Comparison of Molecular and Conventional Methods for Detection of Multidrug-Resistant

Tuberculosis in the United States, 2009 to 2011

Mitchell A. Yakrus, MS, MPHMicrobiologist

Laboratory Branch

16th Annual ConferenceThe Union-North American Region

February 25, 2012

National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention

Division of Tuberculosis Elimination

Background CDC’s DTBE Laboratory Branch provides a service for

the molecular detection of drug resistance (MDDR) to rapidly identify multidrug-resistant isolates of Mycobacterium tuberculosis complex (MTBC)

DNA sequencing of 9 genetic loci associated with resistance to first and second-line drugs

Concurrently perform growth-based conventional drug susceptibility testing (DST) Agar proportion for 12 first and second-line antituberculosis

drugs MGIT 960 for pyrazinamide

MTBC isolates must meet specific criteria before submission for MDDR

Molecular Analysis

Conventional DST

Molecular Results

(Interim Report)

MDDR Isolate Received

Molecular + Conventional DST Results (Final Report)

MDDR Request/Approval

Methods

Received 285 requests for MDDR testing from 43 public health laboratories from September 2009 to February 2011

Analyzed agreement between molecular and conventional results for rifampin (RMP) and isoniazid (INH)

Examined submission criteria as indicated on request forms for possible association with resistance to RMP or INH Isolate known to be multi-drug resistant (MDR) Patient previously treated for tuberculosis (TB) Patient from an area with a high rate of drug resistance

(foreign born)

Agreement Between Molecular and Conventional Testing

Drug No. of Isolates with Both MDDR and Conventional Results

No. of Discordant Results Between MDDR and Conventional Methods

Percent Agreement Between Methods

RMP 229 6 97.4%

INH 228 18 92.1%

Association of RMP or INH Resistance with Submission Criteria

Submission Criterion Association of Detection of RMP or INH Resistance with Testing Method

MDDR Conventional

Known MDR p < 0.001 p < 0.001

Previous Treatment p = 0.585 p = 0.312

Foreign Born p = 0.064 p = 0.271

Summary MDDR was highly concordant with drug resistance

results from growth-based conventional testing

Detection of resistance to RMP or INH was not significantly associated with specific submission criteria with the exception of whether isolates were submitted as known MDR

Further analysis of the association of submission criteria with results is needed to optimize testing algorithms

Need to collect clinical outcome data for patients to measure impact of rapidly identifying drug resistance and discordant results

Acknowledgements Angela M. Starks , PhD, Team Lead,

Laboratory Capacity

Beverly Metchock, Dr. P.H. , D(ABMM), Team Lead, Reference Team