- Email: [email protected]
Fecal leukocytes in stool specimens submitted for Clostridium difficile toxin assay
313
DIAGN MICROBIOL INFECT DIS 1993;16:313-315
Fecal Leukocytes in Stool Specimens Submitted for Clostridium difficile Toxin Assay Christine E. Marx, Arthur Morris, Michael L. Wilson, and L. Barth Reller
To determine their diagnostic utility, fecal leukocytes were sought by methylene blue stain in 502 consecutive stool specimens submitted for Clostridium difficile toxin assay. In addition, the stability of fecal leukocytes was assessed by daily examination of 23 stool specimens stored at 4°C and room temperature. The sensitivity, specificity, and positive and negative predictive values of fecal leukocytes in predicting C. difficile toxin assay results were 28%, 92%, 27%, and 93%, re-
spectively. At 4°C, fecal leukocytes retained morphology for a minimum of 3 days. Leukocytes survived as long in stool specimens containing either C. difficile toxin or an enteric pathogen as they did in stool specimens with neither finding. We conclude that testing stool specimens for fecal leukocytes is not useful for predicting the presence of C. difficile toxin, because 72% of stool specimens positive for C. difficile toxin are negative for fecal leukocytes despite their stability.
INTRODUCTION
m e n s submitted for C. difficile toxin assay for the presence of fecal leukocytes by the m e t h y l e n e blue stain. Additionally, we followed the survival of fecal leukocytes in specimens stored at 4°C a n d r o o m temperature.
Clostridium difficile is the main p a t h o g e n causing antibiotic-associated diarrhea, a n d cytotoxicity neutralization assays are the s t a n d a r d for detection of C. difficile toxin in stool specimens (Bartlett, 1984; Lyerly et al., 1988). Published sources state that 33%50% of patients with C. difficile toxin-induced colitis have fecal leukocytes (Bartlett, 1990; Fekety, 1990; Silva, 1989). These statements, h o w e v e r , are unreferenced and w e have b e e n unable to find supporting data in the literature. Moreover, it is unclear h o w often fecal leukocytes are p r e s e n t in patients with C. difficile toxin-associated diarrhea w h o do not necessarily have p r o v e n colitis. Therefore, to ascertain the utility of the fecal leukocyte test (FLT) in predicting the presence of C. difficile toxin in stool specimens, we prospectively e x a m i n e d 502 stool speciFrom the Clinical Microbiology Laboratory (C.E.M., A.M., M.L.W., L.B.R.), Duke University Medical Center; and the Departments of Pathology (M.L.W., L.B.R.) and Medicine (L.B.R.), Duke University School of Medicine, Durham, North Carolina, USA. Address reprint requests to Dr. C.E. Marx, Clinical Microbiology Laboratory, Box 3879, Duke University Medical Center, Durham, NC 27710, USA. Received 24 August 1992; revised and accepted 4 November 1992. © 1993 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/93/$6.00
MATERIALS AND METHODS
Clostridium difficile Toxin Assay Clostridium difficile toxin was detected with a cytotoxicity neutralization assay that u s e d h u m a n newborn foreskin diploid fibroblasts (Baxter Scientific Products, Deerfield, IL) a n d C. difficile goat antitoxin (Virginia Polytechnic Institute a n d State University, Blacksburg, VA). Stool specimens w e r e diluted as m u c h as I in 4, d e p e n d i n g on stool consistency. The test was p e r f o r m e d in microtiter plates a n d the cells inspected after 24 h of incubation. Stool specimens with a cytopathic change in >10% of the m o n o l a y e r and complete neutralization by antitoxin w e r e considered positive for C. difficile toxin.
