Accuracy of Tuberculosis Infection Diagnosis through IP-10-Based Assays for Immune Detection of Present Mycobacterium tuberculosis: A Cross-Sectional Evaluation

This cross-sectional study in Romania demonstrates that IP-10-based assays (RIDA®QUICK TB and RIDASCREEN® TB) exhibit high sensitivity and excellent positive predictive value for detecting active and latent Mycobacterium tuberculosis infections, often outperforming the standard QuantiFERON-TB Gold Plus assay, though larger studies are needed to fully establish their specificity.

The Gist

🕵️‍♂️ The Big Picture: Hunting for a Hidden Enemy

Imagine Tuberculosis (TB) is a sneaky burglar who can hide in your house (your lungs) for years without you knowing. Sometimes, the burglar is just sleeping (Latent TB), but sometimes, they are actively breaking things (Active TB).

The goal of this study was to test two new security cameras (the RIDA®QUICK and RIDA®SCREEN tests) to see if they are better at spotting this burglar than the old, trusted security system (the QFT-Plus test) that hospitals currently use.

🧪 The New vs. The Old

• The Old System (QFT-Plus): This test looks for a specific signal called IFN-γ. Think of this like a smoke detector that only goes off if there is a lot of smoke. It's reliable, but sometimes, if the fire is small or the smoke is thin, it might miss it.
• The New Systems (RIDA® Tests): These tests look for a different signal called IP-10. Think of IP-10 as a super-sensitive motion sensor. It picks up the burglar's footsteps even when there is no smoke yet. The researchers believe this new signal is produced in much higher amounts, making it easier to spot the infection early.

🏥 The Experiment: Who Did They Test?

The researchers gathered 99 people in Bucharest, Romania, and split them into three groups to see how the cameras performed:

1. The "Active Burglars" (49 people): People who were definitely sick with TB and had the bacteria growing in their sputum (a confirmed diagnosis).
2. The "Neighbors" (30 people): People who lived with or worked closely with someone who had TB. They might have caught the germ but didn't know it yet.
3. The "Sensitive Systems" (20 people): People with autoimmune diseases (where their immune system is a bit confused). This group was important because their immune systems sometimes act weirdly, which can trick old tests.

📊 The Results: Who Caught the Burglar?

When they compared the new cameras to the old one, here is what happened:

• In the "Active Burglar" group:

  • The Old System (QFT-Plus) missed about 3 out of every 10 sick people. It was like a smoke detector that failed to go off for small fires.
  • The New System (RIDA®SCREEN) caught almost everyone (92%).
  • The New System (RIDA®QUICK) also did very well, catching about 86% of the sick people.
  • The Takeaway: The new tests were much better at finding people who were definitely sick.

• In the "Neighbors" and "Sensitive Systems" groups:

  • The new tests agreed very well with the old tests for the neighbors.
  • For the people with autoimmune diseases, the new tests found more positives. The researchers think this is good! It suggests the new tests might be better at seeing the infection even when the immune system is confused or "noisy."

⚠️ A Catch in the Story (The Limitations)

There is one big "but" to this story. To prove a security camera is perfect, you need to test it on houses with burglars and houses with no burglars at all.

In this study, the group of "Active Burglars" was so full of sick people that there were almost no healthy people to compare them against. Because of this, the researchers couldn't perfectly calculate how often the new tests might cry "Wolf!" when there was no wolf (False Positives).

However, because the new tests found more of the confirmed sick people than the old test, the researchers are very optimistic.

💡 The Bottom Line

Think of the RIDA®QUICK and RIDA®SCREEN tests as high-definition, night-vision goggles compared to the old test's standard flashlight.

• The Flashlight (Old Test): Good, but sometimes misses things in the dark or when the signal is faint.
• The Night Vision (New Tests): Brighter, sharper, and catches the enemy even when they are hiding in the shadows or moving quietly.

Conclusion: The study suggests that these new IP-10 tests are excellent tools for finding TB, especially in people who are very sick or have tricky immune systems. They might help doctors catch the disease earlier and treat it before it gets worse. However, we need to test them on more healthy people to make sure they don't get too excited and flag healthy people as sick.

In short: The new tests are a promising upgrade for the medical toolkit, potentially saving lives by spotting the "burglar" sooner.

Technical Summary

1. Problem Statement

The diagnosis of Mycobacterium tuberculosis (Mtb) infection, particularly Latent TB Infection (LTBI), remains a significant challenge. Current standard diagnostics include the Tuberculin Skin Test (TST) and Interferon-Gamma Release Assays (IGRAs) like QuantiFERON-TB Gold Plus (QFT-Plus).

• Limitations of Current Methods: TST suffers from low specificity due to BCG vaccination and non-tuberculous mycobacteria (NTM). IGRAs, while more specific, can yield indeterminate or false-negative results in immunocompromised individuals, early-stage infection, or low-burden disease. They also rely solely on IFN-γ, which may have a lower dynamic range.
• The Gap: There is a need for biomarkers with higher sensitivity and stability. Interferon-gamma-inducible protein 10 (IP-10) is produced in substantially higher concentrations than IFN-γ and is induced by a broader range of cytokines, making it a promising candidate for improved detection. However, robust diagnostic accuracy data for commercial IP-10 assays (specifically RIDA®QUICK TB and RIDASCREEN® TB) against established standards like QFT-Plus and microbiological culture is limited.

