Diagnostic Performance of Different Fecal Lawsonia intracellularis —Specific Polymerase Chain Reaction Assays as Diagnostic Tests for Proliferative Enteropathy in Pigs: A Review

Abstract

Traditionally, diagnosis of Lawsonia intracellularis—associated proliferative enteropathy (PE) has depended on necropsy and histology. Since the establishment of the etiologic role of L. intracellularis, a number of specific polymerase chain reaction (PCR) assays have been developed for the detection of DNA in feces. The present article is a systematic review of peer-reviewed publications on the application of L. intracellularis specific fecal PCR as an antemortem diagnostic test for histologic lesions of PE in pigs. Based on this information, a range of diagnostic sensitivities (36-100%) and specificities (50-100%) of the published tests was calculated. Validity and confidence limits of the estimates varied considerably. The positive and negative predictive values of 6 different PCR assays were calculated for PE prevalence of 15%, 30%, 45%, 60%, 75%, and 90%, using a histologic case definition of PE and based on the reported test sensitivities and specificities. The simulated predictive values suggested that applying the fecal PCR assay as a diagnostic test is more likely to overestimate than underestimate the number of pigs having histologic lesions of PE under field conditions.

Keywords

pigs polymerase chain reaction proliferative enteropathy sensitivity specificity

Introduction

Lawsonia intracellularis is an obligate intracellular bacterium and the causative agent of proliferative enteropathy (PE). Proliferative enteropathy was first described in pigs in 1931, 2 but it was not until 1973 that the presence of bacteria in the lesions was identified. 40 It took another 22 years before this bacterium was given the name Lawsonia intracellularis. 32 Proliferative enteropathy is characterized by proliferation of immature epithelial cells of intestinal crypts, resulting in crypt expansion and elongation, giving the mucosa a thickened appearance. Lawsonia intracellularis can be demonstrated in the proliferating epithelial cells. The pathology of PE has been reviewed previously. 26

Proliferative enteropathy is now recognized as one of the most economically important diseases in the swine industry worldwide. 26 Several different clinical manifestations of PE have been described in pigs. Proliferative hemorrhagic enteropathy (PHE) is classified as an acute form of PE, resulting in bloody diarrhea and sudden death. This type of PE is typically seen in adult pigs aged more than 16 weeks. In contrast, chronic PE is most common in pigs 6-20 weeks old. Chronic PE has traditionally been characterized by anorexia, reduced weight gain, wasting, diarrhea, and occasional death. 31 In 2005, a subclinical form of PE was recognized (subclinical ileitis) and defined as “when pigs become infected and have some intestinal lesions, but without clear diarrhea or weight loss.” 29

The gold standard for diagnosing PE is demonstration of proliferative intestinal lesions associated with intracellular curved bacteria in histologic sections. Several histologic techniques have been described, including staining with hematoxylin and eosin (HE), 39 Warthin—Starry (WS) silver stain, 46 modified Ziehl—Nielsen (ZN) stain, 46 in situ hybridization (ISH), 11 and immunohistochemistry (IHC). 30 Advantages and disadvantages of these and other diagnostic techniques for detection of L. intracellularis have been reviewed previously. 22

In the last 15 years, progress has been made in the development of antemortem diagnostic tests for L. intracellularis, including serologic and polymerase chain reaction (PCR) tests. Both techniques have been applied in veterinary practice and research. Lawsonia intracellularis cannot be cultured by routine methods, and PCR tests for detection of L. intracellularis in feces are now commonly used for routine diagnosis of PE in diagnostic laboratories around the world.

Diagnostic performance of a diagnostic test is characterized by its diagnostic sensitivity and specificity, positive predictive value, and negative predictive value. Diagnostic sensitivity and specificity must be differentiated from analytical sensitivity and specificity. Analytical sensitivity is the same as detection limit, which is the lowest detectable concentration of the substance or pathogen of interest. Analytical specificity is the ability of a test to give negative results in the presence of other substances or pathogens that may be considered potential cross-reactors. Analytical sensitivity and specificity do not have a mathematical definition. Analytical sensitivity and specificity depend on the conditions of the assay and the biologic properties of the test samples such as feces. They both contribute to a test's diagnostic sensitivity and specificity. The diagnostic sensitivity for a test is the proportion of test-positive animals in a population of diseased animals. The diagnostic specificity is the proportion of test-negative animals in a population of healthy animals.

Diagnostic sensitivity and specificity depend on the definition of diseased animals and occurrence of different disease manifestations and covariate factors in the population. 13 This is also the case for fecal PCR detection of L. intracellularis. Diagnostic sensitivity and specificity may be different for PCR detection of pigs subclinically infected with L. intracellularis, or detection of pigs having histologic or gross pathologic lesions of L. intracellularis—associated PE. The diagnostic sensitivity and specificity for detection of disease at the herd level also depends on disease prevalence, number of samples, and the number of test-positive samples before a herd is classified as positive.

