Diagnosing child TB is a challenge in 22 High Burden Countries (HBCs) including Bangladesh. WHO estimates the case detection rate should be 6-10% of total TB cases. Recent publication shows it can be 4-22%, with sporadic reports form HBCs up to 25%. In Bangladesh, case detection was 3.1% in 2011, 2.85% in 2012 and it was 2.74% in 2013. Moreover, case detection rate in the sub-district hospitals (Upazila Health Complex-UHC) were even lower than the national average. One of the important reasons is lack of clinical skills of doctors and awareness of health care workers on child TB at these community hospitals. To develop capacity among these groups of stakeholder, between 2012-13 Bangladesh has developed generic interactive training modules, training video, flip chart and other training aids on Child TB. By using these tools, capacity building program for doctors and health care workers was conducted between November 2013-July 2014. This was done with support from USAID under the guidance of National Tuberculosis Control Program (NTP) with Bangladesh Pediatric Association (BPA) as technical & implementing partner. Total 1181 doctors and 8345 health care workers have been trained on child TB in 17 districts and 122 sub-districts of Dhaka Division. Preand post-test analysis, of doctors trained, showed statistically significant (p<0.001) improvement of knowledge of the participant irrespective of age, qualification and portfolio. Andragogy modular training methodology was applied; and has been highly appreciated and enjoyed by the participant. Health Care worker also showed enthusiasm on the orientation. Case detection rate showed a raising trend in the sub-district hospitals of Dhaka division. The number of cases detected in the project area was 752 in 2013, while 992 cases were detected till 30 September 2014 (132% of 2013) sparing one quarter (October to December) of 2014. Year end data will provide further light on increment. Since no other intervention was undertaken during this period in Dhaka Division, it can be assumed that the increase in the case detection rate was due to this capacity development program of the doctors and health care workers. LINEAGE 7 MYCOBACTERIUM TUBERCULOSIS STRAINS ARE ASSOCIATED WITH LONGER PATIENT DELAY IN PULMONARY TB PATIENTS SA Yimer, G Norheim, A Namouchi, ED Zegeye, W Kinander, T Tønjum,S Bekele, T Mannsåker, G Bjune, A Aseffa, C Holm-Hansen Department of Microbiology, Unit for Genome Dynamics, Oslo University Hospital, PO Box 4950, Nydalen, NO0424 Oslo, Norway, Email: yimsolo@yahoo.com Background This study investigates the genetic diversity of Mycobacterium tuberculosis (Mtb) strains among pulmonary TB patients in Amhara Region, Ethiopia, and the association between specific Mtb linages, sociodemographic and clinical parameters. Methods DNA was isolated from Mtb-positive sputum specimens (n=240) and analyzed by PCR/24-locus MIRU-VNTR and spoligotyping. Bioinformatics was used to assign Mtb genotypes to global lineages. Associations between patient characteristics and genotype were evaluated using logistic regression analysis. Results Mtb strains (n=138) were assigned to seven sub-lineages, four of which were not represented in the MIRUVNTRplus database. The largest sub-lineages (n=60; 26.0%) belonged to Central Asian (CAS), the next (n=36; 15.6%) to lineage 7, and the third (n=35; 15.2%) to Haarlem. The four novel sub-lineages designated NW-ETH3, NW-ETH1, NW-ETH2, NW-ETH4 included 24 (10.4%), 18 (7.8%), 8 (3.5%) and 5 (2.2%) isolates, respectively. Patients infected with lineage 7 strains were highly likely to delay in seeking medical attention compared to patients infected with CAS strains (AOR=4.7, 95% CI 1.6, 13.5). Cases of Harlem infection (OR= 2.8 95% CI 1.2, 6.6) and NW-ETH3 (OR= 2.8 95% CI 1.0, 7.3) appeared in defined clusters. Conclusion The study revealed a high diversity of modern and pre-modern Mtb lineages of which approximately 25% were not previously reported. Infection with Mtb lineage 7 strains is associated with longer patient delay that suggests a possible increased duration of illness among these patients. Intensified active case finding and contact tracing activities in the study region are needed to expedite diagnosis and treatment of TB. RESISTANCE TO PYRAZINAMIDE IN RUSSIAN MYCOBACTERIUM TUBERCULOSIS ISOLATES D.A. Maslov, O.B. Bekker, T.G. Smirnova, Elena E. Larionova, Sofya N. Andreevskaya, L.N. Chernousova, Y. Zhang, V.N. Danilenko d Vavilov Institute of General Genetics, Russian Academy of Sciences, Moscow, Russia; Central TB Research Institute, Russian Academy of Medical Sciences, Moscow, Russia; Division of Global Health Equity, Brigham and Women's Hospital, Harvard University, Boston, MA 02115, USA; NPO SRC «BIOAN», Moscow, Russia. Corresponding author V.N. Danilenko: Vavilov Institute of General Genetics, Gubkin str. 3, 119991, Moscow, GSP-1, Russia; e-mail: valerid@vigg.ru. Tuberculosis (TB) is a major global health problem with an estimated of 9.0 million new TB cases and 1.5 million TB deaths in 2013. Pyrazinamide (PZA) is a key frontline drug in TB chemotherapy, which shortens the lengthy anti-TB therapy to 6 months due to its sterilizing ability against persisters. PZA is recommended for treatment of both drug-susceptible and drug-resistant TB cases. The phenotypic testing for PZA susceptibility can show both false-positive and false-negative results (Piersimoni et al., 2006). Up to 40% of false-positive resistance can be obtained on Bactec MGIT 960 (Chedore et al., 2010). Sequencing of pncA gene is also recommended (Simons et al., 2012). Mutations in gene pncA coding pyrazinamidase (PZase) are the major mechanism of PZA resistance in M. tuberculosis (Scorpio and Zhang, 1996; Scorpio et al., 1997; Cheng et al., 2000). Mutations in rpsA gene, encoding S1 ribosomal protein, a key element of trans-translation, were found in some PZA-resistant M. tuberculosis isolates lacking mutations in pncA (Shi et al., 2011). Another gene involved in PZA-resistance is panD, encoding an aspartate decarboxylase (Zhang et al. 2013; Shi et al., 2014). We analyzed 64 M. tuberculosis clinical isolates from patients from the European part of Russia. The isolates were classified in 3 groups: 20 isolates from patients with long and ineffective chemotherapy; 21 isolates from patients with long and effective chemotherapy; a control group of 23 drug-sensitive isolates from new-TB cases. The collection included 6 monoor poly-drug resistant isolates, 25 MDR and 10 XDR isolates. The genotyping by spoligotyping revealed 70% of the isolates to belong to Beijing lineage. The share of Beijing genotypes varied from 48% in the control group to 95% in the group with long and ineffective treatment. The isolates were tested for phenotypic PZA resistance in Bactec MGIT 960 system and for PZase activity using Wayne method. The genes controlling PZA resistance (pncA, rpsA and panD) from these isolates were sequenced. Twenty-four isolates were found to be PZA-resistant after the first Bactec MGIT 960 DST, twentythree of them carrying mutations in pncA and showing negative PZase activity; five more phenotypically sensitive to PZA isolates also harbored mutations in pncA and had negative PZase activity, three of them confirmed PZA resistance on the second DST. Mutations in rpsA and panD not leading to PZA resistance in clinical diagnostics concentration (100 mkg/ml) were found, which still may lead to low-level PZA resistance. We have found 1 PZA-resistant isolate with no mutations in known genes, which may harbor a new mechanism of