Sven Hoffner chairman of the task force summarised the recommendations and work plan of the task force established following the recommendations of the Technical Advisory group for TB control of the WHO European Region. The task force had its 2^nd meeting on 26-28 May in the same venue. The Task force has eight members and a secretariat at WHO. It acts as a technical advisory board for TB laboratory related issues in WHO European Region and participated in assessment, planning, guideline preparation, training, advocacy. During the meeting the Task force identified 13 priority countries and agreed a work plan:
•To prepare a statement on the optimal laboratory functions tailored to EURO situation
•To revise and adapt the WHO “Global Strategy for the strengthening of TB Diagnostic services –2005-2008”according to EURO situation
•To prepare a Status Report on Laboratory Quality Assurance System
•To prepare Status Reports on needs in Human Resources (staff and consultants) and in Infection Control practices
•To take part in the Task Force for “Wolfheze Guidelines for Infection Control in High Incidence countries and MDR TB settings”
•To prepare the WHO Advanced Regional Workshop for NRL Managers in June 2005 in Riga

A. Infuso (EuroTB) summarised the current status of European surveillance of anti-TB drug resistance (DRS), a project started in the EuroTB network in 1998. By 2003, 41 countries had provided DRS data, mostly through routine reporting of initial DST for all culture positive TB cases. This approach is not currently feasible in the countries of the former Soviet Union, where culture and DST are not routinely used and external quality assurance for DST is rarely implemented. In Western and Central Europe, most countries participate in international quality assurance for DST, while national EQA schemes are less complete. Levels of drug resistance are very high in the Baltic States, where ~20% of all TB cases notified in 2003 were multidrug resistant (MDR), and are much lower in the remaining countries in the EU & West (MDR: 1.8% [0.3-2.5%]). In all these countries the drug resistance is associated with TB treatment history and the prevalence of drug resistance is associated with foreign origin and is particularly high in cases from the FSU. Few countries provide data in the Centre (the Balkans) indicating low prevalence of drug resistance in Croatia and Fed of Bosnia, and intermediate levels in Romania (3% of primary MDR in the preliminary results of a nationwide survey done in 2003-2004). Data from the East remain very incomplete (few surveys conducted) but available evidence suggests that the prevalence of MDR is high throughout the area. Recent trends in primary MDR-TB indicate a relatively stable situation in the Baltics and an increase in the EU & West (0.8% in 1999 and 1.4% in 2003). Corresponding trends in acquired MDR are more difficult to interpret due to definitional changes in the Baltic States (notification of retreated cases other than relapses) and to small numbers and wider yearly variations in the other countries in the EU & West.

K. Kremer (RIVM) presented an overview of the use of DNA fingerprinting to study the epidemiology of M. tuberculosis. The application of DNA fingerprint techniques has greatly facilitated the possibilities to study transmission of tuberculosis at various scales. This has e.g. led to improved detection of nosocomial and institutional outbreaks and laboratory cross-contaminations. Application of DNA typing has also led to discovery that recent transmission, including reinfection, is more important in the epidemiology of tuberculosis than was assumed previously. This finding has important implications for the control of tuberculosis.

Strain typing in different geographic areas has revealed that the population structure of M. tuberculosis differs significantly between settings with a low- and a high incidence of tuberculosis. This has led to extensive research on possible selective advantages of predominant genotypes, such as the ‘Beijing’genotype of M. tuberculosis. By combining information on time trends and comparisons of age groups with data on drug resistance on over 29,000 patients, four distinct patterns were found to Beijing genotype tuberculosis: (1) endemic with no association with drug resistance (at a high level in most of East Asia and at a lower level in parts of the USA); (2) epidemic, sometimes to high levels, associated with drug resistance (in Cuba, the former Soviet Union, Vietnam, South Africa, and, at a lower level, in parts of Western Europe); (3) epidemic but drug sensitive (Malawi, Argentina); (4) very low level or absent (parts of Europe, Africa). Furthermore, in a European study, including IS6110 RFLP typing results and epidemiological data on 683 MDR-TB cases, the Beijing genotype of M. tuberculosis was strongly associated with clustering. In experiments in Macaques monkeys and BALB/c mice, immune responses against infections with Beijing genotype strains were significantly lower, and mortality rates were higher than those after infections with other M. tuberculosis strains. Selection of predominant M. tuberculosis genotypes may have implication for research on the development of new tuberculosis vaccines.

Discussion: the higher M. tuberculosis strain heterogeneity in low incidence countries might reflect both the greater part of reactivation disease but also the great proportion of cases originating from, and infected in, other areas. The great homogeneity observed among M. tuberculosis isolates in most high-incidence countries reflects the high rate of ongoing transmission of strains of certain M. tuberculosis genotypes which probably have a selective advantage over strains of other (rare) M. tuberculosis genotypes.

