Aliment. Pharmacol. Ther. (1992) 6, 549-555.
Comparison between a standard pancreafic supplement and a high enzyme pieparation in cystic fibrosis
G. M O R R I S O N * , J. M. MORRISON", A. 0. B. REDMOND", C. A. BYERS", K. J. McCRACKENt, J. A. DODGE*, S. A. GUILFORDS, M. W. BOWDENS
" The Royal Belfast Hospital for Sick Children, Belfast, t Department of Agriculture, N. Ireland and Queen's University of Belfast, Department of Child Health, Queen's University of Belfast, S Cilag Limited, High Wycombe, Bucks, UK
*
Accepted for publication 14 May 1992
SUMMARY
This study compared the relative effectiveness of a standard pancreatic enzyme supplement (' Creon ', Duphar) and a new preparation ('Pancrease HL', Cilag) containing about 3 times the lipase and more than 5 times the protease activity. Capsule dosage was adjusted to a ratio of approximately 3 :1. Fat balances showed that absorption of fat did not change significantly on conversion to the new high-lipase product, and the coefficient of absorption of total energy was similarly maintained. The coefficient of protein absorption was significantly enhanced with the high enzyme preparation ( P < 0.01),which may explain the reported subjective improvement in stool odour. No adverse effects were recorded. Patient acceptability of the new compound was high; the great reduction in the number of capsules required at each meal was cited by all patients as the reason for their preference.
Correspondence to: Professor J. A. Dodge, Department of Child Health, Institute of Clinical Science, Grosvenor Road, Belfast BT12 6BJ, UK. 549
550
G . M O R R I S O N et al.
INTRODUCTION The cardinal clinical features of cystic fibrosis are progressive suppurative lung disease and malabsorption resulting from pancreatic enzyme insufficiency. While the main cause of reduced life expectancy is pulmonary infection, increasing attention is being paid to the association between good nutrition and improved quality and duration of life.' The maldigestion and malabsorption of nutrients must therefore be corrected as far as possible by the administration of adequate quantities of pancreatic enzymes. The amount needed for any individual patient will reflect the properties of the enzyme supplement used as well as the pathophysiological factors present. Different pancreatic enzyme supplements vary both in quantity of enzyme per capsule and the conditions under which the enzymes are released in viva.The most effective preparations are those in which the enzymes are contained within pHsensitive microspheres, themselves contained within a larger, usually gelatine, capsule. The enteric coating of the microspheres disintegrates only when the pH rises above 5.5, so that the enzymes are protected from acid inactivation during passage through the stomach, but released in the duodenum. Pancreatic enzyme preparations contain stated quantities of lipase, amylase and protease. Different preparations are, however, commonly compared solely on the basis of lipase content. A change from one preparation to another based on lipase content will almost certainly result in a change in the quantity and ratios of the other two enzymes classes. The bioavailability of the enzymes is dependent on the release properties of the preparation, and it is therefore possible for enzyme supplements of widely varying composition to produce indistinguishable clinical results.' Quantities of supplement capsules taken daily can be large, and may exceed 10 per main meal. This number of capsules may become an impediment to treatment, resulting in poor compliance and poor motivation to eat. The need for more compact products which will allow a reduction in capsule intake has been highlighted as an urgent need in treating cystic fibrosis patienk3r4 Recent technological developments have enabled pancreatin to be compressed into 'microtablets' with a corresponding increase in the enzyme activity which can be contained within a single capsule. The purpose of our study was to determine the relative clinical effectiveness of a widely used standard enzyme supplement ('Creon ', Duphar, Southampton, UK) and a new preparation ('Pancrease HL', Cilag, High Wycombe, UK) containing, capsule for capsule, about 3 times the lipase and more than 5 times the protease activity. SUBJECTS A N D M E T H O D S The composition of the two enzyme preparations is given in Table 1. The design was an open, one-way, cross-over study of 38 patients with cystic fibrosis and pancreatic insufficiency who were established on a minimum of 12 capsules of Creon each day. The research protocol was approved by the Research Ethics
C O M P A R I S O N OF PANCREATIC SUPPLEMENTS
551
Table 1. Minimum stated enzyme activities of the two preparations (BP units per capsule)
Lipase Amylase Protease
Pancrease HL Creon
Ratio
25000 22 500 1250
3.1:l 2.5:l 5.9:1
8000 9000 210
Table 2. Subject characteristics (n = 33, males 18, females 15)
Age Weight (kg) Height (cm) Weight for age (centile) Height for age (centile) Shwachman score (%)
Mean (S.D.)
