- Email: [email protected]
Gastrointestinal protein loss in children recovering from burns
Gastrointestinal
Protein Loss in Children
By I. Matoth,
E. Granot,
A. Gorenstein,
Recovering
K. Abu-Dalu,
From Burns
and K. Goitein
Jerusalem, israel 0 Qualitative gastrointestinal protein loss was evaluated in 10 children with second- and/or third-degree burns covering 10% or more of their body surface area (BSA) by using fecal a-1-antitrypsin (FA-l-AT) as a marker. Patients were subdivided according to the extent of the burned area: group I (5 patients) had burns covering less than 20% of BSA; group II (5 patients) had burns covering more than 20% of BSA (mean, 37.2% = 24.9%). Results were compared with those
of random, nondried stool specimen has been validated.“]* This technique has obvious advantages over the ‘Cr-albumin excretion method; it is noninvasive, nonradioactive, does not require prolonged stool collections, and is simple to perform. Using the FA-l-AT as an indicator, we have studied intestinal protein loss in children with bums.
of 12 healthy normal controls. Mean maximal FA-I-AT excretion in group II patients (2.71 -c 1.35 mg/g) was significantly greater than that found in group I children (0.43 f 0.26 mg/g;
MATERIALS
AND
METHODS
P = .006) and in the controls (0.62 + 0.25 mgi’g; P = .004). The mean maximal FA-l-AT excretion positively correlated to
Patients
the percent of BSA covered with burns (r = 0.63). Although the mean septic sco’re (SS) of group I patients (7 2 2.9) was significantly greater than that calculated for group II children (3 2 2.45; P = .047), only 2 patients in group II had positive microbiological caltures. Patients in both groups had received more than the recommended enteral caloric and protein allowance during the 96 hours prior to the maximal FA-I-AT measurements. Within this range, no correlation was found between the amount of FA-l-AT and the number of calories per kilogram protein consumed. By using the method of FA-l-AT quantification, this study provides the first report on postburn intestinal protein loss in children. Copyright 0 199 1 by W.B. 98Und8rs Compen y
Ten children with bums were admitted to the Pediatric Surgery Department and the Pediatric Intensive Care Unit (PICU) of the Hadassah University Hospital, Jerusalem, during a 5-month period from June 1,1987 to November 1,1987. All these patients had second- and/or third-degree burns caused by hot water, covering 10% or more of their body surface area (BSA). The patients were divided according to the extent of the burn area. Five patients (4 boys, 1 girl) were 1.15 to 8 years old (mean, 4.23 2 2.66 years) and had second-degree burns covering less than 20% of BSA (mean, 11.2% f 1.79%). These patients constitute group I. The other 5 children (3 boys, 2 girls) were 1.3 to 8 years old (mean, 3.52 f 2.62 years) and had second- and/or third-degree burns ranging from 20% to 80% of BSA (mean, 37.2% ? 24.88%) and constitute group II. Twelve healthy children (4 girls, 8 boys) aged 1 to 11 years (mean, 5.58 + 4.03 years) served as normal controls. Routine laboratory tests including complete blood cell counts, serum electrolytes, urea, creatinine, glucose, bilirubin, and albumin were performed every 1 to 3 days depending on the patient’s general health condition. If the required blood tests were not available on the days of FA-l-AT measurements, the data obtained nearest to these days were used (in all patients this was obtained within 48 hours of FA-l-AT evaluation). Blood gases were obtained on admission and whenever a deterioration in the patient’s clinical state was noted. Blood and urine cultures were taken if a temperature > 37.5”C was recorded (this occurred at least twice weekly). Cultures from the burn areas were taken whenever a suspicion of a local infection was raised. Chest x-ray was performed when indicated clinically. The data were used to calculate the septic (SS) and infective scores (IS).‘.r3 The SS considers data from the following categories: (1) local effects of tissue infection; (2) secondary effects of infection; (3) pyrexia; and (4) laboratory data (including complete blood cell
INDEX
WORDS:
Burns,
protein-losing
enteropathy;
a-l-
antitrypsin.
