TL;DR
This paper introduces a novel mathematical framework to explain how "preexisting resistance" (drug resistance that's already present before treatment) can persist and even thrive. They demonstrate that frequency-dependent ecological interactions between resistant and sensitive cells, rather than just intrinsic fitness, are crucial for this phenomenon, with implications for cancer therapy and other evolutionary diseases.
What Problem Does This Paper Address?
The persistence of drug-resistant strains before treatment begins—known as preexisting resistance—is a major challenge in medicine, from infectious diseases to cancer. Typically, these resistant strains have a "fitness cost," meaning they perform worse than sensitive strains in the absence of the drug, so we'd expect them to be naturally selected against and disappear. However, they are frequently observed. Previous explanations have included spatial effects or simple mutation rates. This paper tackles the question of how these costly resistant mutants can survive and become prevalent enough to cause treatment failure, proposing that the interactions between the resistant and sensitive cells themselves play a critical, underappreciated role.
Methods
The authors employ a multi-pronged approach:
- Generalized Moran Process: This is a standard model for population genetics that they adapt to include "ecological interactions." They define growth rates that depend on the frequency of ancestor and mutant cells, allowing for positive (mutants benefit from ancestors) or negative (mutants are harmed by ancestors) interactions. This allows for analytical solutions for extinction times.
- Wright-Fisher Simulations: To handle more realistic population sizes and scenarios (like multiple mutants), they use Wright-Fisher simulations. This is a common method for modeling population evolution over generations.
- Analytical Approximations: They derive mathematical formulas (including the Lambert W function) to approximate key quantities like extinction times and the number of mutants, which are crucial for understanding the dynamics across a wide range of parameters.
- Experimental Validation: They genetically engineer common drug-resistance mutations found in non-small-cell lung cancer (NSCLC) and test their growth dynamics against the ancestral cell line in coculture experiments.
Datasets
Not explicitly detailed as "datasets" in the traditional sense, but the study relies on:
- PC-9 human adenocarcinoma cells: Used as the ancestral cell line.
- Genetically engineered NSCLC cell lines: Created to express specific, clinically observed drug-resistance mutations (BRAF-V600E, KRAS-G12V, PIK3CA-E545K).
- Previously characterized evolved resistant NSCLC cell line: Used as a basis for the study.
Key Insights and Ideas
- Ecological Interactions are Key: The most striking finding is that the interactions between resistant and sensitive cells (frequency-dependent selection) are far more important for the persistence of preexisting resistance than previously appreciated.
- Positive Interactions Save Mutants: Even a weak positive ecological interaction can dramatically increase the extinction time of a resistant mutant, making it much more likely to survive until treatment.
- "Rare Mutant Regime" is Critical: Even when the overall mutation rate is low, positive ecological interactions can allow rare mutants to persist and become significant threats.
- Distribution Matters: Ecological interactions don't just affect if a mutant survives, but also which mutants become prevalent. Those with stronger positive interactions are disproportionately represented.
- Real-World Mutations Exhibit Positive Interactions: The experimental validation with engineered cancer cell lines confirms that clinically relevant resistance mutations indeed exhibit positive ecological interactions with their ancestors.
Key Results
- Extinction Times Vary Wildly: Mean extinction times for mutants with the same intrinsic fitness can vary by five orders of magnitude depending on their ecological fitness.
- Expanded Parameter Space for Preexistence: The authors identify a significant parameter space where preexisting resistance is likely only if ecological interactions are considered.
- Positive Interactions Dominate Stationary Distribution: Even when positive ecological interactions are rare, they become overrepresented in the population over time, dominating the distribution of resistant mutants.
- Experimental Confirmation: All three engineered clinically relevant NSCLC resistance mutations showed positive ecological interactions with their ancestral cell line.
Limitations and Future Work
- Distribution of Ecological Effects (DEE) is Unknown: The authors acknowledge that the actual distribution of ecological effects in real populations is not yet experimentally measured, and their models make assumptions (e.g., uniform distribution). Future work should focus on measuring the DEE.
- Model Simplifications: The models simplify complex biological systems. For instance, they don't explicitly account for the immune system, spatial dynamics, or treatment adherence, which can all play roles in real-world scenarios.
- Experimental Scope: The experimental validation was limited to one cell line system (NSCLC). Future work should extend these tests to other cancers, infectious diseases (like HIV), and bacteria.
- Generalizability: While the principles are expected to apply broadly, further testing across diverse model systems is needed to confirm the universality of these findings.
Summary updated