Fecal Leukocyte Testing From S e p t e m b e r to D e c e m b e r 1991, all stool specim e n s submitted for C. difficile toxin testing at D u k e University Medical Center w e r e e x a m i n e d for the
314
presence of fecal leukocytes. Stool specimens were stored at 4°C and examined for fecal leukocytes within 48 h. Most specimens (85%) were examined within 24 h. A small amount of stool or mucus was transferred to a clean glass slide with a wooden stick, mixed thoroughly with a drop of methylene blue stain, and coverslipped. After waiting 2-3 min to permit good nuclear staining, the entire slide was scanned under low power ( x 100) to find areas containing fecal leukocytes. Morphology was confirmed under high power ( x 400). If the results of low-power scanning were negative, at least 30 fields were examined under high power ( x 400). Stool specimens were considered positive if two or more leukocytes were present in at least two separate high-power fields. Only leukocytes with distinct nuclear detail and pale-staining cytoplasm were recorded.
Fecal Leukocyte Survival Aliquots from 23 stool specimens submitted for culture for fecal pathogens and found to contain fecal leukocytes were divided, and separate aliquots were stored at 4°C and at room temperature. Each specimen aliquot was examined daily for at least 3 days to compare the morphology of leukocytes stored at the two temperatures. The result of any fecal pathogen culture (Salmonella, Shigella, and Campylobacter) or C. difficile toxin assay was recorded for each patient.
Patients The following information was recorded for each patient: age, gender, and recent peripheral blood white blood cell (WBC) count. The majority of WBC counts were performed on the same day that the C. difficile test was ordered. A total of 88 specimens from leukopenic patients (peripheral WBC count ---1000 x 106/liter) were excluded from the study, because it was thought that leukopenia might negate the FLT.
RESULTS A N D DISCUSSION Fecal Leukocytes in Clostridium difficile Specimens A total of 502 stool specimens from 403 patients were received for C. difficile toxin assay. Patients had a median age of 52 years (range, 1 month to 95 years); 202 specimens were from male and 300 specimens were from female patients; 56 specimens (11%) were from pediatric patients (-<18 years). The relationship between the FLT and C. difficile toxin assay is given in Table 1. The sensitivity and specificity of the FLT
C.E. Marx et al.
TABLE 1 Frequency of Fecal Leukocytes in Stool Specimens Tested for Clostridium difficile Toxin
C. difficile Toxin Assay Fecal Leukocytes
Positive
Negative
Total
Present Absent
13 33
36 420
49 453
Total
46
456
502
in predicting a positive C. difficile toxin assay were 28% and 92%, respectively. The positive and negative predictive values of the FLT were 27% and 93%, respectively. Ten (18%) of 56 specimens from children (eight children ->10 years of age, one child of 2 years of age, and one child of 2 months of age) versus 36 (8%) of 446 from adults contained C. difficile toxin.
Fecal Leukocyte Survival A total of 23 stool specimens with fecal leukocytes were studied for leukocyte survival: five were C. difficile toxin positive, seven contained a fecal pathogen (three Shigella, three Campylobacter, and one Salmonella), and 11 were either negative for both or negative for one and did not have the other test performed. Stool specimens were examined for 318 days (median, 6 days). All specimens stored at 4°C retained good fecal leukocyte morphology for 3 days, but cell morphology deteriorated markedly after 5 days. Although fecal leukocyte morphology deteriorated - 2 days earlier in stool specimens stored at room temperature, leukocytes were still recognizable after storage at room temperature for 2-3 days. Fecal leukocyte morphology did not deteriorate earlier in stool specimens containing C. difficile toxin or a fecal pathogen than in stool specimens with neither finding. Several authors state that fecal leukocytes are present in 33%-50% of stool specimens from patients with C. difficile toxin-induced colitis (Bartlett, 1990; Fekety, 1990; Silva, 1989). These statements, however, are not supported by primary references to the literature and we know of no specific comment on the presence of fecal leukocytes in C. difficile toxinpositive stools. In our study, only 13 (28%) of 46 specimens positive for C. difficile toxin contained fecal leukocytes. Conversely, only 13 (27%) of 49 specimens with fecal leukocytes were positive for C. difficile toxin. Although neonates often harbor C. difficile and its toxin without clinical sequelae, low C. difficile prevalence rates are observed after 6-12 months when