P
ROTEIN REQUIREMENTS of both adults and children are supernormal following burns.‘,’ This increased need is attributed to the breakdown of somatic protein, redistribution of proteins into an edematous interstitium, and loss through the bum wound.3 Although not systematically documented before, the possibility of protein loss through the intestinal tract was raised by a few investigators.233 Burns are as yet not considered a cause of proteinlosing enteropathy.4 Increased permeability of the intestinal mucosa during the postburn period has been described in both animals and human adults.‘-’ This phenomenon is attributed to a decreased mucosal blood flow with a loss of mucosal cell mass,6 bacterial endotoxin effects,728 and bacterial overgrowth induced by oral antibiotics,9 the increased intestinal mucosal permeability has been documented by studying [15N]glycin excretion,* lactulose/ manitol absorption: and bacterial translocation.7 The reliability of fecal ol-1-antitrypsin (FA-l-AT) determination as an indicator of abnormal gastrointestinal (GI) protein loss has been confirmed in number of studies, and simplification of the method by the use
JournatofPediatric Surgery, Vol26, No 10 (October), 1991: pp 1175-l 178
From the Depattments of Pediatrics and Pediatric Surgery, Hadassah Hebrew UniversityMedical Center, Jerusalem, Israel. Date accepted: September 151990. Supported in part by a grant )?om the Joint Research Fund of the Hebrew University and Haddassah. Address reprint requests to Israel Matoth, MD, Division of Neurology, Children> Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104. Copyright o 1991 by W.B. Saunders Company 0022-3468/9112610-0007$03.00/0
1175
MATOTH ET AL
1176
counts and plasma albumin levels). The IS incorporates a variety of factors related to infection, including the presence of urinary tract infections, glucose tolerance, hemodynamic stability, and the temporal relations between positive blood cultures and clinical infection. These scoring systems were used to determine whether a clinically significant infectious process was present. If such an infection is not present, the individual’s SS and IS are 0 and 1, respectively.5 Cultures taken during the 96 hours preceding FA-I-AT measurement were used for the evaluation of SS and IS. All the children were treated prophylactically by penicillin. Other antibiotic therapy was added when clinical evidence suggesting the presence of infection appeared. Enteral calories and protein intakes were estimated throughout the hospitalization period and were compared to the estimated requirements according to Pasulka and Watchtel.“’
FA-I-AT Determination Stool samples for FA-l-AT determination were collected starting with the first postburn stool and than every 2 to 3 days, as available. Stool samples were collected throughout the hospitalization period, which ranged from 6 to 42 days (mean, 18.1 + 14.8 days). Watery stools were excluded from the study. Determination of FA-l-AT was performed using a previously described method.” Samples were taken and stored at -20°C until analysed. An aliquot of 1 g was diluted 1:lO with 0.9% saline solution and then homogenized. After centrifugation at 2,300g for 15 minutes, 20-pL samples of supernatant were placed into the wells of commercially available immunodiffusion plates (LC Partigen; Behring Werke, Marburg, Germany). The precipitation rings were measured after 3 days. A reference curve was established using a standard serum A-l-AT solution. Results were expressed as milligram of FA-l-AT per gram wet weight of stool.
Statistical Analysis Data analysis was performed on a computer using a statistical software package (Statpak V.3.1; Northwest Analytical, Inc, Portland, OR). Analysis of means, standard deviation, variance, correlations, and t testing were applied as appropriate. A P value of < .OSwas accepted as significant. Values are expressed as mean + SD.
RESULTS
Population
The mean age of group I patients (4.23 f 2.66 years) was not statistically different from that of the group II patients (3.52 + 2.62 years) and that of the controls (5.58 -+ 4.03 years) (P > .l; Table 1). As was expected, there was a significant difference between the percentages of BSA covered with burns in the two patient groups: 11.2% + 1.79% in group I versus 37.2% + 24.88% in group II (P = .048). FA-I-AT Excretion
The mean maximal FA-l-AT excretion in the group II children was 2.71 f 1.35 mg/g (range, 1.35 to 4.5 mg/g). This value was significantly greater than that found in the group I children (0.43 k 0.26 mg/g; range, 0.2 to 0.8 mg/g; P = .006) and in the controls (0.62 2 0.25 mg/g; range, 0.2 to 1 mg/g; P = .004). There was no statistically significant difference in the maximal mean FA-l-AT excretion values between the group I children and the controls (P = .38) (Fig 1). As a whole, a correlation was found between the percent of BSA covered with burns and the mean maximal FA-l-AT excreted (r = .83). The maximal FA-l-AT values were noted on day 6 f 3.39 (range, 2 to 10) postburn in the group I patients and on day 11 + 8.28 (range, 4 to 21) in group II (Fig 2). No significant difference in the day of maximal FA-l-AT excretion was found between the two patient groups (P = .25). Blood Albumin
In all the burned children, blood albumin levels had their nadir on the same days on which the maximal FA-I-AT values were documented. The
Table. 1. Characteristics of Patients and Controls Burn Area
Group
1
Mean f SD Group II
Positive
Postburn
FA-l-AT
Cultures
DW
(msl9)
Age Iv)
(% BSA)
IS
5.5
12
2
6
2
0.6
1.15
10
2
2
-
8
0.8
a
14
1
0
-
3
0.3
2.5
IO
2
2
-
7
4
10
2
5
11.2 f 1.8
1.8 f 0.4
2.3
36
2
4
-
19
3.5
20
2
5
-
4
4.2 k 2.7
ss
Burn exudate
3.0 % 2.4
1.3
20
2
6
2.5
30
3
11
Blood
-
8
80
4
10
Blood & respiratory
0.2
10
0.24
6.0 f 3.4
0.4 ‘t 0.3 1.35 1.5
6
2.6
21
3.6
5
4.5
tract secretion Mean ? SD Controls Mean f SD
3.4 r 2.6
37.2 f 24.9
5.6 2 4
*Mean of FA-l-AT in controls (not postburn).
