MAY 0 7 2001 11 9 9 0 0. SEP 112003 MWESEQ; 1 ABSTRACT AN tnnmomslc sum 10 EVAwA'rE METHODS or nmcmc BOVINE 'msnncumsxs By David Robert Towar The object of this project was to determine if a combination of the presently available tests for tuberculosis would give results that could help differentiate animals infected with Mycobacterium 223;; from non-infected animals. The combination of tests included the comparative cervical intradermal tuberculin test and a passive hem- agglutination test of serums collected before and following the caudal fold intradermal tuberculin test of all animals in all herds in a selected geographical area. Results substantiated previously observed errors inherent in each of these methods used as a single or in a combined test of a single animal. Both supplemental methods reduced the number of animals having a false indication of infection with 5. my, Neither method, however, was totally effective in identifying all animals with 3. P311; infection. There was too high a degree of error for their effective use as single tests for the individual animal. The comparative cervical tuberculin test identified a large percent of the g. 291;; infected animals with signs of progressive disease. It also demonstrated a certain number of probable false positive results. There was a significant increase in post tuberculin serum titers from animals in g. M infected herds. A high percentage of false positive results, however, eliminates this method as a single test for a single animal. As a combination, the David Robert Towar supplemental test methods were found to be effective for differentiating herd infections with Li. REE-.9.- Seventy-one of 92 herds receiving caudal fold tuberculin tests had one or more animals showing a tuberculin response. Supplemental test- ing procedures rapidly reduced the number of suspicious herds to twelve. Five of the 12 herds were found to be infected with 11. 22".}:- A com- bination of a positive comparative tuberculin test and a post tuberculin injection hemagglutination titer of 640 or higher was highly specific for g, bggig_infected animals with demonstrable lesions at post mortem examination. This combination of tests, however, did miss some animals with g. M infection. The use of the supplemental tests resulted in earlier established knowledge on the presence or the absence of l_i. 2911;; infections within a herd. Despite increased effectiveness with these procedures, epide- miological investigative procedures still should play an important role in arriving at a diagnosis. AN BPIDEHIOLDGIC STUDY TO EVAEEATE METHODS OF DETECTING BOVINE TUBERCULOSIS By David Robert Tower A THESIS Submitted to ' Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Microbiology and Public Health 1966 AXY‘M; ~1 \" /\.: .{F /\ V DEDICATED to JUDY ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation and thanks to Dr. C. A. Niles for his efforts in the testing proce- dures, to Dr. C. L. Hendee, Dr. R. M. Scott, and Dr. W. L. Hallmann for their support and guidance, and to Dr. V. B. Hallmann and her group for their able assistance. Thanks also, to the many people in the Michigan Department of Agriculture and the Heat Inspection Division of the Federal government who gave much needed assistance in collecting and proc- essing samples and to Dr. J. L. Isbister, Tuberculosis Control Officer of the Michigan Department of Health, and his group for the associated testing of the human population. Much appreciation for the initial help in public relations is given to the veterinary practitioners in the Deckerville clinic and to Keith Sowerby, Sanilac County Extention Director. Gratitude is expressed to the Michigan and U. S. Department of Agriculture who made this study possible. ii. TABLE OF CONTENTS Page IN ROD" CTION o s o o a o e s s e o e o e o o a a o o e e o e ’ 1 LITERATURE REVIEW . . . . . . . . . . . . . . . . . . . . . 10 DEFINITIONS . . . . . . . . . . . . . . . . . . . . . . . . 20 PROCEDURE . . . . . . . . . . . . . . . . . . . . . . . . . 22 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . 29 Caudal intradermal tuberculin test . . . . . . . . . . 30 Other investigations . . . . . . . . . . . . . . . . . 33 Seventh day observation of caudal fold tests . . . . . 35 Tuberculoid skin lesions . . . . . . . . . . . . . . . 36 Comparative test . . . . . . . . . . . . . . . . . . . 40 Serology . 46 CONCDUSIONS . . . . . . . . . . . . . . . . . . . . . . . . 56 Modifications . . . . . . . . . . . . . . . . . . . . . 63 SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . 65 APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . 68 LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . 74 iii. Table 3a 3b 4a 4b LIST OF TABLES Caudal fold intradermal test results . . . . . . . . Cattle having a caudal fold test response and/or tuberculoid skin lesions . . . . . . . . . . . . Comparative cervical test results . . . . . . . . . Comparative cervical test results . . . . . . . . . Mean increase of post tuberculin injection hemagglutination titers . . . . . . . . . . . . . Hemagglutination (HA) titers from herds with comparative reactors . . . . . . . . . . . . . . . Comparison of testing results from one Mycobacterium bovis infected and one non-infected herd . . . . . Suggested epidemiologic procedure for detecting a. bov18 infection 0 O O O O O O O O O O O O O O 0 iv. Page 31 38 42 42 48 49 59 62 LIST OF CHARTS Figure Page 1 Tuberculosis reactor rate since 1917 . . . . . . . 7 2 Progress in tuberculosis eradication since 1916 as shown on logarithmic graph . . . . . . . 8 3 No gross lesion rate of tuberculin test reactors since 1920 . . . . . . . . . . . . . . 9 4 Marion Township cattle with caudal fold test responses by age . . . . . . . . . . . . . . . . 39 5 Incidence of tuberculoid skin lesions in Marion Township cattle by age . . . . . . . . . 39 6 Comparative cervical test results . . . . . . . . 43 7 Cattle farm locations - Marion Township . . . . . 45 8 Pre and post tuberculin injection hemagglutination titers. All samples . . . . . 52 9 Post tuberculin injection hemagglutination titers Animals from Mycobacterium bovis infected and non'InfECEEd herds s s s a s s s s s s e s e s s 53 10 Post tuberculin injection hemagglutination titers. Animals with caudal test responses and caudal test negative animals . . . . . . . . . . . . . 54 11 Post tuberculin injection hemagglutination titer 1ncre.se8 O O O O O O O O O O O O O O O O O O O 55 LIST OF APPENDICES Appendix 1 USDA, ARS, ANH method of classifying responses to the caudal fold intradermal test. Determined by p‘ lp‘tion O I O O O O O O O O O O O O O O O O 2 x2 Ana lye 18 0f data 0 O O O O O O O O O O O O O 3 Post Mortem and Laboratory Results . . . . . . vi. Page INTRODUCTION Since the initiation of the bovine tuberculosis eradication program and the resulting reduction of bovine tuberculosis, increasing dif- ficulties in the interpretation of the intradermal tuberculin test have been encountered. These difficulties can be classified under two general headings: 1. Some tuberculous animals do not respond to tuberculin. This factor makes the assured elimination of tuberculosis from a large herd of cattle through use of the tuberculin test very difficult. Examination of the tuberculin test history of animals found to be tuberculous at slaughter indicates that the percent failing to respond has remain- ed about the same over the years. This error is referred to as a Type 11 error. 2. Some non-tuberculous animals do respond to tuberculin. These animals are generally reflected by the no-gross lesion rate of slaughtered reactors which has increased over the years. This is a Type I error. This latter problem, the non-specific reactor becomes more apparent as the final phase of bovine tuberculosis eradication approaches. With modern means of rapid transportation, bovine tuberculosis is readily transported. To eradicate bovine tuberculosis, there must be an effec- tive means of detecting all infected animals before spread of the dis- ease occurs. At present, the two major methods of locating infection are dependent upon advanced tuberculosis. Tracing diseased animals found at slaughter means the animal has undoubtedly exposed many others. It is also hampered by a lack of a satisfactory system of identification of the cattle population. The second method, screen- ing the cattle population with the intradermal test, theoretically detects cattle with tuberculosis earlier. If this test were 100% accurate, tuberculosis eradication would be simply a matter of test- ing all cattle and reservoirs of bovine tuberculosis, and eliminating the animals positive to the test. However, development of further criteria is necessary for interpreting the results of the caudal fold tuberculin test, because (a) only a portion of the cattle population is tested by any screening method; (b) a portion of those tested will show positive results due to non-specific causes; (c) not all tuber- culous animals respond to the tuberculin test; (d) not all tuberculous animals will have gross lesions at slaughter.' Completely negative results to the test are not a problem unless occurring in a herd where tuberculosis is known to exist. In all other cases the veterinarian is called upon to decide if the response to the tuberculin test is due to Mycobacterium 221;; or due to some other cause. To do this he must consider the individual animal's and the herd's history, the number and size of responses observed in the herd, the animal's physical condition, and the environment. If a decision cannot be made, a retest 60 days later using the same type test is the only recourse. A further diagnostic test is needed. If such a test could accurately differentiate specific from non-specific responses, it would contribute immeasurably to the eradication program even though it might be a complicated and time consuming procedure. No such method has yet been developed. Many other methods have been developed to aid in the diagnosis of tuberculosis. Some are as accurate as the caudal fold tuberculin test method. None are as simple or sufficiently more accurate than the caudal intradermal tuberculin test and have not been used widely in this country. It could be that the area of error for these various tests do not completely overlap, so that certain results for a group of tests on the individual or herd may improve diagnostic efficiency. In this paper an attempt is made to define the non-specific problem which is defined as sensitization responsible for responses to mammalian tuberculin in animals free of tuberculosis due to M, bggig, and to apply certain other testing methods in an effort to differentiate the specific from the non-specific responses. The intradermal tuberculin test is the most common method used throughout the world and is the official test in the USA. It is used as the basic test in this project. Other methods are added in an attempt to clarify the caudal test results. A comparative intradermal tuberculin test similar to the English single intradermal comparative method is used to retest all animals responding to the caudal fold test. Inclusion of this test in this project was prompted by the high correlation between progressive mycobacterial disease and a positive comparative test (McGavin, 1963). In the study of the patho- genicity of atypical mycobacteria by the tuberculosis project personnel at Michigan State University, it was demonstrated that some Group III mycobacteria can cause a progressive disease comparable to bovine tuberculosis, some cause a progressive disease that terminates in a non-progressive type, while others produce no disease. Comparative tuberculin tests were applied to these animals at two month intervals during the course of the experiment. When the disease was in a progressive state, the reaction to mammalian tuberculin was generally at least 5 mm greater than that of avian tuberculin. ‘When the reaction to mammalian tuberculin was less than 5 mm greater, or less than that for avian tuberculin, the disease was generally in a non-progressive state. Later studies have revealed exceptions to these findings. Serologic procedures are in many ways much more desirable as testing methods. Testing of a sample of blood eliminates any possi- bility of influencing an animal's immunologic response by injection of a foreign substance. Tuberculin may be capable of inducing anti- body formation when injected intrsdermally. Mallmann 35 :1. (1963;) found that a single hemagglutination (EA) test as a diagnostic tool was not satisfactory due to an anamnestic-like response following a tuberculin test. Preliminary results of tests of serums from two gross lesion cows naturally infected and two no-gross lesion cows indicated a four fold increase occurred only in the gross lesion animals. It was not consistently true in calves experimentally infect- ed. If this anamnestic EA test is consistent in this respect in naturally infected animals, it has great potential as a field test for tuberculosis. Disclosure of the presence of an agent in a herd capable of producing a non-specific hypersensitivity can be a great aid in explain- ing tuberculin responses in such a herd when M, Egyig is not evident. Non-specific hypersensitivity has been reported due to M. ggigm, _M_. matuberculosis, M. tuberculosis. These are generally localized factors causing hypersensitivity while the non-specific problem is more ubiquitous