Lab tests and scans, not blood inflammation, separate TB and Covid-19

Tuberculosis (TB) and Covid-19 both attack the lungs, and doctors have worried that when they occur together they might cause worse disease. To explore this, a team led by Felipe Ridolfi performed a cross-sectional study of adults hospitalized with either pulmonary TB, Covid-19, or TB/Covid-19 co-infection. They compared basic patient information, clinical features, routine laboratory tests, and inflammatory blood markers across the three groups. The study included 76 people: 33 with pulmonary TB, 30 with Covid-19, and 13 with TB/Covid-19. Men were the majority in every group. People with TB or TB/Covid-19 had much lower body mass index (median BMI 18.4 and 19.4) than people with Covid-19 (median BMI 25.5). HIV infection was far more common in the TB and TB/Covid-19 groups (63% and 61%) than in the Covid-19 group (20%). Other comorbidities such as diabetes mellitus, chronic obstructive pulmonary disease, and hypertension were concentrated in the Covid-19 group. On chest CT scans, lung cavitation—an imaging sign typical of TB—was common in TB and TB/Covid-19 patients but absent in those with Covid-19.

To probe immune signals, the researchers measured inflammatory soluble factors using MILLIPLEX MAP Human Cytokine/Chemokine Premixed 29 Plex and Luminex Intelliflex xMAP. For statistics they used Chi-square tests for categorical variables and Mann-Whitney U or Kruskal-Wallis tests for continuous measures. A random forest model assessed which variables best separated the groups, using the Mean Decrease Gini Index (≥1.5) to rank importance. Lab differences stood out: participants with Covid-19 had lower platelet counts (p=0.004) and higher creatinine (p<0.001) and urea levels (p=0.004). The random forest identified urea, hemoglobin, hematocrit, and platelets as the best tests to discriminate TB/Covid-19 from the other groups, with Area Under the Curve>0.70. Among the measured cytokines, MCP-1 was significantly higher in Covid-19 than in TB and TB/Covid-19 (p=0.014). Importantly, no overall inflammatory signature in plasma distinguished TB/Covid-19 co-infection from the single infections.

The study shows that TB and Covid-19 can look clinically similar in hospitalized patients, but co-infection with both is relatively uncommon. Measuring a wide panel of inflammatory molecules in plasma did not provide a clear fingerprint to tell the conditions apart. Instead, basic clinical information—presence of comorbidities—radiologic findings such as cavitation on chest CT, and routine blood tests including urea, hemoglobin, hematocrit and platelet counts were the most useful discriminators. The high proportion of people living with HIV/AIDS among TB patients underscores that TB care often overlaps with HIV care. For clinicians, the findings suggest that existing, widely available tests and imaging may offer practical clues to distinguish TB, Covid-19 and co-infection while targeted inflammatory biomarkers like MCP-1 may have limited diagnostic value on their own. The results point to the need for further studies to refine tools for diagnosis and management of patients with overlapping respiratory infections.