolescent children with CF and PI from a multicenter sample. Twelve percent of this sample, all of whom had a diagnosis of PI, had FE levels >15 mg/g stool, indicating some degree of pancreatic function. If this rate were true across the 92% of the approximately 23,000 people with CF taking enzymes in the United States, then approximately 2,500 patients (23,000 × 0.92 × 0.12) have some pancreatic function (2), many of whom may not require pancreatic enzyme medication. This suggests a need for verification of pancreatic status in some patients.
The range of FE activity for the entire sample (<15 to 667 mg/g) is consistent with those described in other CF studies, which have included both pancreatic sufficient and PI subjects (3-12, 29-32). As expected and as reported by others (4,12,13,32), the majority (61%) of subjects with NO-FE were DF508/DF508 homozygous. Only one subject of nine with R-FE was homozygous for ΔF508. Patients with CF who are DF508/DF508 homozygous usually develop pancreatic insufficiency (12,13) by the age of our subjects. It should be noted that the 2789, R334W and R347P alleles have been associated with mild or absent pancreatic disease in patients with CF (12,33,34).
Different cut-off levels for the classification of FE status as pancreatic sufficient or PI in CF have been proposed, and there is no current consensus (29). A FE level of >15 μg/g stool is the level of detectability for the assay. The present data suggest that detectable FE levels (R-FE group) were associated with more positive changes in growth and nutritional status over a 24-month period than were those with low FE levels. Specifically, among 6- to 8-year-old children, those with R-FE attained stature at their genetic potential, had no delay in skeletal maturity and absorbed significantly more dietary fat, based on the %CoA. Children with no detectable FE (NO-FE) levels had suboptimal growth status and delayed skeletal maturation. Over the 24 months of the study, children with R-FE experienced significant improvements in weight status, including both muscle and fat components, whereas children with NO-FE did not improve. Fat absorption improved and may be related to general improvement in care as children get older within this age range. Energy intake increased over time in the sample as a whole as expected because the children increase in body size as they age. Energy intake did not differ between the groups either at baseline or over time. However, fat absorption and serum cholesterol remained significantly higher for the R-FE group throughout the 24-month period. Thus, less dietary energy loss, and not greater energy intake, explained the observed improvement in growth and nutritional status in those children with higher FE. It is also likely that some of the NO-FE subjects had suboptimal pancreatic enzyme regimens and this contributed to their poorer growth and nutritional status. Suboptimal treatment may be related to such issues as difficulty in matching the enzyme dose to dietary intake, suboptimal adherence to a well-designed regimen or enzyme products with decreased pancreatic lipase activity.
There are few studies in which both the FE and the nutritional status of the subjects are described. Walkowiak et al. (10), in a cross-sectional study, described subjects with CF as having adequate nutritional status; however, they did not report nutritional status methodology or relate the FE values to nutritional status. In another cross-sectional study, Walkowiak et al. (12) used weighted Z scores to assess nutritional status in subjects with CF and found similar deficits in weight status to our population. One study, Dorlcohter et al. (29), assessed fat-soluble vitamin profiles in subjects with CF with and without PI, classified by FE levels. No correlation between fat-soluble vitamin status and FE status was found, and no longitudinal measures of clinical status such as growth or pulmonary function were reported.
In this sample of preadolescent children, FEV1% predicted declined over the 2 years and boys had significantly higher FEV1 throughout the study period. There was a tendency (P = 0.20) for this decline to be attenuated among the small sample of children in the R-FE group. Children with CF and pancreatic sufficiency generally experience less disease severity and thus enhanced survival compared with those with PI (33).
In summary, we present data from children from a number of CF centers showing that 12% of preadolescent children who were diagnosed with PI and prescribed pancreatic enzymes were possibly misdiagnosed. Pancreatic status remains difficult to firmly diagnose in some children. Enzyme replacement therapy, although undoubtedly essential for those who are PI, has potential important clinical and psychosocial adverse effects. Consequences may include the clinical assumption that poor growth and nutritional status is a result of inadequately treated PI, and other causes for nutritional failure may be missed. With respect to fat absorption, growth and nutritional and maturity status, our results suggest, as expected, that the clinical course is more favorable in children with higher FE levels. Based on these data, an FE assessment to determine pancreatic status in children with CF should be considered. Further research is needed to determine benefit in enzyme replacement therapy in children with CF and residual FE activity, with a focus on the 15 to 200 μg/g stool range. Until more data are available, we suggest verification of pancreatic status by evaluating two separate stool samples for FE in children older than 2 years of age with non-ΔF508 homozygous genotypes. When the two values are in the normal range, the clinical team may consider discontinuing enzyme therapy under careful observation.
The authors thank the children, their families, and the CF care team of the 13 participating CF Centers for their dedication and cooperation. We also thank the General Clinical Research Center staff and the Nutrition and Growth Laboratory at The Children's Hospital of Philadelphia. We would also like to thank the Center Directors and staff at the 13 Cystic Fibrosis Centers from which the children were recruited for this study: The Children's Hospital of Philadelphia (Philadelphia, PA); Johns Hopkins Children's Center (Baltimore, MD); St. Christopher's Hospital for Children (Philadelphia, PA); Children's National Medical Center (Washington, DC); Schneider Children's Hospital of Long Island (New Hyde Park, NY); Long Island College Hospital (Brooklyn, NY); Children's Hospital of Buffalo (Buffalo, NY); Hershey Medical Center (Hershey, PA); Albany Medical Center (Albany, NY); Emory University (Atlanta, GA); The Children's Medical Center (Dayton, OH); State University New York at Stony Brook (Stony Brook, NY); and University of Florida (Gainesville, FL).