in scope. One factor probably causing hypersensitiv- ity and seemingly also world-wide in occurrence is tuberculoid skin lesions. Many non-specifically hypersensitive reactors have clinical signs of skin lesions. In unpublished studies Willigan,‘McGavin, and Goyings (1963) found microscopic lesions indicative of tuberculosis in the prescapular lymph nodes of reactor animals with skin lesions on the forelegs. Later, Mallmann 55 51. (1963) isolated atypical myco- bacteria from lymph nodes of N.G.L. tuberculin reactors both with and without skin lesions. Some of the Group III organisms were capable of inducing varying stages of disease when introduced as an aerosol or per utero. The isolation of atypical mycobacteria from tuberculin positive cattle that can produce disease and sensitivity to tuberculin in experimental calves, plus the relationship of skin lesions and prescapular lymph node granulomas, introduces a new concept in non- specific hypersensitivity. A skin lesion-lymph node granuloma syndrome may be represented by many stages with clinical skin lesions representing one stage. Almost all reports of a connection between skin lesions and non-specific hypersensitivity are biased. ‘Most reports on incidence of skin lesions are from tuberculosis reactor slaughter reports or from herds, cattle, or areas where a non-specific response or a skin lesion problem has arisen. This bias could be avoided by examining all animals in a geographical area for skin lesions. The tuberculosis research team at Michigan State University proposed a field survey of the incidence of non-specific tuberculin reactions, the application of the comparative test for detecting progressive disease, and the application of the anamnestic hemagglutina- tion test to both tuberculin-positive and negative animals. A con- ference was arranged with Dr. R. M. Scottl, Michigan Bovine Tuberculosis Project Leader, to plan such a survey. Later, a second conference was held with Dr. A. F. Ranneyz who gave his Department's support to such a project to be conducted by the author and Dr. C. A. Niles3 subject to approval of the Federal Veterinarian in Charge, Dr. Asa Winter4, and the State Veterinarian, Dr. John F. Quinn5. The data and conclusions drawn from the resulting project are applicable only to a small portion of Michigan. Conditions throughout the country where tuberculin test problems exist are sufficiently similar that collection of these data could establish a base of normal information that is essential for further epidemiological investiga- tions. This project was undertaken in an effort to answer questions regarding tuberculin hypersensitivity and to establish an improved testing procedure to help guide the veterinarian in the evaluation of situations where the caudal fold test needs to be supplemented. 1R. M. Scott, D.V.M., Assistant Veterinarian in Charge (Michigan), USDA, ARS, Animal Health Division. 2A. r. Ranney, D.V.M., Chief Staff Veterinarian, Tuberculosis Eradica- tion, USDA, ARS, AND. 3c. A. Miles, D.V.M., Tuberculosis Epidemiologist, USDA, ARS, m. 4Asa Winter, D.V.M., Veterinarian in Charge (Michigan), USDA, ARS, AND. 5John F. Quinn, D.V.M., State Veterinarian, Michigan Department of Agriculture. uu_>¢um zumemnom smeeeaeu_xu< mz3e3:u_xe< Io hearerreuo .m.= poem Hmomum coma cam.” . ova." 0mm." ONmA l1|1 ‘ q I w d 1 [H d I mo.o . n H d m r I N 9 u 1. 0 TC 0 . - a u. 1. .l 9 1. 9 B a | v m- . u n .uHmH ooeau ease eeuouue usuoesoeonsa .H enema» muw>zLa rm rsqmw speak 4 neu_¢o< m¢=.330.ec< no creepequma .».3 98 88 88 mam. 82 gm. com. 2.2 2m. 8m. 8m. mom. 82 RE. 92 one one 8m. 82 2a.. _a age—.2 .132122::_::::_::1:“qu22—11 111:. 4.1—2: :3 _o _- 588° ., ll..- / 11.8 nun 0.— / V3 0 $50000 . a w 2 u . 11°. . 88o / / 4|. v 1m" 8. r 1 om. ~ 88 _ #1... III. .11" 83 /‘ _ . _ 1 110mg n :8 A _ 1...: _ 1L... _ g 80.2 m A _ II: II. _ e «.3 . _ IE 8x8; 000.0006. «mm 1!!! .33. 538mm 2 83 228.. 855sz el femoo 80.8. I: @583 80.83 Se :2 138mm z 822%. ml . . e. .8. BBB 0868.0. as $229 2983 NI . as... BEE 80.086. Eu $083“ _I Mn. u“ 8.83. _ e _ , _ a u S. b b b s. . a. 9.. u .103 umsftomo. co c397. no 03— ooc? cotofivoto £33232 cm ”360$ .N 239“. 32>sz :oxqunmm 44.23.330.52 wanzzugg mo ~51»:qu .moa nwmu Own: mama coma auéa ovms moo.— .omma nNmH Ommu 11 e u q u a u q u q .I. O 19 "0st 00! led sou oON¢H ”Or—Hm ”HOUOQUH umflu CfiHn—UhOD—Ju HO QUMH GOHmwfi mmOHw OZ ofi UHDth. LITERATURE REVIEW Tuberculin has been used as a diagnostic agent in the U.S.A. for tuberculosis in cattle since 1892, just two years after its dis- covery by Robert Koch. At that time, Pearson (1892), who had worked in Koch's laboratory, tuberculin tested a herd of cattle in Penn- sylvania. Bis subcutaneous injection of Koch's old tuberculin re- sulted in an elevation of temperature of many animals, all of which showed signs of tuberculosis when slaughtered. Further testing indi- cated a high correlation between an animal's response to tuberculin and its infection with Mycobacterium'bggig. Many other methods of applying the tuberculin test have since evolved; however, for ease and convenience the intradermal test has been shown to be superior to other methods. By 1917, the U. 3. Government was sufficiently convinced of the efficacy of the tuberculin test to inaugurate a bovine tuberculosis eradication program based on this test. The subcutaneous thermal test was used in the early 1920's but the intradermal test applied to the caudal fold was initiated which facilitated a more rapid tuberculin test coverage of cattle herds without sacrificing reli- ability. In 1934, the Bureau of Animal Industry introduced a heat inactivated, concentrated tuberculin, the filtrate of organisms grown on synthetic media. This replaced the animal peptone medium tuberculin and resulted in a more standardised product. This product and method of test is still used today. Before the United States bovine tuberculosis eradication program started in 1917, the incidence of tuberculosis in the cattle 10 11 population was estimated from slaughter reports to be about 52, though some estimates ranged as high as 8 percent. It can be seen from Figure 1 that there was a rapid decline in the reactor rate until the early 1940's. Since that time the rate appears to have assumed an endemic plateau of about 0.2 percent. Figure 2 is a logarithmic graph illustrating numbers of animals retained and condemned due to tuber- culosis at slaughter since 1916. The reactor rate has had little effect on the rate of reduction of animals found with lesions of tuberculosis on slaughter. If infection with M. M were the only cause of tuberculin hypersensitivity and if it could be assumed that the proportion of animals in early and later stages of the disease remained roughly constant, then these two graphs should be quite similar. The reactor rate is based on tuberculin hypersensitivity. How- ever, )1. M is not the sole agent responsible for inducing hypersensitivity, some diseased animals do not react to tuberculin and some animals reacting to tuberculin were found with no-gross lesions (N.G.L.) of tuberculosis. The early concern was with the failure to disclose all diseased animals, but as the N}G.L. rate be- gan to rise (Figure 3), it was improbable that all N.G.L. reactors represented early tuberculosis infection. In 1914, Hastings reported an increase in the N.G.L. rate in Wisconsin reactors when the testing program was widened to include more than just suspect herds. He also noted that the N.G.L. rate was inversely proportional to the number of reactors found in a herd. Bis conclusion at this time was that most of the no-lesion reactors prob- ably had tuberculosis lesions not detected on routine autopsies. By 12 1924, he believed the accumulated evidence pointed to sensitization by something other than'!, bggig, Hagen (1930) reviewed 10 years of the eradication program and stated that one would expect the decrease in bovine tuberculosis to be accompanied by a similar decrease in N.G.L. reactors but instead, the latter was increasing. The N.G.L. rate was not consistent throughout the U.S.A. It was quite low in some areas and as high as 322 in others. By 1965, the reactor rate had decreased to 0.081.while the N.G.L. rate for these reactors had risen to over 75 percent. This latter figure is misleading. Mbrehouse (1963) reports that about 302 of the lesions from tuber- culin reactors submitted to the National Animal Disease Laboratories are due to causes other than tuberculosis. Many workers have thought that the increased numbers of N.G.L. reactors indicated an increased rate of non-specific hypersensitivity or a change in the test. To the contrary, Robertson (1963) has shown that the proportion of tests negative for animals actually infected with M. 1321;; (Type 11 error) and tests that are positive for non-infected animals (Type I error), is similar to that seen in early phases of the program. The increased "percent" of N.G.L. reactors is an incorrect conclusion based on improper statistical sampling. With the proper procedure the Type I error, the proportion of non-tuberculous animals that react, can be defined as a percent of the normal (non-tuberculous) population. Thus, it is evident that even after the last _M_. my; infected animal is slaughtered there will be a certain percent of the cattle popula- tion that will still react to tuberculin. This has been observed by Stenius (1949) in Finland where M. 22113 has been considered eradi- cated, yet 32 of the cattle population still react to tuberculin. 13 Determination of the cause or causes of the Type I error is necessary for determining the nature of non-specific hypersensitivity but not necessary for the eradication of tuberculosis. As already mentioned, Finland can claim eradication of bovine tuberculosis with an unexplain- ed 31 level of hypersensitivity. Many possible causes of non-specific hypersensitivity have been suggested. Hagen (1945) points out that infection with;!. agigg, _M_. Muberculosis, M. tuberculosis have all been shown to produce cross reactions with mammalian tuberculin. Granulomatous lesions with acid fast organisms found in the skin and subcutaneous tissues of the legs were at one time thought due to M, bggig. Later they could not be associated with M. m, were labeled tuberculoid skin lesions and incriminated as a cause of non-specific hypersensitivity (Traum 1959). Karlson (1959) in a review of the literature found at least 25 different mycobacteria capable of sensitizing animals to tuberculin. Mallmann, g£_;l, (1964) isolated many unclassified mycobacteria from bovine and swine tissues. Inoculation of experimental animals with these organisms led them to the conclusion that tuberculosis or tuberculosis-like disease in domestic animals may be caused by a range of acid-fast organisms. Furthenywhen infection occurred, sensitivity was induced which confirmed the heterogeneity of the isolants. Peterson (1956) gives an excellent review of reports of suspected causes of non- specific hypersensitivity but evidence incriminating organisms other than the mycobacteria is scant and circumstantial at best. Consider- ing the opportunities cattle have of contacting the genus MycobaCteria, it is no wonder that there is a certain portion that react to the wide spectrum antigenicity of mammalian tuberculin. Attempts to develop a 14 more specific agent such as purified protein derivative (P.P.D.) tuberculin have fared little better. Francis (1958) points out that no tuberculin has so far been developed that will detect all tuber- culous animals without eliciting a non-specific reaction in a varying but high proportion of animals. The level of non-specific hypersensitivity in the cattle popula- tion reported in the literature varies. McLaughlin (1962) tested 2,686 cattle in 86 herds in a Wisconsin township and found 6.12Hwith a detectable response to tuberculin in 601 of the herds tasted. Tervola and Berman (1959) found 12.62 responding cattle in 751 of the herds tested in six townships. In Outagammie County, Wisconsin, 1.52 of 42,850 cattle were reactors. Ranney (1966) reports that in Cache County, Utah, 591 (32) of 20,440 cattle tested were classified reactors or suspicious by the use of the tuberculin test. Chaloux (1964) reports that 18,202 cattle tested in 23 states by regulatory veterinarians disclosed 3.22.with tuberculin responses. Reports from many areas of the world present similar findings. These data are difficult to com- pare as the populations tested were usually biased and frequently not fully defined. In many of these samplings, only adults were tested or calves below 6 months of age were excluded. In most, the definition of a reaction is nebulous as it takes into account a qualitative judgement by many different veterinarians and frequently, the popula- tion tested is selected due to an apparent high degree of tuberculin hypersensitivity. Undeniably the best answer would be the development of a simple test capable of detecting all stages of M. bovis infection. This goal has thus far eluded the grasp of research workers. Personnel 15 responsible for conducting the program, meanwhile, must make the best use of the tools already available. This is difficult due to the conflicting data and claims for the different testing methods. Many different forms of retesting questionable animals have arisen. In the U.S.A. a second caudal fold intradermal test, 60-120 days after the first test is the primary method of