Blood Assay for Tuberculosis: Initial Findings in a Preplacement Surveillance Program

To The Editor: New blood assays for Mycobacterium tuberculosis (BATB) can assist in the diagnosis of tuberculosis (TB) infection and disease,1,2 and QuantiFERON-TB Gold (QFT-G) was recently approved for use in the United States.3 It indirectly identifies M. tuberculosis complex (ie, M. tuberculosis, Mycobacterium bovis, Mycobacterium africanum, Mycobacterium microcoti, and Mycobacterium canetti) in whole blood specimens. Individuals with M. tuberculosis complex infections usually have T-lymphocytes in their blood that recognize early secretory antigenic target-6 (ESAT-6), culture filtrate protein-10 (CFP-10), and other mycobacterial antigens and then generate and secrete the cytokine interferon-γ, which is detected and quantified by enzyme-linked immunosorbent assay (ELISA) in the QFT-G test. These proteins are absent from all bacille Calmette-Guérin (BCG) strains and so the interferon-γ assay for M. tuberculosis complex should be negative in persons whose TB skin test is positive solely due to BCG vaccine.1,4 We summarize our earliest experience using a BATB under program conditions and highlight certain procedural issues. Methods All new hires get a preplacement medical examination that includes TB skin testing. All newly hired employees with a documented positive Mantoux TB skin test were required to get baseline chest radiographs and were offered voluntary QFT-G as part of the clinical TB evaluation. Persons with a reported, but not documented, positive TB skin test were required to have either a TB skin test or a QFT-G test. Specimens for QFT-G were processed Monday through Friday, serum aliquots were stored at 6°C, and the ELISA QFT-G interferon-γ assay was performed weekly. All protocols followed the QFT-G package insert.5 Demographic data were abstracted from medical records and summarized in an excel spreadsheet. The UMDNJ Newark campus Institutional Review Board approved the review of the BATB clinical data on December 15, 2005. As a result of preliminary findings, an on-site review of the laboratory procedures was undertaken on January 5, 2006, with representatives from the manufacturer of QFT-G. Results From August 25 to December 31, 2005, 159 persons were tested with QFT-G; 136 (85%) were born outside the United States, and 123 (77%) reported receiving BCG with another 15 (9%) uncertain of their BCG history. Thirty-one individuals (19%) had a QFT-G that was positive. No individual had symptoms compatible with TB disease and all had normal physical examinations and chest radiographs. They were diagnosed with latent TB infection (LTBI) and referred for treatment. Ninety-six (60%) had a negative test and were counseled. Thirty-two (20%) had a QFT-G that was indeterminate; 19 were from high background responses, and 13 had low mitogen responses. They were counseled and offered a voluntary repeat test. Because of the unexpectedly high percentage of results that were indeterminate, laboratory procedures were reviewed with the manufacturer. The following procedural changes were instituted: 1) a new microplate shaker was purchased to enable more consistent and increased shaking of the initial processing plate; 2) the incubator for the assay was switched to one to be maintained at 37°C, an increase of 2°C; 3) technologists were retrained to hold the antigen droppers absolutely vertical for the most accurate drop volumes; and 4) the filter on the microplate washer was changed. After these changes, during the next 5 months, the indeterminate results fell two thirds to 7% of the next 86 tests. Because of our findings and concern regarding possible laboratory procedures, all 128 employees who had negative or indeterminate test results in 2005 were notified and offered voluntary retesting. Of 43 persons (34%) who volunteered for retesting after the change in procedures, five (12%) persons had a positive QFT-G result; initially, two were negative and three were indeterminate; 32 were negative, including 15 who were initially indeterminate; and six were indeterminate. One additional person who was initially positive requested retesting and was negative on retest. Comment We report on our initial use of a BATB in an urban academic medical center employee health service. Although the QFT-G kit includes controls for the interferon-γ ELISA portion of the assay, there is no quality control method of the critical initial stimulation steps of the QFT-G assay. Only our subjective determination that the level of indeterminate results appeared high alerted us to a possible technical problem. At present, hospitals and other clinical laboratories need to strictly adhere to the protocols for performing the QFT-G with special emphasis on the initial stimulation steps of the assay. Methods to ensure adequate quality control during these early steps of the protocol need to be developed. Alternatively, a newer version of the QFT-G test is anticipated that may simplify its handling. Finally, we chose to use the QFT-G as a confirmatory test to the TB skin test rather than as a primary test, because early published reports indicated that QFT-G, although more specific, might be less sensitive for the detection of LTBI than the TB skin test.6 However, this is in disagreement with the Centers for Disease Control and Prevention (CDC) guidelines, which were issued after we began. Although the CDC also notes that the QFT-G test may be less sensitive than the TB skin test for LTBI, it recommends that TB surveillance programs in healthcare facilities should choose either a BATB or TB skin testing for detecting LTBI.1 In addition, the CDC and the company that developed QFT-G state that BATB testing may be more cost-effective than the TB skin test for employee testing1 without evidence to support this assertion. The company refers to a report that solely looked at TB skin testing.7 However, before an institution switches to this assay for testing employees on a cost basis alone, it should do its own cost comparison. For example, the administrative advantages for BATB noted by the CDC during the preplacement evaluation1 may not be present during the annual surveillance process. Therefore, we believe that using the BATB as a secondary test to determine whether a positive TB skin test is due to BCG vaccination or TB infection was a prudent approach to using this new technology.8,9 Of course, more research is needed to determine the best approach to employee screening for TB. Lawrence D. Budnick, MD, MPH Michele Burday, PhD Gwen Brachman, MD, MPH Carolina T. Mangura, MD Donald DeBlock, RN Alfred Lardizabal, MD University of Medicine and Dentistry of New Jersey–New Jersey Medical School Newark, New Jersey