1. Dominici R, Franzini C. Fecal elastase-1 as a test for pancreatic function: a review. Clin Chem Lab Med
2. Cystic Fibrosis Foundation Patient Registry, 2002 Annual Data Report to the Center Directors. Bethesda, MD: Cystic Fibrosis Foundation, 2003.
3. Beharry S, Ellis L, Corey M, et al. How useful is fecal pancreatic elastase 1 as a marker of exocrine pancreatic disease? J Pediatr
4. Cade A, Walters M, McGinley N, et al. Evaluation of fecal pancreatic elastase-1 as a measure of pancreatic exocrine function in children with cystic fibrosis. Pediatr Pulmonol
5. Carroccio A, Verghi F, Santini B, et al. Diagnostic accuracy of fecal elastase 1 assay in patients with pancreatic maldigestion or intestinal malabsorption. Dig Dis Sci
6. Gullo L, Graziano L, Babbini S, et al. Faecal elastase 1 in children with cystic fibrosis. Eur J Pediatr
7. Lüth S, Teyssen S, Forssmann K, et al. Fecal elastase-1 determination: ‘gold standard’ of indirect pancreatic function tests? Scand J Gastroenterol
8. Soldan W, Henker J, Sprössig C. Sensitivity and specificity of quantitative determination of pancreatic elastase 1 in feces of children. J Pediatr Gastroenterol Nutr
9. Walkowiak J. Faecal elastase-1: clinical value in the assessment of exocrine pancreatic function in children. Eur J Pediatr
10. Walkowiak J, Cichy W, Herzig K. Comparison of fecal elastase-1 determination with the secretin-cholecystokinin test in patients with cystic fibrosis. Scand J Gastroenterol
11. Walkowiak J, Herzig K, Strzykala K, et al. Fecal elastase-1 is superior to fecal chymotrypsin in the assessment of pancreatic involvement in cystic fibrosis. Pediatrics
12. Walkowiak J, Herzig K, Witt M, et al. Analysis of exocrine pancreatic function in cystic fibrosis: one mild CFTR mutation does not exclude pancreatic insufficiency. Eur J Clin Invest
13. Walkowiak J, Nousia-Arvanitakis S, Agguridaki C, et al. Longitudinal follow-up of exocrine pancreatic function in pancreatic sufficient cystic fibrosis patients using the fecal elastase-1 test. J Pediatr Gastroenterol Nutr
14. Tanner JM, Healy MJR, Goldstein H, Cameron N. Assessment of Skeletal Maturity and Prediction of Adult Height. TW3 Method
, 3rd ed. Philadelphia: WB Saunders; 2001.
15. Gardner RMC. Report of Snowbird workshop on standardization of spirometry. Am Rev Respir Dis
16. Morris A, Kanner R, Crapo R, et al. Clinical Pulmonary Function Testing
, 2nd ed. Salt Lake City, UT: Intermountain Thoracic Society; 1984:1-24.
17. Herzog H, ed. Lung Function in Children and Adolescents-Methods, Reference Values-Progress in Respiration Research
. Pratteln, Switzerland: Offsetdruck; 1987:1-220.
18. Lohman TG, Roche AR, Martorell R. Anthropometric Standardization Reference Manual
. Champaign, IL: Human Kinetics; 1988.
19. Himes JH, Roche AF, Thissen D, et al. Parent-specific adjustments for evaluation of recumbent length and stature of children. Pediatrics
20. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States. Advance Data from Vital and Health Statistics. Hyattsville, MD: National Center for Health Statistics; 2000: no. 314.
21. Frisancho A. New norms of upper limb fat and muscle areas for assessment of nutritional status. Am J Clin Nutr
22. Frisancho A. Anthropometric Standards for Assessment of Growth and Nutritional Status
. Ann Arbor, MI: University of Michigan Press; 1990.
23. Institute of Medicine. Energy. In: Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids
. Food and Nutrition Board. Washington, DC: National Academy Press; 2002:5-1 to 5-114.
24. Tomezsko JL, Stallings VA, Kawchak DA, et al. Energy expenditure and genotype of children with cystic fibrosis. Pediatr Res
25. Van de Kamer JH. Total fatty acids in stool. In: D. Seligson, ed. Standard Methods of Clinical Chemistry
, Vol. 2. New York: Academic Press; 1958:34-39.
26. Laird N, Ware JH. Random-effects models for longitudinal data. Biometrics
27. Zhang J, Quan H, Ng J, Stepanavage ME. Some statistical methods for multiple endpoints in clinical trials. Control Clin Trials
28. Tukey JW, Ciminera JL, Heyse JF. Testing the statistical certainty of a response to increasing doses of a drug. Biometrics
29. Dorlochter L, Aksnes L, Fluge G. Faecal elastase-1 and fat-soluble vitamin profiles in patients with cystic fibrosis in Western Norway. Eur J Nutr
30. Hamwi A, Veitl M, Maenner G, et al. Pancreatic elastase 1 in stool: variations within one stool passage and individual changes from day to day. Wien Klin Wochenschr
31. Walkowiak J, Nousia-Arvanitakis S, Cade A, et al. Fecal elastase-1 cutoff levels in assessment of exocrine pancreatic function in cystic fibrosis. J Cystic Fibrosis
32. Wallis C, Leung T, Cubitt D, et al. Stool elastase as a diagnostic test for pancreatic function in children with cystic fibrosis. Lancet
33. Borowitz D. Evidence for the diagnosis of pancreatic insufficiency. Pediatr Pulmonol
34. Kristidis P, Bozon D, Corey M, et al. Genetic determination of exocrine pancreatic function in cystic fibrosis. Am J Hum Genet
© 2005 Lippincott Williams & Wilkins, Inc.