retesting question- able animals. The veterinarian is required to determine subjectively if the response to the second test is similar to, or larger than that of the first test. If this criterion is met, the animal is consider- ed infected with M. 39133. This second test, being the same as the first, gives little additional information to the investigator as to the nature of the hypersensitivity. Gregory (1949) suggested that the short thermal test could be used as a check on intradermal reactors. This method consists of the subcutaneous injection of tuberculin and a subsequent body temperature elevation in.M§ 222i; infected animals. Robertson (1963) found the short thermal test ancillary to the sub- caudal to be superior to a repeat subcaudal test given 60 days after the initial test. Larson 5 Ropecky (1964) developed a modified thermal test in which a quantity of tuberculin is injected directly into the vein of an animal. A rise in temperature is considered indicative of specific M. Loy}; infection. They concluded that this method could not replace the standard skin test but was useful as a supplemental tool in detecting TB-carrier cattle. Northern Ireland favors the Stormont technique, a double inject- ion, using P.P.D. tuberculin. A second intradermal injection is made in the site of the first, seven days later. It is their belief that an M. bovis infected animal will show a further increase in skin 16 thickness within 24 hours after the second injection. Kerr, Lamont and McGirr (1949) reported a more reliable specificity by this pro- cedure than by either single or comparative intradermal tests. In the U.S.A. an attempt was made to reduce non-specific re- sponses by decreasing the dose of tuberculin in units given per dose. It was found that many tuberculous cattle would not respond to these decreased doses due to a low level of hypersensitivity. For similar reasons, selecting animals with responses over a certain size as being specifically infected gives no better results. The degree of hypersensitivity in both the specifically and non-specifically infect- ed animals varies widely. Many animals actually infected with M. 9.91.12. but with a low degree of hypersensitivity are not identified, while too many highly hypersensitive non-specifically infected animals will be slaughtered. The English use what they call a single intradermal comparative test which is based on the assumption that despite cross reactions, homologous antigens elicit a larger response in infected animals than those elicited by heterologous antigens. In their test a mammalian P.P.D. is compared with an avian P.P.D. for capacity in producing responses in the same animal. It is assumed that the antigens re- sponsible for non-specific reactions are more closely related to avian P.P.D. than to the mammalian P.P.D. Thus, a specifically infected animal will respond more to mammalian P.P.D. while the non-specifically infected will respond more to avian P.P.D. or equally to both. This is the method used in many countries today; however, just as with other intradermal methods it is evident that is also confuses specific with non-specific responses. 17 Van waveren (1953) using the single intradermal comparative test in Holland found that 96.5% of cattle subsequently found to be tuber- culous, had responses to the mammalian P.P.D. When mammalian and avian tuberculin reactions were compared, only 86.32.wou1d have been detect- ed by the comparative test criteria. Be concluded that the comparative test should not be used in heavily infected herds or on animals of unknown origin, but that it is very valuable in clean herds as it pre- vents the condemnation of many no-lesion reactors. During studies of the pathogenicity in cattle of some Group III mycobacteria, McGavin, ggugl. (1964) noted a relationship between progression of a mycobacterial disease and the single intradermal com- parative test. Animals in which the Group III organisms caused a progressive disease indistinguishable from bovine tuberculosis, reacted to the comparative test while those with non-progressive infect- ions did not. These animals with non-progressive lesions at time of slaughter had reacted early after the introduction of the organism but became non-reactors on later tests. Many serologic tests have been applied to the diagnosis of tuber- culosis in cattle with variable success. Middlebrook & Dubos (1948) described a hemagglutination test subsequently modified by Smith and Scott (1950) and by Rheins and Thurston (1955) which detected specific antibodies in the serums of tuberculous cattle. Vardman (1962) found the hemolytic modification of this hemagglutination test to be superior. Grey (1953) meanwhile, reported that neither the B.A. test or its hemolytic modification were superior to the intradermal test. 18 Many researchers have worked on a complement-fixation test. In 1945 Johnson found 712 of tuberculosis reactors with lesions posi- tive to the C.F. test. There were also 75% of reactors with only skin lesions and 51% of N.G.L. reactors that had positive C.F. tests. Seven percent of normal cattle were positive. vardman and Larsen (1964) after testing over 2,000 cattle for C.F. antibodies against tuberculosis concluded that this test was of no value for detecting in- fected cattle on an individual basis; however, they did find a good correlation on a herd basis between Mg‘gggig infection and high serum titers. They found only 2 of 28 tuberculosis free herds (7.12) had animals with diagnostic titers, while 14 of 15 M} 223;; infected herds (93.3%) had such animals. In recent years much work has been done on gel and double gel diffusion precipitation tests. These seem to offer more hope as a diagnostic test than other serologic tests. The double-diffusion gel precipitin test could, when fully exploited by use of multiple anti- gens in serial dilution, be of critical value in distinguishing tuberculosis from non-tuberculous disease (Alshabkhoun g£._l 1960). Parlett (1959) found the double-diffusion method to be highly specific with only a 2.1% non-specific response rate. He noted that most of these false positives actually resulted fnam the presence of antibodies against mycobacterial antigens in the serum of these reactors but that some may have been precipitation artifacts. 19 Mallmann, g£_§1_(l963) found the hemagglutination methods of Takahashi, and of Middlebrook and Dubos to have little value in differentiating those animals reacting to the tuberculin test due to M. 223$; infection from non-specifically reacting animals. They did note an anamnestic-like rise in antibody titers in tuberculin positive animals from gross lesion herds, three to fifteen days after tuberculin testing. They suggested that a differential test might be possible utilizing this anamnestic-like response. It can be readily noted in a brief review of the literature that comparisons of similar testing methods is difficult if not impossible. Various details of the testing procedure change, the population tested is frequently not adequately defined, and what each author means by reactor, suspect or deviator is subject to question. In 1956, Paterson suggested, "A valuable study from the epizootilogi- csl aspect would be to carry out standardized tests in different world areas on randomly selected tuberculosis-free herds of compar- able productivity and geographical status. Such investigations would give little information on causes of sensitization, but would permit quantitative comparison of reactivity in a way at present impossible, and eliminate much subjective opinion on the incidence of non- specific sensitization." ADULT CATTLE: CAUDAL FOLD TEST: CERVICAL TEST: COMPARATIVE TEST: GROSS LESION: H.A. TEST: HERD: HYPERSENSITIVITY: N.G.L. NON-SPECIFIC: DEFINITIONS Two or more years of age An intradermal tuberculin test applied to the fold of skin at the base of the tail. An intradermal tuberculin test applied to the skin of the lateral aspect of the neck. The simultaneous intradermal injection of the same animal with two different tuberculins for the purpose of comparing the degree of hyper- sensitivity of one to the other. In this report A.R.S. mammalian and avian tuberculins were used. Visible signs of tuberculosis at post mortem inspection. Hemagglutination test. In this report refers to Mallmann,g_t_; L1. (1963) modification of Middlebrook and Dubos (1948) procedure. A group of cattle belonging to one owner. Any regction elicited at the site of the intra- dermal injection of an amount of tuberculin inocuous to the normal individual visible or palpable at the official time of reading. No Gross Lesion. The lack of any visible signs of tuberculous disease at post mortem inspection. In this report, refers to sensitization which is responsible for responses to mammalian tuberculin in the animals believed to be free of tuberculosis due to M. bovis. 20 P.P.D. REACTOR: RESPONSE: SKIN LESIONS: TYPE I ERROR: TYPE II ERROR: 21 Purified protein derivative tuberculins. P.P.D.-S is an antigen made from M. tuberculosis, while P.P.D.-B is made from Batty type organisms. Lydia Edwards (1960). An animal with a comparative cervical test having a mammalian response as defined below, 5 mm or more than the avian response. Any visible or palpable sign of hypersensitivity at the reading time for the test. This was at 72 hours for the caudal test and 48 hours for the comparative test. Minimal response--those classi- fied less than Pl, X2 in size. See Appendix.l. Clinical signs of a nodular disease of the skin or subcutaneous tissues usually involving the lower forelimbs of cattle. Atypical mycobacteria have been incriminated as the causative agent. The error of a test that denotes bovine tuber- culosis in an animal or herd when it does not actually exist. The error of a test failing to detect bovine tuber- culosis when it actually occurs. PROCEDURE The initial tuberculin testing and investigation were con- ducted under actual field conditions. The purpose being that re— sults obtained would be applicable to the present testing methods in the U.S.A. Further, the additional differential methods used in this project are familiar procedures and if successful, could be easily included in the present bovine tuberculosis eradication program. To obtain results without the bias that would be introduced by spot selection of certain types of herds all resident herds were tuberculin tested in one entire township. Herds representing all three types of tuberculin test problems were to be found in the area to be investigated., These three types were (1) herds infected with M, Mgggg, (2) herds with animals with tuberculoid skin lesions, and (3) herds with animals with tuberculin hypersensitivity not caused by M, 221;; infection (non-specific hypersensitivity). The selection of this area was governed primarily by the presence of known M, 223;; infected or possibly infected herds. The other two types of herd problems are ubiquitous. In only one area of Michigan were there several adjacent townships that seemed to meet these criteria. One of these townships was chosen. It was Marion Township, in Sanilac County, 36 square miles of undulating to hilly land in the center of Michigan's primary dairy region. The general soil description is well to imperfectly drained, slightly acid to neutral, sandy loans to loam. One small town, Deckerville, with a population of about 22 23 800 is located in the southwest corner of the township. The total human papulation in the township is about 1550. At the time of this investigation there were two herds under quarantine due to M. 22113 infection. Herd histories indicated that tuberculin reactors had been removed from 39 of the herds in this township in the previous 12 year period. Gross lesions, characteristic of bovine tuberculosis, had been found in animals from 14 of these herds. Tuberculoid skin lesions were present in some herds, and some herds had animals with tuberculin responses not due to M. Egg-Lg infection. The proposed program was submitted to the State and Federal disease control officials and to the Director of the Michigan Depart- ment of Agriculture for approval. Upon receipt of this approval, the cooperation of State and Federal stockyard and meat inspection personnel was obtained for aid in post mortem examinations and collection of tissue samples. The hemagglutination test on all blood samples and bacteriologic and histologic examination of tissue samples were made at the Michigan State University Tuberculosis Research Laboratories. The local veterinary practitioners and the County Extension Director were enlisted to acquaint the herd owners with the project and obtain their cooperation. Dr. J. L. Isbister of the Michigan Department of Health1 con- ducted tests on the human population using a comparative test comparable to that used for the cattle testing. The results of these tests might reveal some relationship or lack of relationship between human and animal non-specific tuberculin hypersensitivity 1J. S. Isbister, M,D., Chief Cardio-Pulmonary Diseases Section 24 and would give the incidence, if any, of active human tuberculosis in the township. A double Mantoux test, incorporating P.P.D.-S and P.P.D.