Streptococcus agalactiae - an increasing problem in Scandinavia Report from Denmark, Jørgen Katholm. Knowledge Centre for Agriculture, Cattle

Streptococcus agalactiae - an increasing problem in Scandinavia Report from Denmark, Jørgen Katholm. Knowledge Centre for Agriculture, Cattle NMSM - The Nordic Dairy Association's Committee for Milk Quality, Mastitis symposium, Rebild, Denmark, 9 June 2010 I will focus on the Streptococcus agalactiae (S. agalactiae.) program in Denmark since I was employed to be in charge of the program as 1 October, 2007. Although a thorough report on the program in 2004 (Pedersen et al 2004) stated that the purpose had to be reformulated, secured for the future and the visibility should be increased, the words at my employment were: "You will have to look into the need for continuing this program". Since the program started in 1963, it has undergone several changes and there have been a series of reports on the program (Jensen 1976, Andersen and Huda 1995, Andersen et al 2003, Pedersen et al 2004, Katholm and Rattenborg 2009, Katholm and Rattenborg 2010). Figure 1 Prevalence in percent of herds in the B-register as of 31 December As can be seen from figure 1, the increase in the prevalence of infected herds is rather severe, ending with 6.1% of milking herds being registered as infected with S. agalactiae. A beginning increase with a rise to 3% in 2003 was also reported by Pedersen et al (2004). Knowledge from the past From the earlier reports I would like to draw attention to the following observations. In total there had been 4.85 herd investigations in 35 herds, which had been infected for a longer period before they were cleaned of the infection ( Jensen 1976). Our experience is that we can find a 200 cow herd infected in a Bulk Tank Milk (BTM) sample (inoculum 120µl) although only one cow is found infected in the testing of all quarters. 1

The new infection rate is difficult to reduce even in a period with an intensified eradication program (Andersen and Huda 1995). In repeated bulk tank sampling of 100 herds 7 times every 2 weeks S. agalactiae was found in 4 herds known to be infected before the project, and 8 new herds. S. agalactiae was not found in all seven samplings in any of the eight new herds. Additionally, 3 herds were tested positive but suspected to be caused by cross-contaminated as there was a marked coincidence between the same ribotype and a higher number of colony forming units in the bulk tank sample from herds just previously sampled in the same examination round (Andersen et al 2003). The collection valve in the lorry at that time was of Spentrup automatic sampling system (Mark IV, Spentrup machine woks, Spentrup Denmark). This has now been changed to VM OVP valve (company VM Tarm A/S). First 30-40 l, no sampling, thereafter 1.5 ml 40 times at interval depending on last herd milk load, total sample 60 ml, thereafter cleaning of the device with pressure air. In 2004 an evaluation was made of historical data from 1989 to 2002 (Pedersen et al 2004). 1299 herd investigations in 1076, new infected herds included quarter samples from 56.133 cows. Overall, 10% of the cows in a new infected herd were infected. 37% of the herds had less than 5% infected cows, and 1.2% of the herds had more than 50% infected. Lactating cows in first lactation were 5.8% infected, whereas 16.6% of cows in fifth lactation were infected. Of the infected lactating cows 65.0% had only one quarter infected. Among the 8268 positive cows 25%. 8% of the cows hat CMT value 1, and 11.8%, 11.9% 18.2% and 32.2%, respectively, had the CMT values 2,3,4 and 5. I will also draw attention to some other Danish publications from the past stating that cows infected with human strains were demonstrated to have a higher self-cure rate than cows infected with bovine strains (Jensen, 1982),. Agger et al (1994) found that their results indicated that purchase of cows or heifers increases the risk of a new Strep. a herd infection, and as early as 1965 Livoni and Mølgaard presumed that more than half of the new infections were introduced by human carriers. Investigations in 2008 and 2009 General information to farmers and advisors In the past 2 years we have intensified the focus on information about S. agalactiae. We continue to forward information to the farmers of their results in the yearly testing included in the weekly information papers from the Danish Cattle Database. As an extra information, all herds that have changed status the past year receive a letter after the completion of the annual testing of BTM, commonly in January, with information n the new status and how to act. All veterinarian practices will at the same time receive information about all the farms they visit and which are registered in the B-register. On top of that we produced farmer information in the Weekly Cattle News and latest in "Danske Mælkeproducenter" in April 2010. Information to veterinarians and farmers is available from our website www.maelkekvalitet.dk. 2