Median
Minimum
Maximum
30.4 (16.3) 125.6 (25.7) 45.7 (27.1) 40.8 (26.8) 83.2 (11.6)
8 years 24.3 122.8 50 50 86
15 months 11.4 80 3 3 58
27 years 80.0 174 97 97 97
Committee of the Faculty of Medicine, Queen’s University of Belfast, and written informed consent was given by the subjects or their parents. The age range of patients was 15 months to 27 years, median 8 years (Table 2). The study was run in three separate cohorts within a 3-month period. During a &week run-in period, patients continued on their usual dose of Creon, the second week being a test week during which the degree of control of malabsorption was assessed by fat, protein and energy balances. Assessment of dietary intake was done according to the methods described by Marr.’ Each patient or carer was issued with a dietary record book and a set of electronic scales, and received comprehensive written instructions from a qualified dietitian (J. M.) on how to measure all food and drink. Each patient was also contacted at home at least once during the period of the weighed intake in case any problems had arisen. Each item of food on the completed diaries was coded according to the tables of McCance & Widdowson‘ and total nutrient intakes calculated using a computer program (’ Microdiet ’, University of Salford). Each patient was required to complete a 7-day weighed dietary inventory and a 3-day stool collection. To confirm completeness of the stool collections, each patient took 3 capsules each containing 8 radio-opaque pellets every day during the 7 days of the weighed intake period and the number of radio-opaque pellets If 72 pellets were visible, present in the stools was determined by X-ray the sample was regarded as complete, but in samples deemed incomplete the wet weight was multiplied by a factor of 72 and divided by the number of pellets counted. All stools passed during the final 3 days of the test week were collected in polythene bags and stored on ice. After delivery to the laboratory they were stored at -20 O C until analysis was carried out. Stools were weighed, freeze-dried and 31
BAP 6
552
G. M O R R I S O N ef al.
weighed again. Milled samples were analysed for dry matter (forced-draught oven at 100 "C for 24 h), ash (mufflefurnace at 550 "C overnight), protein (Kjeldahl), fat (Soxhlet) following acid hydrolysis, and energy (adiabatic bomb calorimeter, Gallenkamp). These methods are used routinely in the Department of Food and Agricultural Chemistry, The Queen's University of Belfast. Coefficients of absorption were then calculated by dividing the gross intake less the faecal losses, by the gross intake, in respect of total energy, fat and protein. After 2 weeks, when the above procedure was completed, Pancrease HL was substituted for Creon, at a daily dose of one-third the number of capsules. The patients or carers were advised to adjust the pattern of capsule-taking with meals and snacks in order to achieve the recommended reduction. When the number normally taken could not neatly be divided by 3, patients were advised to round down to a smaller number. In the fourth week, the test procedures described above were repeated. A 7-day weighed dietary intake was again carried out, with the patients strongly encouraged to follow the same diet in this second test week. Three-day stool collections and laboratory analyses were carried out as before. After an interim analysis of the first cohort, subsequent patients were advised to round up to the nearest larger number, Capsule intakes of 1 or 2, common with snacks, could not be decreased below a single capsule of the new compound. All statistical significance tests were carried out at the 5 % level and were twosided. Digestibility ratios were summarized, and treatment differences were analysed using the analysis of variance. Data processing and statistical analysis were carried out using SAS statistical software (release 6.04).