-B (purified protein derivative tuberculins, 8 made from M, tuberculosis and B from the "Battey" Mycobacterium) injected on opposite arms of each individual was used. The physician reading these tests recorded all changes and called any reaction over 2 mm in diameter a response. The success of any field test program depends upon the coopera- tion of all individuals in the subject area. To gain the acceptance of the herd owners that would be involved, a meeting was held in the Deckerville High School to which all township residents were invited. In addition, articles were released to local newspapers explaining the program and its possible long range benefits to all cattle owners. There were estimated to be about 100 cattle herds in this town- ship, largely dairy operations. The area was divided in half. Each half containing approximately equal numbers of herds was assigned to a veterinarian trained in epidemiology who performed all tests and investigations in his respective half. All the herds in each section of the township were tested before proceeding to the next section, to avoid introducing bias from early selection of the more easily in- vestigated herds or more cooperative farmers. A minimum of two visits to each farm was required. The first for injection of tuberculin and initiation of other investigations and the second to read the results of the test. All information was gathered during these two visits unless tuberculin hypersensitivity was found in the herd. 25 The first testing procedure was as follows: All cattle one day and older in age were inoculated intradermally in the caudal fold with 0.1 ml of A.R.S. standard mammalian tuberculin. ‘Multiple dose, cali- brated tuberculin syringes and 3/8 inch 26 gauge needles were used. Before this injection was made the animal's identification, ear tag number or tattoo was recorded and a 5 ml blood sample obtained from the adults. Three days later (72 hours) the animals were examined visually and by palpation for signs of tuberculin hypersensitivity. All changes from normal were recorded as a response. These responses were measured and recorded in millimeters across the greatest thickness with the Hauptner dermal thickness gauge. They were also classified by the A.R.S. method of P, hard and circumscribed, or X, soft and diffuse with no distinct demarcation.* All animals in the herd were visually examined and the legs of the adults palpated for signs of tuberculoid skin lesions. When found these were classified in one of three groups. These three groups were: I. Moderate to extensive lesions occurring in chains following the lymphatic drainage of the legs. 11. One or a few well described typical nodules in the same locations. 111. Not typical or questionable lesions occurring on the legs and those occurring high on the shoulder, flank or other body area. Due to the questionable nature of the latter, they were not used in compiling skin lesion data. All investigations were completed at this time for those herds in which no cattle with tuberculin response were found unless tests for other species were required. During these two visits the epidemiologist gathered and record- ed information about the animals with tuberculin hypersensitivity. Information about the herd was also gathered. General management, sanitation, herd health, feed practices, purchase and sales practices, use of communal pastures and possible associations with local past or present known M. bovis infected herds were noted. *See Appendix 1 26 Arrangements were made to test all chickens and hogs in lots of 40 or less and at least 101 of larger lots unless they were very large brooder type operations. Chickens were tested with 0.05 ml of A.R.S. standard avian tuberculin injected intradermally into the wattle. Responses at 48 hours were recorded. Hogs were inoculated with 0.1 m1 of both avian and mammalian tuberculins in the skin over either ear. When possible other livestock and pets were tested. Blood samples collected during the first visit and samples taken seven days after the caudal injection were tested by the hem- agglutination test for tuberculo-polysaccharide specific antibodies in the Michigan State University Tuberculosis Research Project laboratory. The second phase of work involved only those herds in which one or more of the animals had a caudal fold tuberculin hyper- sensitivity: One week after the caudal test injection of tuberculin all animals that demonstrated tuberculin hypersensitivity were again blood sampled. An equal number of non- responding animals from each of these herds were also bled. All caudal responding animals then received the cervical intradermal comparative test. Hair was clipped from two sites on the same side of the neck. These sites were located on a plane parallel to the shoulder, in the middle third of the neck. The upper site was placed 4 to 5 inches below the crest of the neck and the lower site about 5 inches below this. 27 A fold of skin was lifted in each site, measured and recorded in millimeters thickness with the Hauptner dermal thickness gauge. Avian tuberculin (0.1 ml) was always injected into the upper site and mammalian (0.1 ml) into the lower. Two days (48 hours) later these two sites were observed and palpated for signs of hypersensitivity. Responses at both sites were measured by lifting the response and determining in millimeters the greatest increase in skin thickness. The normal skin measurement at each site was then subtracted from its respective response measurement and the results for the two sites compared for a differential measurement. These were recorded as a mammalian response or avian response predominance by number of millimeters difference. Animals with a mammalian response 5 mm or larger than its avian were considered reactors if other circumstances indicated the possibility of M, bovis infection and sent to slaughter. Reactors were consigned to slaughter at Federally approved slaughter houses where a Meat Inspection Division vet- erinarian and an assigned stockyard veterinarian conducted a post mortem examination for the presence of lesions compatable with tuberculosis. Samples of all lesions were submitted to the M.S.U. tuberculosis research project. When no lesions were found, representative lymph node samples were submitted for complete histologic and bacteriologic examinations. Agricultural Research Service avian tuberculin was used as the antigen to compare with mammalian tuberculin primarily because of its availability. The comparative test is based on the premise that an infected animal will react more to its homologous antigen and to a lesser degree to other antigens. If the animal's hyper- sensitivity is due to some agent other than those from which the two antigens are made, then its degree of sensitivity to both should be nearly equal and/or minimal. Reactor, suspect, or negative status was assigned to the animals with caudal fold responses after the herds' history and caudal test results, individual animal's history, and results of the cervical comparative test were reviewed. Only the animals that the veterinarian considered likely to be infected with M, bovis could 28 be condemned in this project. To obtain slaughter data on the non- reacting cattle, the owners were given pre-addressed postal cards with which they were asked to report prior to the slaughter of animals from their herd. A reporting system already existed in that most of the animals culled from herds in this area were sent to slaughter houses with meat inspection services. These services are required to report the occurrence of all tuberculous lesions and submit samples for confirmation. As long as no reports are received, theoretically, no tuberculous animals have been slaughtered. This gives a check on most of the herds that periodically cull some animals. RESULTS AND DISCUSSION Veterinarians conducting intradermal tuberculin tests agree that most tuberculous and some non-tuberculous animals will react to tu- berculin. What constitutes a reaction is debatable. The size and consistency have been used as criteria in the past. However, some tuberculous animals have only a minimal or no response to the intra- dermal tuberculin test. Further, some non-tuberculous animals show more than minimal responses. Early in an eradication program it is the non-responding tuberculous animals that prolong and spread the disease. When a very low incidence of disease is reached it is the non-tuberculous animals that respond that show or defeat attempts of eradication by camouflaging responses due to bovine tuberculosis. It is not economically sound or feasible to remove a large number of non- specifically sensitive animals while searching for the bovine tuberculous animal or for a more ideal test for the presence of tuberculosis. Instead, present diagnostic methods and techniques should be further refined, or new tests devised, to give the veterinarian more aid in making a sound judgement. One factor hampering the veterinarians' use of many of the diagnos- tic methods is a lack of organized data from the use of the tests. Non- specific reaction rates for the caudal intradermal test are ill-defined because minimal responses are frequently not recorded. As the difference between a minimal and full response is a subjective decision, one frequently cannot be sure what is reported. 29 30 In this thesis to avoid much subjective bias, any deviation from' 'normal at the time of reading was considered a response. These were measured and recorded in millimeters skin thickness and were also recorded by the Agricultural Research Service method of X or P types*. CAUDAL INTRADERMAL TUBERCULIN TEST All cattle one day and older were tested in one township in Michigan. The area in which this township is located was not noted for being either especially high or low in non-specific responses. In the previous 12 years, 252 reactors had been removed from 41 herds in this township. Gross lesions had been found in 14 of these herds and two were still under quarantine at the start of this project. There were 92 herds containing 3,972 cattle; seventy-two were dairy herds, ten were breeding beef herds and ten were feeder herds. Cattle less than two years of age numbered 1,441, more than 1/3 of the total number. Those six months and under comprised about 16% of the total. The results of the caudal test on these animals are recorded in table 1. Over 8% of the cattle responded to some degree to the caudal fold test. Only 0.251 were under two years of age and none were less than eight months old. The response rate for those cattle over two years of age is similar for each age group. See figure 4, page 31, If only cattle two years and older were considered, as in the U. S. Uniform Method and Rules for area testing and the area test programs of some states, 12.81 responded which is more representative of the response rate likely to be encountered by veterinarians *Appendix 1 31 269 Table 1. Caudal told intradermal test results. Number of Number Number % Number of % Percent cattle of with of caudal of of caudal Total - 3972 herds caudal herd responses herd responses Adult- 2531 response totals P1, x2 or totals P1, x2 or larger larger From g. bovis infected . herds '3 . E 3 235 5 61 25.95 24 10.21 39.34 5'2 m o. 193 31.60 12.43 s m H o ‘5 F From non- 3 f, infected 3‘3 herds 0,“ 3241 63 249 7.68 87 2.68 34.93 m a 'o E a 2033 12.24 4.27 a o 55 Others u-a In 3 54 3 12 22.22 4 7.40 33.33 36 33.33 11.11 From caudal test negative herds 442 21 32 conducting routine tuberculin tests. On a herd basis, responses were observed in 77.01 of the herds. Over 87% of the dairy herds, 642 of the breeding beef herds, and 20% of the feeder herds had one or more animals with a caudal response. It is a tuberculin test rule of thumb that the large response is more likely to be due to !, bgglg, At one time in the nation- wide testing program all responses X2 or P1 and larger were to be considered reactors and slaughtered unless the possibility of g, bggig infection was highly improbably. In this township, 36% of all animals with caudal response were classified P1 or X2 and larger. Over 50% of the total herds in the township had one or more of these animals. They represented 3.31 of all cattle tested and 4.5% of those two years and older. The above figures included five herds that proved to be infected with y. 22:13, All five had past histories of reactors with gross lesions of tuberculosis and included the two under quarantine at the start of the program. Mycobacterium bovis was isolated and identified at the M.S.U. tuberculosis research project, from animals slaughtered as the result of this program testing. Table 1 shows the caudal test results for the 5. 