This material was presented in part at the 24th Annual ACVIM Forum in Louisville, KY and the 16th ECVIM-CA Congress in Amsterdam, The Netherlands.
Evaluation of Fecal Elastase and Serum Cholecystokinin in Dogs with a False Positive Fecal Elastase Test
Article first published online: 15 MAR 2010
Copyright © 2010 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 24, Issue 3, pages 643–646, May/June 2010
How to Cite
Steiner, J.M., Rehfeld, J.F. and Pantchev, N. (2010), Evaluation of Fecal Elastase and Serum Cholecystokinin in Dogs with a False Positive Fecal Elastase Test. Journal of Veterinary Internal Medicine, 24: 643–646. doi: 10.1111/j.1939-1676.2010.0489.x
- Issue published online: 7 MAY 2010
- Article first published online: 15 MAR 2010
- Submitted October 20, 2009; Revised December 22, 2009; Accepted February 5, 2010.
- Exocrine pancreatic insufficiency;
- Pancreatic function
Background: An assay for the measurement of pancreatic elastase in dog feces has been introduced.
Hypothesis/Objectives: The goal of this study was to evaluate the rate of false-positive fecal-elastase test results in dogs with suspected exocrine pancreatic insufficiency (EPI) and to assess serum cholecystokinin (CCK) concentrations in dogs with a false positive fecal elastase test result.
Animals: Twenty-six fecal and serum samples from dogs suspected of EPI, for which samples had been submitted to a commercial laboratory (Vet Med Labor) for analysis.
Methods: Prospective study. Serum trypsin-like immunoreactivity (TLI) was measured in 26 dogs with a decreased fecal elastase concentration of <10 μg/g feces. Serum CCK concentrations were measured in 21 of these dogs.
Results: Of 26 dogs with a decreased fecal elastase concentration, 6 (23%) had serum TLI concentrations within or above the reference range. Serum CCK concentrations were significantly higher in dogs with a true positive fecal elastase test result (median: 1.1 pmol/L; range: 0.1–3.3 pmol/L) than in those with a false positive fecal elastase test result (median: 0.1 pmol/L; range: 0.1–0.9 pmol/L; P value = .0163).
Conclusions and Clinical Importance: The rate of false positive fecal elastase test results was high in this group of dogs, suggesting that diagnosis of EPI must be confirmed by other means. The decreased CCK concentration in dogs with a false positive fecal elastase test result could suggest that false positive results are because of decreased stimulation of exocrine pancreatic function caused by other conditions.
canine trypsin-like immunoreactivity
exocrine pancreatic insufficiency
Exocrine pancreatic insufficiency (EPI) is characterized by a lack of pancreatic digestive enzymes in the small intestine, leading to maldigestion with associated signs of loose stools and weight loss. Serum canine trypsin-like immunoreactivity (cTLI) has been established as the gold standard for a diagnosis of EPI.1,2 While the current reference range for serum cTLI concentration is 5.7–45.2 μg/L (5.0–35.0 μg/L at the time this study was conducted), a value of ≤2.5 μg/L is considered diagnostic for EPI. Sensitivity and specificity for serum cTLI concentration for the diagnosis of EPI are high.1 In rare instances, serum cTLI concentration might be within the reference range in dogs with isolated lipase deficiency or EPI because of obstruction of the pancreatic duct. Until recently, measurement of serum cTLI concentration was only available through a radioimmunoassay. Because of tight regulatory restrictions, operation of radioimmunoassays is difficult in some European countries. Therefore, a new test for the diagnosis of EPI, fecal pancreatic elastase concentration (cE1), has been developed and validated for use in dogs.2 Previously, a similar assay had also been developed for use in humans. False positive test results occur in both species.2,3
However, only 1 study evaluating potential reasons for a high rate of false positive results of measurement of fecal elastase concentration is available for either species in the literature.4
The 1st goal of this study was to identify the rate of false positive test results for fecal elastase in a group of dogs with severely decreased fecal elastase concentrations. The 2nd goal of this study was to evaluate whether false positive fecal pancreatic elastase concentrations in dogs without EPI could be because of a lack of stimulation of pancreatic enzyme secretion by cholecystokinin (CCK), by measuring serum CCK concentrations in dogs with a false positive fecal elastase test result.