2.6 2 0.9
7.0 2 2.9
11.0 k 8.3
2.7 r 1.4 0.6 t 0.3’
1177
GI PROTEIN LOSS IN BURN PATIENTS
Blood cultures were positive for bacteria in only 2 of the 5 patients in group II. In one of these patients, Haemophilus influenzae was isolated from the respiratory tract secretions. No positive blood caltures were obtained from the group II patients. In one child, Staphylococcus aureus was isolated from the burns exudate. When calculated for the whole patient group, a correlation was found between the IS and SS and the amount of mean maximal FA-l-AT excretion (r = -86 and r = .85, respectively).
4, z -P P
2
7
2
Calories and Protein
0 Controll Fig 1.
Mean maximal FA-l-AT excretion and burn area (% BSh$
1 mean minimal blood albumin level of grou (34.6 ? 3.36mg/dL) was significantly higher than t at Pi of group II (23.4 ? 0.55 mg/dL; P = .OOOl). Infection
Because 9 of the 10 patients had prolonged peric IS of fever during their hospitalization (including dur g the 24 hours prior to the maximal FA-l-AT measu :ments), only one patient had an IS of 1. According 0 Ziegler et a1,5an IS of 2 or more is considered to e evidence of clinically significant infection. The rn< IS of both groups (group I, 1.8 + 0.45; group 2.6 + 0.89) was not significantly different (P = .ll) statistically significant difference between the t 0 groups was noted when the mean SS were calcula (7 2 2.92 in group II v 3 & 2.45 in group I; P = O.OL
r” 4: ;t
5-
4g & 5
3-
I
10 Post Burn Day Fig 2.
Day of maximal FA-l-AT excretion.
I
20
Enteral feeding (orally or through a feeding tube) was initiated early in the course of hospitalization in all the patients. The children in group I were able to consume food orally starting either on the day of admission or on the following day and were served with at least 100% of the calories and protein recommended by Pasulka and Watchtel.14 The children in group II were fed through a feeding tube (alone or in combination with oral feeding) starting 2 to 3 days postburn. During the 96 hours prior to the maximal FA-l-AT measurements, these patients had received an average of 118.22% + 53.97% (range, 62.47% to 196.36%) of the calories and 101.36% + 49.98% (range, 55.85% to 167.58%) of the protein recommended by Davis. No correlation was found between the amount of FA-l-AT and the number of calories per kilogram of protein consumed. DISCUSSION
This study demonstrates the existence of protein loss through the intestinal mucosal barrier in children having burns of 20% or more of their BSA. This protein loss is up to fivefold higher than that noted in children having burns of less than 20% of their BSA, or than that found in healthy children. Protein loss was positively correlated with the extent of the burn area. Maximal protein loss, as determined by FA-l-AT, was found to occur on day 6 ‘_ 3.39. Zeigler et al,’ based on a study of lactulose/manitol absorption in burn adults, noted that maximal mucosal permeability occurred on day 18 ? 5 postburn. The present results are similar to those of Majeima et al, who found that in 40% burned rats, translocation of viable Escherichia coli (of lo-fold more than in controls) occurs on days 2 to 4 postburn.’ Zeigler et al’ emphasize the association between the increased GI mucosal permeability and the presence of infection. Although they were unable to determine whether the increased permeability caused
1178
MATOTH ET AL
the infection or vice versa, they suggested that local infection contributed to the observed permeability changes. In the present study, when all patients’ data are grouped together, we could establish a correlation between the SS and IS and the amount of FA-l-AT excreted. When using the conventional scoring system, 9 of the 10 patients have SS and IS of above 0 and 1, respectively, and, therefore, fit the criteria of a significant clinical infection. Despite repeated blood and urine cultures and stool examinations for cultures and parasites, we could document an overt bacterial infection in only 3 patients. The SS IS scoring system is based on multiple clinical variables. In burn patients, a number of these parameters may in effect actually be influenced by factors other than infection. The low blood albumin level can be caused by the redistribution of proteins into the edematous interstitium and from protein loss across burn wounds.3 Moderately elevated white blood cells counts can be attributed to the secretion of stress hormones,15 and the fever could be the result of the increased heat production in these patients.’ Because the present patients had low scoring, it might be that only a few were infected. In this study, the mean maximal FA-l-AT correlated with the extent of the burned area. Was it caused by the decreased GI mucosal blood flow, which tends to be more evident with the increase of the burned area? Or was this a result of the higher susceptibility to infection? Of great interest is the finding that maximal FA-l-AT occurred concomi-