22113 infected herds. and herds from which Li. _bgv_i_s_ was not isolated. A third group of three herds with responding animals was not included in either set of data as they had probably been recently exposed to g. M. Nearly 267. of the animals in the 21. Mg infected herds responded to the tuberculin test, while the average response rate in the non-infected response herds was 7.7%. In the former 33 herds 12.42 of the animals had responses classed Pl, X2 or larger. Only 4.3% had similar size responses in the non-infected responder herds. There is a highly significant difference in the caudal response rate between.!, 22:35 infected herds and non-infected herds, but no significant difference between the two in the percent of the caudal responses that are classified P1, X2 or larger in size. Appendix 2 contains X2 values for these tests. The ranges of each group overlap considerably so that classification by the caudal test alone is very difficult. The range for the caudal response rate in y} Eggig infected herds was from 14 to 361 of the animals in the herd. In the non- infected herds this was 1.4% to 54%. The ranges of P1, X2 or larger response rates for both groups were from about two to 19% of the herds. OTHER INVESTIGATIONS Other epidemiologic observations were made in addition to the cattle testing to detect possible causes of non-specific sensitization. There were 20 flocks of chickens, ranging in size from one to 7,800 birds associated by farm.with 20 of the cattle herds. All or a representative sample from each flock were tuberculin tested. An injection of 0.05 ml of A.R.S. standard avian tuberculin was made into the wattle of each bird and the test was read at 48 hours. There were three reactors found in one flock. These birds and all birds in other flocks with any questionable test results and those with a physical appearance suspicious of avian tuberculosis were slaughtered. Tissue samples were examined bacteriologically and histologically at the M.S.U. tuberculosis research project. No indication of avian tuberculosis was found in any bird or flock. 34 There.were only four pigs on two farms in this township. These were tested with both avian and mammalian tuberculins. The skin over the back of the ears was the testing site with 0.1 ml of mammalian tuberculin injected into one ear and 0.1 ml avian tuberculin into the opposite ear. All tests were negative. Other livestock and pets were tuberculin tested when convenient or when the results of the tests of the cattle were highly suspicious. These tests included horses, ponies, dogs, and cats. All tests were negative. No paratuberculosis has been reported from this area. There was no evidence of this disease being present in the cattle examined during the testing in this township. One animal with chronic diarrhea was slaughtered and examined for signs of paratuberculosis. No gross or histological pathology of paratuberculosis was found. A tuberculin test survey of the human population of this township was conducted by the Tuberculosis Control Division of the Michigan Department of Health. A comparative type test was used on 1553 indi- viduals. About 700 of these lived outside of the town of Deckerville. Each individual was injected with P.P.D.-S on one forearm and P.P.D.-B on the other forearm. A physician read the results of the test at 48 hours and recorded all responses observed at the injection sites. All changes from normal skin appearance over 2 mm in diameter were called responses. About 201 of those tested had a response at the P.P.D.-S site and about 171 at the P.P.D.-B site. A large number of those responding to one antigen also responded to the other. No evidence of active tuberculosis was found in the human population. 35 The preceding findings indicate that the non-specific hyper- sensitivity in cattle probably was not caused by u. avium g, par - tuberculosis, or g. tuberculosis. The responses to both antigens in individuals without other evidence of tuberculosis demonstrates a similar problem of hypersensitivity in man and other animals. SEVENTH DAY OBSERVATION OF CAUDAL FOLD TESTS Consideration was given to the possibility that a portion of the responses seen at 72 hours might be due to other than a delayed type hypersensitivity. Duboczy (1965) reports that "residual" non-specific early or local sensitization both of which peak early between 24-48 hours might extend through 72 hours. If this were to account for a large portion of the non-specific responses seen at 72 hours, a later reading time might be beneficial. In some areas of the country, human tuberculin tests are being examined at seven days in an attempt to avoid some of the non-specific response problems. A 96 hour interval to observation has been recommended by some as just as accurate as a 72 hour interval in finding E} 22312 infected animals without as high a non-specific response rate. Approximately half of the cattle with caudal responses were examined at seven days post injection. About 202 of the responses seen at 72 hours had disappeared by this reading time, almost all of which were minimal responses. It would appear that further delay of the time of observation (until the seventh day) would give little help in differentiating between large responses caused by E, 22211 infected and those resulting from other causes. TUBERCULOID SKIN LESIONS All animals were examined for the presence of tuberculoid skin lesions which were classified as follows: 1. Moderate to extensive lesions occurring in chains following lymphatic drainage of the legs. 11. One or a few well described typical nodules in the same locations. III. Not typical or questionable lesions occurring on the legs and those occurring high on the shoulder or flank or other body areas. This latter group, due to their more questionable nature were excluded from the skin lesion data. Skin lesions of Group I and II were found on 117 animals in 44 herds. There were nine animals in the third group. An average of 52 of the animals in the 44 herds were affected with a range of 0.4 to 33.3%. All of these herds had one or more animals with a caudal response, and a herd response rate average of 10% with a range of two to 53.52. See Table 2. Nearly 801 of the animals with skin lesions, from 901 of the skin lesion herds, elicited a caudal response. These animals rep- resented 291 of the total cattle with caudal fold response. No tuberculoid skin lesions were found on any animals less than two years of age. Nearly eight percent of the adult population in the affected herds had skin lesions. The percentage affected for each age group over two years of age were approximately equal. (Figure 5.) Approximately 14% of the cattle p0pulation were beef cattle. 36 37 No skin lesions were found on any beef breed animals. Personal com- munications with veterinarians in Michigan confirmed that skin lesions are either non-existent or are a rare occurrence in the beef breeds. It is interesting to speculate, if this is true, as to whether this is due to an environmental difference between beef and dairy types such as housing or feeding practices or to a genetic difference resulting in an increased resistance or susceptibility to the etiologic agent of skin lesions. It has been stated by some that skin lesions cause only a low level hypersensitivity to tuberculin. Over 40% of the skin lesion animals in this study had caudal responses that were classified Pl, X2 or larger. This represents a significantly higher rate of large responses in animals with a caudal fold test response and skin lesions than in animals with a caudal fold test response but with no skin lesions. Fincher (1959) and others have reported a lack of skin lesions in M, 22315 infected herds. It is quite likely in any infected herd that has been under tuberculin test for a period of time that better than 802 of the skin lesion animals will have been removed as reactors. There were 14 skin lesion animals found in four M5 bggig infected herds in this study. These represented over 7% of the adults and only 23% of those with caudal responses in these herds. This can be compared with 5% of adults and 39% of the caudal fold responding cattle affected with skin lesions in their respective herds. In this township, M, ngig infection in the herd did not preclude skin lesions from occurring on some animals in the herd or even reduce *The author has since observed typical skin lesion chains on two grade Angus cows in another part of the State. 38 n I mn.ao «m.be bu.>b «one» «o n mm.mm «m.vm an me up «a seas: I a va.« em.~ nn.m asuou no « av.me Ho.mm mm baa mam «ban easlan< oncogene was» seaweed «no» anon oaou mousse name must Houses one o» s mngsea ends agendas and essences a anon anon anew-ea name uncommon anon on. encased Houses one seat «assess eaou Houses same even unease” swan op ounce-oh no anoosom no «escape seas hoses: nuds.soassz a news soalsz uHIuOF .enOHuoH sane muonsouonsv uo\ons use» case Houses on» on essences s undead ouuuso .u cease 39 Figure 4. Marion township cattle with a caudal fold test response by age. 3° 60 50 4o .. o ‘” n c E o z 8 30 z o a. 20 IO 2 3 4 5 5 7 8 9 10 1f Age in years Figure 5. Incidence of tuberculoid skin lesions in Marion township cattle by age. 2 5 25 "\ I‘s ”I \\ *5 ,’ u o 215 l 152 a 33 ’ :3 I z 10 10 5 5 2 3 4 5 6 ’F s 9 10 Age in years Percent of animals. Number of animals. 40 the likelihood. Two of the 14 skin lesion animals were found to have gross lesions at slaughter and therefore an M, bogig infection in an animal does not preclude skin lesions. There is a highly significant relationship between the occurrence of skin lesions and caudal fold tuberculin hypersensitivity. Over one- third of the caudal test responding animals had clinical skin lesions. More than 80% of the samples of lymph nodes from N.G.L. reactors from Michigan, submitted to the tuberculosis research project at M.S.U. had microscOpic granulomas, the majority were from pre-scapular lymph nodes. The suggested relationship between the granulomatous lymph node and clinical skin lesions suggests that non-specific hypersensitivity in cattle may be due to a skin lesion syndrome of mycobacterial etiology which involves the lymphatic drainage of the legs primarily. COMPARATIVE TEST The cervical comparative intradermal test using standard A.R.S. mammalian and avian tuberculins was used to retest all of the 322' animals responding to the caudal test. Results for this test are listed in Table 3a and 3b. Nearly three-fourths (722) of the animals tested with the com- parative procedure had a response to mammalian tuberculin that was larger than the comparative avian tuberculin response. Twenty-six animals in twelve herds had a response at least 5 mm larger to the mammalian tuberculin than to the avian tuberculin. These are called comparative test reactors in the remainder of this thesis. Fifteen were from the five M. 2913;. infected herds. Of the latter group, one reactor was found in each of the two herds that had been under 41 quarantine and tested at the start of this program. In a third herd, the one comparative reactor was originally from one of the quarantined herds and has been the only gross lesion, M, ngig isolation case from this herd. The other 12 came from the two herds which had not been tuberculin tested during the previous two years. Three of the com- parative reactors from.M,‘Eg!is infected herds had minimal caudal responses. It is probable that the previous tuberculin tests in the quarantined herds had removed many of the animals that would have shown a positive comparative test. The comparative test would have great value in testing herds in which M. 2.915. is suspected and which have not been recently tested, but less relative value in retesting long quarantined herds which have had a number of tuberculin tests. The comparative test was also run on 45 caudal fold negative adults in three of the M, Egyig infected herds. One comparative reactor was found in one of the previously unquarantined herds. The animal had thoracic lesions from which M. 2932.2 was isolated. Gross lesions were found in 442 of the comparative test reactors in M, Mggig herds. M132- bacterium bovis was isolated from each. This gives much support to a claim that this testing method is more efficient than other testing or retesting methods in locating the tuberculous animal. Comparative test reactors while associated with M. _b_9_v_i_s_ infection, unfortunately, are not limited to M. M infections. Eleven other comparative reactors were found in seven other herds. One herd had three reactors, two others had two each, and the remainder, one each. There was no history or other evidence of bovine tuberculosis in any of these seven herds. No association between the reacting cattle or herds and other M, bovis infected animals or herds could be established. Table 3a. 42 Comparative cervical test results Number of Number Number of % Number of % animals in of comparative caudal test comparative herds reactors responding reactor 1 cattle in herds these herds M. bovis infected 235 5 16 6.8 61 25.96 Non- infected 332 7 11 3.3 31 9.33 Table 3b. Comparative cervical test results Caudal test Number of % of Number of comp. Comparative responding comparatfiye comp. reactors with H.A. reactors with cattle also reactors reactors titer increases H.A, titers comparative with skin 4 8 16+ 640 or test reactors lesions fold fold fold greater 15 3 20.0 3 4 3 5 ll 7 63.6 1 3 1 O 1. Comparative test reactor: An animal that had a comparative cervical test response 5mm or larger to mammalian tuberculin than to avian tuberculin. 43 Comparative cervical test results. Figure 6. e I 25. 4" ‘ l O O l '0 20 " I ' l a ,3 I , | j, , e \ 5 . .‘ ‘. 2 15 '- o | : | 8 o I 3. ,1 . \ a \ '0 \ I. “ 10 ,' e ‘\ . \ a “ e ' 0 I \ - I. ' \ . \ .F s 5 L X ‘s‘ +“ /’ e e \(x a». I a A a a l l# P a s ' . A: ' . 5+- 5 4 3 2 1 1 2 3 ‘ ' 5 5+ Mammalian predominance Avian predominance measured in millimeters — Animals from M. bovis infected herds. Animals from non-infected herds. 