The predictive values are very important test characteristics because they express how likely the test result reflects the true status of the animal. Positive predictive values represent the proportion of test-positive animals that are true positives. Negative predictive values represent the proportion of testnegative animals that are true negatives. Positive and negative predictive values depend on disease prevalence in the test population and the test's diagnostic sensitivity and specificity. The different aspects of diagnostic performance have to be taken into consideration for fecal PCR detection of L. intracellularis.

In clinical practice and epidemiologic field investigations, PCR detection of L. intracellularis in feces can be applied for several reasons. These include assessment of L. intracellularis as a possible cause of low productivity in a herd. Pigs can sometimes shed PCR-detectable levels of L. intracellularis without reduction in average daily gain. 6 This implies that interpretation of a positive fecal PCR result can be difficult in terms of importance to the pig or herd. The likely cause of the negative effect of L. intracellularis has been suggested as functional changes in the thickened intestinal mucosa. 31 Based on this suggestion, it can be hypothesized that histologic recognizable proliferative lesions of PE must be present for L. intracellularis to have a negative effect in a pig. In this context, it becomes important to know whether pigs with PCR-detectable levels of L. intracellularis in feces always have histologic lesions of PE or not.

The objective of the current study was to make a systematic review of diagnostic performance for PCR tests used for antemortem fecal detection of L. intracellularis—associated histologic lesions of PE in pigs. The diagnostic performance was reviewed in terms of diagnostic sensitivity, specificity, and positive and negative predictive values using a histologic case definition of PE and simulated PE prevalence. Aspects of analytical sensitivity and specificity were also reviewed.

Review methodology

Publications

A review of studies concerning PCR detection of L. intracellularis was conducted by searching available databases (up until December 13, 2009). The databases included Agricola (1970 to November 2009), Agris (1975 to June 2009), Biosis previews (1969-2009), CAB Abstracts (1973-2009 week 49), and Medline (1950 to November week 3, 2009). The search terms were “Lawsonia or proliferative enteropathy” and “PCR or polymerase chain reaction.”

The reference list of included publications was checked for other relevant publications. Duplicate records and non-peer-reviewed publications were removed.

Data extraction. Data concerning PCR evaluation were extracted from the publications. The data included the following:

use of PCR for detection of L. intracellularis in feces from pigs,

primers used,

PCR principle,

field or experimental samples,

number of samples tested,

use of spiked feces samples or purified DNA for analytical evaluation,

analytical sensitivity or detection limit,

bacteria species used to evaluate analytical specificity,

diagnostic criteria for PE or infection with L. intracellularis,

10)

number of pigs in study,

11)

number of PCR-positive test results among PE-positive pigs,

12)

number of PCR-negative test results among PE-negative pigs,

13)

age of pigs in the study,

14)

statistical data analysis, and

15)

authors and year of publication.

Inchusion criteria. To be included in the review, a publication had to contain data on analytical and/or diagnostic sensitivity and/or specificity for PCR tests applied for detection of L. intracellularis in fecal samples from pigs. For analytical sensitivity, only data for PCR tests applied on fecal samples are reported because feces may contain PCR inhibitors. 17 This may lead to a lower analytical sensitivity when the PCR test is applied on fecal samples, compared with samples containing purified DNA or samples from intestinal tissues. 18,28,44 For diagnostic sensitivity and specificity, only data for PCR tests applied on fecal samples and using one or more histologic technique as part of the diagnostic criteria for PE are reported.

Data analysis. To evaluate the performance of the published PCR tests, a histologic case definition of PE was applied to calculate diagnostic sensitivity and specificity with 95% confidence intervals. The necessary data were extracted from the publications. For the calculations, PE-positive pigs were defined as pigs with L. intracellularis-associated histologic lesions of PE. Various histologic techniques with different diagnostic abilities have been applied for diagnosis of PE. Currently, IHC detection of L. intracellularis and associated proliferative lesions are the reference method for histologic examinations. 16 This technique has not been applied for all PCR test evaluations. For the calculations in the current review, histology was defined as any histologic technique. The calculations for each PCR test were performed using the histologic technique applied in each PCR test evaluation study. The histologic technique used in each study is reported in the review.

The PCR test result of fecal samples obtained immediately before or at the time of necropsy was used in the calculations. Diagnostic sensitivity, specificity, and 95% confidence limits were calculated using the “Clinical Calculator 1” at VassarStats: Website for Statistical Computation (http://faculty.vassar.edu/lowry/VassarStats.html).