Infection with S. agalactiae somatic cell count (SCC) and total bacterial count (TBC) We have produced various results concerning S. agalactiae in the past two years, and I will emphasize some of these findings. It is well-known that infection with S. agalactiae. will increase the SCC at cow level and herd level. On the other hand, the variation in number of infected cows in a herd and the cyclic shedding of bacteria from infected cows make it difficult to predict the influence of the infection on the herd SCC and thereby also the possibility to reduce the Bulk Tank SCC (BTSCC) after eradication. Figure 2: The geometric mean of SCC each week for 261 milking herds in the B-register compared to the milking herds not registered in the B-register. For Arla suppliers the geometric mean will be based on BTSCC at each delivery of milk whereas for non-arla suppliers this will be of 1 testing each week. Figure 3: The geometric mean of TBC each week for herds in the B-register compared to milking herds not registered in the B-register. The values are based on one measurement every second week, and for problem herds once a week. From figure 2 it can be seen that for the whole year 2009 a constant difference of around 25.000 higher BTSCC for milking herds in the B-register can be expected. From figure 3 it can be seen that for the whole year 2009 a constant difference of around 2.000 3.000 higher TBC for milking herds in the B-register can be expected. The influence of S. agalactiae infection on the TBC is also known from the literature. Samples of BTM from infected herds contain bacterial counts in the range of 20.000 100.000 CFU/ml (Hoogan and Smith 1992). At the beginning of the infection shed up to 10 8 CFU/ml, one single cow can bring the TBC in a herd with 1000 cows to 100.000 (Guterbock and Blackmer 1984). Infected herds were 5.48 times as likely to be penalized as uninfected herds. Penalties are imposed when standard plate count exceeds 50.000 CFU/ml (Keefe et al 1997). 3

Relation to herd size and milking system and producer of AMS We have compared the number of cows in the dairy herds with the risk of being registered in the B- register, cf. figure 4. Herds with more than 300 cows have around 20% possibility of being registered in the B-register. Figure 4: The percentage of herds registered in the B-register in relation to herd size in 2009. Table 1: The percentage of herds in the B register in relation to milking system and AMS producer In the past two year many veterinarians have phoned us about problems with S. agalactiae in infected herds, and often they referred to AMS herds. One of the veterinarians, Michael Graves, asked us to investigate the influence of the producers on AMS as he found that the Lely herds seemed to be the herds with problems with the S. agalactiae infections in his area. Mr. Jørgen Nielsen, our Statistics Manager, has investigated this problem and found that, overall the risk of being registered in the B-register on 21 December 2009 was 5.7% (193 out of 3198) for a conventionally milked herd, whereas an AMS herd had a risk of 7.6% (65 of 795) This difference is statistically significant despite a weak, P=0.04. We know that AMS herds often have more than 100 cows and that herds with more cows are more often registered in the B-register. If we take these two factors into account with a logistic regression with the two factors in an additive model, the statistic analyses cannot detect any significant effect of AMS. It was thereafter investigated if the different producers of AMS had an influence on the possibility of a herd being registered in the B-register. A simple comparison of the percentage of herds in the B register in relation to milking system and AMS producer is shown in table 1. A logistic regression with number of cows in the herd and the AMS producer show no statistic difference between VMS and herds without AMS, whereas the Lely herds are statistically different to the herds without AMS P=0.04 and to VMS P=0.01. The estimated possibilities of being registered in the B-register compared to herd size are shown in figure 5, and the difference with VMS being lower and Lely and other producers being higher than the herds without AMS is constant for all herd sizes. 4