44 Eight of these reactors were slaughtered. No gross lesions were found and no M. by}; was isolated from tissue samples from these animals. This would not seem to support the work of McGavin £5 21 that a positive comparative test indicates progressive mycobacterial disease. Seven of the eleven reactors (64%) had clinical tuberculoid skin lesions. Three of the 16 comparative reactors (19%) in M, bggi§_herds also had skin lesions. Thus over 601 of the comparative reactors had visible lesions of mycobacterial disease. Ten had tuberculoid skin lesions and seven had thoracic lesions from which M. m was isolated. It could be, as reported by McGavin'g£.gl that a comparative reaction is specifically indicative of a progressive mycobacterial disease. As previously mentioned, the pathologists, Willigan, McGavin and Goyings, in the tu- berculosis research group at M.S.U. found that a high percent of N.G.L. reactors had microscopic granulomas. It is possible to speculate that a large proportion of the non-specific caudal fold sensitivities are due to some phase of atypical mycobacterial infection, then it is also possible to speculate that the comparative reactors could be animals with presently active or progressive mycobacterial disease. The research group found that animals with primary complex disease, i.e. granulomas in prescapular lymph nodes only, were not comparative test reactors. It may be that these lesions were in a static or regressive state and no comparative differentiation would be expected. Unfortunately, complete histologic and bacteriologic examinations were not possible on all responding animals in this project and the number with or without progressive mycobacterial disease is not known. 45 Figure 7. Cattle farm locations - Marion township. Herds with one or more animals with a caudal test response. Herds with one or more animals with a caudal test response and one or more animals with skin lesions. Herds infected with Mycobacterium bovis. A Caudal test negative herds. Herds with comparative test reactors 46 A total of 6.8% of the animals in M, bggig herds were comparative test reactors which represented over 262 of these herds' caudal respond- ing animals. Only 3.3% of the animals in non-infected, comparative test reactor herds had comparative reactions, but represented over 35% of the animals with caudal fold test responses in these herds. The actual caudal response rate of these two groups shows 261 responses with a range of 14 to 362 for M. Mg infected herds and 117. with a range of 5 to 19% for the non-infected, comparative reactor herds. Thus, there are significantly more caudal test responding animals in My 221;; in- fected herds than in non-infected herds which contain false comparative test reactors.* The comparative test method, like the caudal test, has a high degree of both Type I and Type II error when used on individual animals. These errors also occur on a herd basis but the test greatly reduces the number of herds which are considered suspicious by the caudal fold test without missing M. bovis infected herds. SEROLOGY Blood samples were obtained before and after tuberculin testing from 518 adult cattle. Approximately half of the animals had re- sponded to the caudal test and the other half were negative animals from the same herds. The first sample was collected prior to the caudal fold injection of tuberculin and the second seven days later. Serums from these samples were tested by the Michigan State University (MSU) Tuberculosis Research Group for tuberculopolysaccharide-specific antibodies by the modified passive hemagglutination test (Mallmann et 1 *Appendix 2 47 1964). The titers were recorded as the reciprocal of the highest dilution of the serum which yielded a positive test. The pre and post tuberculin injection titers for each animal were compared for an anamnestic-like response as indicated by a 4 fold or greater increase in titer. There were 109 tests from M. Magi; infected herds and 409 from the non-infected. Titers for 215 caudal test responders were also com- pared with 202 caudal test negative animals (Tables 4a and 4b). Preinjection titers for all animals averaged slightly more than the 1/40 dilution (Figure 8). Only 12.1% of the samples had no detect- able antibodies at the 1/20 dilution. There was no significant difference in preinjection titers among animals from M. 9232.2. infected herds and non-infected herds or from subsequent caudal test responding and caudal test negative animals. Animals with tuberculoid skin lesions had significantly higher pre and post tuberculin injection titers than did those without skin lesions. In all of these groups there was a post injection titer increase ranging from 1.8 to 5.8 times the preinjection titer level (Thble 4a). Post injection titers are compared for M. 9.22.1.9. infected herds and non- infected herds, and for animals with and without caudal responses (Figures 9, 10). The mean titer increase for animals from M. Magi; infected herds with caudal responses was nearly six fold. This com- pared with a mean increase of 2.4 fold for animals from non-infected herds with caudal responses. Statistically this increase is significant only to 101. All five M. _b_o_v_i_§_ infected herds and seven non-infected herds had comparative test reactors. Titers from these two groups were compared and the results were approximately the same for both. 48 Table 43. Mean increase of post tuberculin injection hemagglutination titers. Number ' Mean increase of 1 of samples post injection titers All samples 518 2.5 Caudal test 267 2.9 responding animals Caudal test negative 251 2.0 M, bovis infected 101 4.4 herds Caudal test responding animals 52 5.7 Caudal test negative 49 2.6 Non-infected herds 417 2.2 Caudal test 215 2.4 responding animals Caudal test negative 202 1.8 Animals with skin lesions 82 3.1 Animals with no skin 436 2.4 lesions 1. Increase in multiples of the mean pre tuberculin injection titer. Table 4b. Hemagglutination (H with comparative test reactors Number of Increases in H.A. 49 1A.) titers from herds titer after Number with a post animals tuberculin test tuberculin titer of 640 or more 4 8 16+ fold fold fold M. bovis infected herds 52 14 10 5 6 Z 26.92 19.23 9.61 11.53 Non-infected herds 26 6 3 4 3 2 23.07 11.53 15.38 11.53 1. H.A. titer: reciprocal of highest serum dilution in which tuberculo polysaccharide specific antibodies were detected. Comparative reaction: Cattle which had a cervical 5 mm response of or greater to mammalian tuberculin than to avian tuberculin. 2. Animals had responded to tuberculin injected into the caudal fold and the pre and post tuberculin H.A. titers determined. 50 Nearly 30% of the animals bled had titer increases of four fold or greater (Figure 11). There were a significantly greater number with a four fold or greater increase in the M.'§ggig infected herds. Com- paring the number of titer increases that were 16 fold or greater, however, shows no significant differences in high titers between M. M infected and non-infected herds. Due to the lack of a clear cut separation, a four fold anamnestic-like increase in titer has little value as a differential test under field conditions. However, 6 of the 21 animals with 16 fold or greater titer increases were from M. M infected herds and 4 of these were found to have gross lesions at slaughter. A number of non-infected animals will have these high titers, but their occurrence in a herd should alert the veterinarian that there is a greater possibility of M. m infection. Table 4a shows that the mean increase of post tuberculin injection titers of animals with skin lesions falls near the middle of the range of the other means listed. Appendix 2, however, shows it to be a significant difference. As the preinjection titers also demonstrate a significant difference, we conclude that skin lesion disease itself is responsible to some degree for the higher initial H.A.titers but that the anamnestic increase is actually no greater than it is for the other samples. Only 9.5% of animals with very high post injection titers had skin lesions. This compares with 7.82 skin lesion incidence rate in the total population. The lack of a high titer increase would seem to indicate that skin lesion disease is not responsible for the very high levels of circulating, tuberculopolysaccharide-specific antibodies. A decrease in titers between the first and second samples was observed in 5.202 of all samples. In M. bovis infected herds only 1.9% 51 of the samples decreased, while in non-infected herds, 61 decreased. Investigation of non-infected comparative reactor herds showed 6.52 of their samples decreased in titer. Statistical analyses indicated there was no significant difference in these reversal rates between caudal test responding animals and caudal test negative animals or between M. 13.9119. infected and non-infected herds. Five of 16 animals with post injection sample titers of 640 or higher, were also comparative test reactors. Their caudal test responses were all x2, P1 or larger in size. These five were all found to have gross lesions at slaughter, and M, Eggig_was isolated from each. Non- infected animals while meeting one or two of these criteria failed to meet all three. Further investigation of this facet may indicate that the combined use of the three tests will effectively increase the detect- ion of M, bovis infection with less time, effort, and error. Number of samples 150 100 ”P:- e-tuberculin ant—.— P03 _f I "20 20 80 160 Reciprocal of injection HA titers t-tuberculin injection HA titers 320 titer 53 Figure 9. Post tuberculin injection hemagglutination titers. Animals from Mycobacterium bovis infected and non-infected herds. I I — 30 120- I . a '| I h E m- : g d .c: a I '8 > o I 3 .0 I g . H ml I I: 3 -20 I 8 so- I-I I In E ' 3 m I g ‘6 .' . I. I "‘ _ 2 t . ° 5 I =3 z I ‘5’ I , z I I -10 I 40— ’I I I / / \ \ l l l I l l l L- - "T -20 20 40 80 160 320 640 1280 1280 #- Reciprocal of titer — Animals from M.bovis infected herds. ---- Animals from non-infected herds. Figure 10. Post tuberculin injection hemagglutination titers. Animals with caudal test responses and caudal test negative animals. 100 - I Number of animals a I 0' o» I l a 1 A I l l 1 l -20 20 4O 80 160 320 640 1280 1280+- Reciprocal of titer Caudal test positive. _ Caudal test negative. 55 Figure 11. Post tuberculin injection hemagglutination titer increases. 160- 120- 80- animals Number of 40- I I I I I I I I 0 2 4 8 10 32 64 128 Titer increase In multiples of original titer CONCLUSIONS The caudal fold test of M, Eggig infected herds in this township was found to miss over 36% of the animals with tuberculous lesions (Type II error). Not all of these lesions have been proved by lab- oratory work to be due to M, ngig, so this figure may be high. Other workers have reported as much as a 19% Type 11 error (Robertson 1936b). There were over 8% of non-specific positive tests (Type I error) which is approximately in the middle of a range from other reports of Type I error. It is obvious that such a test is ineffective as an individual animal test in detecting less than 0.01% infection in a population of 100 million. None of the tests for tuberculosis presently available, when used singly, are any more reliable. Further, it is unlikely that any ante-mortem test will be devised that is totally reliable. However, the effects of both types of error can be minimized by considering M, ngig infection on a herd wide basis. A high percentage of animals responding in a herd tested by the caudal test indicates possible M, Mggig infection. There is a portion of non-specifically infected herds that will also show similar results to this test. Therefore, an adequate history of the herd and individual animal is invaluable, although not infallible, in differentiating specific from non-specific infection. A veterinarian finding caudal fold suspicious results while test- ing a herd has at present four alternatives: 1. He can decide that there is M, 223;; infection in the herd and remove all responding animals. This will probably remove a large percent of the productive animals. The owners economic loss is great, and perhaps needless if the infection is non-specific. 56 57 2. Some of the responding animals can be slaughtered. If they have lesions of tuberculosis, the other animals can be removed. However, an infected herd can be sampled many times without disclosure of lesions at slaughter. 3. The herd can be retested in 60-90 days using again the caudal fold intradermal test. The results of such a retest will be no more definitive than the original test. 4. The responses can be considered to be non-specific due to the large number of skin lesions and the herd called negative. The danger here, of course, is the possibility of leaving an animal infected with M, Mgzi§_which can be spread from the animal or herd to Others. Supplemental tests, when their limitations are recognized, can be used with a higher degree of confidence to help differentiate those herds with suspicious caudal fold test results. Individually, the two supplemental tests used in this project gave overall results similar to those of the initial caudal test. Each test had some degree of Type 11 error but the errors did not entirely overlap. Thus, the use of the combined results from the three tests provided a more specific diagnosis than possible with any of the tests individually. The comparative test had a high degree of both Type I and Type 11 error but as a follow-up to the caudal fold test, reduced the Type I error appreciably. The sizes of comparative test responses cannot be used as a single determinant. However, the occurrence of a number of mammalian responses in a herd approaching or exceeding the 5 mm measurement indicates that an M, bovis infection more probably exists. 