Positive and negative predictive values were simulated applying the calculated diagnostic sensitivities and specificities. Only those PCR tests where both diagnostic sensitivity and specificity could be calculated were included in the simulations. Assumed true prevalence for PE of 15%, 30%, 45%, 60%, 75%, and 90% was used. The positive and negative predictive values were simulated in Excel and calculated as described elsewhere 37 :

\begin{array}{l} P P V = (T p \times S e) / (T p \times S e + (1 - S e) \times (1 - S p)) \\ N P V = ((1 - T p) \times S p) / (T p \times (1 - S e) + S p \times (1 - T p)), \end{array} ​

where PPV = positive predictive value, NPV = negative predictive value, Tp = true population prevalence of pigs with histologic lesions of proliferative enteropathy, Se = diagnostic sensitivity, and Sp = diagnostic specificity.

Finding

PCR principles applied for detection of L. intracellularis in feces

A total of 21 publications reported on the development of new PCR tests for detection of L. intracellularis in feces of pigs.

Primers. Seven different sets of primers have been applied for detection of L. intracellularis in feces of pigs. The first 2 sets of primers were based on sequences from a 375-base pair (bp) segment of a DNA fragment from the L. intracellularis DNA clone p^78.18 These primers were later used in development of other new PCR tests for detection of L. intracellularis in feces of pigs. 1,3,8,10,17,21,25,28,33,34,36,38,47,48 The third set of primers was also directed towards a chromosomal DNA sequence. 44 Three sets of primers have been directed towards regions of the 16S ribosomal DNA gene. 7,24,27 The last reported primer set was directed towards a gene coding for a methylase involved in the ubiquinone/menaquinone biosynthesis. 35

Test principles. Five different PCR principles have been applied for detection of L. intracellularis in feces of pigs. The first principles were conventional and nested PCRs. 18 Conventional (simple) PCR refers to the original PCR principle, while nested PCR assay involves a second set of primers used to re-amplify the reaction products of the first PCR reaction. 18 These principles were followed by multiplex PCR assays having the ability to detect the DNA from a number of different intestinal pathogens, concurrently, in the one test sample. 10 Later, a PCR-enzyme-linked oligosorbent assay (PCR-ELOSA) was developed. 47

The ELOSA involved labeling PCR products with biotin during amplification followed by hybridization with an amine-modified internal oligonucleotide probe immobilized in microwell plates, addition of peroxidase–streptavidin complex, and measurement of spectrophotometric signal, reported as an optical density value. Finally a real-time PCR was described using a specific oligonucleotide probe labeled with 2 fluorescent dyes. This allows amplification and detection of an emission signal to occur simultaneously. 27 The real-time PCR technique has recently been applied to develop a method for quantification of L. intracellularis in porcine fecal samples. 35

Analytical sensitivity and specificity of PCR assays used for detection of L. intracellularis in feces

A total of 8 publications reported analytical sensitivity and 16 publications reported analytical specificity for PCR tests applied on fecal samples (Table. 1). Microorganisms used to evaluate analytical specificity were Actinobacillus pleuropneumoniae, Brachyspira spp., Campylobacter spp., Clostridium perfringens types A–C, Desulfovibrio desulfuricans, Enterococcus faecalis, hemolytic and nonhemolytic Escherichia coli, Haemophilus parasuis, Mycoplasma hyopneumoniae, Myxococcus xanthus, Proteus mirabilis, Salmonella spp., Streptococcus suis.

Estimates of diagnostic sensitivity, specificity, and predictive values

After extraction of data, both diagnostic sensitivity and specificity from 7 publications representing 6 different PCR tests were calculated. These included 4 conventional PCRs, 1 nested PCR, and 1 real-time PCR. Three of these publications investigated cases and controls from natural sources, 25,27,44 3 publications investigated experimental cases and controls, 14,19,21 and 1 publication investigated a combination. 18 The age of the tested pigs ranged from 3 weeks to slaughter (approximately 25 weeks old). Results of the diagnostic sensitivity and specificity calculations, including 95% confidence limits, are displayed in Table 2. Diagnostic sensitivity and/or specificity could not be calculated from a total of 5 publications, though one or more histologic techniques were part of the diagnostic criteria for PE (Table. 2). Positive and negative predictive values were simulated using the reported combinations of diagnostic sensitivities and specificities. Results are displayed in Table 3.

Discussion

To the authors' knowledge, the current study is the first review of the analytical and diagnostic sensitivity and specificity of L. intracellularis–specific fecal PCR tests for PE in pigs. In the review, a clear distinction between analytical sensitivity and/or specificity and diagnostic sensitivity and/or specificity was made.