Figure 5: The estimated possibility of being registered in the B-register related to number of cows in the herd and AMS producer on 21 December 2009 We have investigated if this difference in risk of being registered in the B-register is related to trade with animals, and we could not find any difference in the attitude of trading animals in VMS and the Lely herds that could explain this variation. Another explanation could be status in the B- register before changing to AMS. We are looking at this, but at moment we have no results. Questionnaire to farmers and veterinarians about infected herds In July 2008, we send a questionnaire to 233 milking herds registered in the B-register as of 2 July 2008. The questionnaire was answered by 77 (33%) of the farmers. Of those, 21% stated that the infection had caused increased problems with clinical mastitis, 29% stated that the infection had caused an increase in BMSCC, and 10% noted other milk quality problems. Milk quality advisors had visited 42% of the herds since 2000 for extra control of milking procedures and milking equipment. Only 97% of the farmers have regular control of the milking equipment. Post-milking teat dipping (PMTD) was only used in 75% of the herds. Four owners (5%) had been treated by the physician because of the infection in the herd. Figure 6: Questionnaire to 233 milking herds in B-register as of 2 July 2008. Owner answer in % from 77 herds (33%) Figure 7: Questionnaire to 233 milking herds in B-register as of 2 July 2008. Veterinary answer in % from 82 herds (35%) 5

The questionnaire was returned by 82 (35%) veterinarians. The veterinarians stated that they had diagnosed S. agalactiae infections in individual cows in 18 (22%) of the herds since 1 January 2007, typically from 1 to 8 cases in the period. In 3 of the 233 dairy herds S. agalactiae was isolated from a considerable number of individual milk samples. Milk sample in relation to all mastitis treatments were taken by 54% of the veterinarians. New calving cows were tested for CMT reaction in 19% of the herds. According to veterinarians dry cow therapy was only used in 74% of the herds. Only in 10 (12%) herds the veterinarians were involved in eradication programs. In only one of these programs segregation in groups was carried out. In 22 (27%) cases the veterinarians had recommended the farmers and staff to consult a physician. In 2 cases this consultation resulted in medical treatment. One of these cases was also represented in the questionnaire returned by the farmer. A preliminary investigation of trade patterns since 2000 among infected farms showed that among the 233 infected dairy herds, 24 herds had not bought cattle at all. Of these, 17 herds had infection introduced after 2000 as they all had negative herd samples in the period 1991-2000. The remaining 7 had been infected previously. Further, 104 herds had only bought cattle from herds not registered in the B-register. However, 105 herds had been supplied with cattle from other infected farms. One herd was supplied with cattle from 16 different herds in the B-register, and one herd had bought new cattle from 12 herds in the B-register. Investigations of serotypning and Multi Locus Sequence Typing (MLST) For a period we have not focused on the difference in serotypes in the herds. In 2008, however, we had comments from veterinarians of types with severe clinical expression which was new to me as we usually think of S. agalactiae infections as subclinical and chronic infections. Jensen 1982 found in experimental trails that affection of the general condition of the cows in the initial phase of the disease is seen more often in cases of mastitis caused by S. agalactiae of human origin than in cases caused by S. agalactiae of bovin origin. Another conspicuous feature in the infection with human strains of S. agalactiae is its tendency to relatively fast, spontaneous elimination. In the spring 2008, we started a collaboration with Lotte Munch Lambertsen, Statens Serum Institut (SSI), Copenhagen. We started with 21 samples for serotyping. We found that, in two herds with severe symptoms, all 4 cows in one herd had serotype V, and in the other 2 cows and the bulk tank all had serotype NT, 7 samples serotype Ia, 5 samples serotype NT, and 2 samples serotype III. In all cases with more samples from a herd we found the same serotype. We decided to investigate further the relationship to chronic or acute symptoms in the herd in relation to MLST of the bacteria. Again, all serotyping and MLST were performed at SSI, Copenhagen. Table 2. Clinical appearance, MLST and serotype of S. agalactiae isolates BTM samples from 25 selected Danish herds in 2009 Clinic MLST Serotype Total number of Ia III V NT 1 Isolates Acute 1 3 3 8 1 1 10 1 1 2 19 1 1 6