RESULTS Two patients withdrew their consent before the study began, and 3 further patients failed to complete necessary procedures at various times and did not therefore complete the protocol. Statistical analysis was not performed on the data from 6 further patients, 3 of whom developed gastroenteritis while another 3 deviated significantly from the study protocol. The results presented therefore refer to the 27 patients who fully completed the study. The diary records were used to test between periods for symptoms commonly considered relevant to the dose of pancreatic enzyme supplements. There was no significant difference between the number, colour or consistency of stools during the two periods nor in the incidence or severity of abdominal pain. Stool odour was regarded as markedly less during the Pancrease HL period, and the difference on subjective coding was highly significant ( P < 0.001). No significant differences emerged in appetite or eating habits. Food intake was comparable in both test weeks (Table 3 ) . The number of capsules ingested was calculated both from the number issued less the number returned, and the number recorded on the diary during the test weeks. Some discrepancies emerged, due to the fact that very large numbers of
C O M P A R I S O N OF PANCREATIC SUPPLEMENTS 553 Table 3. Mean food intake (S.D.) during study weeks ( n = 27)
Energy, mJ kCal Fat, g Protein, g
Pancrease' HL
Creon
9.21 (3.01) 2202 (720) 89.9 (33.9) 65.8 (23.8)
9.09 (2.49) 2173 (594) 87.8 (26.8) 66.2 (27.9)
Table 4. Coefficients (%) of apparent digestibility for fat, energy, protein and dry matter ( n = 27)
Fat Energy Protein Dry matter
Pancrease' HL
Creon
Mean (S.E.M.) difference (%)
P
83 (9) 90 (4) 86 (6) 92 (3)
84 (10) 89 (4) 82 (7) 91 (3)
-1 (2) I(1) 4 (0 I(1)
0.640 0.140 0.007 0.120
capsules were issued, while patients previously taking Creon sometimes took capsules from their pre-existing supplies. Sibling pairs (there were 3 ) sometimes shared capsule pots. Coefficients of digestibility are shown in Table 4. Seven patients did not reduce capsule intake to the prescribed level, for a variety of reasons, e.g. consumption of excessive snacks, which precluded reduction of capsules below 1.The remaining 20 attained a ratio reasonably close to the planned 0.33: 1 Pancrease HL:Creon ratio (mean 0.39, range 0.27-0.50). When this core group of 20 were analysed alone, control of digestibility was maintained for fat (85% us. 84%), energy (89% us. 91%) and dry matter (91% us. 93%), while digestibility of protein improved from 82% to 86% ( P < 0.007). Separate analysis of the 7 patients who reduced their capsule intake less than instructed showed no significant difference in digestibility between the periods on Creon and Pancrease HL (fat 83 % us. 82 YO,energy 89% us. 90 %, protein 83 % us. 85 % and dry matter 91% us. 91%). No adverse effect attributable to the medication was recorded. Patient and parent acceptability of the new compound was high. Only one of the 27 patients whose results have been presented preferred Creon. Of the entire 33 patients who were able to compare both preparations, only 3 expressed a preference for Creon. Two of these were among those excluded from analysis because of coincidental reported 'gastroenteritis ', which was not supported by bacteriological evidence and had occurred in one while on the new compound and in the other while on 31 2
554
G. MORRISON ef a/.
Creon. The remaining patient thought that he had fewer abdominal symptoms on his usual Creon, although 4 others reported fewer abdominal pains on Pancrease HL. The great reduction in capsule number was cited in all cases as the sole or main reason for preferring Pancrease HL.