Focal WBCs and Clostridium difficile Toxin
normal fecal flora is established (Bartlett, 1990; Donta and Myers, 1982). Since only one of our pediatric toxin-positive specimens was obtained from an infant, the low frequency of fecal leukocytes in C. difficile toxin-positive specimens is not due to asymptomatic toxin carriage in children. Fecal leukocytes were found to survive for several days at either 4°C or room temperature. The presence of common fecal pathogens or C. difficile toxin did not shorten the time that fecal leukocytes retained their morphology. We conclude that stool specimens stored at 4°C for up to 3 days provide reliable FLT results, and that the low sensitivity for the FLT in predicting toxin-positive stool specimens is not due to fecal leukocyte deterioration. The examination of stool specimens stored at room temperature for up to 3 days may provide less reliable FLT results, especially when leukocytes are present as either single cells or small aggregates rather than in large sheets. These data are consistent with a previous study in which good cellular detail was observed in stool specimens stored at room temperature for as long as 24 h (Harris et al., 1972). Fecal leukocytes are often found in specimens containing common fecal pathogens such as Cam-
315
pylobacter (70%-85%) (Blaser et al., 1982; Sazie and Titus, 1982), Salmonella (36%-90%) (Harris et al., 1972; Pickering et al., 1977; Wolff, 1969), and Shigella (69%100%) (Anderson, 1921; Harris et al., 1972; Korzeniowski et al., 1979; Peirce et al., 1974; Pickering et al., 1977; Wolff, 1969). This suggests that the presence of fecal leukocytes increases the probability of recovering a fecal pathogen. This association does not appear to hold for predicting the presence of C. difficile toxin. We did not undertake a clinical study, however, and it is u n k n o w n whether patients with pseudomembranous colitis are more likely to have fecal leukocytes than those with toxin-associated diarrhea only. This point needs investigation. Also, the duration and magnitude of the diarrhea in our patients were not studied. In our hospital, most FLTpositive stool specimens contain neither C. difficile toxin nor a fecal pathogen. In conclusion, the diagnostic utility of fecal leukocytes in C. difficile-associated disorders is minimal. We thank Nancy G. Henshaw, PhD, and the staff of the Pediatric Virology Laboratory for access to the C. difficile toxin assay results and the stool specimens.
REFERENCES Anderson J (1921) A study of dysentery in the field. Lancet 2:998-1002. Bartlett JG (1984) Antibiotic-associated colitis. Dis Mon 30:154. Bartlett JG (1990) Clostridium difficile: clinical considerations. Rev Infect Dis 12(Suppl):S243-S251. Blaser M, Reller LB, Luechtefeld N, Wang W-N (1982) Campylobacter enteritis in Denver. West J Med 136:287290. Donta ST, Myers MG (1982) Clostridium difficile toxin in asymptomatic neonates. J Pediatr 100:431-434. Fekety R (1990) Antibiotic-associated colitis. In Principles and Practice of Infectious Diseases, 3rd ed. Eds, GL Mandell, RG Douglas, and JE Bennett. New York: Churchill Livingstone, p 865. Harris JC, DuPont HL, Hornick RB (1972) Fecal leukocytes in diarrheal illness. Ann Intern Med 76:697-703.
Korzeniowski O, Burada F, Rouse J, Guerrant R (1979) Value of examination for fecal leukocytes in the early diagnosis of shigellosis. Am J Trop Med Hyg 28:10311035. Lyerly DM, Krivan HC, Wilkins TD (1988) Clostridium difficile: its diseases and toxins. Clin Microbiol Rev 1:1-18. Peirce J, DuPont H, Lewis K (1974) Acute diarrhea in a residential institution for the retarded. Am J Dis Child 128:772-775. Pickering L, DuPont H, Olarte J, Conklin R, Ericsson C (1977) Fecal leukocytes in enteric infections. Am J Clin Pathol 68:562-565. Sazie E, Titus A (1982) Rapid diagnosis of Campylobacter enteritis. Ann Intern Med 96:62-63. Silva J (1989) Update on pseudomembranous colitis. West J Med 151:644-648. Wolff M (1969) The fecal smear in the therapy of diarrhea. Trop Geogr Med 21:427-435.