Specificity estimates were not clearly defined in a proportion of the published literature. Evaluation of a PCR test using negative fecal samples with or without additional DNA from other bacteria or samples of purified DNA from other bacteria species should be reported as analytical specificity. Diagnostic specificity should be restricted for evaluation of PCR tests in a population of healthy animals classified using a standard reference test.

Seven of the reviewed publications contained data from which it was possible to estimate both diagnostic sensitivity and specificity using histologic techniques for definition of case and control animals. Five publications reported studies where estimation would be possible based on the study design. However, diagnostic sensitivity and/or specificity were not reported in those publications, and the necessary data for calculation could not be extracted.

Other publications on PCR test evaluations used study designs without any histologic technique included in the diagnostic criteria for PE. Among these were agreement studies, where the evaluated PCR test was compared with another PCR test. In the reviewed publications, a large range in the estimates of both diagnostic sensitivity (36–100%) and specificity (50–100%) was observed. Different study designs may also explain some of the differences in diagnostic performance between the individual PCR tests.

One important explanation for observed differences may be the application of different histologic reference techniques in the different studies. The various histologic techniques have different diagnostic abilities. Hematoxylin and eosin examination may only pick up moderate to severe cases of PE in contrast to IHC. 16 Further, demonstration of L. intracellularis in tissues using IHC alone is different from demonstration of proliferative lesions alone and different from demonstration of lesions in combination with L. intracellularis. These differences between histologic techniques have to be taken into consideration in the interpretation of positive and negative test results for each PCR test. Application of different histologic techniques may explain those differences reported for the PCR tests that were evaluated in more than 1 study. However, large differences in diagnostic sensitivities and specificities for different PCR tests were observed between studies applying the same histologic techniques.

Four studies evaluated the PCR tests under experimental conditions. The experimental settings may not have reflected the situations where the tests will be applied in epidemiologic field studies and in veterinary practice. Therefore, the estimates of diagnostic sensitivity and specificity from those studies should be interpreted with caution. Studies using field samples are necessary for correct estimates for the field application of the PCR tests. These studies are most likely to include animals representing different disease manifestations of PE and include covariates with influence on diagnostic performance. That is, these studies may be closer to the situations where the tests will be applied in practice and research.

Table 1

Analytical sensitivity (Se) and specificity (Sp) of fecal polymerase chain reaction (PCR) detection of Lawsonia intracellularis. *

		Analytical sensitivity		Analytical specificity
PCR principle	Primer reference	Sample type	Sef†	Sample‡ type	Sp §	Reference no
Conventional	18	Fecal samples spiked with L. intracellularis	10³	1,2	0	18
	18	NR	NR	1	0	33
	18	Fecal samples spiked with L. intracellularis	103	1	0	28
	18	Fecal samples spiked with L. intracellularis	10³	NR	NR	5
	18	NR	NR	1	0	3
	7	NR	NR	3	0	7
	44	NR	NR	2,3	0	44
	18	NR	NR	3	0	17
	18	Fecal samples spiked with L. intracellularis	1.1 × 10⁵	1	0	48
Nested	18	Fecal samples spiked with L. intracellularis	10³	1,2	0	18
	18	Fecal samples spiked with L. intracellularis	10³	NR	NR	5
	18	Fecal samples spiked with L. intracellularis	2 × 10²	3	0	34
	18	Fecal samples spiked with mimic DNA molecules	10³ ‖	3	0	17
	Fecal samples spiked with L. intracellularis	1.1 × 10³	1	0	48
PCR-ELOSA	18	One clinical sample	NR	1	0	47
Real-time	27	Fecal samples spiked with L. intracellularis	4 ×104	1,3	0	27
	35	NR	NR	1,3	0	10
Multiplex	18	NR	NR	3	0	10
	18	NR	NR	3	0	1
	18	Fecal samples spiked with L. intracellularis	1.1 × 10⁵	1,3	0	48
24	Fecal samples spiked with L. intracellularis	10²−10³	1,3,4	0	24
	18	NR	NR	3	0	36

NR = not reported; ELOSA = enzyme-linked oligosorbent assay.

†

Cells per gram of feces.

‡

1 = L. intracellularis–free fecal samples; 2 = purified swine DNA; 3 = purified DNA from other microorganisms; 4 = fecal samples spiked with other microorganisms.

Number of positive reactions.

‖

DNA molecules per gram of feces.