tantly with the nadir of blood albumin levels. Because our method is only qualitative, we are not able to prove that the increased protein loss can in effect be partially responsible for the decreased albumin levels. Although diminished enteral nutrition may enhance the GI mucosal permeability,6@ this was not the case in the present patients because most of them were maintained on an adequate enteral diet. On average, the children were receiving 118% of the recommended caloric allowance with a range of 62.47% to 196.36%. Within this range, no correlation could be demonstrated between the enteral caloric intake and the excretion of FA-l-AT. FA-l-AT has previously been used as an indicator of protein loss in diseases characterized by GI mucosal lesions (eg, celiac disease, Crohn’s disease, and transient GI protein Ioss).~~ Although the possibility that intestinal protein loss occurs in burn patients has been previously mentioned,2.3 to our knowledge, this has not been measured. By using the method of FA-l-AT quantification, this study provides the first report on postburn intestinal protein loss in children. These findings are intriguing, and we believe that they warrant further studies to quantify the amount of protein loss and the various factors that influence it. ACKNOWLEDGMENT The authors are grateful to Irena Gal for her expert technical assistance.
REFERENCES 1. Wilmore DW: Nutrition and metabolism following thermal injury. Clin Plast Surg 1:603-609,1974 2. Kien GL, Young VR, Rohrbaugh AM, et al: Increased rates of protein synthesis and breakdown in children recovering from bums. Ann Surg 187:383-391,1978 3. Waxman K, Rebello T, Pinderski L, et al: Protein loss across burn wounds. J Trauma 27:136-140,1987 4. Schmitz F: Protein-losing enteropathies, in Milla PJ, Muller PR (eds): Harries’ Pediatric Gastroenterology. London, England, Churchill-Livingstone, 1988, pp 260-271 5. Ziegler TR, Smith RJ, O’Dwyer ST, et al: Increased intestinal permeability associated with infection in bum patients. Arch Surg 123:1313-1319,1988 6. Mochizuki H, Trocki 0, Dominioni L, et al: Mechanism of prevention of postbum hypermetabolism and catabolism by early enteral feeding. Ann Surg 200:297-310,1984 7. Maejima M, Deitch EA, Berg RD: Bacterial translocation from the gastrointestinal tract of rats recovering thermal injury. Infect Immun 43:6-lo,1984 8. Deitch EA, Berg RD, Specian R: Endotoxin promotes the translocation of bacteria from the gut. Arch Surg 122:185-190,1987 9. Berg RD: Promotion of translocation of enteric bacteria from
the gastrointestinal tract of mice by oral treatment with penicillin, clindamycin, or metronidazole. Infect Immun 33:854-861,1981 10. Thomas DW, Sinatra FR, Merritt RJ: Random fecal alpha1-antitrypsin concentration in children with gastrointestinal disease. Gastroenterology S&776-782,1981 11. Magazzu G, Jacono G, Di Pasquale G, et al: Reliability and usefulness of random fecal alpha-1-antitrypsin concentration: Further simplification of the method. J Pediatr Gastroenterol Nutr 4:402-407,1985 12. Catassi C, Cardinali E, D’Augelo G, et al: Reliability of random fecal alph-l-antitrypsin determination on nondried stool. J Pediatr 109:5OQ-502,1986 13. Elebute EA, Stoner HB: The grading of sepsis. Br J Surg 70:29-31,1983 14. Pasulka PS, Watchtel TL: Nutritional considerations for the burned patient. Surg Clin North Am 67:109-131,1987 15. Wolfe R, Goodenough RD, Burke JF, et al: Response of protein and urea kinetics in bum patients to different levels of protein intake. Ann Surg 197:163-171,1983 16. Deitch EA, Winterton J, Li M, et al: The gut as a portal of entry for bacteria: Role of protein malnutrition. Ann Surg 205:681682,1987
Protein Loss in Children
By I. Matoth,
E. Granot,
A. Gorenstein,
Recovering
K. Abu-Dalu,
From Burns
and K. Goitein
Jerusalem, israel 0 Qualitative gastrointestinal protein loss was evaluated in 10 children with second- and/or third-degree burns covering 10% or more of their body surface area (BSA) by using fecal a-1-antitrypsin (FA-l-AT) as a marker. Patients were subdivided according to the extent of the burned area: group I (5 patients) had burns covering less than 20% of BSA; group II (5 patients) had burns covering more than 20% of BSA (mean, 37.2% = 24.9%). Results were compared with those
of random, nondried stool specimen has been validated.“]* This technique has obvious advantages over the ‘Cr-albumin excretion method; it is noninvasive, nonradioactive, does not require prolonged stool collections, and is simple to perform. Using the FA-l-AT as an indicator, we have studied intestinal protein loss in children with bums.