DISCUSSION The dose of pancreatin capsules taken by cystic fibrosis patients is usually determined simply by subjective reports of effect on the bowel habits. The effect of inadequate lipase dosage is apparent as steatorrhoea, but signs of poor digestibility of other nutrients are less obvious, apart from the unpleasant odour associated with uncontrolled protein excretion. Repeated nutrient balances, although regarded as the best measure of effectiveness, are rarely performed except for research purposes. The drawbacks of such balance studies are well documented.8 Faecal output is not necessarily a reliable guide to absorption. Apart from the human errors associated with faecal collection, it has been calculated that approximately 30% of the stool energy is derived from colonic bacteria and endogenous secretions, which may be particularly enhanced in cystic fibrosis children.' Although most objective comparisons have concentrated on fat balance, it is likely that consideration of energy balance will be more informative. Moreover, the importance of protein deficiency in failure of growth in childhood is well known. In this study we therefore measured protein and energy intake and losses in addition to those of fat. The results suggest that the higher concentration of protease in the Pancrease HL (Table I) enhanced protein absorption. Although dietary protein deficiency would not be expected in British patients with cystic fibrosis there is evidence that even stable (but chronically infected) patients are often in a state of protein catabolism." Although the group analyses gave results close to those predicted, they mask considerable variations between patients. Moreover, the relative efficiency of fat, protein and energy digestion varied widely within individual patients. Not all patients were optimally controlled at the beginning of the study, although we eliminated from analysis any in whom there was a substantial difference in intake or variation from the protocol between the 2 study weeks. We conclude that the potencies of the two preparations are broadly those which are indicated by their composition, and as a starting point for conversion 1 high-lipase capsule should be substituted for 3 of the standard preparation. However, we would recommend that capsule numbers are rounded up rather than down when the number of standard capsules is not divisible by 3. Because of the overwhelming patient preference for the new formulation we would regard products of this type as being a significant advance in the nutritional management of pancreatic insufficiency. Our preliminary follow-up observations suggest that the greater compliance which follows reduction in capsule numbers is reflected in better appetite and in some cases a consequent improvement in weight gain.
C O M P A R I S O N OF PANCREATIC SUPPLEMENTS
555
ACKNOWLEDGEMENTS
This study was supported by a grant from Cilag Ltd, High Wycombe, UK. REFERENCES 1 Corey M, McLaughlin F J, Williams M,
Levison H. A comparison of survival, growth and pulmonary function in patients with cystic fibrosis in Boston and Toronto. J Clin Epidemiol 1988; 41: 583-91. 2 Williams J, MacDonald A, Weller P H, Fields J, Pandov H. Two enteric coated microspheres in cystic fibrosis. Arch Dis Child 1990; 65 : 594-7. 3 Robinson P J, Olinsky A, Smith A L,
Chitravanshi S B. High compared with standard dose lipase pancreatic supplements. Arch Dis Child 1989; 64: 143-45. 4 Owen G, Peters T J, Dawson S, Goodchild M C. Pancreatic enzyme supplement dosage in cystic fibrosis. The Lancet 1991; 338: 1153. 5 Marr J. Individual dietary surveys : purposes
and methods. World Review Nutr Diet 1971; 13: 105-64. 6 Paul AA, Southgate D A T. McCance and
Widdowson’s the Composition of Foods. London: HMSO, 1978. 7 Simpson F G, Hall G P, Kelleher J, Losovosky M S. Radio-opaque pellets as faecal markers for faecal fat estimation in malabsorption. Gut 1979; 20: 581-4. 8 Holmes G K T, Hill P G. Do we still need to measure faecal fat? Br Med J 1988; 296: 1552-3. 9 Murphy J L , Wootton S A , Bond S A , Jackson A A . Energy content of stools in
normal healthy controls and patients with cystic fibrosis. Arch Dis Child 1991; 66: 495-500 10 Holt T L, Ward L C, Francis P J, Isles A,
Cooksley W G E, Shepherd R W. Whole body protein turnover in malnourished cystic fibrosis patients and its relationship to pulmonary disease. Am J Clin Nutr 1985 ; 41 : 1061-6.
