Smear-negative TB: C-reactive protein may provide useful screening method

Screening HIV-positive people with smear-negative pulmonary TB for high levels of C-reactive protein (CRP) can detect the presence of active TB with a fairly high degree of accuracy, suggesting that C-reactive protein could provide the basis for a point-of-care test to detect active TB in smear-negative cases in high-burden settings, according to findings from a study presented at the Fourth South African AIDS Conference in Durban in early April.

Smear microscopy is the first line of diagnosis in people with suspected TB. Sputum is added to a slide, stained with a dye that shows up TB bacilli and then viewed under a microscope to determine if TB bacilli are present.

However, in people with HIV, pulmonary infection with TB is more likely to produce a smear-negative result, resulting in delayed treatment while further diagnostic tests are carried out, or – in some cases – a complete failure to treat active TB.

Both lead to increased mortality in HIV-positive people.

A simple point-of-care diagnostic test to overcome these problems is urgently needed.

C-reactive protein, a marker of inflammation that is easily measured in a blood sample, is elevated in untreated smear-negative pulmonary TB.

Previous studies in HIV-negative patients suggest that C-reactive protein is more likely to be elevated where more serious tissue damage has occurred in the lungs as a result of TB.

In order to investigate the relationship between CRP levels and the presence of active TB in smear-negative individuals with suspected pulmonary tuberculosis, Douglas Wilson from the University of KwaZulu-Natal and colleagues conducted a sub-group analysis on a prospective cohort of people with suspected smear-negative tuberculosis, recruited from primary health care clinics in the Edendale Hospital catchment area between 2005 and 2007.

For each patient recruited into the study, mycobacterial culture was performed on induced sputum and other clinically relevant material. The two groups were divided into those with confirmed pulmonary TB and those who had pulmonary TB excluded.

Inclusion criteria for both groups were that individuals had to be HIV-positive or have clinical evidence of HIV infection, have been coughing for more than two weeks and have two AFB-negative sputum smears, as well as having been evaluated by a primary health care clinician, including with a chest X-ray.

For the confirmed pulmonary TB group, laboratory evidence of TB through culture testing was necessary and patients would be initiated on TB treatment. For the pulmonary TB-excluded group, all laboratory evidence would have to confirm the exclusion of pulmonary TB.

Of the 504 people with suspected TB who were screened, 421 were enrolled into the cohort. Of these:

• 105 patients (24.9%) were confirmed with smear-negative pulmonary TB
• 102 patients were confirmed culture-positive
• two patients tested AFB smear-positive
• one patient was lymph-node histology positive for TB
• 46 patients (10.9%) were diagnosed with smear-negative pulmonary TB
• 67 patients (63.8%) received an oral antibiotic.

At baseline, 88% of the pulmonary TB cases were experiencing night sweats and 89% experienced severe weight loss, compared to 61% experiencing night sweats and 76% severe weight loss in the pulmonary-TB-excluded group.

The median C-reactive protein level in the confirmed pulmonary TB group was significantly higher than in the pulmonary-TB-excluded group (86.5 mg/L (95%: CI 47.7 to 126) in the PTB group versus 5.5 (95%: CI 2.9 to 31.1) in the excluded group). Elevated C-reactive protein was found to have 79% sensitivity and 85% specificity for detecting pulmonary TB, indicating that the test would miss active TB in around one in five people and wrongly diagnose a person as having active TB in about one in seven cases.

While the findings have positive implications for the diagnosis of pulmonary TB in HIV-positive populations, as this study is a sub-analysis which needs validation, more research into the topic is necessary. Further research also needs to be conducted on diagnosing extrapulmonary TB in HIV-positive people.