58 A four fold increase in hemagglutination titer was not found to be specific for M, 22313 infection in individual animals. There was a significant four fold rise in titer seven days after the injection of tuberculin overall samples collected from animals in M, Mgzis infected herds. There was, however, no significant difference between M, Mggig infected herds and others in the occurrence of post tuberculin titers 16 fold or greater. A specific rise in titer cannot be used as a determinant. As the comparative test, this test has some degree of Type I error, yet is helpful in supplementing the results of the caudal fold test. To illustrate the value of these two supplemental tests, the testing results from two herds in this project are presented in Table 5. The caudal fold testing of all animals in each herd revealed responses in 33% of herd A, and 53% of herd B's animals. If the caudal test had been conducted as in the routine screening method only on animals two years and older, then responses would have been found on 48% of the animals tested in herd A, and 65% of those in herd B. Further, 24% of the adults in both herds had responses classified Pl, X2 or larger. Examination for tuberculoid skin lesions revealed 17% of herd A and 38% of herd B had clinical skin lesions. Reviewing the history of each herd revealed that herd A had extensive tuberculosis infection several years before and had been released from quarantine two years before this test. Herd B had two negative tuberculin tests in the preceding six years and had no indication of M, ngig infection at any time in its history. The veterinarian conducting these tests should have little trouble in concluding that herd A was probably infected with M, bovis. The tuberculin test results indicated that 59 mm.om o o m.m~ m.mm a R mm o o m mm me m v.o~ N.e o.eH o.v~ m.mm m n N h n pm we < uneamem mused» somsea equ eouuecausmwmeeom mucuoeoa so «x .m& uncommon new? use: newsman no pace amp» uncommon Hmpseo use: as senssz w some soaesz o>dueseasoo Hsoseo some somssz ea nomssz use: ew>on Esuneuomnoohx eco Seem muasmeu wswumeu we somwmmaeou .vue; pouoewcuIcom one one neuuomow .m edema 60 M, ngig infection is present in Herd B also. However, the history and presence of a large number of skin lesions indicate that more probably a non-specific sensitivity exists. This is substantiated by the results of cervical comparative tests of animals in herds A and B responding to the caudal fold test. There were six reactors in Herd A and none in Herd B. All adults in both herds were tested with the comparative procedure. Over one-third of the adults in Herd A had mammalian responses 4 mm or greater than the avian response, while no animal in Herd B had more than 3.3 mm difference. Blood samples were collected from all adults in both herds and the serums tested by the hemagglutination test. In Herd A, 23% of the samples from animals responding to the caudal fold test had anamnestic- like titer increases of 8 fold or greater. Only 5% of the corresponding animals in Herd B were this high. More significantly, three animals of Herd A and none in Herd B had post tuberculin injection titers of 640 or greater. These three in Herd A each had comparative test results with a mammalian response 4 mm or more larger than the avian response and all three had gross lesions of tuberculosis from which M, Eggig was isolated. Of two caudal fold negative animals in Herd A, later demonstrated to have lesions of tuberculosis, one had a positive comparative test and one had a 16 fold titer increase to 640. The results were not this clear cut in all instances. Nearly half (11 of 26) of the caudal fold test responding animals that were found to be comparative test reactors came from herds not infected with M, bgzig. Comparison of the combined results for each of the testing methods did reveal a significant difference in caudal fold test responses and 61 post tuberculin H.A. titers of 4 fold or more between M, 22215 infected herds and non-infected herds that had comparative test reactors. Therefore, it can be seen that the combination of tests gives more accurate information about the herd than any of the methods singly. The most definitive combination was a caudal response P1, X2 or larger, a comparative test mammalian response 5 mm or more larger than the avian, and a seven day post tuberculin H.A. titer of 640 or greater. There was no Type I error for M, 22213.1“ this small sample. It did not detect all those in which tuberculous lesions were found or from which M. SELLS. was isolated. Its major value is in identifying the most likely tuberculous animals in a suspicious herd. Table 6 outlines a suggested procedure for determining the presence or absence of M, Eggig infection.within an established herd of cattle. Some modifications and other supplemental tests may further increase the accuracy of the procedure. In addition to the supplemental test procedure, the epidemiologic observations and interpretations of the veterinarian conducting the test must play a major role in the final decision. Once it is established that a herd is infected with M, Mggig and the responsing animals removed, the effectiveness of epidemiological investigations is reduced. Subsequent efforts to remove all animals infected with M, ngig from the herd involves repeated testing and removal of new responders. State laws and regulations generally specify a given number of negative herd tests as the basis on which the herd can be released from quarantine. The knowledge that TypeII error occurs and may extend through several tests, and that such herds may be released from quarantine demonstrates a potential threat to the cattle industry. Table 6. Suggested epidemiologic procedure for detecting Mycobocterium bovis infection CATTLE POPULATION CAUDAL FOLD TEST BY HERD MINIMAL SIGNIFICANT NO RESPONSES RESPONSES RESPONSES Fewer than average More than expected number of minimal responses. large reSponses or a high percent Previous history of of total reSponses. Inadequate no M. bovis infec- history or history of _M_. Bovis tion. ¢ infection. RELEASE RELEASE Herd and animal history. Examination for skin lesions. Comparative cervical testing. Post tuberculin hemegglutinins Possible occurrence of other mycobacterial diseases on form. DIAGNOSIS DIAGNOSIS DIAGNOSIS NON DECISIVE _M_. BOVIS INFECTION NO _M_. BOVIS INFECTION Further Slaughter reactors for confir- supplemental """"" motion of diagnosis. testing. RELEASE ELIMINATE HERD RELEASE 62 63 Once a herd infection with y, 22! g has been established, it is more economical both to the herd owner and to the testing agency to eliminate the herd at that time. If the herd is allowed to continue, reactors are replaced with new purchases. Some may become infected, are infected, or may later respond non-specifically further compound- ing the problem. The savings achieved by the immediate elimination of the herd are recognized in time and further testing expense both to herd owner and veterinarian, and in good will between the testing agency and the cattle industry. MODIFICATIONS The described procedure has value and should be used; however, further studies should be made. Berman, 25 31 (1959) showed in a similar project in Wisconsin, that substitution of Weybridge P.P.D. tuberculins for the ARS tubercu- lins resulted in a decrease in the non-specific reactor rate without missing more !, 22312 infected animals. Substitution of P.P.D. tuberculins for the ARS tuberculins in the comparative testing of herds with suspicious caudal fold test results should further reduce the Type II error of the caudal fold test and add to the effectiveness of this testing procedure. Davenport (1963) demonstrated a closer antigenic relationship between skin lesion isolants and some antigens from other than those from _M_. _b_9_\_r_i_s_, L1. tuberculosis, or 13. Elli—um. Skin lesions may be responsible for a large percent of the non-specific hypersensitivity. Substitution of antigens obtained from skin lesion isolants for the avian tuberculin may reduce the Type II error of the comparative test. 64 The English use the comparative test but read the tests at 72 and 96 hours. It has been demonstrated that delayed hypersensitivity in some E, bggig infected animals does not arise until 72 hours or later and persists for some period of time. Some non-specific responses are due to an immediate type hypersensitivity that may still be evident through 48 hours. Therefore, a later reading time for the comparative and perhaps even the caudal test should be investigated. More work is required to determine the interval after the injection of tuberculin at which the maximum hemagglutination titer increase occurs and whether or not the cervical intradermal test a few days after the caudal test has any further effect on the tubercu- lopolysaccharide specific antibody levels. The results of this study indicate that the caudal test may be effectively supplemented with the comparative test and the hem- agglutination test; however, further studies are needed. SUMMARY All cattle in one Michigan dairy farming township were tubercu- lin tested using the method required by the State-Federal bovine tuberculosis eradication program. A large number of the responses to this test could not be attributed to infection with g, bggig. Two supplemental testing procedures were examined both as an indi- vidual test and as herd test procedures. Combined results of caudal tests, comparative tests and hemagglutination tests were examined to determine if the combined results would differentiate caudal test hypersensitivity due to yycobacterium bovis infection from hyper- sensitivity due to other undetermined causes (non-specific sensitivity). All tests and procedures were used in an actual trial situation in one township. The non-specific response rate for the caudal tuberculin test was approximately 7% of the township's cattle population or 11.2% of cattle two years or over in age. Over one-third of the caudal responses were more than minimal size (classified as Pl, X2 or larger). All animals were examined for tuberculoid skin lesions. Signs of skin lesions were detected in 117 animals of which 93 were hypersensi- tive to tuberculin. There was a significant relationship between the occurrence of skin lesions and tuberculin hypersensitivity. All caudal test responding animals were retested one week after the caudal test injection with a comparative cervical test using mammalian and avian tuberculins. The animals on.which the response was at least 5 millimeters larger to the mammalian tuberculin than to the avian tuberculin were slaughtered unless the epidemiologic investigation had indicated that most probably no g, bgyig infection existed. There were 26 (8%) of the caudal test responding animals 65 66 -that were comparative test reactors. The 26 animals came from 12 herds, five of which.were subsequently demonstrated to be infected with !, bovis. Fifteen of the 26 animals were from the five infected herds and half of these were found to have lesions of tuberculosis when slaughtered. The comparative test can be used after the caudal test to reduce its Type I error (indication that y, Egyig infection exists when it does not). The Type II (indication that E} 22333 infection does not exist when in actuality it does), and I errors of the comparative test are too high to use as an initial individual animal test. However, a number of com- parative test reactors within a herd is strongly indicative of specific infection. Blood samples were collected from all animals prior to the caudal test. One week after the injection of tuberculin blood samples were collected from all animals responding to the caudal test and an equal number of caudal test negative animals in the same herds. The serums were tested by the passive hemagglutination test for tuberculopoly- saccharide specific antibodies. There was no significant difference in pre-tuberculin injection titers between animals from.