Comparison between a standard pancreafic supplement and a high enzyme pieparation in cystic fibrosis
G. M O R R I S O N * , J. M. MORRISON", A. 0. B. REDMOND", C. A. BYERS", K. J. McCRACKENt, J. A. DODGE*, S. A. GUILFORDS, M. W. BOWDENS
" The Royal Belfast Hospital for Sick Children, Belfast, t Department of Agriculture, N. Ireland and Queen's University of Belfast, Department of Child Health, Queen's University of Belfast, S Cilag Limited, High Wycombe, Bucks, UK
*
Accepted for publication 14 May 1992
SUMMARY
This study compared the relative effectiveness of a standard pancreatic enzyme supplement (' Creon ', Duphar) and a new preparation ('Pancrease HL', Cilag) containing about 3 times the lipase and more than 5 times the protease activity. Capsule dosage was adjusted to a ratio of approximately 3 :1. Fat balances showed that absorption of fat did not change significantly on conversion to the new high-lipase product, and the coefficient of absorption of total energy was similarly maintained. The coefficient of protein absorption was significantly enhanced with the high enzyme preparation ( P < 0.01),which may explain the reported subjective improvement in stool odour. No adverse effects were recorded. Patient acceptability of the new compound was high; the great reduction in the number of capsules required at each meal was cited by all patients as the reason for their preference.
Correspondence to: Professor J. A. Dodge, Department of Child Health, Institute of Clinical Science, Grosvenor Road, Belfast BT12 6BJ, UK. 549
550
G . M O R R I S O N et al.
INTRODUCTION The cardinal clinical features of cystic fibrosis are progressive suppurative lung disease and malabsorption resulting from pancreatic enzyme insufficiency. While the main cause of reduced life expectancy is pulmonary infection, increasing attention is being paid to the association between good nutrition and improved quality and duration of life.' The maldigestion and malabsorption of nutrients must therefore be corrected as far as possible by the administration of adequate quantities of pancreatic enzymes. The amount needed for any individual patient will reflect the properties of the enzyme supplement used as well as the pathophysiological factors present. Different pancreatic enzyme supplements vary both in quantity of enzyme per capsule and the conditions under which the enzymes are released in viva.The most effective preparations are those in which the enzymes are contained within pHsensitive microspheres, themselves contained within a larger, usually gelatine, capsule. The enteric coating of the microspheres disintegrates only when the pH rises above 5.5, so that the enzymes are protected from acid inactivation during passage through the stomach, but released in the duodenum. Pancreatic enzyme preparations contain stated quantities of lipase, amylase and protease. Different preparations are, however, commonly compared solely on the basis of lipase content. A change from one preparation to another based on lipase content will almost certainly result in a change in the quantity and ratios of the other two enzymes classes. The bioavailability of the enzymes is dependent on the release properties of the preparation, and it is therefore possible for enzyme supplements of widely varying composition to produce indistinguishable clinical results.' Quantities of supplement capsules taken daily can be large, and may exceed 10 per main meal. This number of capsules may become an impediment to treatment, resulting in poor compliance and poor motivation to eat. The need for more compact products which will allow a reduction in capsule intake has been highlighted as an urgent need in treating cystic fibrosis patienk3r4 Recent technological developments have enabled pancreatin to be compressed into 'microtablets' with a corresponding increase in the enzyme activity which can be contained within a single capsule. The purpose of our study was to determine the relative clinical effectiveness of a widely used standard enzyme supplement ('Creon ', Duphar, Southampton, UK) and a new preparation ('Pancrease HL', Cilag, High Wycombe, UK) containing, capsule for capsule, about 3 times the lipase and more than 5 times the protease activity. SUBJECTS A N D M E T H O D S The composition of the two enzyme preparations is given in Table 1. The design was an open, one-way, cross-over study of 38 patients with cystic fibrosis and pancreatic insufficiency who were established on a minimum of 12 capsules of Creon each day. The research protocol was approved by the Research Ethics
C O M P A R I S O N OF PANCREATIC SUPPLEMENTS
551
Table 1. Minimum stated enzyme activities of the two preparations (BP units per capsule)
Lipase Amylase Protease
Pancrease HL Creon
Ratio
25000 22 500 1250
3.1:l 2.5:l 5.9:1
8000 9000 210
Table 2. Subject characteristics (n = 33, males 18, females 15)
Age Weight (kg) Height (cm) Weight for age (centile) Height for age (centile) Shwachman score (%)
Mean (S.D.)