DIAGN MICROBIOL INFECT DIS 1993;16:313-315
Fecal Leukocytes in Stool Specimens Submitted for Clostridium difficile Toxin Assay Christine E. Marx, Arthur Morris, Michael L. Wilson, and L. Barth Reller
To determine their diagnostic utility, fecal leukocytes were sought by methylene blue stain in 502 consecutive stool specimens submitted for Clostridium difficile toxin assay. In addition, the stability of fecal leukocytes was assessed by daily examination of 23 stool specimens stored at 4°C and room temperature. The sensitivity, specificity, and positive and negative predictive values of fecal leukocytes in predicting C. difficile toxin assay results were 28%, 92%, 27%, and 93%, re-
spectively. At 4°C, fecal leukocytes retained morphology for a minimum of 3 days. Leukocytes survived as long in stool specimens containing either C. difficile toxin or an enteric pathogen as they did in stool specimens with neither finding. We conclude that testing stool specimens for fecal leukocytes is not useful for predicting the presence of C. difficile toxin, because 72% of stool specimens positive for C. difficile toxin are negative for fecal leukocytes despite their stability.
INTRODUCTION
m e n s submitted for C. difficile toxin assay for the presence of fecal leukocytes by the m e t h y l e n e blue stain. Additionally, we followed the survival of fecal leukocytes in specimens stored at 4°C a n d r o o m temperature.
Clostridium difficile is the main p a t h o g e n causing antibiotic-associated diarrhea, a n d cytotoxicity neutralization assays are the s t a n d a r d for detection of C. difficile toxin in stool specimens (Bartlett, 1984; Lyerly et al., 1988). Published sources state that 33%50% of patients with C. difficile toxin-induced colitis have fecal leukocytes (Bartlett, 1990; Fekety, 1990; Silva, 1989). These statements, h o w e v e r , are unreferenced and w e have b e e n unable to find supporting data in the literature. Moreover, it is unclear h o w often fecal leukocytes are p r e s e n t in patients with C. difficile toxin-associated diarrhea w h o do not necessarily have p r o v e n colitis. Therefore, to ascertain the utility of the fecal leukocyte test (FLT) in predicting the presence of C. difficile toxin in stool specimens, we prospectively e x a m i n e d 502 stool speciFrom the Clinical Microbiology Laboratory (C.E.M., A.M., M.L.W., L.B.R.), Duke University Medical Center; and the Departments of Pathology (M.L.W., L.B.R.) and Medicine (L.B.R.), Duke University School of Medicine, Durham, North Carolina, USA. Address reprint requests to Dr. C.E. Marx, Clinical Microbiology Laboratory, Box 3879, Duke University Medical Center, Durham, NC 27710, USA. Received 24 August 1992; revised and accepted 4 November 1992. © 1993 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/93/$6.00
MATERIALS AND METHODS
Clostridium difficile Toxin Assay Clostridium difficile toxin was detected with a cytotoxicity neutralization assay that u s e d h u m a n newborn foreskin diploid fibroblasts (Baxter Scientific Products, Deerfield, IL) a n d C. difficile goat antitoxin (Virginia Polytechnic Institute a n d State University, Blacksburg, VA). Stool specimens w e r e diluted as m u c h as I in 4, d e p e n d i n g on stool consistency. The test was p e r f o r m e d in microtiter plates a n d the cells inspected after 24 h of incubation. Stool specimens with a cytopathic change in >10% of the m o n o l a y e r and complete neutralization by antitoxin w e r e considered positive for C. difficile toxin.