of 12 healthy normal controls. Mean maximal FA-I-AT excretion in group II patients (2.71 -c 1.35 mg/g) was significantly greater than that found in group I children (0.43 f 0.26 mg/g;
MATERIALS
AND
METHODS
P = .006) and in the controls (0.62 + 0.25 mgi’g; P = .004). The mean maximal FA-l-AT excretion positively correlated to
Patients
the percent of BSA covered with burns (r = 0.63). Although the mean septic sco’re (SS) of group I patients (7 2 2.9) was significantly greater than that calculated for group II children (3 2 2.45; P = .047), only 2 patients in group II had positive microbiological caltures. Patients in both groups had received more than the recommended enteral caloric and protein allowance during the 96 hours prior to the maximal FA-I-AT measurements. Within this range, no correlation was found between the amount of FA-l-AT and the number of calories per kilogram protein consumed. By using the method of FA-l-AT quantification, this study provides the first report on postburn intestinal protein loss in children. Copyright 0 199 1 by W.B. 98Und8rs Compen y
Ten children with bums were admitted to the Pediatric Surgery Department and the Pediatric Intensive Care Unit (PICU) of the Hadassah University Hospital, Jerusalem, during a 5-month period from June 1,1987 to November 1,1987. All these patients had second- and/or third-degree burns caused by hot water, covering 10% or more of their body surface area (BSA). The patients were divided according to the extent of the burn area. Five patients (4 boys, 1 girl) were 1.15 to 8 years old (mean, 4.23 2 2.66 years) and had second-degree burns covering less than 20% of BSA (mean, 11.2% f 1.79%). These patients constitute group I. The other 5 children (3 boys, 2 girls) were 1.3 to 8 years old (mean, 3.52 f 2.62 years) and had second- and/or third-degree burns ranging from 20% to 80% of BSA (mean, 37.2% ? 24.88%) and constitute group II. Twelve healthy children (4 girls, 8 boys) aged 1 to 11 years (mean, 5.58 + 4.03 years) served as normal controls. Routine laboratory tests including complete blood cell counts, serum electrolytes, urea, creatinine, glucose, bilirubin, and albumin were performed every 1 to 3 days depending on the patient’s general health condition. If the required blood tests were not available on the days of FA-l-AT measurements, the data obtained nearest to these days were used (in all patients this was obtained within 48 hours of FA-l-AT evaluation). Blood gases were obtained on admission and whenever a deterioration in the patient’s clinical state was noted. Blood and urine cultures were taken if a temperature > 37.5”C was recorded (this occurred at least twice weekly). Cultures from the burn areas were taken whenever a suspicion of a local infection was raised. Chest x-ray was performed when indicated clinically. The data were used to calculate the septic (SS) and infective scores (IS).‘.r3 The SS considers data from the following categories: (1) local effects of tissue infection; (2) secondary effects of infection; (3) pyrexia; and (4) laboratory data (including complete blood cell
INDEX
WORDS:
Burns,
protein-losing
enteropathy;
a-l-
antitrypsin.
