🛡️ immunology

A mechanistic basis for CD8+ T cell expansion sensitivity as a predictor of HIV post-treatment control

This study uses mechanistic modeling to demonstrate that the expansion sensitivity of CD8+ T cells provides the underlying biological basis for why high levels of Ki-67+ and TCF-1+ cells predict successful post-treatment control of HIV by establishing lower viral set points.

Original authors: Phan, T., Pagane, N., Kreig, J. A. F., Marc, A., Locke, M., Peluso, M. J., Sandel, D. A., Deitchman, A. N., Rutishauser, R. L., Deeks, S. G., Ke, R., Ribeiro, R. M., Perelson, A. S.

Published 2026-09-01

📖 5 min read🧠 Deep dive

CC0 1.0

Original authors: Phan, T., Pagane, N., Kreig, J. A. F., Marc, A., Locke, M., Peluso, M. J., Sandel, D. A., Deitchman, A. N., Rutishauser, R. L., Deeks, S. G., Ke, R., Ribeiro, R. M., Perelson, A. S.

Original paper dedicated to the public domain under CC0 1.0 (https://creativecommons.org/publicdomain/zero/1.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

For decades, the goal of HIV research has been to find a way to stop the virus from returning after a person stops taking their daily medication. When treatment is paused, the virus usually wakes up from its hiding places in the body and multiplies rapidly, forcing the person back onto medication. However, a small number of people, known as post-treatment controllers, are able to keep the virus in check on their own for months or even years after stopping therapy. Scientists have long suspected that the key to this natural control lies in the immune system, specifically in a type of white blood cell called the CD8+ T cell, which acts as a hunter to find and destroy infected cells. Recent studies have identified two specific markers on these cells that seem to appear more often in people who successfully control the virus: one marker, called Ki-67, indicates that the cells are actively dividing and growing, while another, called TCF-1, suggests the cells have a "stem-like" quality that allows them to persist and respond effectively over time. The mystery has been why these markers matter and how exactly they help the body fight the virus.

A team of researchers at Los Alamos National Laboratory and other institutions decided to use mathematical modeling to uncover the hidden mechanics behind these observations. They focused on data from a recent clinical trial where ten people with HIV received a combination of immunotherapies before stopping their medication. Nine of these participants experienced a return of the virus, but six of them managed to keep the viral levels low, while the other three saw the virus surge to high levels. The researchers built a computer simulation of the battle between the virus and the immune system inside the human body. This simulation tracked how the virus spreads, how infected cells die, and how the immune system's CD8+ T cells expand to fight back. Crucially, they did not use the Ki-67 or TCF-1 measurements to build or tune their model; they only fed the model the raw data on viral load and total CD8+ T cell counts. The model's job was to figure out what internal rules of the immune system would produce the patterns seen in the patients.

The simulation revealed a single, powerful factor that separated the successful controllers from those who could not control the virus. This factor was the "expansion sensitivity" of the CD8+ T cells. In simple terms, this measures how quickly the immune system's soldiers can start multiplying when they detect even a tiny amount of virus. The researchers found that in the people who controlled the virus, their immune cells were incredibly sensitive; they began to expand and multiply almost immediately upon detecting the virus, even when the viral load was very low. In contrast, the people who could not control the virus had immune cells that were much slower to react, requiring a much larger amount of virus to trigger the same response. This difference in sensitivity was so distinct that it completely separated the two groups in the model. The model showed that this rapid, early response allowed the controllers to keep the number of infected cells and the amount of virus in the blood much lower than in the non-controllers.

The most striking discovery came when the researchers compared the model's findings with the real-world measurements they had not used. They found that the "expansion sensitivity" calculated by the computer matched perfectly with the biological markers observed in the patients. Specifically, the people whose immune cells showed the highest sensitivity were the same people who had the highest levels of Ki-67 and TCF-1 at the moment the virus first started to rebound. This provided a clear, mechanistic explanation for the earlier clinical observations: the presence of these markers was not just a coincidence, but a sign that the immune system possessed the specific ability to react quickly to low levels of virus. The study suggests that the "stem-like" nature of the TCF-1 positive cells allows them to respond to the virus with a lower threshold, meaning they do not need to wait for a massive infection to begin their defense. This rapid reaction is what keeps the viral load low and prevents the virus from taking over.

The researchers also explored whether the size of the hidden virus reservoir was the main reason for the different outcomes, but their model suggested otherwise. They found that the initial size of the reservoir did not significantly change the final viral levels once the virus started to rebound. Instead, the outcome was determined almost entirely by how the immune system responded to the virus as it emerged. This finding shifts the focus from simply trying to reduce the size of the reservoir to enhancing the quality of the immune response. The study indicates that the immunotherapies used in the trial may have helped by priming the immune system to create a larger pool of these highly sensitive, stem-like cells. While the study was limited to a small group of participants and lacked a control group, the mathematical consistency of the results offers a strong, testable explanation for why some people can control HIV after treatment stops. It transforms the understanding of these immune markers from simple correlations into a causal mechanism, suggesting that the ability of CD8+ T cells to expand rapidly and sensitively is the key to keeping the virus in check.

