Quantitative interferon gamma responses to Mycobacterium tuberculosis in a community-based survey of adolescents and adults in Blantyre, Malawi

This study of adolescents and adults in Blantyre, Malawi, demonstrates that analyzing quantitative interferon gamma release assay (IGRA) responses rather than binary results reveals critical age- and sex-specific variations in *Mycobacterium tuberculosis* immunoreactivity and transmission risks that are otherwise obscured, particularly highlighting the impact of HIV and the potential for refined public health strategies in high-burden settings.

The Gist

Imagine your immune system is a highly trained security team guarding a fortress (your body) against a sneaky intruder called Mycobacterium tuberculosis (TB bacteria). Usually, when this team spots the intruder, they sound a loud alarm.

For decades, doctors have used a test called the IGRA (specifically the QFT-Plus) to check if this alarm has ever been triggered. Think of this test like a volume meter for the alarm.

The Old Way: The "On/Off" Switch

Traditionally, doctors treated this volume meter like a simple light switch.

• If the volume is above a certain line (0.35 IU/mL): The light turns ON. You are "positive." You've seen the bacteria.
• If the volume is below the line: The light stays OFF. You are "negative." You haven't seen the bacteria.

The problem? This "On/Off" switch throws away a lot of useful information. It's like saying a car is either "moving" or "stopped," ignoring whether it's crawling at 1 mph or speeding at 100 mph. A low-level hum might mean you were exposed recently and the bacteria are fresh, while a loud roar might mean you were exposed years ago and your body has settled into a long-term defense.

The New Study: Listening to the Whole Symphony

This study, conducted in Blantyre, Malawi (a place where TB is very common), decided to stop just looking at the light switch. Instead, they listened to the entire volume of the alarm for nearly 3,000 teenagers and adults. They wanted to see if the loudness of the response told them something new about who was getting infected recently and who was at risk.

Here are the key discoveries, explained simply:

1. The "Volume" Doesn't Change Much with Age or Gender
You might think that as people get older, their immune "volume" gets louder because they've been exposed to TB more times over the years. Or that men might have a louder alarm than women.

• The Finding: Once the alarm is triggered (the light is ON), the volume is surprisingly similar for everyone, regardless of age or gender.
• The Analogy: It's like a group of people hearing a siren. Once they hear it, they all scream at roughly the same volume. The study suggests that age and gender determine whether you hear the siren (get infected), but not how loud you scream once you do.

2. The "HIV Twist"
The study looked at people living with HIV. Usually, HIV weakens the immune system, making it harder to fight infections.

• The Finding: People with HIV were less likely to show a specific type of "loudness" pattern (where the TB2 tube is much louder than the TB1 tube) that scientists thought indicated a very recent infection.
• The Analogy: Imagine a security team that is tired or understaffed (due to HIV). Even if a burglar just broke in, the tired team might not scream as loudly or as distinctly as a fresh, well-rested team. This makes it harder to spot new infections in this group using this specific "volume" trick.

3. The "Threshold" Problem
The study played a game: "What if we lowered the line for turning the light ON?"

• The Finding: If they lowered the threshold (made the test more sensitive), they found many more people with low-level alarms. Crucially, this changed the picture of who was getting infected. At the standard "high" threshold, men and women looked similar. But at a "lower" threshold, it became clear that men were getting infected slightly earlier in their teenage years than women.
• The Analogy: Imagine a fishing net with big holes. You only catch big fish (high immune responses). If you switch to a net with tiny holes, you catch many more small fish (low immune responses). Suddenly, you realize you were missing a whole school of small fish that were actually very important.

Why Does This Matter?

This study is like upgrading from a black-and-white security camera to a high-definition, color camera with audio.

• Old View: "Someone has TB exposure. Someone doesn't."
• New View: "We can see how the exposure is happening. We can see that men might be getting infected slightly earlier than women, and that HIV changes how the immune system reacts to new infections."

The Bottom Line:
By listening to the volume of the immune response instead of just checking if the light is on, scientists can get a much clearer map of how TB is spreading in a community. This could help health workers target their vaccines and treatments more precisely, catching the "fresh" infections before they turn into full-blown disease, rather than just guessing based on a simple yes/no test.

Technical Summary

1. Problem Statement

Current diagnostic approaches for Mycobacterium tuberculosis (Mtb) infection rely on binary classification (positive/negative) of Interferon-Gamma Release Assays (IGRAs), such as the QuantiFERON-TB Gold Plus (QFT-Plus). This dichotomization, typically using a fixed threshold of 0.35 IU/mL, discards valuable quantitative data regarding the magnitude of the immune response.

• Limitations: Binary thresholds cannot distinguish between recent and remote infections, nor do they fully capture the spectrum of exposure intensity.
• Knowledge Gap: While higher IGRA magnitudes have been correlated with recent exposure in contact studies, there is a lack of data on how quantitative IGRA responses are distributed in a general, high-burden community population. Furthermore, it is unclear how alternative positivity thresholds might alter inferences about age- and sex-specific transmission dynamics.