Some of the reviewed studies were characterized by a small sample size. This is reflected in the wide 95% confidence limits for the estimates of diagnostic sensitivity and specificity. Interpretation of these studies should be done with caution. The small sample sizes could lead to underestimation or overestimation of the diagnostic sensitivity and/or specificity because of random error. Such random error concerning test evaluations should be taken into consideration when applying the tests for prevalence estimation or risk factor analysis. 12

Predictive values were calculated for prevalence ranging from 15% to 90% for the PCR assays where both diagnostic sensitivity and specificity were available. The 3 PCR tests with diagnostic specificity less than 1 had low positive predictive values at lower prevalence. Applying these tests in a herd with prevalence below 15% means that test-positive animals are roughly more than 40% likely not to have histologic lesions of PE.

The 4 PCR assays with diagnostic sensitivity less than 1 had low negative predictive values at higher prevalence. Three of these PCR tests had diagnostic sensitivity of 50% or less. Applying one of these tests in a herd with prevalence above 60% means that testnegative animals in fact are roughly more than 50% likely to have histologic lesions.

Previous studies in Denmark have reported within-herd prevalence for L. intracellularis of 5–100% with mean/mode around 15–30%. 42 From the simulated predictive values, it can be observed that the risk of false positives will usually exceed the risk of false negatives. This applies for the reported PCR tests used for antemortem detection of L. intracellularis–associated histologic lesions of PE under Danish conditions.

In practice, where the prevalence is unknown, a PCR test with high sensitivity could be combined with a PCR test with high specificity, increasing the validity of the diagnosis. At the herd level, another possibility is to raise the cutoff value for herd diagnosis (i.e., allow more positive individuals before classifying herds as positive).

Table 2.

Diagnostic sensitivity (Se) and specificity (Sp) of fecal polymerase chain reaction (PCR) detection of Lawsonia intracellularis-associated histologic lesions of proliferative enteropathy (PE).*

					Test results
PCR principle	Study	PE diagnosis	Cases (n)	Controls (n)	T+/C+ T-/C−	Se (%)	95% Cl	Sp (%)	95% Cl	Reference no.
Conventional E	HE + WS-stained sections of ileum, cecum, colon	9	19	9	17	100	(63-100)	89	(65-98)	19
	E+F	HE + WS-stained sections of ileum, cecum, colon	7	5	7	5	100	(56-100)	100	(46-100)	18
	F	Histology of ileum	3	2	3	2	100	(31-100)	100	(20-100)	28
	E	HE + WS-stained sections of ileum, colon	10	47	5	47	50	(20-80)	100	(91-100)	21
	E	IHC of ileum, colon	14	43	5	43	36	(14-64)	100	(90-100)	21
	F	HE + WS-stained sections of ileum	13	4	13	2	100	(72-100)	50	(9-91)	44
	E	IHC of ileum + PCR-positive fecal sample at any time	‡	‡	‡	‡	‡	‡	‡	‡	16
	E	HE + IHC†	20	4	10	4	50	(28-72)	100	(40-100)	14
	E	WS silver stain†	0	8	‡	8	‡	‡	100	(60-100)	6
	F	HE + IHC + ISH + WS-stained sections of ileum	29	167	‡	‡	‡	‡	‡	‡	25
Nested	E+F	HE + WS-stained sections of ileum, cecum, colon	7	5	7	5	100	(56-100)	100	(46-100)	18
	NR	Histology†	0	31	†	31	‡	‡	100	(86-100)	34
Real-time	F	IHC of ileum	98	106	95	90	97	(91-99)	85	(76-91)	27
Multiplex	NR	Histology†	1	0	1	†	100	(5-100)	‡	‡	10

Cl = confidence limit; T+ = PCR positive; C+ = positive PE diagnosis; T- = PCR negative; C− = negative PE diagnosis; E = experimental; HE = hematoxylin and eosin; WS = Warthin-Starry silver stain; F = field; IHC = immunohistochemistry; ISH = in-situ hybridization; NR = not reported.

†

Type of intestinal sections not reported.

‡

Not reported and could not be calculated.

The reported estimates of diagnostic sensitivity/ specificity and predictive values only imply for a case definition of PE based on histologic techniques as the standard reference tests. Other values may be obtained using other case definitions, such as occurrence of pigs having gross PE lesions, detection of pigs shedding L. intracellularis, or detection of pigs infected with L. intracellularis.

For detection of infected pigs, the diagnostic sensitivity will depend on the level of L. intracellularis cells per gram of feces in the positive pigs and the analytical sensitivity for the PCR test. The diagnostic specificity will depend on the analytical specificity. Analytical sensitivity has been reported in the range of 10²–10⁵ cells per gram of feces. Nested PCR techniques are in general considered to be the most sensitive. However, on average, the reported nested PCRs were equally sensitive compared with the conventional PCRs. This finding may reflect differences in the applied DNA extraction methods or PCR reaction conditions.