§. ngis and non-infected herds or caudal test responding and caudal test negative animals. Animals with tuberculoid skin lesions had significantly higher titers than those without. Serums from animals in each of these groups had an increase in post injection titer levels. Significant differences were found only between animals with skin lesions and those without. A combination of the three tests reduced both types of error. It is believed that fewer infected animals were missed and fewer non-infected 67 animals were erroneously condemned using this procedure than would have occurred using routine tuberculin test procedures. The results of the testing in this township indicate that a combination of certain results can identify the infected herd and some diseased animals. In six instances, g, bgzig diseased animals were identified by a P1, X2 or larger caudal test, a comparative test mammalian response 4 millimeters larger than their avian response, and a post injection H.A. titer of 640 or greater. In the small sample tested, the combined tests failed to detect some diseased animals but identified all of the infected herds. It also falsely identified six non-infected herds as being infected with g, bggig. Despite this latter error, this testing procedure reduces greatly the number of epidemiologic investigations of herds that the veterinarian would be required to make, reduces the number of return trips for retesting the herd, and shortens considerably the time required to establish the presence or absence of infection in a herd. The results of this study emphasize the characteristic short- comings of all the present tuberculin tests. In view of the multi- tude of these shortcomings, and as the number of tuberculosis infected herds in the U.S.A. are quite low, it is recommended that once infection has been established within a herd that the entire herd be eliminated. Further investigative procedures could be more efficiently concentrated on finding the source of infection and possible routes of dissemination from the infected herd. 68 Appendix 1 USDA-ARS-ANHW Method of classifying responses to the caudal fold intradermal tuberculin test. Determined by palpation. Circumsgribed Swelling P = Pea size 3/16" diameter. P2 3 2 times pea size. p3 a 3 times pea size, etc. pp a pea type less than 3/16 dia. a ., . ' (($33: f: T}; :2“ Diffuse Swelling x a thickness or caudal fold. X2 3 2 times thickness. x3 - 3 times thickness, etc. Caudal fold Ventral side of tail 69 2 Appendix 2. X Analysis of data Surveys cannot be regarded as satisfactory substitution for experiment; nevertheless. they are of value in situations inwwhich experiments are difficult or impossible. though in such cases all conclusions must be tentative. They are also of value as.a preliminary to experimental work, since they frequently indicate the factors that are likely to be most worth investigating. Yates (1934) X2 analysis of the relationship between HA titers from animals in M. ,bovis infected herds and those in non-infected herds. Differences-in the number of classes is due to grouping to obtain sufficiently large expected values. Pre injection Post injection Difference between pre G HA titers HA titers post injection titers All x2 8.78 x2 10.47 x2 198.22 Samples N 5.518 N 518 N 1036 df 4 df 6 df 8 Sig. N.S. Sig. N.S. Sig ** S 2 2 2 amples from X 3.66 x 8.93 x 112.48 animals with N 267 N 267 N 534 caudal fold test df 4 df 6 df 7 response Sig. N.S. Sig. N.S. Sig. ** S 2 2 2 amples from X 8.01 x 4.91 X 65.18 animals with no N 251 N 251 N 502 caudal fold test df 4 df 5 df 6 response Sig. N.S. Sig. N.S. Sig. ** .2 X analysis of the relationship between HA titers lesions and animals without skin lesions. from animals with skin .02 X N df Size 12.43 82 I. H x2 21.54 N 82 df 6 Sig. ** x N df Sig. 44.08 164 5 es 70 Appendix 2 (continued) 1:2 ANALYSIS as THE asunmsmr BENEEN _11. bovis infected herds Animals with M. bovis infected Non-infected com- sk_i_1;_l_gg_i_g_ns One degree M parative reactor Animals with of freedom Non-infected herd: herds no skin lesions Number of x? 86.12 x 26.65 x? 815.12 caudal N 3550 N 557 N 3972 responses Sig. ** Sig. ** Sig. ** Number of x? 0.41 x2 0.13 x2 3.97 caudal N 310 N 92 N 322 responses Sig. N.S. Sig. N.S. Sig. * Pl, x2 and larger Number of x? 27.70 x? 3.37 x? 1.81 comparative N 322\ N 557 N 380 reactors Sig. ** Sig. N.S. Sig. N.S. Number of NA titer x2 4.13 x2 7.32 x2 2.29 increases N 518 N 147 N 518 of 4 fold 'Sig. * Sig. ** Sig. N.S or greater Number of NA x2 1.16 x? 2.27 titer increase N 518 N 322 16 fold or Sig. N.S. Sig. N.S. greater ' Number with X? 5.49 x2 0.75 skin N 3530 N 557 lesions Sig. ** Sig. N.S. Number of NA X2 2.65 X? 0.03 titers that N 518 N 518 decreased Sig. N.S. Sig. N.S. * .05 ** .02 71 QH>O£ .2 ocoAuermA 02 20.82H 91 he 14 oA>om mummm eA>om mammm Iflifllflifl enoooun ca HHH aaouo chAueroA oz sex so qoz mo HHH useeeoue sea 92H ownuAsu anon oA mocAeuao uoz eowwuuaaaueuum ounuAnu m H hAno .oueA2 hAao .ouewm oooz 2 0:02 0x5 2 0 < oueAd no m< A moueroH one: mauanewuo seas: Baum nonmeAa €000 HOB H< HQZIOHOOflx A2212u02 eewoz :20-..— uehu coda weaeuxuaeu soon no: m1 222 come no: m¢aA2 A22 soon you ma~-a .m owonnu u .ueoe-H H H H 29024 oAnaom 02 oAnaem o2 oAnaem o2 oAnaem o2 oAmaem o2 oAaaem 02 no .Aooa n-< 20:4 AenomveuH aon< Azneaaeaeuoneva mom< oAnsem 02 mammom< mmmnom< woeeoooum one-ya #01moA consoA «ormoA «cuNoA «onAoA concoA scram «oun¢ «once «cine ee-~e AmeNNV Aheaev Assume AemeNV AeennNV AmmomV ne-eus ne-nes mouAm nounm «ouamw Aeaoev ee-~s «ouAm eeaueuauaueuen Aeawn< an am an 0A «A A 0°: when euAsnom muoueuoneA one aeuuo2 noon 1 n savannae 72 nosono mAmem Horou0tvuae2 coma» uonnvoxAE amH.eoauquua oz :Sz .5 eosuugeua oz.guz_mo ama.euueruH 2.32 mm 0.263.: oAnenonn .20 SH re Strum emu A: .8 $23.2 yen-.aoua .m.< so amm eoaueeaeaueuun an oao2 0:02 20< 02< ue>AA1H A 0:02 hAco.oDeA2 oao2 oooz oco2 eevonun ocoz eaves-H voueAOeH one: ounuAno naeAaewuo nuAnz scum ensueAH AH ca 056 on you men use enereA oAnsem o2 oAmaem o2 oAnaew o2 oAnaem o2 oAnaem oz 2 SEES: A2..neueeun A22 neon A0 AenAumeAvo2 2222 A221movo2 A212 .Ezbeoz .ELSS A2 A22 A22 Nwvonuem AekovoueAm A0 A02 A02 hAno eoeroA nAxm hAao enoueeA oAxm A02 A02 A02 A212 A01H A02 A02 A02 A02 umou A0 neoeoum A22 neon v02_a < A02 A02 A0 Auzuooooz A02- novoz A010 A0 Auzwm Act-evo2 A02 as: Numaeeuum enouu 0 oAasem o2 eAanem o2 eAnflem 02 oAmaem 62 oAmaem 02 hAno oueAsammnumd 209024 uo>AA-A .a neoconnra mmmom< eeouene A mmnon< manom< ne>AAnHmmaom< ovonuH omonrn eomocuH nova: .0 ue>AA1A .eovon eowon.ouednon eev©2 eovoa . u veeeouonm one-AH ee-en Amounv Aeoaev “Afloev Aoeeuv Aesoov «ormAN nouwmA normMA nonomA noomNA nonoA mounA «cram «010% ecumAA «enuAA «ouAAA AcnoAA «0100A «ouooA (vaneu ee-emw eeaueuauaueuen AeEAn< mm mm on No ¢¢ 0m .02 when venoAunoo .n auvnomm< 73 92H uw>on .2 uAnanoum A0um0 sun 3332 3 £380 .00 mu .52 .8 AHA H0 0A>On.2 0002a u02 coqwmuamfiuaomm uuuouquooha AHA anon» 002000 0AA 000A90 luau-mac unam 0A0< now-0A ounouuouuw: aoAuuA canoououowx 02 2092< 04 0 A 2 vouuAmaoo 000 unauAnu mung ~00 A< vouquou uuum «unuAno naoAauwuo noAnz aoum owauuAH A22 At: 0000 A2n< A2 A22 A02 A22 Nquonu-M A00A00A000A2 50:3 HHH A02 Ag unau A0n< A02 A0 2 No A02 A0 302 0 0 hwvonunm 000u0 aowuuA n.0u0 u A0 0033 000.5 02 .. A02 wcwmhu wow 5030“» uaoauAumanan u 92H 36 - H0300 no naouqu 0030032 summon 000A» humus-x 063A n Anton 0am and! .00m Eon van 0 2002 owe-non a uuuuuao-u: uuuuuonh #002 a 0002 u A¢0A>uoo II 2020-02 I AAuH no-oo~ unnon< 00-non 3:38. animusm 0500.3. Alla: “ 88 mm 00 N0 .02 than 0000A0000 .n xAvaoam¢ LITERATURE REVIEW Alshabphoun, A., Chapman, P. T., White, M. P. and Degroat, A. 1960. A study of the double-diffusion gel precipitation test in tuber- culous patients. Am. Rev. Tuberc. Pulmonary Dis. 81:704. Berman, D. T., Tervola, C. A., and Erdmann, A. A. 1959. Preliminary report of investigations of tuberculin sensitivity in Wisconsin cattle. Proc. U. S. Livestock San. Ass'n., 63rd Ann. Meeting, p. 197-204. Bodie, G. P. 1962. Diagnostic methods in veterinary medicine. 5th ed., p. 289. J. B. Lippincott Co., Philadelphia. Chaloux, P. 1964. Editorial. Tuberculosis Review of the USDA, ARS, An. Disease Erad. Div. '1: p. 8. Crawford, A. B. 1936. Tuberculin reactions in cattle showing no visible tuberculosis lesions on post mortem. J. Am. Vet. Med. Assoc. ‘82: 562-583. Davenport, C. A. 1963. The precipitinogen relationships among myco- bacteria of bovine, porcine, human, and soil origins. Ph.d. Thesis, Michigan State University, East Lansing. Duboczy, Ela. 1965. Nonspecific early type of tuberculin reaction. Am. Rev. Resp. Dis. 22: 1 55-63. Edwards, P. Q. & Edwards, L. B. 1960. Story of the tuberculin test. Am. Rev. Resp. Dis. §lfi1): 1-47. Fincher, M. G. 1959. Proc. Tuberculosis Eradication Conference, Kansas State Univ. 142-144. Franchis, John. 1958. Tuberculosis in animals and man. lst ed., Cassell & Co. Ltd., London. Gregory, T. S. 1949. The accuracy of diagnostic methods used in the detection of tuberculous cattle. Aust. Vet. J. 25(7): 138-152. Grey, D. P. 1953. The significance of hemagglutination titers in bovine tuberculosis. Aust. Vet. J. 22; 293. Hagan, W. A. 1931. The no-lesion case problem in the tuberculosis eradication campaign. Corn. Vet. gl;163. Hagan, W. A., Peldman, W., Traum, J. Schoening, H. W., Night, A. 3., Hastings, E. C. 1945. Proc. U. S. Livestock San. Ass'n. 49th Ann. aceting. Report 1944 of committee on tuberculosis. p. 142-144. 74 75 Hastings, E. C., Wisnicky, W., Beach, B. A., & McCarter, J. 1933. A detailed study of no-lesion tuberculin reacting cattle. J. Am. Vet. Med. Ass'n. 8;: 565. Hastings, E. G. 1914. What has been done with the tuberculin test in Wisconsin. Bull. Univ. Wisconsin Agr. Exper. Station. No. 243. Hastings, E. C., Beach, B. A., and Weber, C. W. 1924. No-lesion and skin-lesion tuberculin reacting cattle. J. Am. Vet. Med. Assoc. 96:36. Hebert, C. N. & Paterson, A. B. 1955. Tuberculin reactivity in a sample of attested herds. Vet. Rec. 67:1143. Henley, R. R. 1951. History of the bureau's work with tuberculin. Proc. 55th Ann. Meeting. p. 273. Johnson, R. B. 1945. Studies on the complement-fixation reaction as applied to bovine tuberculosis. Corn. Vet. 35: 52-60. Karlson, A. G. 1959. Reciprocal or cross sensitivity reactions to tuberculin in cattle. Tuberculosis Brad. Conference, Kansas State University. 48-60. Kerr, W. R., Lamont, H. G. & MeGirr, J. L. 1949. Further studies on tuberculin sensitivity in the bovine. Vet. Rec., 61: 466. Larson, A. B., Ropecky, K. E. Sept. 1964. New bovine tuberculosis test. Ag. Res. p. 14. Mallmann, V. B., Robinson, P., McGavin,'M. D. 1963a. An anamnestic like reaction elicited by tuberculin. Am. J. Vet. Res. g; (106): 693-698. Mallmann, V. H., Mallmann, W. L., & Robinson, P. (1964) Relation- ship of atypical bovine and porcine mycobacteria to those of human origin. Health Laboratory Science, 1(1): 11-20. Mallmann, W. L., Mallmann. V. H., Ray, J. A., McGavin, M. D. Ellis, D. L. 1963b. Infectivity of Atypical Group III mycobacteria of NGL cattle, Swine 6 Soil origin. Proc. Am. Vet. Med. Aas'n. 129: 265-267 McGavin, M. D., ggflgl. 1964. Experimental inoculation of calves with unclassified mycobacteria by the intradermal route. Final report of the Michigan State University bovine tuberculosis research contract with the USDA. 9-46. McLaughlin, A. R. 1962. A study of tuberculin sensitivity in Wis- consin cattle. 'M. S. Thesis, Univ. Wisconsin, Madison. 76 Middlebrook, G., & Dubos, R. J. 1948. Specific serum agglutination of erythrocytes sensitized with extracts of tubercule bacilli. J. Exptl. Med. §§_: 521. Moorehouse, L. G.,‘g£.'al 1963. Examination of bovine specimens suspected of being tuberculous, Proc. U. S. Livestock San. Ass'n. 67th Ann. Meeting. pp. 430-437. Munce, T. E. 1920. Tuberculosis eradication in Pennsylvania. J. Am. Vet. Med. Ass'n. 2; 14. Parlett, R. C., Youmans, G. P. 1959. An evaluation of the specifi- city and sensitivity of a gel double-diffusion test for tuber- culosis. ART Res. 82: 153-166. Paterson, A. B. 1956. The incidence and causes of tuberculin reactions in non-tuberculous cattle. Adv. Tuberc. Res. 1: 101-129. Plum, N. 1937. Studies on tuberculosis and tuberculin tests. Skand. Vet. Tidskr. 21; 589. Ranney, A. P., Chaloux, P. A., & Schwindaman, D. 1964. Comparative field studies using dilute and standard tuberculins. Tubercul- osis Review 1; p. 1. Ranney, A. P. 1966. Personal communication. Rheins, M. S. and Thurston, J. R. 1955. Additional tuberculous anti- bodies demonstrated with erythrocytes sensitised by successive adsorptions of OT and further studies on the adsorptive capacities of tanned cells. Am. Rev. tuberculosis. 72: 210-217. Ritchie, J. H. 1959. Tuberculosis eradication, p. 713-739. In Stableforth & Galloway (ed.), 1959. Diseases due to bacyeria. Vol. 2. Academic Press, Inc. N. Y. Robertson, T. G. 1963a. Evaluation of the short thermal and repeat 'caudal fold tests used as ancillary tests on initial reactors to the caudal fold test. New Zealand Vet. Journ. 11: 11-15. Robertson, T. G. 1963b. The valuation of tuberculin tests. New Zealand Vet. Journ..ll: 6-10. Sjollema, P. 1954. Causes of non-specific reactions to the tuber- culin test in cattle. Tydschr. Diergeneesk. 18, p. 761. 1953. Smith, D. T. and Scott, N. B. 1950. Clinical interpretation of the Middlebrook-Dubos hemagglutination test. Am. Rev. Tuberculosis 62: 121. ' Stableforth, A. W. & Galloway, I. A. 1959. Diseases due to bacteria Vol. 2. Academic Press, Inc. New York. 77 Stenius, R. 1949. Experience gained in the control of bovine tuber- culosis in Finland. Proc. 15th Int. Veterinary Congress, ;; 221-224. Tervola, C. A. 1960. A study of the transitory nature of tuberculin hypersensitivity in Wisconsin cattle. Unpublished report. USDA ARS, ADED-Wisc. Traum, J. 1959. Skin lesions and their significance in tuberculosis eradication. Proc. Tuberculosis Eradication Conference, Kansas State University. pp. 60-67. USDA, ARS, ANH. 1965. Tuberculosis eradication progress report. p. 39. VanWaveren, G. M. 1953. Proc. XVth Int. Vet. Congre. Stockholm, Part I, p. 140. Vardman, T. H. & Larsen, A. B. 1962. A comparison of the hema- gglutination, hemolytic, and complement-fixation tests on serums from intradermal bovine tuberculosis reactors. Am. J. Vet. Res. 23(93): 274. Vardman, T. H. & Larsen, A. B. 1964. The complement fixation test for detecting tuberculosis in cattle. Am. J. Vet. Res. 25 (106): 690. Walker, C. C. 1920. Tuberculin Tests. J. Am. Vet. Med. Ass'n. 9:5-13. Wight, A. E. 1939. Tuberculosis in livestock. U.S.D.A. Farmers bulletin No. 1069. p. 16. Williams, J. E. 1958. Our present tuberculin. Proc. Tuberculosis eradication conference, Michigan State University. 41-44. Willigan, D. A., McGavin, M. D. & Goyings, L. 3., Unpublished data. MSU Tuberculosis Project. Yates, F. 1949. Sampling methods for censuses and surveys. p. 128. Chas. Griffin & Co. Ltd. London.