Materials and Methods
Fecal samples from 26 dogs that had been submitted to a commercial laboratory (Vet Med Labor) because of a clinical suspicion of EPI or because of chronic diarrhea of unexplained etiology and had shown a fecal elastase concentration considered diagnostic for EPI (< 10 μg/g feces; reference range: >40 μg/g) were selected for this study. Fecal elastase was measured by a species-specific ELISA for the measurement of canine pancreatic elastase (cE1) in dog feces.a A blood sample was collected from all 26 dogs either on the day of the collection of the fecal sample (n = 18) or within a few days of fecal sample collection (n = 8; number of days: 1, 1, 2, 3, 4, 5, 5, and 7 days, respectively). As per standard operating procedures, veterinarians were asked to collect the blood sample after with holding food from the dogs for at least 10–12 hours. However, fasting could not be verified, though only 1 serum sample appeared grossly lipemic. Serum cTLIb concentration (reference range 5.0–35.0 μg/L) was measured in serum samples from all 26 dogs and a serum cTLI concentration of ≤2.5 μg/L was considered diagnostic for EPI. Serum CCK concentrations were measured in 21 of the dogs with a decreased fecal elastase concentration suggestive of EPI. These dogs were part of the original 26 dogs with severely decreased fecal elastase concentrations and were those dogs for which sufficient serum was available for CCK analysis. Serum CCK concentrations were measured by use of a specific radio-immuno- assay that has previously been shown to not have any cross-immunoreactivity with canine gastrin or gastrins from other mammalian species.5 For comparison, serum CCK concentrations were also measured in 8 healthy dogs. Median serum CCK concentrations were compared by a statistical software package (GraphPad Prism 5.0c). First, all 3 datasets were evaluated for normal distribution of the data by a D'Agostino & Pearson omnibus normality test. Then the 2 datasets for dogs with a true positive fecal elastase test result and those with a false positive fecal elastase test result were compared with a Mann-Whitney test. Finally, all 3 datasets were compared by a Kruskal-Wallis test with a Dunn's post-test.
Twenty (77%) of the 26 dogs with a fecal elastase concentration suggestive of EPI had serum cTLI concentrations diagnostic of EPI, while 6 (23%) had serum cTLI concentrations within or above the reference range (cTLI values: 6.4, 8.0, 9.5, 10.9, 42.3, and 66.4 μg/L; [Fig 1]). Most dogs had serum cTLI and fecal elastase determined on samples collected on the same day, but in 8 dogs there was a time difference for sample collection between 1 and 7 days. Of the dogs with a false positive fecal elastase test result for the diagnosis of EPI, 4 of them had samples collected on the same day, while for 2 dogs there was a time difference of 1 and 3 days, respectively.
Sufficient serum for measurement of serum CCK concentration was available for 15 of the 20 dogs with a true positive fecal elastase test result and for all 6 dogs with a false positive fecal elastase test result. Serum CCK concentrations were significantly higher in the 15 dogs with a true positive fecal elastase test result (median: 1.1 pmol/L; range: 0.1–3.3 pmol/L) than in the 6 dogs with a false positive fecal elastase test result (median: 0.1 pmol/L; range: 0.1–0.9 pmol/L; P value = .0163; [Fig 2]). When all 3 groups were compared, there was a significant difference between the median serum CCK concentrations between the 3 groups (P= .0346). Median serum CCK concentration was significantly higher in dogs with EPI than in dogs with a false positive fecal elastase test result (P < .05), but was not significantly different between healthy dogs (median: 0.6 pmol/L; range: 0.2–1.9 pmol/L) and dogs with EPI or between healthy dogs and dogs with a false positive fecal elastase test result (Fig 2).