Median
Minimum
Maximum
30.4 (16.3) 125.6 (25.7) 45.7 (27.1) 40.8 (26.8) 83.2 (11.6)
8 years 24.3 122.8 50 50 86
15 months 11.4 80 3 3 58
27 years 80.0 174 97 97 97
Committee of the Faculty of Medicine, Queen’s University of Belfast, and written informed consent was given by the subjects or their parents. The age range of patients was 15 months to 27 years, median 8 years (Table 2). The study was run in three separate cohorts within a 3-month period. During a &week run-in period, patients continued on their usual dose of Creon, the second week being a test week during which the degree of control of malabsorption was assessed by fat, protein and energy balances. Assessment of dietary intake was done according to the methods described by Marr.’ Each patient or carer was issued with a dietary record book and a set of electronic scales, and received comprehensive written instructions from a qualified dietitian (J. M.) on how to measure all food and drink. Each patient was also contacted at home at least once during the period of the weighed intake in case any problems had arisen. Each item of food on the completed diaries was coded according to the tables of McCance & Widdowson‘ and total nutrient intakes calculated using a computer program (’ Microdiet ’, University of Salford). Each patient was required to complete a 7-day weighed dietary inventory and a 3-day stool collection. To confirm completeness of the stool collections, each patient took 3 capsules each containing 8 radio-opaque pellets every day during the 7 days of the weighed intake period and the number of radio-opaque pellets If 72 pellets were visible, present in the stools was determined by X-ray the sample was regarded as complete, but in samples deemed incomplete the wet weight was multiplied by a factor of 72 and divided by the number of pellets counted. All stools passed during the final 3 days of the test week were collected in polythene bags and stored on ice. After delivery to the laboratory they were stored at -20 O C until analysis was carried out. Stools were weighed, freeze-dried and 31
BAP 6
552
G. M O R R I S O N ef al.
weighed again. Milled samples were analysed for dry matter (forced-draught oven at 100 "C for 24 h), ash (mufflefurnace at 550 "C overnight), protein (Kjeldahl), fat (Soxhlet) following acid hydrolysis, and energy (adiabatic bomb calorimeter, Gallenkamp). These methods are used routinely in the Department of Food and Agricultural Chemistry, The Queen's University of Belfast. Coefficients of absorption were then calculated by dividing the gross intake less the faecal losses, by the gross intake, in respect of total energy, fat and protein. After 2 weeks, when the above procedure was completed, Pancrease HL was substituted for Creon, at a daily dose of one-third the number of capsules. The patients or carers were advised to adjust the pattern of capsule-taking with meals and snacks in order to achieve the recommended reduction. When the number normally taken could not neatly be divided by 3, patients were advised to round down to a smaller number. In the fourth week, the test procedures described above were repeated. A 7-day weighed dietary intake was again carried out, with the patients strongly encouraged to follow the same diet in this second test week. Three-day stool collections and laboratory analyses were carried out as before. After an interim analysis of the first cohort, subsequent patients were advised to round up to the nearest larger number, Capsule intakes of 1 or 2, common with snacks, could not be decreased below a single capsule of the new compound. All statistical significance tests were carried out at the 5 % level and were twosided. Digestibility ratios were summarized, and treatment differences were analysed using the analysis of variance. Data processing and statistical analysis were carried out using SAS statistical software (release 6.04).
