Johne's Disease in Beef Cattle: A Fresh Look at Diagnosis, Control, and Regulatory Issues
Allen J. Roussel, DVM, MS, DACVIM, DECBHM
Johne's disease has become more important as both a clinical entity and a regulatory topic over the past 10 years. The purpose of this presentation is to provide an update on the scientific and political issues surrounding Johne's disease which have occurred in recent years. Some of this information is documented in the scientific literature, while some of it is new information or is "nonscientific" information that I have accumulated by attending meetings and training sessions, and from participating in conversations with colleagues in the field. Therefore, statements that are not referenced should be considered anecdotal or simply my personal opinion. Because this presentation is an "update" being presented in the "specialist" session, it is by design, more like a series of bullet points rather than a complete review of the topic. Note also that the title specifies beef cattle, but by necessity, much of the work cited comes from dairy cattle.
One of the most frustrating facets of Johne's disease control is the relative weakness of the testing platforms available for diagnosis. The most frequently-used serologic test, the ELISA, offers the advantage of wide availability, relative technical ease, rapid turnaround time, and low-cost. Because cattle usually do not respond serologically in the early and middle stages of infection, the ELISA suffers from poor sensitivity. Because cattle respond serologically to other mycobacteria in the environment, the ELISA can suffer from poor specificity in certain situations.1,2 Absorbed ELISAs display superior test performance in these situations. In the absence of cross-reacting environmental mycobacteria, ELISAs have relatively high specificity. Unfortunately, even tests with high specificity have low positive predictive value when the prevalence of the disease is low. This is particularly a problem in beef herds, and the problem can be compounded by the presence of cross reacting mycobacteria. The antigen detection tests (fecal culture and PCR) virtually eliminate the problem of imperfect specificity. The sensitivity of fecal culture and of certain PCR kits is higher than that of ELISA.3,4 However the cost is usually 4 or more times greater than that of ELISA. Also, variations in reagents and in laboratory technique have resulted in widely disparate results from fecal culture, and especially PCR, performed in different laboratories. The turnaround time for PCR is comparable to that of ELISA, while fecal culture requires 6 to 16 weeks.
Recent advances have improved these testing platforms. In the middle of 2006, IDEXX changed some of the reagents in their ELISA kit which substantially improved specificity without sacrificing sensitivity. IDEXX laboratories is currently in the process of making the transition in some countries to a different ELISA which was developed by Pourquier, a French company. In side-by-side comparisons, the Pourquier kit has the highest sensitivity and specificity of all the major commercial ELISA kits.5 If the new kit performs as well in beef herds plagued by high false positive rates as it did in the battery of sera used in the published comparisons, it will provide an important tool in the control of Johne's disease in beef cattle. At present, at least in the southern tier of states, culling or segregation of purebred beef cattle on the basis of ELISA is currently very expensive, even prohibitive, in some herds due to the high false positive rate. A different testing platform, flow cytometry, was proposed recently.6 Subsequently, the same group has developed an ELISA that showed great promise until they encountered herds with high cross-reactivity.7
Over the past five years, the major advance in fecal culture has been the proliferation of liquid culture systems in diagnostic laboratories. The advantages of the liquid culture systems are both a reduction in laboratory space and labor, and a reduction in the time to positive results. One such system made by TREK can detect heavy shedders within a few weeks. Samples are called negatives if no growth has occurred by 6 weeks. Another manufacture, Becton Dickinson, has introduced an improved liquid culture system called MGIT which has improved sensitivity over its previous, system, BACTEC. The net result to the end-user, the producer and veterinarian, is a more accurate and quicker test system, but for a higher cost.
