Solvents, Ethanol, Car Crashes & Tolerance
How risky is inhalation of organic solvents?
Acute Effects and the Dose Metric
Our work with rats involved assessing their behavior or visual function while they inhaled a solvent. In these studies, a vapor is generated by evaporating the liquid solvent, diluting the vapor in clean air, and passing it through a chamber in which a rat is either working for food (see Figure 2, top) or viewing a video screen on which visual stimuli are presented (see Figure 2, bottom). For the behavioral studies, the rats are trained to press levers for food before they are exposed to any solvents. In these hour-long tests, a series of trials is presented in which a light flashes briefly during half of the trials (a “signal”). The lever that produces food depends on whether or not a signal occurred, and because the occurrence of the signal is unpredictable, we posit that the rat must attend to that light to respond accurately. Performance is quantified by counting the numbers of correct and incorrect responses (accuracy), and the time between insertion of the levers in each trial and the lever press (response time). An analogous procedure for human subjects has shown remarkable consistency between the two species across several important task parameters.
For electrophysiological studies of visual function, rats are first prepared surgically with electrodes on the surface of the skull. These electrodes detect the electrical activity over the visual cortex, which is driven by alternating patterns of visual stimuli on the video screen. Cortical neurons generate electrical signals that can be used to quantify visual function. Similar functions can be obtained from humans wearing scalp electrodes.
The effects of solvents on behavior and visual function are reliable, robust and directly related to the amount of the solvent inhaled. Behaviorally, accuracy decreases and response time increases with increasing exposure; visually, the amplitude of the recorded signal decreases with increasing exposure. At very high concentrations, animals become anesthetized; indeed, some solvents have historically been used as surgical anesthetics. These effects reflect steps along a general pathway often characterized as “narcosis.”
Of course, the amount of the solvent entering the brain depends on its concentration in the air and the duration of the exposure. In addition, this “internal dose” of the chemical is a complex function of the individual’s physiology (for example, breathing rate and cardiac output) and the solvent’s physicochemical properties. Computational methods have been developed to estimate internal doses of chemicals under a wide range of exposure conditions. Such physiologically based pharmacokinetic (PBPK) models incorporate these parameters into a set of differential equations representing the flux of the compound in blood and tissues. PBPK models for several solvents, including trichloroethylene and toluene, have been essential for understanding the effects of solvents on the CNS and the implications of inhaled solvents on public health.
Using the toluene PBPK model for rats, we examined the relation between exposure concentration (C), duration of exposure (t), and effects on signal detection behavior and visual function to determine the measure of dose (“dose metric”) that best explains the observed pattern of effects of several solvents, including toluene. This analysis revealed clearly that the concentration of the solvent in the brain at the time of measurement provided an accurate estimate of the magnitude of its effects on performance of the task, in comparison to the concentration and duration of exposure and two metrics of cumulative dose. The results for response time in the behavioral test are shown in Figure 3.
The left most panel of Figure 3 shows that response time remains almost constant when rats perform the test in clean air, but it rises with increasing C and t of exposure to toluene. Because both C and t affect performance, both variables must be known to determine the magnitude of the increase. Because the cumulative dose determines the effect of some chemicals, we plotted the same response-time data against the total amount of toluene inhaled at each time point (C x t) and the total amount in the brain (estimated from the PBPK model) to see whether these metrics improved prediction of the effect. These single metrics did improve the predictions (middle panels of Figure 3), but neither metric predicted the effect as accurately as did the concentration of toluene in the brain at the time the behavior was measured (right most panel of Figure 3). Here it is clear that this single metric provides an accurate and unambiguous estimate of the magnitude of the slowing of the rats’ responses. Similarly, the momentary concentration of the solvent in the brain also best determines its effects on visual function.
What are the implications of this relation between the concentration of a solvent in the brain of a rat and its effects on these esoteric measures of CNS function? Certainly the effects are large and reproducible, and may be relevant to Darel’s neuropathy and Lars’s confusion, but can they address Marcie’s worry about her exposure to gasoline in the car driving home? The rats were exposed to very high concentrations of solvents, whereas Marcie’s exposure was orders of magnitude lower. To explore these questions, my colleagues and I examined published data from other experiments with laboratory animals and human subjects. We wanted to find out how general these effects are, and whether similar dose-effect relations could reveal important information about short-term exposure to solvents at concentrations of concern to Marcie.