RESULTS Two patients withdrew their consent before the study began, and 3 further patients failed to complete necessary procedures at various times and did not therefore complete the protocol. Statistical analysis was not performed on the data from 6 further patients, 3 of whom developed gastroenteritis while another 3 deviated significantly from the study protocol. The results presented therefore refer to the 27 patients who fully completed the study. The diary records were used to test between periods for symptoms commonly considered relevant to the dose of pancreatic enzyme supplements. There was no significant difference between the number, colour or consistency of stools during the two periods nor in the incidence or severity of abdominal pain. Stool odour was regarded as markedly less during the Pancrease HL period, and the difference on subjective coding was highly significant ( P < 0.001). No significant differences emerged in appetite or eating habits. Food intake was comparable in both test weeks (Table 3 ) . The number of capsules ingested was calculated both from the number issued less the number returned, and the number recorded on the diary during the test weeks. Some discrepancies emerged, due to the fact that very large numbers of
C O M P A R I S O N OF PANCREATIC SUPPLEMENTS 553 Table 3. Mean food intake (S.D.) during study weeks ( n = 27)
Energy, mJ kCal Fat, g Protein, g
Pancrease' HL
Creon
9.21 (3.01) 2202 (720) 89.9 (33.9) 65.8 (23.8)
9.09 (2.49) 2173 (594) 87.8 (26.8) 66.2 (27.9)
Table 4. Coefficients (%) of apparent digestibility for fat, energy, protein and dry matter ( n = 27)
Fat Energy Protein Dry matter
Pancrease' HL
Creon
Mean (S.E.M.) difference (%)
P
83 (9) 90 (4) 86 (6) 92 (3)
84 (10) 89 (4) 82 (7) 91 (3)
-1 (2) I(1) 4 (0 I(1)
0.640 0.140 0.007 0.120
capsules were issued, while patients previously taking Creon sometimes took capsules from their pre-existing supplies. Sibling pairs (there were 3 ) sometimes shared capsule pots. Coefficients of digestibility are shown in Table 4. Seven patients did not reduce capsule intake to the prescribed level, for a variety of reasons, e.g. consumption of excessive snacks, which precluded reduction of capsules below 1.The remaining 20 attained a ratio reasonably close to the planned 0.33: 1 Pancrease HL:Creon ratio (mean 0.39, range 0.27-0.50). When this core group of 20 were analysed alone, control of digestibility was maintained for fat (85% us. 84%), energy (89% us. 91%) and dry matter (91% us. 93%), while digestibility of protein improved from 82% to 86% ( P < 0.007). Separate analysis of the 7 patients who reduced their capsule intake less than instructed showed no significant difference in digestibility between the periods on Creon and Pancrease HL (fat 83 % us. 82 YO,energy 89% us. 90 %, protein 83 % us. 85 % and dry matter 91% us. 91%). No adverse effect attributable to the medication was recorded. Patient and parent acceptability of the new compound was high. Only one of the 27 patients whose results have been presented preferred Creon. Of the entire 33 patients who were able to compare both preparations, only 3 expressed a preference for Creon. Two of these were among those excluded from analysis because of coincidental reported 'gastroenteritis ', which was not supported by bacteriological evidence and had occurred in one while on the new compound and in the other while on 31 2
554
G. MORRISON ef a/.
Creon. The remaining patient thought that he had fewer abdominal symptoms on his usual Creon, although 4 others reported fewer abdominal pains on Pancrease HL. The great reduction in capsule number was cited in all cases as the sole or main reason for preferring Pancrease HL.