Creating pools of 5-10 individual fecal samples has been documented to be relatively sensitive in dairy cattle.8,9,10 This technique has not been well studied in beef cattle, but offers promise to reduce the cost of fecal culture and PCR. Of course the disadvantage in beef cattle is that the cattle are usually not readily accessible for follow-up collection of individual samples from animals in the positive pool. Storing all fecal samples until the results from the pools are obtained requires extensive freezer space in the laboratory. It is imperative that these samples be stored at -70°C. because storage in at conventional home freezer temperature (-20°C) results in substantial loss of viability of Mycobacterium avium subspecies paratuberculosis (unpublished data)
The Johne's Disease Control Program in the USA
This year Tavornpanich, et al., published a paper using a sophisticated modeling technique to show what many people working in the field had already suspected; that the USDA Voluntary Bovine Johnes Disease Control Program protocol for the Test Negative Program was terribly insensitive at the herd level.11 The testing strategy upon which the program was based, ELISA screening followed by fecal culture of the ELISA positives, suffered from an extremely low herd sensitivity. For example, using the minimum 30 randomly selected animals, fewer than 10% of herds with 5% true prevalence would be detected. Only about 60% of herds with a 35% true prevalence would be detected. Testing 150 animals was only predicted to uncover 38% of herds with a true prevalence of 5%. The cost effectiveness of the strategy was also abysmal. Based on the perceived inadequacy of the testing protocol, new information concerning the sensitivity and specificity of the tests involved, and the misunderstanding of the term "Test Negative Program," the National Johne's Working Group and the Johne's Committee of the United States Animal Health Association requested that a study be commissioned to suggest ways to improve the efficiency and cost-effectiveness of the program. The USDA funded the project and 5 individuals (Wells, Fosler, Gardener, Roussel, Tarpanovich) produced a draft program for consideration by the NJWG Johne's Committee of the USAHA. The draft proposes a classification scheme based on maximum estimated prevalence. Whole herd or statistical subset testing is required, but flexibility in the type of testing done is allowed. Depending on the tests used and the test positive prevalence, herds are placed into one of several classifications. A major difference between this system and the Test Negative Program is that large herds with very low test prevalence can move substantially up the classification scheme. No confirmed positive animals (with antigen detection test) were allowed in the Test Negative Program. The proposed program focuses on helping producers estimate the risk of buying infected cattle. It is statistically no less risky to purchase from a small herd that has tested a few times and found no culture positive animals than to purchase from a large extensively tested herd with a documented low prevalence of infection. It is hoped that the new program will increase participation from large herd owners who are actively controlling the disease and who have low herd prevalence. Under the new system they can be recognized for their efforts. At the current time this proposed system is under review, but it has not been officially accepted. Another publication written by a panel of experts provides a framework for designing testing programs for producers with different types of cattle operations (beef, dairy, commercial, seed stock). This consensus statement suggests the "best test" to use to achieve the desired goal in a particular situation.12
Monensin and Johne's Disease
The treatment of Johne's disease risk traditionally relied upon anti-tuberculosis drugs. In the 1990s, Gordon Brumbaugh came across a statement saying that monensin inhibited Mycobacterium phlei. Theorizing that it might be effective against MAP, he led a research group through a series of studies beginning with a challenge model in mice in which the drug had a protective effect.13 Then we administered monensin to naturally infected cattle and showed that it reduced lesion scores.14 He then went on to demonstrate in vitro activity of monensin against MAP.15 Subsequent to those studies, Whitlock performed a challenge experiments which showed that monensin reduced the level of infection in calves.16 Recent studies in Canada shed some light of the effect of monensin in lactating dairy cattle. In one study, cows administered a monensin bolus had about an 80% reduction in the number of MAP organisms shed in the feces.17 The odds of being fecal culture positive after being treated with monensin was less in the group that was fecal culture positive at the beginning of the study and greater in those that were culture negative at the beginning. The same research group showed in another study that cattle in herds which fed monensin to lactating cows were about 1/5 as likely to be ELISA positive as those in herds which did not feed monensin. In herds with a history of previous disease, cows from herds which fed monensin to breeding age heifers were about half as likely to be ELISA positive.18Obviously, we do not know the full impact of feeding monensin on Johne's infection, clinical disease, control programs or diagnostic tests.