Fecal elastase measurements in these dogs were associated with a high rate (23%) of false positive test results when serum cTLI concentration was used as the gold standard. This rate of false positive results is much higher than the one previously reported for normal dogs (8%), normal Beagles (9%), or dogs with enteropathies (8%), but is lower than has previously been reported for normal German Shepherd Dogs and Rough-Coated Collies (30%).2
Based on previous reports, serum cTLI concentration is highly sensitive for EPI and thus dogs with a cTLI concentration ≥2.5 μg/L were not considered to have EPI.1 There have been isolated reports of dogs with EPI with a serum cTLI concentration >2.5 μg/L.1,6,7 One dog reported in the literature was suspected of having an isolated pancreatic lipase deficiency.7 However, this condition is suspected to be extremely rare in dogs. Other dogs with EPI that did not have a serum cTLI concentration below the diagnostic cut-off value for EPI had serum cTLI concentrations in the questionable range, but rarely within the reference range or even above the reference range.1,6 It should be pointed out that several dogs with EPI, had concurrent chronic pancreatitis and had serum cTLI concentrations within the reference range.8,9 However, these dogs had clinical signs compatible with pancreatitis during those episodes of pancreatitis and a normal or even increased serum cTLI concentration.8,9 This was not the case for any of the dogs in this study as the dogs of this study were examined because of a clinical suspicion of EPI or chronic diarrhea where an alternative diagnosis could not be confirmed. Also, feeding has previously been shown to cause a transient increase in serum cTLI concentration (D.A. Williams unpublished data collected at the University of Florida, personal communication, 1994).d However, these increases are not clinically important and do not lead to serum cTLI concentrations within the normal range, suggesting that not with holding food in the dogs studied here, against what was requested from the veterinarians submitting the samples for this study, would not likely have resulted in false negative serum cTLI concentrations.d Theoretically, there would have been a chance that some of the dogs might have had an obstruction of the pancreatic duct, for example, because of pancreatic cancer. Dogs with EPI and an obstructed pancreatic duct might be expected to have a normal or even increased serum cTLI concentration, but could have clinically significant EPI. However, to the authors' knowledge, this situation has not yet been described in dogs. While the authors are aware of a single case where such a scenario has been hypothesized (S. Hill, personal communication, 2007) and thus would consider it reasonable to assume that such cases could occur in dogs, such a scenario can be assumed to be rare in the dog. Thus, it is reasonable to consider it highly unlikely that this situation was present in any of the 6 dogs with a false positive fecal elastase test result.
In summary, while a normal exocrine pancreas could not be confirmed definitely, for example, by sequential sectioning of the entire pancreas, in the dogs with a normal serum cTLI concentration and a decreased fecal elastase concentration, based on the high sensitivity of serum cTLI concentration previously reported, it is reasonable to conclude that all 6 dogs had a false positive diagnosis of EPI based on results of fecal elastase testing.1,2
Given the high cost of treatment of EPI, especially in large breed dogs, and the associated rate of euthanasia in dogs diagnosed with this disease, a false positive rate of 23% is of concern. Therefore, a diagnosis of EPI based on a decreased fecal elastase concentration of <10 μg/g feces should be confirmed by other diagnostic means, such as measurement of serum cTLI concentration, repeated measurement of fecal elastase concentration on at least 3 fecal samples, or a treatment trial with pancreatic enzymes, before a definitive diagnosis is made.
The definitive cause of decreased fecal elastase concentrations in dogs that do not have EPI remains to be determined. The current findings would suggest that 1 possible reason for a false positive fecal elastase test result for a diagnosis of EPI could be a decreased stimulation of elastase secretion from pancreatic acinar cells through lack of CCK release. Such decreased CCK release could further be because of chronic small intestinal disease, leading to damage of intestinal endocrine cells. We showed that in the dogs with a false positive fecal elastase test result studied here, serum CCK concentrations were significantly lower than in those with a true positive fecal elastase test result.
One limitation of the current study is that we cannot definitively exclude that serum CCK concentration could have been increased in dogs with EPI rather than decreased in dogs with a false positive fecal elastase test result. However, as mentioned above, mean serum CCK concentrations are not significantly different between healthy dogs and dogs with EPI, both after with holding food (EPI dogs: 2.5 pmol/L; healthy dogs: 2.4 pmol/L) and after feeding a standardized meal (EPI dogs: 4.4 pmol/L; healthy dogs: 4.2 pmol/L).d Unfortunately, these data have not been reported in a peer-reviewed manuscript and the assay used for the previous study was not the same as the one used for this study, so these results might not be directly comparable.d
While the authors are not aware of any evidence that would suggest that EPI is associated with increased serum CCK concentrations in dogs, the data provided here do not allow for a definitive exclusion of this possibility and thus this needs to be further studied before a more definitive statement can be made. However, the hypothesis that intestinal disease could be associated with damage of I cells in the small intestine, which in turn would lead to decreased secretion of CCK and in turn decreased pancreatic stimulation, is intriguing and deserves further study.
aFecal Elastase 1, ScheBo Tech, Giessen, Germany
bCanine TLI-RIA, Siemens Medical Solutions, Malvern, PA
cGraphPad Prism version 5; GraphPad, San Diego, CA
dWashabau RJ, Callan MB, Williams DA, et al. Cholecystokinin secretion is preserved in canine pancreatic insufficiency. J Vet Interm Med 1995;9:193 (abstract)
- 2Canine faecal pancreatic elastase (cE1) in dogs with clinical exocrine pancreatic insufficiency, normal dogs and dogs with chronic enteropathies. Eur J Comp Gastroenterol 2000;5:1–6., , , et al.