Technical Summary: A Mechanistic Basis for CD8+ T Cell Expansion Sensitivity as a Predictor of HIV Post-Treatment Control

Problem Statement
A primary objective in HIV-1 cure research is understanding why certain individuals, known as post-treatment controllers (PTCs), maintain viral suppression after the interruption of antiretroviral therapy (ART), while others experience rapid viral rebound. Recent clinical studies, specifically the UCSF-amfAR combination immunotherapy trial, identified that the frequency of non-naïve CD8+ T cells in the cell cycle (Ki-67+) co-expressing the transcription factor TCF-1 at the time of viral rebound correlates with post-intervention control. However, the mechanistic link between these immunological markers and the ability to control viral rebound remains unclear. Furthermore, while mathematical models have previously identified "effector cell expansion sensitivity" as a determinant of control, they lacked a connection to measurable clinical biomarkers.

Methodology
The authors employed a mechanistic within-host mathematical modeling framework based on the Conway-Perelson model to analyze longitudinal viral load and CD8+ T cell data from nine participants in the UCSF-amfAR trial following ART interruption.

Data Source: The study utilized data from 10 participants, excluding one who remained below the limit of quantification. The cohort included three non-controllers (NCs) with rapid, high-level rebound and six post-intervention controllers (PICs) who maintained low viral set points.
Model Structure: The model tracks uninfected target cells ( $T$ ), latently infected cells ( $L$ ), productively infected cells ( $I$ ), free virus ( $V$ ), and effector cells ( $E$ ). A key feature is the effector cell expansion term, modeled as $\frac{bEI}{K_B + I}$ , where $K_B$ represents the infected-cell density required to reach half-maximal expansion. The inverse of this parameter, $1/K_B$ , is defined as the effector cell expansion sensitivity.
Fitting Strategy: Using a non-linear mixed-effects modeling framework (Monolix), the authors simultaneously fitted the model to viral load and total CD8+ T cell dynamics. The model incorporated a parameter $t_a$ to account for the stochastic time to rebound and absorbed residual drug effects into kinetic parameters. Crucially, the model was fitted only to viral and CD8+ T cell dynamics; Ki-67 and TCF-1 measurements were reserved for post-hoc validation and were not used in the fitting process.
Analysis: The authors compared fitted parameters between PICs and NCs and performed analytical derivations to establish the relationship between expansion sensitivity and the viral set point.

Key Results

Expansion Sensitivity Distinguishes Controllers: The parameter $K_B$ (and its inverse, expansion sensitivity) was the single most significant differentiator between PICs and NCs. PICs exhibited a $K_B$ value over an order of magnitude lower than NCs, indicating their effector cells are far more sensitive to low levels of infected cells.
Correlation with Clinical Biomarkers: The model-inferred expansion sensitivity ( $1/K_B$ $1/ K_{B}$ ) showed a strong linear correlation with clinical measurements taken at the first post-rebound time point (post-R1), despite these measurements not being used for fitting:
- Positive correlation with the percentage of Ki-67+ non-naïve CD8+ T cells ( $r \approx 0.8$ ).
- Positive correlation with the percentage of TCF-1+ cells within the Ki-67+ population ( $r \approx 0.8$ ).
Mechanism of Control: Analytical results demonstrated that the post-rebound viral set point is inversely proportional to the expansion sensitivity. Individuals with higher sensitivity (lower $K_B$ ) initiate effector expansion at lower antigen burdens. This rapid early response allows the effector population to curb infected cell growth more effectively, resulting in a significantly lower viral set point.
TCF-1 as a Stronger Predictor: When the viral set point was defined as the average viral load over the final two weeks of observation (rather than averaging all post-peak data), the correlation between the viral set point and the %TCF-1+ fraction within responding CD8+ T cells was exceptionally strong ( $r \approx -0.9$ ), suggesting TCF-1 is a potent predictor of long-term control.
Reservoir Size: The study found that the initial size of the latent reservoir ( $L_0$ ) did not significantly influence the post-rebound set point in this cohort, reinforcing that immune-mediated containment is the critical factor for sustained control once rebound occurs.

Significance and Claims
The paper claims to provide a mechanistic explanation for the clinical observation that Ki-67+ and TCF-1+ CD8+ T cells predict post-treatment control. The authors argue that these markers are not merely correlative but serve as proxies for a specific immune capability: effector cell expansion sensitivity.

Reframing "Capacity": The authors propose reframing the concept of "expansion capacity" to focus on "expansion sensitivity"—specifically, the threshold at which effector cells begin to respond to antigen.
Biomarker Validation: The study links the model-derived parameter $K_B$ directly to measurable immunological markers (Ki-67 and TCF-1), validating these markers as early indicators of immunotherapy success.
Generalizability: The authors suggest that this mechanism, previously identified in non-intervention settings, likely generalizes across different rebound contexts, including combination immunotherapy trials.
Therapeutic Implication: The findings elevate TCF-1 and Ki-67 from correlative biomarkers to potential therapeutic targets, suggesting that interventions should aim to increase the pool of stem-like, highly responsive CD8+ T cells to lower the antigen threshold required for effective viral control.

The study concludes that the ability of CD8+ T cells to undergo rapid, robust expansion in response to low antigen levels is the critical determinant of HIV remission, providing a quantitative basis for future cure strategies.

More like this