Genetics and Johne's Disease
Possible heritability of resistance or susceptibility to Johne's disease in Holstein cattle has been suggested by several authors in the past decade. Until very recently there was virtually nothing in the literature about the heritability of resistance or susceptibility in beef cattle. An increased likelihood of a positive ELISA in Brahman cattle compared to Bos taurus breeds was shown by our group as well as researchers in Florida.19 A series of studies by Osterstock suggests a hereditary predisposition to being sero-positive for Johne's disease. The appearance of certain cattle in the pedigree of Texas Longhorn cattle significantly increased the odds of being seropositive.20 In a cross-sectional study, the odds of an animal being classified as "suspect" or greater on the ELISA was 4.6 times greater if the dam was classified as "suspect" or greater.20 However in another study when herd of origin and seroprevalence was included in the model, there was a very weak association between the status of the dam and the calf.21 Taken together, these two studies suggest that if the status of the herd of origin is unknown, the status of the dam may be important in making a decision about purchase of an animal. However the second study suggests that within a herd, there is not a strong relationship between dam status and calf status. In his last published a study, Osterstock used microsatellite analysis to define 11 genetically related clusters in a population of tested animals.22 Cattle in 3 of the clusters has significantly greater odds of being positive (36.4, 7.4, and 5.9 times) than those in the referent cluster. One of the clusters containing about 15% of the cattle contained 5/9 fecal culture positive cattle.
Crohn's Disease and Map
The issue of MAP as a cause of Crohn's disease has not been resolved. Recently there have been a few noteworthy publications. In 2007, the much awaited results of a large-scale double-blind study of antibiotics in Crohn's patients in Australia were published. The conclusion of the paper was that the combination of clarithromycin, rifabutin, and clofazimine did not have long-term positive effects on Crohn's patients.23 As expected, the paper has received criticism from those with a different opinion. A systematic review of the published literature on the possible link between Crohn's disease and MAP was published in 2007. The authors considered 28 publications worthy of review. In studies looking for MAP in tissues with PCR, and comparing patients with Crohn's disease to those without inflammatory bowel disease, the pooled odds ratio was 7·01 (95% CI 3·95-12·4). They concluded that the association between Crohn's disease and MAP was specific, but its role in the etiology remained undetermined.24
1. Roussel AJ, et al. JAVMA 2007;230(6):890.
2. Osterstock JB, et al. JAVMA 2007:230(6):896.
3. Whitlock RH, et al. Vet Micro 2000;77:387.
4. Whitlock RH. 2007; Personal communication.
5. Collins MT et al. Clin Diag Lab Immuno 2005;12(6):685.
6. Speer CA, et al. Clin Vacc Immun 2006;13(5):535.
7. Eda S, et al. Food Path Dis 2005;2(4).
8. Kalis CHJ, et al. J Vet Diag Invest 2000;12:547.
9. van Schaik G, et al. Prev Vet Med 2007;82:159.
10. Wells SJ, et al. AJVR 2002; 63(8)
11. Tavornpanich S, et al. Prev Vet Med 2008;83:65
12. Collins MT, et al. JAVMA 2006;229(12):1912.
13. Brumbaugh GW, et al. Am J Vet Res 1992;53(4) 544.
14. Brumbaugh GW, et al. J Comp Path, 2000;123(1):22.
15. Brumbaugh GW, et al. Can J Vet R 2004;68:175.
16. Whitlock RH, et al. 8th Internat Coll Paratuber 2005, Copenhagen.
17. Hendrick SH, et al. Prev Vet Med 2006;75:206.
18. Hendrick SH, et al. Prev Vet Med 2006;76:237.
19. Elzo, MA, et al. J. Anim. Sci. 2006;84:41.
20. Osterstock JB, et al. Vet Micro 2007(in press).
21. Osterstock JB, et al. Prev Vet Med 2007(in press).
22. unpublished data.
23. Selby W, et al. Gastroenterology 2007;132(7):2313.
24. Feller M, et al. Lancet Infect Dis 2007;7:607.