S. Popov asked whether the Beijing study was a multi-centre study and if the same method had been used to compare the strains? He specified the necessity to take those aspects into account according to his experience in the Russian Federation. Also the stability of the strain when developing in the human body should be considered. K. Kremer answered that the Beijing survey was indeed a multi-centre study, which included many members of the previous EU concerted action on TB project, contacts of these members, and contacts of previously published studies. Most laboratories used either IS6110 RFLP typing or spoligotyping to identify Beijing strains, according to the standardized, published, definition. Regarding the stability, there are indeed examples of gain or loss of one or few bands in the RFLP patterns of isolates associated to outbreaks. Also there might be diversity of bands within a specified bacterial population from the same patient: some strains can differ by a weak extra band corresponding to a mixture of bacterial populations. In general only a minority of the isolates show such variation and these slight differences will not interfere with the determination of the genotype.

Another question was about the principle of the VNTR technique. K. Kremer answered that the VNTRs (variable number of tandem repeats) are located on different positions on the chromosome. The number of repeats at each locus can be determined after DNA-amplification by PCR using primers directed to the flanking regions of these loci, and subsequent fragment-sizing on an agarose gel. The chairman asked what level of discrimination could be achieved with VNTR typing. K. Kremer replied that the discriminatory power depends on the strain collection investigated and the VNTR loci used. VNTR typing using the 12 published MIRU-VNTR loci is somewhat lower than that of IS6110 RFLP typing. The discriminatory power of the method increases when more VNTR loci are included in the analysis. Finally, the role of BCG vaccination in relation to the selection of the Beijing genotype and the impact of BCG vaccination policies was discussed.

A. Infuso presented the general characteristics of the proposed RIVM/EuroTB protocol for this new component of European surveillance. Participation in the epidemiological component will be proposed to all countries in the Region, if EQA for DST in place with good results. National surveillance authorities will provide quarterly updated anonymous individual information on all individuals with a MDR-TB isolate. Participation in the genetic typing component will be possible if one national laboratory can perform IS6110 RFLP according to the recommended protocol. RIVM will receive quarterly typing results for strains of MDR-TB patients. Typing results will be analysed on pooled European typing data to identify international cluster of cases with identical genotypes. For strains with less than six RFLP bands spoligotyping an MIRU-VNTR will be used as additional methods to improve discriminatory power. The result of European cluster analysis will be returned to the countries to be matched with epidemiological information, and to EuroTB for data validation. EuroTB will issue a quarterly report with description of MDR-TB cases and of international clusters. Gradually, principles for investigation of international clusters will be developed.

Discussion: The proposal was judged favourably overall and participants from ~20 countries manifested their intention to participate. The increased value to extend this type of surveillance beyond MDR-TB, to describe general patterns of TB transmission and links of MDR to non MDR transmission was recognised as clustering patterns of MDR-TB cases would not apply to other strains. However, collecting very large numbers of strains would be difficult initially and typing is not done routinely for all TB strains in many European countries. Data confidentiality issues (patients’consent, indirect identification or self recognistion of the patient due to small numbers) was considered a potentially sensitive issue in several countries. It was pointed out that the type of demographic information collected for MDR-TB cases would not differ much from that already collected for all TB cases at the European level. More detailed information might be collected in the framework of cluster investigation. While it is planned to have a patient code and a record number code for reporting the same patient in different calendar years, the system will not be able to detect reports of the same patient diagnosed in different countries. The countries will need to establish a specific communication system for this purpose.

V. Kuyvenhoven reported on the results of a forum discussion on TB policy development for low/middle incidence countries which had been organised two days before as part off the 11th Wolfheze Workshops on TB control. Based on the need to go beyond passive case finding and adequate treatment of TB cases, the recent WW paper “Framework for TB control and elimination in low incidence countries”stressed the need for active case finding in risk group and for treatment of both TB infection and disease. The forum identified the need expressed by many participants in continuing the policy development work initiated in the Wolfheze Workshops and identified areas for policy development, such as development of national TB control guidelines, fulfilling “unmet needs”(eg, free access to TB treatment and care, including 2^nd line drugs, legal tools to enforce treatment for non compliant patients) and pooling experiences to improve policy making in risk group management (eg screening and active case finding in risk groups, contact investigation and outbreak management). These issues should be discussed and lead to consensus recommendations and guidelines.

Karl Ekdahl from the ECDC could not attend the meeting. A. Infuso presented selected ECDC slides to describe the current structure and perspectives of ECDC, which is based in Stockholm, Sweden, and became officially operational on 27 May 2005. The staff will be of 50 persons in 2006 with a budget of 16 M euros. Until 2007, the ECDC will focus on infectious disease surveillance, alert and response and in preparedness for health threats of unknown origin. It has 3 technical units Scientific Advice (director J Giesecke), Surveillance & Communication (A. Ammon), Preparedness & Response (D. Coulombier). Surveillance activities (head:, Germany) will build on existing network and support activities (Eurosurveillance, EPIET) strengthen coordination of network and integrate some, and avoid duplication of activities. Participation of third countries in Europe is welcome as part of the neighbourhood policy of the EU. The ECDC will not have own laboratories but build on existing networks and national resources. A. Infuso informed that the ECDC has already received baseline documentation on EuroTB and all public databases. EuroTB will answer specific ECDC requests upon time availability and will be evaluated in 2006.