23 1 1 2 103 2 2 4 ST 462 1 1 ST 463 1 1 Chronic 1 3 3 19 1 1 23 1 1 1 3 41 1 1 ST 460 1 1 ST 461 1 1 Total number 5 2 2 16 25 We found no ST-17, which is a human type of bovine origin that is found to be highly (30%) associated with neonatal infections. Bisharat et al (2004) found ST-17 in 29% of their human types from both neonatal infections and carriers. Among the other human types they found ST-1 in 14%, ST-19 in 13% and ST-23 in 11% as the most common. They only found 1 bovine type ST-1 and one ST- 23, whereas they found ST-67 in 66% of the bovine types. Some of their bovine types were collected back in 1955. We found ST 1 in 24% (6 of 25), ST 19 in 8% (2 of 25) and ST 23 in 20% (5 of 25) indicating that a high proportion of S. agalactiae in Danish Dairy herds is of human origin. Testing of BTM samples BTM samples have been taken with different intervals since 1964. Since 1995, the BTM survey has been carried out once a year and usually in the fourth quater. BTM sampling is carried out during milk collection. Within 24 hours, the samples are shipped to Eurofins Laboratory, Holstebro, Denmark. Selective agar mixed with 120µl of milk is cultured for S. agalactiae (Agger et al 1994). Herds infected with S. agalactiae are recorded in the so-called B- register on the basis of positive individual cows and BTM testing. Free herds found BMT positive are tested again. If the second sample is positive, the herd is recorded in the B-register. If the second sample is negative, a third test settles the status of the herd. Calculating the prevalence of herds in the B-register, the numerator also includes infected herds having ceased milk production until the last cattle of the milking breed has left the herd as well as herds with regular movements of animals from infected herds. This procedure of calculation has been carried out since the start of the program. Since 2005, the prevalence has also been calculated as the proportion of infected active dairy herds. The total number of active dairy herds in Denmark is calculated every year in May by the Danish Dairy Board. Results of these investigations is seen in figure 1. In 2009, we decided also to test these BTM samples from each farm with the PathoProof PCR test. This test was performed on the same milk sample as the culture for S. agalactiae. In table 3 we have compared the result of the standard culture procedure with the result for S. agalactiae in the PCR test. 7

Table 3: Results of culture and PCR on 4258 Danish BTM samples in 2009 Numbers in bracket are herds between 37 og 39,9. *11 herds in B-register. 9 herd samples were positive **28 herds in B-register The PathoProof Mastitis PCR Assay (Finnzymes Oy, Espoo, Finland) was used for direct DNA extraction and real-time PCR, i.e. without any culture procedures. The test took approximately 4 hours from start to finish. A total of 350 µl of milk was used as a starting volume for DNA extraction. The DNA extraction protocol involved an enzymatic lysis step, disrupting the cell walls of Gram+ and Gram- bacteria, as well as spin column based DNA purification and elution steps (produced for use with, and included in, the PathoProof Mastitis PCR Assay by Qiagen gmbh, Hilden, Germany). The assay targeted a total of 11 most important bacteria causing IMI, as well as the staphylococcal beta-lactamase penicillin resistance gene. From table 1 a clear tendency appears towards more positive BTM when the PCR test is used. This could be a reason for culturing not finding the bacteria because they are dead and because the inoculum is higher with the PCR, 350 µl compared to 120 µl. There is a rather high number of herds in the B-register in both the culture positive and PCR negative, i.e. 11 out of 20 herds, and in the PCR positive and culture negative 28 out of 133 herds. We have tested 17 isolates that were purified from the 20 culture positive and PCR negative herds. They are all confirmed as S. agalactiae. The fluctuation in BTM in culture is discussed by Andersen et al (2003), and Zadoks (2005) also discussed the possibility of human contamination of the BTM samples as she found a BTM sample positive, but all 25 cows in the dairy herd were negative. In April 2010 we had the opportunity to test BTM from all 30 milking herds on the Faroe Islands, and although the herds had no knowledge of Infection with S. agalactiae, we found that 7 (23%) herds were positive for S. agalactiae in the PCR test. All 7 samples were then also tested positive by culture. Testing of individual cows Testing of individual cows for S. agalactiae infection has also traditionally been performed by culture. Both sampling of individual quarters and composites samples from all 4 quarters have been used in test programs for S. agalactiae. infection. With the PathoProof PCR, the possibility of testing individual cow samples taken by the milk control program in the herd (DHI) became a possibility. Although bacterial growth is stopped by adding bronopol to those samples, the PCR is able to detect the bacteria anyway. So, we tested this in 3 different milk control situations - the Lely robot, the ordinary TrueTest, and the DeLaval collecting tool. The herd sizes were 183, 135 and 126 samples respectively. We compared the results from culture of composite milk samples taken by our quality advisers and PCR at the DHI samples. Results of the culture samples were respectively 23 (13%), 5 (4%) and 18 (14%) positive samples 8