P
ROTEIN REQUIREMENTS of both adults and children are supernormal following burns.‘,’ This increased need is attributed to the breakdown of somatic protein, redistribution of proteins into an edematous interstitium, and loss through the bum wound.3 Although not systematically documented before, the possibility of protein loss through the intestinal tract was raised by a few investigators.233 Burns are as yet not considered a cause of proteinlosing enteropathy.4 Increased permeability of the intestinal mucosa during the postburn period has been described in both animals and human adults.‘-’ This phenomenon is attributed to a decreased mucosal blood flow with a loss of mucosal cell mass,6 bacterial endotoxin effects,728 and bacterial overgrowth induced by oral antibiotics,9 the increased intestinal mucosal permeability has been documented by studying [15N]glycin excretion,* lactulose/ manitol absorption: and bacterial translocation.7 The reliability of fecal ol-1-antitrypsin (FA-l-AT) determination as an indicator of abnormal gastrointestinal (GI) protein loss has been confirmed in number of studies, and simplification of the method by the use
JournatofPediatric Surgery, Vol26, No 10 (October), 1991: pp 1175-l 178
From the Depattments of Pediatrics and Pediatric Surgery, Hadassah Hebrew UniversityMedical Center, Jerusalem, Israel. Date accepted: September 151990. Supported in part by a grant )?om the Joint Research Fund of the Hebrew University and Haddassah. Address reprint requests to Israel Matoth, MD, Division of Neurology, Children> Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104. Copyright o 1991 by W.B. Saunders Company 0022-3468/9112610-0007$03.00/0
1175
MATOTH ET AL
1176
counts and plasma albumin levels). The IS incorporates a variety of factors related to infection, including the presence of urinary tract infections, glucose tolerance, hemodynamic stability, and the temporal relations between positive blood cultures and clinical infection. These scoring systems were used to determine whether a clinically significant infectious process was present. If such an infection is not present, the individual’s SS and IS are 0 and 1, respectively.5 Cultures taken during the 96 hours preceding FA-I-AT measurement were used for the evaluation of SS and IS. All the children were treated prophylactically by penicillin. Other antibiotic therapy was added when clinical evidence suggesting the presence of infection appeared. Enteral calories and protein intakes were estimated throughout the hospitalization period and were compared to the estimated requirements according to Pasulka and Watchtel.“’
FA-I-AT Determination Stool samples for FA-l-AT determination were collected starting with the first postburn stool and than every 2 to 3 days, as available. Stool samples were collected throughout the hospitalization period, which ranged from 6 to 42 days (mean, 18.1 + 14.8 days). Watery stools were excluded from the study. Determination of FA-l-AT was performed using a previously described method.” Samples were taken and stored at -20°C until analysed. An aliquot of 1 g was diluted 1:lO with 0.9% saline solution and then homogenized. After centrifugation at 2,300g for 15 minutes, 20-pL samples of supernatant were placed into the wells of commercially available immunodiffusion plates (LC Partigen; Behring Werke, Marburg, Germany). The precipitation rings were measured after 3 days. A reference curve was established using a standard serum A-l-AT solution. Results were expressed as milligram of FA-l-AT per gram wet weight of stool.
Statistical Analysis Data analysis was performed on a computer using a statistical software package (Statpak V.3.1; Northwest Analytical, Inc, Portland, OR). Analysis of means, standard deviation, variance, correlations, and t testing were applied as appropriate. A P value of < .OSwas accepted as significant. Values are expressed as mean + SD.
RESULTS
Population
The mean age of group I patients (4.23 f 2.66 years) was not statistically different from that of the group II patients (3.52 + 2.62 years) and that of the controls (5.58 -+ 4.03 years) (P > .l; Table 1). As was expected, there was a significant difference between the percentages of BSA covered with burns in the two patient groups: 11.2% + 1.79% in group I versus 37.2% + 24.88% in group II (P = .048). FA-I-AT Excretion
The mean maximal FA-l-AT excretion in the group II children was 2.71 f 1.35 mg/g (range, 1.35 to 4.5 mg/g). This value was significantly greater than that found in the group I children (0.43 k 0.26 mg/g; range, 0.2 to 0.8 mg/g; P = .006) and in the controls (0.62 2 0.25 mg/g; range, 0.2 to 1 mg/g; P = .004). There was no statistically significant difference in the maximal mean FA-l-AT excretion values between the group I children and the controls (P = .38) (Fig 1). As a whole, a correlation was found between the percent of BSA covered with burns and the mean maximal FA-l-AT excreted (r = .83). The maximal FA-l-AT values were noted on day 6 f 3.39 (range, 2 to 10) postburn in the group I patients and on day 11 + 8.28 (range, 4 to 21) in group II (Fig 2). No significant difference in the day of maximal FA-l-AT excretion was found between the two patient groups (P = .25). Blood Albumin
In all the burned children, blood albumin levels had their nadir on the same days on which the maximal FA-I-AT values were documented. The
Table. 1. Characteristics of Patients and Controls Burn Area
Group
1
Mean f SD Group II
Positive
Postburn
FA-l-AT
Cultures
DW
(msl9)
Age Iv)
(% BSA)
IS
5.5
12
2
6
2
0.6
1.15
10
2
2
-
8
0.8
a
14
1
0
-
3
0.3
2.5
IO
2
2
-
7
4
10
2
5
11.2 f 1.8
1.8 f 0.4
2.3
36
2
4
-
19
3.5
20
2
5
-
4
4.2 k 2.7
ss
Burn exudate
3.0 % 2.4
1.3
20
2
6
2.5
30
3
11
Blood
-
8
80
4
10
Blood & respiratory
0.2
10
0.24
6.0 f 3.4
0.4 ‘t 0.3 1.35 1.5
6
2.6
21
3.6
5
4.5
tract secretion Mean ? SD Controls Mean f SD
3.4 r 2.6
37.2 f 24.9
5.6 2 4
*Mean of FA-l-AT in controls (not postburn).