D. Falzon gave an update on TB surveillance in Europe. There were a number of important developments since last year in this respect, namely the securing of EC funding for EuroTB up to end 2007, the establishment of the ECDC and the expansion of the EU to 25 countries. The increasing divergence in tuberculosis notification rates between western (mean: 14/105) and eastern (104/105) countries over time is a result of decreasing overall rates in the West and continuing transmission and improved case-detection in the East. In the West, TB is mainly a disease of risk groups and it is decreasing in most countries, except for Italy, Norway and the United Kingdom, where much of the increase is due to disease in cases of foreign origin. The average age of foreign TB cases is younger than for native cases and most originate from Africa (38%), Asia (25%) and Centre & East (21%). HIV co-infection in TB cases is highest in Portugal (16% in 2003) and Spain (10%) and is under 5% in most other countries. In Estonia and Latvia, a small increase has been noted between 1999 and 2003. In 2003, representative drug resistance data were largely limited to countries in the EU & West where the percentage of combined multi-drug resistance was below 4% in all countries except Israel (6%) and the Baltic States (15-23%). Recent surveys in the East show levels of MDR comparable to those observed in the Baltic States. Success ratios tend to be comparable in the EU & West, Centre and East but death tends to be commoner in the West (7% in new cases in 2002), as a result of a more aged population, while failure is frequent in the East (9%). Overall tuberculosis mortality rates increase steadily as one moves from the west to the East.

There are still a number of limitations in TB surveillance in Europe. The intra-country variation in rates of disease is not addressed by current European surveillance. Data from certain regions (like Kosovo, Abkhazia and S. Ossetia) are missing. The definition of a notifiable case differs between countries and sometimes over time, making international comparisons and trend analyses difficult to interpret. Among others, the definitions of treatment history, foreign origin and treatment outcome are also non-standardised. Data completeness is problematic, particularly in culture confirmation, HIV sero-status and outcome reporting in particular.

Discussion: In countries with major underreporting presentation of notification data without due commentary, EuroTB may be furthering the political ends of the authorities in certain countries wishing to under or overstate their tuberculosis situation. Would it be possible to develop an estimate or a range for country indicators in EuroTB reports based on available knowledge about these factors? Rather than developing estimates, it is suggested that anyone doing a country analysis would also list the possible sources of bias (eg incomplete reporting, low laboratory activity, underestimated denominator populations etc). Drug resistance data in migrants to Europe may be useful to policy and care in this group but do not represent resistance in migrants in the countries of origin, where, moreover, data they are frequently incomplete or lacking (e.g. China, India, Russian Federation).

A. Infuso presented preliminary results of a recent survey based on 29/36 countries invited to participate. The survey revealed that BCG policies have changed in the majority of the countries since 2000. BCG remains recommended for all newborns or young children in all countries with an incidence > 20 per 100,000 while under 20 per 100,000 policies vary widely from systematic vaccination of newborns to targeted vaccination of children at risk (origin from or travel to a high TB incidence country, contact with a TB case / family history of TB) to no use of BCG in children. Five countries recommend vaccination of all older children Four countries still recommend revaccination. Vaccine coverage is very high (>90%) in most countries using BCG for all newborns, whereas it is lower (60% to 90%) or not available in those targeting children with risk origin. In many countries, information on BCG status and eligibility is not routinely collected through TB case notifications or is incomplete. Unexpectedly, surveillance of severe side effects (disseminated BCGitis) is reported to be in place in less than half of the countries. Very few countries have data on incidence of mycobacterioses other than TB. Enhanced surveillance and evaluation of current recommendations could contribute to improve/harmonise current BCG policies in Europe. The final survey results will be published in the Eurosurveillance journal.

Discussion: A summary of policies should be made available and regularly updated on the EuroTB website. Information from paediatric TB notifications cannot be used reliably to compare BCG policies.

This was the first time EuroTB correspondents were invited to submit abstracts to the annual meeting. All nine abstracts received were offered oral presentation and eight were presented (see list below).

The new electronic TB case reported system adopted in the Netherlands was presented. Hungary described the introduction of laboratory reporting and its beneficial effect for improving completeness of reporting and quality and completeness of bacteriological information. A set of two studies from Scotland prompted discussions on the contribution of molecular surveillance to TB control in low incidence urban settings. In Slovakia, the epidemiology of TB in the Rroma population was described as well as measures to improve case holding and adherence to treatment in group. In the Czech Republic, a tuberculin survey of health care workers showed a high prevalence of large positive skin tests, and is expected to provide useful indications for prevention in this group in spite of the continuing use of BCG revaccination. Finally overviews of national programme organisation and outcomes were presented for Slovakia and Bulgaria. Each presentation has been followed by discussions showing the need to further strengthen links between surveillance, studies and control actions.