Eye Tracking as a Biomarker of Cannabis Effects

During a piloting phase with our staff, the technology/eye-tracking device did not collect outcomes as expected.

Biomarkers of recent drug use and intoxication have societal relevance, in that they are used by law enforcement and other agencies to detect drug impairment. For instance, a breathalyzer can quickly and accurately detect blood alcohol content (BAC) to indicate if a person is under the influence of alcohol; however, there is currently no similar way to quickly detect if a person is under the influence of cannabis. In light of increasing cannabis use, it is important to define a quantitative, objective method of determining recent use and intoxication. The link between changes in eye characteristics (e.g. movement, pupil dilation) and cannabis use is documented (Peragallo et al. 2013), but insufficiently characterized. Certain outcomes of eye behavior are known to be affected by recent cannabis use (e.g. the eyes' ability to converge on a target; Stapleton et al 1986), while findings are mixed regarding other outcomes (e.g. the eyes' ability to smoothly follow a target; Fant et al. 1998). Thus, the goal of this study is to identify a characteristic pattern of eye behavior, defined by performance on a battery of four eye tasks, as a function of recent cannabis use (7% vs. 0% THC). Using 30 healthy cannabis users (15 men, 15 women), this study will be one of the first to assess changes in eye behavior as a function of recent cannabis use within a quantified virtual reality (VR) environment. This study will examine the effect of smoked cannabis (7% vs. 0% THC) on individual eye movements, with the goal of defining the utility of the eyes as potential objective indicators of cannabis use and intoxication. Four eye tests (nystagmus, smooth pursuit, convergence, and pupillary light response; outlined below), which previous literature has defined as effective in detecting recent drug use (including opioids and alcohol; Murillo et al. 2004), have been compiled into a 5-minute task battery using a VR headset environment equipped with high frequency infrared eye trackers (the HTC Vive with Pupil Labs Tracking). This 5-minute VR battery of four eye tests will be administered prior to cannabis consumption as a baseline, and then at 0, 15, 30, 45, 60, 75, 90, 105, 120, and 165 min after cannabis, with the goal of comparing baseline values to the ten post-cannabis timepoints to detect changes in eye behavior as a function of cannabis intoxication. The study will also utilize a battery of subjective-effects and mood visual analogue scales (0-100 mm; e.g. 'Good Drug Effect') prior to the eye test battery at each timepoint, allowing us to correlate each outcome of the eye tasks to subjectively reported cannabis impairment and mood. In addition to measuring eye behavior as a function of cannabis use, the training session of this study will be used to also collect exploratory data on the relationship between pupil dilation and experimental pain. Using Quantitative Sensory Testing (Medoc TSA-II NeuroSensory Analyzer), thermal pain threshold and tolerance will be induced using a cold stimulus (4.0°C; induced with a 30 x 30 mm Peltier thermode, which is 1.5" square metal applicator that is connected to the TSA-II NeuroSensory Analyzer device and software, and produces an ongoing cold sensation applied to the lower palm of the participant's non-dominant hand). Participants will indicate first feelings of pain (pain threshold), and when the pain becomes too much to bear (pain tolerance) by pressing a button on a controller connected to the TSA-II. Throughout exposure to the cold stimulus, changes in pupil size to the patient's subjectively reported pain latencies will be recorded.

Each volunteer will participate in two sessions in a cross-over fashion (active cannabis vs. inactive placebo cannabis)

In this session, the participant will smoke two-thirds of one active cannabis cigarette (7% THC) according to our paced-puff procedure (Foltin et al. 1987). They will complete an eye task battery (5 minutes per battery) 15 minutes prior to smoking as a baseline measure in each session, and again at the following timepoints: 0, 15, 30, 45, 60, 75, 90, 105, 120, and 165 minutes post-cannabis. Baseline assessments will be compared to those at post-cannabis timepoints.

In this session, the participant will smoke two-thirds of one inactive placebo cannabis cigarette (0% THC) according to our paced-puff procedure (Foltin et al. 1987). They will complete an eye task battery (5 minutes per battery) 15 minutes prior to smoking as a baseline measure in each session, and again at the following timepoints: 0, 15, 30, 45, 60, 75, 90, 105, 120, and 165 minutes post-cannabis. Baseline assessments will be compared to those at post-cannabis timepoints.

Using a Virtual Reality headset, a stimulus (a fixation cross) moves from 12 inches to 2 inches away from the participant's nose at a continuous speed over 5 seconds, then is held at that point for 10 seconds. The accuracy with which the participant completes this task will be measured by comparing pupil position with the target position within the virtual reality environment, down to the millimeter. Speed will also be assessed by measuring pupil position as a function of target fixation, down to the millisecond.

% Change from baseline during each of two sessions (measurements taken at baseline, 0, 15, 30, 45, 60, 75, 90, 105, 120, and 165 minutes post-cannabis)

Inclusion Criteria: Males/non-pregnant females Report smoking cannabis ≥1 day per week Able to perform all study procedures Exclusion Criteria: Meeting DSM-V criteria for severe substance use Use of illicit drugs ≥1 day/week in the prior 4 weeks Abnormality with the eyes which may affect the eye tracking technology such as color blindness, naturally occurring nystagmus, amblyopia, strabismus, age-related macular degeneration (AMD), cataract, diabetic eye disease, glaucoma, dry eye, extreme refractive error, bacterial or viral infections of the eye 7. User of supplemental oxygen