DISCUSSION The dose of pancreatin capsules taken by cystic fibrosis patients is usually determined simply by subjective reports of effect on the bowel habits. The effect of inadequate lipase dosage is apparent as steatorrhoea, but signs of poor digestibility of other nutrients are less obvious, apart from the unpleasant odour associated with uncontrolled protein excretion. Repeated nutrient balances, although regarded as the best measure of effectiveness, are rarely performed except for research purposes. The drawbacks of such balance studies are well documented.8 Faecal output is not necessarily a reliable guide to absorption. Apart from the human errors associated with faecal collection, it has been calculated that approximately 30% of the stool energy is derived from colonic bacteria and endogenous secretions, which may be particularly enhanced in cystic fibrosis children.' Although most objective comparisons have concentrated on fat balance, it is likely that consideration of energy balance will be more informative. Moreover, the importance of protein deficiency in failure of growth in childhood is well known. In this study we therefore measured protein and energy intake and losses in addition to those of fat. The results suggest that the higher concentration of protease in the Pancrease HL (Table I) enhanced protein absorption. Although dietary protein deficiency would not be expected in British patients with cystic fibrosis there is evidence that even stable (but chronically infected) patients are often in a state of protein catabolism." Although the group analyses gave results close to those predicted, they mask considerable variations between patients. Moreover, the relative efficiency of fat, protein and energy digestion varied widely within individual patients. Not all patients were optimally controlled at the beginning of the study, although we eliminated from analysis any in whom there was a substantial difference in intake or variation from the protocol between the 2 study weeks. We conclude that the potencies of the two preparations are broadly those which are indicated by their composition, and as a starting point for conversion 1 high-lipase capsule should be substituted for 3 of the standard preparation. However, we would recommend that capsule numbers are rounded up rather than down when the number of standard capsules is not divisible by 3. Because of the overwhelming patient preference for the new formulation we would regard products of this type as being a significant advance in the nutritional management of pancreatic insufficiency. Our preliminary follow-up observations suggest that the greater compliance which follows reduction in capsule numbers is reflected in better appetite and in some cases a consequent improvement in weight gain.
C O M P A R I S O N OF PANCREATIC SUPPLEMENTS
555
ACKNOWLEDGEMENTS
This study was supported by a grant from Cilag Ltd, High Wycombe, UK. REFERENCES 1 Corey M, McLaughlin F J, Williams M,
Levison H. A comparison of survival, growth and pulmonary function in patients with cystic fibrosis in Boston and Toronto. J Clin Epidemiol 1988; 41: 583-91. 2 Williams J, MacDonald A, Weller P H, Fields J, Pandov H. Two enteric coated microspheres in cystic fibrosis. Arch Dis Child 1990; 65 : 594-7. 3 Robinson P J, Olinsky A, Smith A L,
Chitravanshi S B. High compared with standard dose lipase pancreatic supplements. Arch Dis Child 1989; 64: 143-45. 4 Owen G, Peters T J, Dawson S, Goodchild M C. Pancreatic enzyme supplement dosage in cystic fibrosis. The Lancet 1991; 338: 1153. 5 Marr J. Individual dietary surveys : purposes
and methods. World Review Nutr Diet 1971; 13: 105-64. 6 Paul AA, Southgate D A T. McCance and
Widdowson’s the Composition of Foods. London: HMSO, 1978. 7 Simpson F G, Hall G P, Kelleher J, Losovosky M S. Radio-opaque pellets as faecal markers for faecal fat estimation in malabsorption. Gut 1979; 20: 581-4. 8 Holmes G K T, Hill P G. Do we still need to measure faecal fat? Br Med J 1988; 296: 1552-3. 9 Murphy J L , Wootton S A , Bond S A , Jackson A A . Energy content of stools in
normal healthy controls and patients with cystic fibrosis. Arch Dis Child 1991; 66: 495-500 10 Holt T L, Ward L C, Francis P J, Isles A,
Cooksley W G E, Shepherd R W. Whole body protein turnover in malnourished cystic fibrosis patients and its relationship to pulmonary disease. Am J Clin Nutr 1985 ; 41 : 1061-6.