2.6 2 0.9
7.0 2 2.9
11.0 k 8.3
2.7 r 1.4 0.6 t 0.3’
1177
GI PROTEIN LOSS IN BURN PATIENTS
Blood cultures were positive for bacteria in only 2 of the 5 patients in group II. In one of these patients, Haemophilus influenzae was isolated from the respiratory tract secretions. No positive blood caltures were obtained from the group II patients. In one child, Staphylococcus aureus was isolated from the burns exudate. When calculated for the whole patient group, a correlation was found between the IS and SS and the amount of mean maximal FA-l-AT excretion (r = -86 and r = .85, respectively).
4, z -P P
2
7
2
Calories and Protein
0 Controll Fig 1.
Mean maximal FA-l-AT excretion and burn area (% BSh$
1 mean minimal blood albumin level of grou (34.6 ? 3.36mg/dL) was significantly higher than t at Pi of group II (23.4 ? 0.55 mg/dL; P = .OOOl). Infection
Because 9 of the 10 patients had prolonged peric IS of fever during their hospitalization (including dur g the 24 hours prior to the maximal FA-l-AT measu :ments), only one patient had an IS of 1. According 0 Ziegler et a1,5an IS of 2 or more is considered to e evidence of clinically significant infection. The rn< IS of both groups (group I, 1.8 + 0.45; group 2.6 + 0.89) was not significantly different (P = .ll) statistically significant difference between the t 0 groups was noted when the mean SS were calcula (7 2 2.92 in group II v 3 & 2.45 in group I; P = O.OL
r” 4: ;t
5-
4g & 5
3-
I
10 Post Burn Day Fig 2.
Day of maximal FA-l-AT excretion.
I
20
Enteral feeding (orally or through a feeding tube) was initiated early in the course of hospitalization in all the patients. The children in group I were able to consume food orally starting either on the day of admission or on the following day and were served with at least 100% of the calories and protein recommended by Pasulka and Watchtel.14 The children in group II were fed through a feeding tube (alone or in combination with oral feeding) starting 2 to 3 days postburn. During the 96 hours prior to the maximal FA-l-AT measurements, these patients had received an average of 118.22% + 53.97% (range, 62.47% to 196.36%) of the calories and 101.36% + 49.98% (range, 55.85% to 167.58%) of the protein recommended by Davis. No correlation was found between the amount of FA-l-AT and the number of calories per kilogram of protein consumed. DISCUSSION
This study demonstrates the existence of protein loss through the intestinal mucosal barrier in children having burns of 20% or more of their BSA. This protein loss is up to fivefold higher than that noted in children having burns of less than 20% of their BSA, or than that found in healthy children. Protein loss was positively correlated with the extent of the burn area. Maximal protein loss, as determined by FA-l-AT, was found to occur on day 6 ‘_ 3.39. Zeigler et al,’ based on a study of lactulose/manitol absorption in burn adults, noted that maximal mucosal permeability occurred on day 18 ? 5 postburn. The present results are similar to those of Majeima et al, who found that in 40% burned rats, translocation of viable Escherichia coli (of lo-fold more than in controls) occurs on days 2 to 4 postburn.’ Zeigler et al’ emphasize the association between the increased GI mucosal permeability and the presence of infection. Although they were unable to determine whether the increased permeability caused
1178
MATOTH ET AL
the infection or vice versa, they suggested that local infection contributed to the observed permeability changes. In the present study, when all patients’ data are grouped together, we could establish a correlation between the SS and IS and the amount of FA-l-AT excreted. When using the conventional scoring system, 9 of the 10 patients have SS and IS of above 0 and 1, respectively, and, therefore, fit the criteria of a significant clinical infection. Despite repeated blood and urine cultures and stool examinations for cultures and parasites, we could document an overt bacterial infection in only 3 patients. The SS IS scoring system is based on multiple clinical variables. In burn patients, a number of these parameters may in effect actually be influenced by factors other than infection. The low blood albumin level can be caused by the redistribution of proteins into the edematous interstitium and from protein loss across burn wounds.3 Moderately elevated white blood cells counts can be attributed to the secretion of stress hormones,15 and the fever could be the result of the increased heat production in these patients.’ Because the present patients had low scoring, it might be that only a few were infected. In this study, the mean maximal FA-l-AT correlated with the extent of the burned area. Was it caused by the decreased GI mucosal blood flow, which tends to be more evident with the increase of the burned area? Or was this a result of the higher susceptibility to infection? Of great interest is the finding that maximal FA-l-AT occurred concomi-