The level of L. intracellularis cells per gram of feces in positive pigs are to some extent unknown. Pigs with lesions of PE have been reported to excrete in the range of 5 × 10⁴ to 7 × 10⁸ L. intracellularis per gram of feces. 7,34,41 Recently, a minimum excretion of 2.2 × 10³ L. intracellularis per gram of feces in a pig without clinical signs under natural conditions has been detected by quantitative real-time PCR (Marie Stalh, unpublished data, 2009). These excretion levels are below the analytical sensitivity for some PCR tests. Others have reported that some pigs are PCR negative in feces but PCR positive in samples from intestinal mucosa. 21 More research is needed in this area to estimate diagnostic sensitivity for fecal PCR detection of pigs infected with L. intracellularis.

It seems that the analytical specificity is close to perfect for the PCR tests. In the reviewed publications, different strains of bacteria species and negative fecal samples were tested. No false-positive test results were reported. This could imply that the diagnostic specificity for detection of pigs infected with L. intracellularis is close to 100% (in contrast to detection of pigs having histologic lesions of PE). In 6 publications, analytical specificity was only evaluated using purified DNA from a number of different bacteria. This may not be sufficient when taking the large number of different microorganisms in a fecal sample into consideration. Testing a large number of negative fecal samples from different pigs could be the best evaluation of analytical specificity for future PCR tests.

Table 3

Positive (PPV) and negative (NPV) predictive values of fecal polymerase chain reaction (PCR) detection of Lawsonia intracellularis–associated histologic lesions of proliferative enteropathy.*

	Diagnostic
True prevalence (%)†	Se (%)	Sp (%)	PPV (%)	NPV (%)	PCR reference no(s).
15	100	100	100	100	18, 28
	100	89	62	100	19
	100	50	26	100	44
	50	100	100	92	14, 21
	36	100	100	90	21
	97	85	53	99	27
30	100	100	100	100	18, 28
	100	89	80	100	19
	100	50	46	100	44
	50	100	100	82	14, 21
	36	100	100	78	21
	97	85	73	99	27
60	100	100	100	100	18, 28
	100	89	93	100	19
	100	50	75	100	44
	50	100	100	57	14, 21
	36	100	100	51	21
	97	85	91	95	27
75	100	100	100	100	18, 28
	100	89	96	100	19
	100	50	86	100	44
	50	100	100	40	14, 21
	36	100	100	34	21
	97	85	95	90	27
90	100	100	100	100	18, 28
	100	89	99	100	19
	100	50	95	100	44
	50	100	100	18	14, 21
	36	100	100	15	21
	97	85	98	76	27

Se = diagnostic sensitivity; Sp = diagnostic specificity.

†

True population prevalence of pigs with histologic lesions of proliferative enteropathy.

A case definition where the PCR test and the histologic examination were applied to the same animal within a few days was applied in the current review. In this case definition, the PCR tests were evaluated for the ability to reflect the animal's disease status at the time of sampling. L. intracellularis can be detected in feces using PCR from day 7 to 8 after infection. Shedding of L. intracellularis continues on average for approximately 15 days. 4,14,15,19,21,23,43,45

Further, some have reported that pigs may shed L. intracellularis intermittently over a prolonged period of time (up to 8–12 weeks). 6,15,41 This implies that other estimates of diagnostic sensitivity/specificity and predictive values may be obtained if the PCR tests are evaluated for the ability to reflect the animal's disease history or future disease development. Such anamnestic and prognostic application for the PCR tests is important for interpretation of a positive test result's importance for the pig or herd. These aspects need further evaluation.

Conclusion

Estimates of the diagnostic sensitivity and specificity of L. intracellularis–specific fecal PCR as an antemortem diagnostic test for pigs with histologic lesions of PE were obtained from 7 publications. These represented 6 different PCR tests. A range in both diagnostic sensitivity (36–100%) and specificity (50–100%) were reported. Validity and confidence limits of the estimates varied considerably. From the simulated predictive values, it was concluded that the risk of false positives will usually exceed the risk of false negatives when applying the PCR tests under conditions where up to 30% of the pigs in a herd have histologic lesions of PE. The histologic technique applied for evaluation of each PCR test has to be taken into consideration in the interpretation of positive and negative test results for each PCR test.

References

Baccaro

Moreno

Shinya

Dotto

: 2003, Identification of bacterial agents of enteric diseases by multiplex PCR in growing-finishing pigs. Braz J Microbiol 34:225–229.

Biester

Schwarte

: 1931, Intestinal adenoma in swine. Am J Pathol 7:175–185.

Biksi

Kacskovics

Mandoki

Ivan

: 1998, Detection of Lawsonia intracellularis in Hungarian swine herds by polymerase chain reaction. Acta Vet Hung 46:415–420.