Fecal Leukocyte Testing From S e p t e m b e r to D e c e m b e r 1991, all stool specim e n s submitted for C. difficile toxin testing at D u k e University Medical Center w e r e e x a m i n e d for the
314
presence of fecal leukocytes. Stool specimens were stored at 4°C and examined for fecal leukocytes within 48 h. Most specimens (85%) were examined within 24 h. A small amount of stool or mucus was transferred to a clean glass slide with a wooden stick, mixed thoroughly with a drop of methylene blue stain, and coverslipped. After waiting 2-3 min to permit good nuclear staining, the entire slide was scanned under low power ( x 100) to find areas containing fecal leukocytes. Morphology was confirmed under high power ( x 400). If the results of low-power scanning were negative, at least 30 fields were examined under high power ( x 400). Stool specimens were considered positive if two or more leukocytes were present in at least two separate high-power fields. Only leukocytes with distinct nuclear detail and pale-staining cytoplasm were recorded.
Fecal Leukocyte Survival Aliquots from 23 stool specimens submitted for culture for fecal pathogens and found to contain fecal leukocytes were divided, and separate aliquots were stored at 4°C and at room temperature. Each specimen aliquot was examined daily for at least 3 days to compare the morphology of leukocytes stored at the two temperatures. The result of any fecal pathogen culture (Salmonella, Shigella, and Campylobacter) or C. difficile toxin assay was recorded for each patient.
Patients The following information was recorded for each patient: age, gender, and recent peripheral blood white blood cell (WBC) count. The majority of WBC counts were performed on the same day that the C. difficile test was ordered. A total of 88 specimens from leukopenic patients (peripheral WBC count ---1000 x 106/liter) were excluded from the study, because it was thought that leukopenia might negate the FLT.
RESULTS A N D DISCUSSION Fecal Leukocytes in Clostridium difficile Specimens A total of 502 stool specimens from 403 patients were received for C. difficile toxin assay. Patients had a median age of 52 years (range, 1 month to 95 years); 202 specimens were from male and 300 specimens were from female patients; 56 specimens (11%) were from pediatric patients (-<18 years). The relationship between the FLT and C. difficile toxin assay is given in Table 1. The sensitivity and specificity of the FLT
C.E. Marx et al.
TABLE 1 Frequency of Fecal Leukocytes in Stool Specimens Tested for Clostridium difficile Toxin
C. difficile Toxin Assay Fecal Leukocytes
Positive
Negative
Total
Present Absent
13 33
36 420
49 453
Total
46
456
502
in predicting a positive C. difficile toxin assay were 28% and 92%, respectively. The positive and negative predictive values of the FLT were 27% and 93%, respectively. Ten (18%) of 56 specimens from children (eight children ->10 years of age, one child of 2 years of age, and one child of 2 months of age) versus 36 (8%) of 446 from adults contained C. difficile toxin.
Fecal Leukocyte Survival A total of 23 stool specimens with fecal leukocytes were studied for leukocyte survival: five were C. difficile toxin positive, seven contained a fecal pathogen (three Shigella, three Campylobacter, and one Salmonella), and 11 were either negative for both or negative for one and did not have the other test performed. Stool specimens were examined for 318 days (median, 6 days). All specimens stored at 4°C retained good fecal leukocyte morphology for 3 days, but cell morphology deteriorated markedly after 5 days. Although fecal leukocyte morphology deteriorated - 2 days earlier in stool specimens stored at room temperature, leukocytes were still recognizable after storage at room temperature for 2-3 days. Fecal leukocyte morphology did not deteriorate earlier in stool specimens containing C. difficile toxin or a fecal pathogen than in stool specimens with neither finding. Several authors state that fecal leukocytes are present in 33%-50% of stool specimens from patients with C. difficile toxin-induced colitis (Bartlett, 1990; Fekety, 1990; Silva, 1989). These statements, however, are not supported by primary references to the literature and we know of no specific comment on the presence of fecal leukocytes in C. difficile toxinpositive stools. In our study, only 13 (28%) of 46 specimens positive for C. difficile toxin contained fecal leukocytes. Conversely, only 13 (27%) of 49 specimens with fecal leukocytes were positive for C. difficile toxin. Although neonates often harbor C. difficile and its toxin without clinical sequelae, low C. difficile prevalence rates are observed after 6-12 months when