tantly with the nadir of blood albumin levels. Because our method is only qualitative, we are not able to prove that the increased protein loss can in effect be partially responsible for the decreased albumin levels. Although diminished enteral nutrition may enhance the GI mucosal permeability,6@ this was not the case in the present patients because most of them were maintained on an adequate enteral diet. On average, the children were receiving 118% of the recommended caloric allowance with a range of 62.47% to 196.36%. Within this range, no correlation could be demonstrated between the enteral caloric intake and the excretion of FA-l-AT. FA-l-AT has previously been used as an indicator of protein loss in diseases characterized by GI mucosal lesions (eg, celiac disease, Crohn’s disease, and transient GI protein Ioss).~~ Although the possibility that intestinal protein loss occurs in burn patients has been previously mentioned,2.3 to our knowledge, this has not been measured. By using the method of FA-l-AT quantification, this study provides the first report on postburn intestinal protein loss in children. These findings are intriguing, and we believe that they warrant further studies to quantify the amount of protein loss and the various factors that influence it. ACKNOWLEDGMENT The authors are grateful to Irena Gal for her expert technical assistance.
REFERENCES 1. Wilmore DW: Nutrition and metabolism following thermal injury. Clin Plast Surg 1:603-609,1974 2. Kien GL, Young VR, Rohrbaugh AM, et al: Increased rates of protein synthesis and breakdown in children recovering from bums. Ann Surg 187:383-391,1978 3. Waxman K, Rebello T, Pinderski L, et al: Protein loss across burn wounds. J Trauma 27:136-140,1987 4. Schmitz F: Protein-losing enteropathies, in Milla PJ, Muller PR (eds): Harries’ Pediatric Gastroenterology. London, England, Churchill-Livingstone, 1988, pp 260-271 5. Ziegler TR, Smith RJ, O’Dwyer ST, et al: Increased intestinal permeability associated with infection in bum patients. Arch Surg 123:1313-1319,1988 6. Mochizuki H, Trocki 0, Dominioni L, et al: Mechanism of prevention of postbum hypermetabolism and catabolism by early enteral feeding. Ann Surg 200:297-310,1984 7. Maejima M, Deitch EA, Berg RD: Bacterial translocation from the gastrointestinal tract of rats recovering thermal injury. Infect Immun 43:6-lo,1984 8. Deitch EA, Berg RD, Specian R: Endotoxin promotes the translocation of bacteria from the gut. Arch Surg 122:185-190,1987 9. Berg RD: Promotion of translocation of enteric bacteria from
the gastrointestinal tract of mice by oral treatment with penicillin, clindamycin, or metronidazole. Infect Immun 33:854-861,1981 10. Thomas DW, Sinatra FR, Merritt RJ: Random fecal alpha1-antitrypsin concentration in children with gastrointestinal disease. Gastroenterology S&776-782,1981 11. Magazzu G, Jacono G, Di Pasquale G, et al: Reliability and usefulness of random fecal alpha-1-antitrypsin concentration: Further simplification of the method. J Pediatr Gastroenterol Nutr 4:402-407,1985 12. Catassi C, Cardinali E, D’Augelo G, et al: Reliability of random fecal alph-l-antitrypsin determination on nondried stool. J Pediatr 109:5OQ-502,1986 13. Elebute EA, Stoner HB: The grading of sepsis. Br J Surg 70:29-31,1983 14. Pasulka PS, Watchtel TL: Nutritional considerations for the burned patient. Surg Clin North Am 67:109-131,1987 15. Wolfe R, Goodenough RD, Burke JF, et al: Response of protein and urea kinetics in bum patients to different levels of protein intake. Ann Surg 197:163-171,1983 16. Deitch EA, Winterton J, Li M, et al: The gut as a portal of entry for bacteria: Role of protein malnutrition. Ann Surg 205:681682,1987