Boesen

Jensen

Schmidt

: 2004, The influence of diet on Lawsonia intracellularis colonization in pigs upon experimental challenge. Vet Microbiol 103:35–45.

Chang

: 1997, Prevalence of Lawsonia intracellularis in swine herds in Taiwan. Vet Rec 26:103–104.

Collins

Love

: 2007, Re-challenge of pigs following recovery from proliferative enteropathy. Vet Microbiol 120:381–386.

Collins

Love

Pozo

: 2000, Studies on the ex vivo survival of Lawsonia intracellularis. J Swine Health Prod 8:211–215.

Dias

Gonzalez

Escatel

: 1998, Diagnóstico de ileítis porcina por medio de la reacción en cadena de la polimerasa [Polymerase chain reaction for diagnosis of porcine ileitis]. Vet Mex 29:263–267. In Spanish.

Dünser

Schweighardt

Krassnig

: 1997, Nachweis von Lawsonia intracellularis, des Erregers der Proliferativen Enteropathien des Schweines (Adenomatose-Komplex), mittels Polymerase-Kettenreaction (PCR) in Oberösterreich [Detection in Austria of Lawsonia intracellularis, the causal agent of proliferative enteropathy in pigs, by the polymerase chain reaction]. Wien Tierarztl Monatsschr 84:307–313. In German.

10.

Elder

Duhamel

Mathiesen

: 1997, Multiplex polymerase chain reaction for simultaneous detection of Lawsonia intracellularis, Serpulina hyodysenteriae, and salmonellae in porcine intestinal specimens. J Vet Diagn Invest 9:281–286.

11.

Gebhart

McOrist

Lawson

: 1994, Specific in situ hybridization of the intracellular organism of porcine proliferative enteropathy. Vet Pathol 31:462–467.

12.

Greiner

Gardner

: 2000, Application of diagnostic tests in veterinary epidemiologic studies. Prev Vet Med 45:43–59.

13.

Greiner

Gardner

: 2000, Epidemiologic issues in the validation of veterinary diagnostic tests. Prev Vet Med 45:3–22.

14.

Guedes

RMC

Gebhart

: 2003, Comparison of intestinal mucosa homogenate and pure culture of the homologous Lawsonia intracellularis isolate in reproducing proliferative enteropathy in swine. Vet Microbiol 93:159–166.

15.

Guedes

RMC

Gebhart

: 2003, Onset and duration of fecal shedding, cell-mediated and humoral immune responses in pigs after challenge with a pathogenic isolate or attenuated vaccine strain of Lawsonia intracellularis. Vet Microbiol 91:135–145.

16.

Guedes

RMC

Gebhart

Winkelman

:2002, Comparison of different methods for diagnosis of porcine proliferative enteropathy. Can J Vet Res 66:99–107.

17.

Jacobson

Aspan

Heldtander Konigsson

: 2004, Routine diagnostics of Lawsonia intracellularis performed by PCR, serologic and post mortem examination, with special emphasis on sample preparation methods for PCR. Vet Microbiol 102:189–201.

18.

Jones

Ward

Murtaugh

: 1993, Enhanced detection of intracellular organism of swine proliferative enteritis, ileal symbiont intracellularis, in feces by polymerase chain reaction. J Clin Microbiol 31:2611–2615.

19.

Jones

Ward

Murtaugh

: 1993, Relationship between ileal symbiont intracellularis and porcine proliferative enteritis. Infect Immun 61:5237–5244.

20.

Jordan

Knitted

Schwartz

: 2004, A Lawsonia intracellularis transmission study using a pure culture inoculated seeder-pig sentinel model. Vet Microbiol 104:83–90.

21.

Knittel

Jordan

Schwartz

: 1998, Evaluation of antemortem polymerase chain reaction and serologic methods for detection of Lawsonia intracellularis-exposed pigs. Am J Vet Res 59:722–726.

22.

Kroll

Roof

Hoffman

: 2005, Proliferative enteropathy: a global enteric disease of pigs caused by Lawsonia intracellularis. Anim Health Res Rev 6:173–197.

23.

Kroll

Roof

McOrist

: 2004, Evaluation of protective immunity in pigs following oral administration of an avirulent live vaccine of Lawsonia intracellularis. Am J Vet Res 65:559–565.

24.

Collins

Phillips

: 2006, Development of a multiplex-PCR for rapid detection of the enteric pathogens Lawsonia intracellularis, Brachyspira hyodysenteriae, and Brachyspira pilosicoli in porcine feces. Lett Appl Microbiol 42:284–288.

25.

Ladinig

Sommerfeld-Stur

Weissenböck

: 2009, Comparative evaluation of diagnostic methods for Lawsonia intracellularis infection in pigs, with emphasis on cases lacking characteristic lesions. J Comp Pathol 140:140–148.