Focal WBCs and Clostridium difficile Toxin
normal fecal flora is established (Bartlett, 1990; Donta and Myers, 1982). Since only one of our pediatric toxin-positive specimens was obtained from an infant, the low frequency of fecal leukocytes in C. difficile toxin-positive specimens is not due to asymptomatic toxin carriage in children. Fecal leukocytes were found to survive for several days at either 4°C or room temperature. The presence of common fecal pathogens or C. difficile toxin did not shorten the time that fecal leukocytes retained their morphology. We conclude that stool specimens stored at 4°C for up to 3 days provide reliable FLT results, and that the low sensitivity for the FLT in predicting toxin-positive stool specimens is not due to fecal leukocyte deterioration. The examination of stool specimens stored at room temperature for up to 3 days may provide less reliable FLT results, especially when leukocytes are present as either single cells or small aggregates rather than in large sheets. These data are consistent with a previous study in which good cellular detail was observed in stool specimens stored at room temperature for as long as 24 h (Harris et al., 1972). Fecal leukocytes are often found in specimens containing common fecal pathogens such as Cam-
315
pylobacter (70%-85%) (Blaser et al., 1982; Sazie and Titus, 1982), Salmonella (36%-90%) (Harris et al., 1972; Pickering et al., 1977; Wolff, 1969), and Shigella (69%100%) (Anderson, 1921; Harris et al., 1972; Korzeniowski et al., 1979; Peirce et al., 1974; Pickering et al., 1977; Wolff, 1969). This suggests that the presence of fecal leukocytes increases the probability of recovering a fecal pathogen. This association does not appear to hold for predicting the presence of C. difficile toxin. We did not undertake a clinical study, however, and it is u n k n o w n whether patients with pseudomembranous colitis are more likely to have fecal leukocytes than those with toxin-associated diarrhea only. This point needs investigation. Also, the duration and magnitude of the diarrhea in our patients were not studied. In our hospital, most FLTpositive stool specimens contain neither C. difficile toxin nor a fecal pathogen. In conclusion, the diagnostic utility of fecal leukocytes in C. difficile-associated disorders is minimal. We thank Nancy G. Henshaw, PhD, and the staff of the Pediatric Virology Laboratory for access to the C. difficile toxin assay results and the stool specimens.
REFERENCES Anderson J (1921) A study of dysentery in the field. Lancet 2:998-1002. Bartlett JG (1984) Antibiotic-associated colitis. Dis Mon 30:154. Bartlett JG (1990) Clostridium difficile: clinical considerations. Rev Infect Dis 12(Suppl):S243-S251. Blaser M, Reller LB, Luechtefeld N, Wang W-N (1982) Campylobacter enteritis in Denver. West J Med 136:287290. Donta ST, Myers MG (1982) Clostridium difficile toxin in asymptomatic neonates. J Pediatr 100:431-434. Fekety R (1990) Antibiotic-associated colitis. In Principles and Practice of Infectious Diseases, 3rd ed. Eds, GL Mandell, RG Douglas, and JE Bennett. New York: Churchill Livingstone, p 865. Harris JC, DuPont HL, Hornick RB (1972) Fecal leukocytes in diarrheal illness. Ann Intern Med 76:697-703.
Korzeniowski O, Burada F, Rouse J, Guerrant R (1979) Value of examination for fecal leukocytes in the early diagnosis of shigellosis. Am J Trop Med Hyg 28:10311035. Lyerly DM, Krivan HC, Wilkins TD (1988) Clostridium difficile: its diseases and toxins. Clin Microbiol Rev 1:1-18. Peirce J, DuPont H, Lewis K (1974) Acute diarrhea in a residential institution for the retarded. Am J Dis Child 128:772-775. Pickering L, DuPont H, Olarte J, Conklin R, Ericsson C (1977) Fecal leukocytes in enteric infections. Am J Clin Pathol 68:562-565. Sazie E, Titus A (1982) Rapid diagnosis of Campylobacter enteritis. Ann Intern Med 96:62-63. Silva J (1989) Update on pseudomembranous colitis. West J Med 151:644-648. Wolff M (1969) The fecal smear in the therapy of diarrhea. Trop Geogr Med 21:427-435.