2. Methodology

Study Design and Setting:

• Type: Secondary analysis of a community-based cross-sectional survey.
• Location: Blantyre City, Malawi (a high TB burden urban setting).
• Population: 2,895 participants aged 10–40 years, recruited from randomly selected households in 33 high-burden neighborhoods.
• Data Collection: QFT-Plus assays were performed on venous blood samples. The study utilized the manufacturer's protocol, measuring IFN-γ release in TB1 (CD4+ specific) and TB2 (CD4+ and CD8+ specific) tubes.

Statistical Analysis:

• Data Processing: Nil-subtracted values were calculated for TB1 and TB2. Values exceeding the assay limit (10 IU/mL) were capped at 10.
• Modeling Approach:
  • Hurdle Lognormal Models: Used to analyze response magnitude. This two-part model estimated:
    1. The probability of a response falling below the manufacturer's threshold (<0.35 IU/mL) (the "hurdle").
    2. The mean and standard deviation of log-transformed responses for those above the threshold (the "lognormal" component).
  • Logistic Regression: Used to estimate the odds of a TB2-TB1 differential >0.6 IU/mL (a proposed marker for recent exposure).
  • Alternative Threshold Analysis: Researchers reclassified positivity across hypothetical thresholds (0.1 to 0.5 IU/mL) to model age- and sex-specific immunoreactivity prevalence and the Annual Risk of Immunoreactivity Conversion (ARIC).
• Software: Bayesian regression models were fitted using the brms package in R (v4.3.3) with Hamiltonian Monte Carlo sampling.

3. Key Contributions

• Quantitative Distribution in High-Burden Settings: Provides the first detailed characterization of quantitative QFT-Plus responses in a general, community-based population in a high-burden setting, moving beyond contact-tracing cohorts.
• Re-evaluation of the TB2-TB1 Differential: Investigates the utility of the TB2-TB1 >0.6 IU/mL differential as a marker for recent infection in a general population, finding limited utility for distinguishing recent vs. remote infection in this context.
• Threshold Sensitivity Analysis: Demonstrates how arbitrary changes in positivity thresholds significantly alter epidemiological inferences regarding sex-specific transmission patterns and the timing of divergence in infection risk between males and females.
• Methodological Framework: Establishes a Bayesian hurdle modeling approach for analyzing IGRA data that preserves continuous information rather than discarding it via binary classification.

4. Key Results

Prevalence and Demographics:

• Overall Positivity: 17.4% (503/2,895) of participants were QFT-Plus positive (TB1 or TB2 ≥0.35 IU/mL).
• Age/Sex Trends: The probability of a positive result increased with age and diverged by sex from late adolescence, with males showing higher prevalence in older age groups.

Response Magnitude:

• Distribution: Quantitative responses were highly right-skewed.
• Associations: Among participants with positive responses, the magnitude of TB1 and TB2 responses was not significantly associated with age, sex, HIV status, previous TB treatment, or household TB contact. This suggests that cumulative exposure determines whether a response is elicited, but not necessarily its strength once established.

TB2-TB1 Differential (Recent Exposure Marker):

• Prevalence: Only 3.8% (109/2,895) of participants had a TB2-TB1 differential >0.6 IU/mL.
• Associations: This differential was not associated with age, sex, or household contact.
• HIV Impact: Participants with HIV had significantly reduced odds of having a TB2-TB1 >0.6 IU/mL differential (Adjusted OR: 0.37, 95% CrI: 0.14–0.93), likely due to HIV-associated immune suppression affecting CD8+ T-cell responses (measured by TB2).

Impact of Alternative Thresholds:

• Prevalence Estimates: Lowering the threshold to 0.1 IU/mL increased estimated immunoreactivity prevalence by 8.0 percentage points.
• Sex-Specific Trends: At higher thresholds (e.g., 0.5 IU/mL), the predicted male-to-female prevalence ratio crossed 1.0 (indicating higher male prevalence) at an earlier age (around 19 years) compared to lower thresholds. This suggests that binary thresholds may obscure the true timing of when males begin to experience higher transmission risks than females.

5. Significance and Implications

• Beyond Binary Classification: The study argues that dichotomizing IGRA results obscures critical epidemiological data. Quantitative responses offer a more nuanced view of transmission dynamics, particularly regarding age and sex gradients.
• Public Health Strategy: In high-burden settings, relying solely on the 0.35 IU/mL threshold may exclude individuals with true, albeit lower-magnitude, Mtb sensitization who are at risk of progression or sustaining transmission.
• Targeted Interventions: The findings suggest that quantitative data could help refine targeting for post-infection vaccines and preventive therapy, particularly for young males who show earlier divergence in transmission risk at higher response thresholds.
• Limitations: The cross-sectional design prevents causal inference regarding recent infection vs. disease progression. Additionally, the lack of a gold standard for latent TB infection limits the ability to validate the quantitative thresholds against true infection status.

Conclusion:
Quantitative IGRA responses contain rich epidemiological information that is lost when data is reduced to binary outcomes. In high-burden settings like Malawi, leveraging this continuous data can clarify transmission patterns, improve the understanding of sex-specific risks, and guide more effective, targeted public health interventions to accelerate TB elimination.