Does High-Dose Antimicrobial Chemotherapy Prevent the Evolution of Resistance?

Abstract

High-dose chemotherapy has long been advocated as a means of controlling drug resistance in infectious diseases but recent empirical studies have begun to challenge this view. We develop a very general framework for modeling and understanding resistance emergence based on principles from evolutionary biology. We use this framework to show how high-dose chemotherapy engenders opposing evolutionary processes involving the mutational input of resistant strains and their release from ecological competition. Whether such therapy provides the best approach for controlling resistance therefore depends on the relative strengths of these processes. These opposing processes typically lead to a unimodal relationship between drug pressure and resistance emergence. As a result, the optimal drug dose lies at either end of the therapeutic window of clinically acceptable concentrations. We illustrate our findings with a simple model that shows how a seemingly minor change in parameter values can alter the outcome from one where high-dose chemotherapy is optimal to one where using the smallest clinically effective dose is best. A review of the available empirical evidence provides broad support for these general conclusions. Our analysis opens up treatment options not currently considered as resistance management strategies, and it also simplifies the experiments required to determine the drug doses which best retard resistance emergence in patients.

Fig 1. Hypothetical plots of resistance hazard H(c) as a function of drug concentration c.

The lowest effective dose and the highest tolerable dose are denoted by c _L and c _U respectively. The therapeutic window is shown in green. (a) and (b) drug concentration with the smallest hazard is the lowest effective dose. (c) and (d) drug concentration with the smallest hazard is the highest tolerable dose.

Fig 2. Example where conventional strategy of high-dose chemotherapy best prevents the emergence of resistance.

(a) The dose-response curves for the wild type in blue (r(c) = 0.6(1−tanh(15(c−0.3)))) and the resistant strain in red (r _m(c) = 0.59(1−tanh(15(c−0.45)))) as well as the therapeutic window in green. Red dots indicate the probability of resistance emergence. Probability of resistance emergence is defined as the fraction of 5000 simulations for which resistance reached a density of at least 100 (and thus caused disease).(b) and (c) wild type density (blue), resistant density (red), and immune molecule density (black) during infection for 1000 representative realizations of a stochastic implementation of the model. (b) treatment at the smallest effective dose c _L, (c) treatment at the maximum tolerable dose c _U. Parameter values are P(0) = 10, P _m(0) = 0, I(0) = 2, α = 0.05, δ = 0.05, κ = 0.075, μ = 10⁻², and γ = 0.01.

Fig 3. Example where low-dose strategy best prevents the emergence of resistance.

(a) The dose-response curves for the wild type in blue (r(c) = 0.6(1−tanh(15(c−0.3)))) and the resistant strain in red (r _m(c) = 0.59(1−tanh(15(c−0.6)))) as well as the therapeutic window in green. Red dots indicate the probability of resistance emergence. Probability of resistance emergence is defined as the fraction of 5000 simulations for which resistance reached a density of at least 100 (and thus caused disease).(b) and (c) wild type density (blue), resistant density (red), and immune molecule density (black) during infection for 1000 representative realizations of a stochastic implementation of the model. (b) treatment at the smallest effective dose c _L, (c) treatment at the maximum tolerable dose c _U. (d) The probability that a resistant strain appears by mutation is indicated by grey bars for low and high dose. The probability of resistance emergence is indicated by the height of the red bars for these cases. The probability of resistance emergence, given a resistant strain appeared by mutation, can be interpreted as the ratio of the red to grey bars. Parameter values are P(0) = 10, P _m(0) = 0, I(0) = 2, α = 0.05, δ = 0.05, κ = 0.075, μ = 10⁻², and γ = 0.01.

Fig 4. Frequency distribution of resistant strain outbreak sizes for the simulation underlying Fig 3.

Each distribution is based on 5000 realizations of a stochastic implementation of the model. (a) Low drug dose. (b) High drug dose. Insets show the same distribution on a different vertical scale.