With a border line of positive cows in PCR being cows with a Ct value lower than 37, the number of positive cows with the PCR test in the respective herds were 63 (35%), 12 (9%) and 56 (44%). Totally, the results from the three herd investigations is shown in table 4 Table 4: Comparison of culture results from 442 composite milk samples and PCR performed at DHI samples The 8 culture positive and PCR negative samples had the following CFU/ml and Ct values: (3 NoCt), (8 NoCt), (2-NoCt), (20-39,67), (645-NoCt), (2-37,45), (5 NoCt), (32 37.11) Again, compared to the results from the BTM samples in table 3, there is a tendency towards the PCR test finding a lot more positive results although also here some are missed by the PCR. In these testings, the inoculums for culture are 620 µl compared to the 350 µl for the PCR, so in the composite milk testing the inoculums were higher for the culture compared to the PCR. After the first testings I was convinced that by the PCR tool we now had a chance to get better test results and success in the eradication of S. agalactiae in the Danish Dairy herds. These eradications had practically stopped since the support for payment of the laboratory costs were stopped in 1995. Mostly due to lack of success and as Jensen (1976) found it was necessary to test a herd 5 times in average to have a herd free of the infection, in this period the testings were performed as individual quarter samples. During the test period treatments and follow-up results we managed to reach a free bulk tank in all three herd, but the final results were that only one of the three herds reached a free status. In the following period voluntary testing that has been taken up by some veterinarians, I have advised that all cows with a Ct value for Strep. a. of 37 39.9 are considered to be an infected cow and segregated along with the cows with lower Ct values. Cows with lower Ct values are considered being infected and are advised to be culled or treated in 3 days with penicillin if the udder and teats are found perfectly fit for treatment. Ongoing eradication programs in S. agalactiae infected herds Since these three attempts to eradicate the S. agalactiae herd infection I have been in contact with farmers and veterinarians in 14 herds that have started eradication programs, cf. Table 4. The major reason to start the eradications has been milk quality both SCC and TBC and a high number of treatment for mastitis. As expected from the onset of segregation treatments and culling there was a marked effect at the BTSCC. 9

Table 4: Results from 14 herds trying to make eradications of the infection with S. agalactiae since august 2009 Chr dato opdelinganimals Culture PCR positiv Prevalens Date Result 4 19/8 09 303 281 95 34 26/4 10 33.9 5 10/9 09 166 27 7 4 FRI 7/4 2010 6 1/10 09 175 159 29+13 26 9/5 10 21.6 7 25/11 09 125 107 8+3 10 6/5 10 27.3 8 16/12 09 181 149 79+12 61 8/5 10 NoCt 9 23/12 09 103 94 17+3 21 24/4 10 NoCt 10 26/1 10 50 48 1 2 Fri 18/5 2010 11 1/2 10 42 Pool 3 pool - 3 cows 7 4/5 10 Pos culture 12 8/2 10 75 68 0+2 3 10/5 10 Pos culture 13 18/2 10 78 75 16 21 13/4 10 NoCt 14 9/3 10 301 Pool 15 5 25/3 10 3.culture neg 15 3/3 10 105 102 9 9 25/5 10 Pos culture 16 8/3 10 270 275 12+4 6 16/3 10 1. Culture neg 17 20/4 10 105 Pool 4 4 2/5 10 1. Culture neg A typical result of the SCC in such a herd is shown in figure 8, although it is unusual that the time between known test results and time of action is as long as in this herd. The reason was an infection rate of 61% and therefore the actions have to be accepted by the bank. Figure 8: A typical effect of reduction in SCC in a herd with S. agalactiae infection The first arrow is at time of known test results the second arrow is at time of treatments and culling In figure 3 it was shown that the geometric mean of TBC for the herds in the B-register in 2009 was around 2.000-3.000 CFU/ml higher. Infected herds will also experience a huge variation in the role of the infection on TCB. Many herds with low infection prevalence and few new infections will often not find any influence on the TBC. On the other hand we have found several herds with huge variation in the TBC that often reach values between 100.000 300.000 CFU/ml in periods where many new infections can be expected. Figure 9 is an example from a herd that often had visits from our quality advisers due to penalties on milk quality. Although some smaller changes on the wash such as water amount, temperature and disinfection concentration could be optimized, the TBC was not stable and low until after the test, treatments and culling of cows in the S. agalactiae eradication program. 10