26.

Lawson

GIIK

Gebhart

: 2000, Proliferative enteropathy. J Comp Pathol 122:77–100.

27.

Lindecrona

Jensen

Andersen

Moller

: 2002, Application of a 5' nuclease assay for detection of Lawsonia intracellularis in fecal samples from pigs. J Clin Microbiol 40:984–987.

28.

McCormick

Hasse

Monckton

: 1995, Detection of ileal symbiont intracellularis in porcine fecal samples by polymerase chain reaction. Vet Microbiol 47:387–393.

29.

McOrist

: 2005, Defining the full costs of endemic porcine proliferative enteropathy. Vet J 170:8–9.

30.

McOrist

Boid

Lawson

GHK

McConnell

: 1987, Monoclonal antibodies to intracellular campylobacter-like organisms of the porcine proliferative enteropathies. Vet Rec 121:421–422.

31.

McOrist

Gebhart

: 1999, Proliferative enteropathies. In: Diseases of swine, ed. Straw

Zimmerman

D' Allaire

Taylor

, 8th ed., pp. 727–738. Blackwell Publishing, Ames, IA.

32.

McOrist

Gebhart

Boid

Barns

: 1995, Characterization of Lawsonia intracellularis gen. nov., sp. nov., the obligate intracellular bacterium of porcine proliferative enteropathy. Int J Syst Bacteriol 45:820–825.

33.

McOrist

Gebhart

Lawson

GHK

: 1994, Polymerase chain reaction for diagnosis of porcine proliferative enteropathy. Vet Microbiol 41:205–212.

34.

Moller

Jensen

Jorsal

: 1998, Detection of Lawsonia intracellularis, Serpulina hyodysenteriae, weakly betahaemolytic intestinal spirochaetes, Salmonella enterica, and haemolytic Escherichia coli from swine herds with and without diarrhea among growing pigs. Vet Microbiol 62:59–72.

35.

Nathues

Holthaus

Grosse Beilage

: 2009, Quantification of Lawsonia intracellularis in porcine faeces by real-time PCR. J Appl Microbiol 107:2009–2016.

36.

Nathues

Oliveira

CJB

Wurm

: 2007, Simultaneous detection of Brachyspira hyodysenteriea, Brachyspira pilosicoli and Lawsonia intracellularis in porcine feces and tissue samples by multiplex-PCR. J Vet Med A Physiol Pathol Clin Med 54:532–538.

37.

Nielsen

Houe

Ersboell

Toft

: 2004, Evaluating diagnostic tests. In: Introduction to veterinary epidemiology, ed. Houe

Ersboell

Toft

, 1st ed., pp. 133–151. Biofolia, Denmark.

38.

Pejsak

Zmudzki

Stankevicius

: 2001, Detection of Lawsonia intracellularis by one tube nested-PCR. Medycyna Wet 57:723–726.

39.

Rowland

: 1978, Necrotic enteritis and regional ileitis in pigs at slaughter. Vet Rec 100:338–339.

40.

Rowland

Lawson

Maxwell

: 1973, Intestinal adenomatosis in the pig: occurrence of a bacterium in affected cells. Nature 247:417.

41.

Smith

McOrist

: 1997, Development of persistent intestinal infection and excretion of Lawsonia intracellularis by piglets. Res Vet Sci 62:6–10.

42.

Stege

Jensen

Moller

: 2000, Prevalence of intestinal pathogens in Danish finishing pig herds. Prev Vet Med 46:279–292.

43.

Stege

Jensen

Moller

: 2004, Infection dynamics of Lawsonia intracellularis in pig herds. Vet Microbiol 104:197–206.

44.

Suh

Lym

Bae

: 2000, Detection of Lawsonia intracellularis in diagnostic specimens by one-step PCR. J Vet Sci 1:33–37.

45.

Walter

Knittel

Schwartz

: 2001, Treatment and control of porcine proliferative enteropathy using different tiamulin delivery methods. J Swine Health Prod 9:109–115.

46.

Ward

Winkelman

: 1990, Recognizing the three forms of PE in swine. Vet Med 85:197–203.

47.

Zhang

Gebhart

Burden

Duhamel

: 2000, Improved diagnosis of porcine proliferative enteropathy caused by Lawsonia intracellularis using polymerase chain reaction-enzyme-linked oligosorbent assay (PCR-ELOSA). Mol Cell Probes 14:101–108.

48.

Zmudzki

Osek

Stankevicius

Pejsak

: 2004, Rapid detection of Brachyspira hyodysenteriae and Lawsonia intracellularis in swine fecal and mucosal specimens by multiplex PCR. Bull Vet Inst Pulawy 48:207–214.