2. Methodology

Study Design:

• Type: Cross-sectional diagnostic accuracy study.
• Setting: Marius Nasta Institute of Pneumophtiziology, Bucharest, Romania.
• Population: 99 adults (after one exclusion) divided into three groups:
  1. Active TB: 49 patients with culture-confirmed active pulmonary TB.
  2. TB Contacts: 30 close contacts of smear-positive TB cases (presumptive LTBI).
  3. Autoimmune: 20 patients with documented autoimmune diseases (evaluated for LTBI).
• Interventions: All participants underwent three tests:
  1. RIDA®QUICK TB: A lateral flow assay detecting IP-10.
  2. RIDASCREEN® TB: An ELISA detecting IP-10.
  3. QFT-Plus: The standard comparator IGRA (IFN-γ based).
• Reference Standards:
  • Primary: Mtb culture from sputum (for Active TB group).
  • Secondary: GeneXpert and smear microscopy.
  • Comparator: QFT-Plus was used as an immunological comparator, not a definitive gold standard for active disease.
• Handling Indeterminate Results: Indeterminate results were reclassified using a composite reference standard (culture, GeneXpert, microscopy) to minimize exclusions. Indeterminate results in confirmed TB were coded as positive; in contacts/autoimmune without active disease evidence, they were coded as negative.
• Statistical Analysis: Sensitivity, specificity, PPV, NPV, Kappa ($\kappa$) agreement, and ROC curve analysis were calculated. Notably, ROC analysis was performed against QFT-Plus to measure immunological concordance rather than diagnostic accuracy against culture, due to the lack of true-negative controls in the active TB group.

3. Key Contributions

• Comparative Evaluation: Provides one of the first direct comparisons of two distinct IP-10 platforms (Lateral Flow vs. ELISA) against QFT-Plus across three distinct clinical phenotypes (Active TB, Contacts, Autoimmune).
• Sensitivity Superiority: Demonstrates that IP-10 based assays detect a higher proportion of culture-confirmed active TB cases compared to QFT-Plus.
• Biological Insight: Highlights that IP-10 assays may capture immune activation missed by IFN-γ assays, particularly in early disease or low-bacillary burden scenarios, suggesting IP-10 is a more robust marker of antigen-specific cell-mediated immunity.
• Contextualizing Discordance: Clarifies that "false positives" in IP-10 assays (when compared to QFT-Plus or culture) may actually represent true biological immune responses in cases where culture is negative due to sampling issues or low bacterial load.

4. Key Results

A. Diagnostic Performance in Active TB (N=49, Culture-Positive)

• Sensitivity:
  • RIDA®SCREEN TB (ELISA): 91.8% (45/49 detected).
  • RIDA®QUICK TB (Lateral Flow): 85.7% (42/49 detected).
  • QFT-Plus: 71.4% (35/49 detected).
• Conclusion: Both IP-10 assays outperformed QFT-Plus in detecting culture-confirmed active TB.
• Limitation: Specificity and NPV could not be calculated for this subgroup as there were no culture-negative individuals (disease-enriched cohort).

B. Agreement with QFT-Plus

• Overall Cohort: Moderate to good agreement ($\kappa = 0.47 - 0.93$).
• Directional Disagreement: IP-10 assays consistently yielded more positive results than QFT-Plus.
  • In Active TB, RIDA®QUICK TB was positive in 8 cases where QFT-Plus was negative (McNemar $p=0.039$).
  • RIDA®SCREEN TB showed similar trends (17 vs 3 discordant pairs).
• TB Contacts: Agreement was almost perfect ($\kappa = 0.933$ for Quick, $0.798$ for Screen), indicating high concordance in lower-risk populations.
• Autoimmune Group: Agreement was moderate to poor ($\kappa = 0.50$ and $0.30$), likely due to immune dysregulation affecting cytokine release differently.

C. ROC Analysis

• ROC curves were generated against QFT-Plus (not culture) to assess immunological discrimination.
• RIDA®QUICK TB: AUC = 0.828 (Good discrimination).
• RIDASCREEN® TB: AUC = 0.767 (Good discrimination).
• TB Contacts Subgroup: Excellent discrimination (AUC = 0.971 for Quick).

D. Predictive Values

• PPV: Extremely high (97.7% - 97.8%) for both IP-10 assays against the bacteriological standard in the active TB group.
• NPV: Could not be reliably estimated in the active TB group due to the absence of true negatives.

5. Significance and Clinical Implications

• Enhanced Detection: IP-10 based assays (RIDA®QUICK TB and RIDASCREEN® TB) demonstrate superior sensitivity over QFT-Plus for detecting active TB and LTBI, potentially reducing missed diagnoses in high-risk or immunocompromised populations.
• Complementary Tool: The study supports the use of IP-10 assays not as replacements, but as complementary tools to existing IGRAs. Their ability to detect immune responses when IFN-γ assays are negative suggests they can identify early or low-burden infections.
• Operational Utility: The RIDA®QUICK TB (Lateral Flow) offers a rapid, point-of-care alternative with performance comparable to the ELISA (RIDA®SCREEN TB), which is significant for resource-limited settings.
• Future Directions: The authors emphasize the need for larger, heterogeneous cohorts (including pediatric, HIV+, and rural populations) to accurately define specificity and NPV. Longitudinal studies are required to determine if the increased positivity of IP-10 assays correlates with progression to active disease.

Conclusion: The study validates IP-10 as a highly sensitive biomarker for Mtb infection. The RIDA® assays show promise for improving early detection and managing TB in complex clinical scenarios where standard IFN-γ assays may fail.