Figure 9: A typical effect of reduction in TBC in a herd with S. agalactiae infection The first arrow is at time of known test results. The second arrow is at time of treatments and culling This herd also experienced a drop in the treatment for mastitis in the herd, and this is also found in other herds after test, segregation, treatments and culling, cf. figure 10. Figure 10: A typical effect of reduction in numbers of mastitis and treatments for mastitis in a herd with S. agalactiae infection. The first arrow is at time of known test results. The arrow is at time of treatments and culling Test, segregation culling and treatment The program advised from the NMSM is to be found later in this material. 11

References Agger J.F., Priou C., Huda A.,Aagaard C.: Risk factors for transmission of Streptococcus agalactia infection between Danish dairy herds, Vet. Res., 1994, 25, 234-238. Andersen H.J. og Huda A.: B-streptokokbekæmpelsen 1992-1994, DVT, 1995, 78, 112-1130. Andersen H.J., Pedersen L.H., Aarestrup F.M., Chriél M.: Evaluation of the surveillance program of Streptococcus agalactia in Danish dairy herds, J. Dairy Sci., 2003, 86, 1233-1239. Bishart N, Crook D.W., Leigh J., Harding R.M., Ward P.N., Coffey T.J., Maiden M.C., Peto t., Jones N.: Hyperinvasive neonatal group B streptococcis has arisen from a bovine ancestor, J Clin Microbiol, 2004, 42, 5, 2161 2167. Guterbock W.M. and Blackmer P.E.: Veterinary interpretation of bulk-tank milk, Veterinary Clinics of North America: Large Animal Practice, 1984, 6, 2, 257 268. Hoogan S.J. and Smith K.L.: Using bulk tank cultures in a dairy practice. Nat. Mastitis Council, Mastitis Microbiology Diagnostic Workshops, Arlington, Virginia, 1992 Jensen N. E.: Bekæmpelse af gruppe-b streptokokmastitis. DVT. 1976, 59, 325-330. Jensen N.E.: Distribution of serotypes of group-b streptococci in herds and cows within an area of Denmark. 1980, Acta vet. scand, 21, 354-366. Jensen N.E: Experimental Bovine Group-B Streptococcal Mastitis Induced by Strains of Human and Bovine Origin. 1982 Nord Vet Med, 34, 441-450 Katholm J. and Rattenborg E: Overvågningen af B- streptokok infektion i danske malkekvægsbesætninger. Dansk Veterinær Tidsskrift, 2009, 19, 92, 24-31 Katholm J. and Rattenborg E.:The surveillance program of Streptococcus agalactiae in Danish dairy herds 1989-2008. 5 th. IDF Mastitis conference, Christchurch, New Zeeland 2010, 241-246 Keefe G.P., Dohoo I. R., Spangler E.: Herd prevalence and incidence of Streptococcus agalactiae in dairy industry of Prince Edward Island, J. Dairy Sci., 1997, 80, 464-470. Livoni P og Mølgaard A.: Gruppe B-Streptokokinfektioner hos mennesker og hos kvæg. Medlemsblad for Den danske Dyrlægeforening, 1965, 48, 871-880. Pedersen L. H., Nielsen J., Jepsen Ø. R. Evaluering af det danske Streptococcus agalactia (GBS) overvågningsprogram af malkekvægsbesætninger. Dansk Kvæg Rapport, 2004, 1 60. Zadoks R.N.: Molecular tools enhance the value of Bulk Tank Milk Monitoring. 2005, NMC Annual meeting Proceedings 86-93 http://www.maelkekvalitet.dk/ 12