Deleterious Effects of Cold Air Inhalation on Coronary Physiological Indices in Patients With Obstructive Coronary Artery Disease

Background Cold air inhalation during exercise increases cardiac mortality, but the pathophysiology is unclear. During cold and exercise, dual‐sensor intracoronary wires measured coronary microvascular resistance (MVR) and blood flow velocity (CBF), and cardiac magnetic resonance measured subendocardial perfusion. Methods and Results Forty‐two patients (62±9 years) undergoing cardiac catheterization, 32 with obstructive coronary stenoses and 10 without, performed either (1) 5 minutes of cold air inhalation (5°F) or (2) two 5‐minute supine‐cycling periods: 1 at room temperature and 1 during cold air inhalation (5°F) (randomized order). We compared rest and peak stress MVR, CBF, and subendocardial perfusion measurements. In patients with unobstructed coronary arteries (n=10), cold air inhalation at rest decreased MVR by 6% (P=0.41), increasing CBF by 20% (P<0.01). However, in patients with obstructive stenoses (n=10), cold air inhalation at rest increased MVR by 17% (P<0.01), reducing CBF by 3% (P=0.85). Consequently, in patients with obstructive stenoses undergoing the cardiac magnetic resonance protocol (n=10), cold air inhalation reduced subendocardial perfusion (P<0.05). Only patients with obstructive stenoses performed this protocol (n=12). Cycling at room temperature decreased MVR by 29% (P<0.001) and increased CBF by 61% (P<0.001). However, cold air inhalation during cycling blunted these adaptations in MVR (P=0.12) and CBF (P<0.05), an effect attributable to defective early diastolic CBF acceleration (P<0.05) and associated with greater ST‐segment depression (P<0.05). Conclusions In patients with obstructive coronary stenoses, cold air inhalation causes deleterious changes in MVR and CBF. These diminish or abolish the normal adaptations during exertion that ordinarily match myocardial blood supply to demand.

curves using a Fermi-function constrained deconvolution algorithm using Matlaboratory (MathWorks Inc., Natick, Massachusetts) as described by Jerosch-Herold et al. 2 This generated absolute baseline and stress myocardial blood flow values (ml/g/min) for all 6 segments of the mid-ventricular slice.
Transmural perfusion gradients (TPG) were obtained using dedicated software (EasyScil prototyping, Philips Medical Systems, Best, Netherlands). Using the same contours delineated for deconvolution analysis, subendocardial, Iendo(α,t), and subepicardial signal-intensity curves, Iepi(α,t), were obtained. Ten myocardial layers and sixty radial segments were sampled and a signal-intensity curve generated at each point. TPG curves G(a,t) were calculated based on the difference in subendocardial and subepicardial signal-intensity values over time, t (each dynamic), at a particular myocardial location, α, and normalized to transmural signal-intensity values, Itransm(α,t). 3,4 Peak intensity of the TPG was expressed as a percentage of transmural flow redistribution. A high-pass threshold of TPG was set at 5% to reduce the effect of noise of the measured perfusion gradients. The radial extent of the TPG in angular degrees ( o ) was also calculated. A schematic representation of TPG analysis is shown in figure 2..

Cardiac Catheterization Laboratory Data Analysis
All pressure and flow signals were sampled at 200 Hz and stored on disk for off-line analysis. Ensemble averages of the selected cardiac cycles were performed for distal coronary artery pressure (Pd) and coronary blood flow (estimated with average peak Doppler flow velocity (APV)). Savitzky-Golay filters were applied to preserve peaks in the pressure and flow data while smoothing. 5 These filters fit a polynomial of a chosen order to a number of points around the centre point using least squares. They have the advantage of smoothing data whilst preserving data peaks, which dramatically improves the clinical applicability of coronary wave intensity analyses.
Net wave intensity (WInet: W.m -2 .s -2 ) was calculated to be the product of time Waves are also defined as accelerating when dU increases, and decelerating when dU decreases. In addition waves can be defined as compression waves when associated with an increase in dPd, or defined as expansion waves when associated with a decrease in dPd. Four dominant waves were calculated: the accelerating forward compression wave (FCW); the decelerating forward expansion wave (FEW); the decelerating backward compression wave (BCW); and the accelerating backward expansion wave (BEW). 8 In the coronary circulation forward travelling waves are generated by increases (FCW) and decreases (FEW) in aortic pressure at the inlet and backward travelling waves are generated by changes in the microcirculation due to cardiac contraction (BCW) and relaxation (BEW). [9][10][11] Of these four waves, the acceleratory waves produced the greatest magnitude. Therefore only the two acceleratory waves, FCW and BEW, are reported in the manuscript.
In order to separate WI into the forward (WI+) and backward (WI_) components these respective formulae were applied to collected data: Where = density of blood, c = wavespeed, and WI was defined as the product of the first time derivatives of coronary artery pressure and flow velocity so that the analysis is independent of the sampling interval used, as previously described. 12 The density of blood was assumed to be constant at 1050kgm -3 . 7 Wavespeed was calculated using the single-point technique as previously described. 7 Rate pressure product (RPP), a measure of myocardial oxygen demand, 13 was calculated as systolic blood pressure multiplied by heart rate. Diastolic time fraction was defined as the fraction of the duration of diastole with respect to duration of the cardiac cycle. Systolic duration was defined from the upslope of the arterial pressure trace to the dichrotic notch, and diastolic duration defined from the dichrotic notch to the upslope of the arterial pressure trace. Augmentation index, a measure of central systolic blood pressure augmentation thought to principally arise from pressure-wave reflection, was calculated as the difference between the first and second aortic systolic pressure peaks expressed as a percentage of the pulse pressure (in turn calculated as systolic blood pressure minus diastolic blood pressure).  We observed significant increases in peak TPG intensity and radial TPG extent (both p<0.05). Peak TPG intensity reflects the subepicardial-subendocardial redistribution of myocardial blood flow, and radial extent TPG reflects the amount of myocardium affected. Increases in both parameters suggest an insufficiency of subendocardial perfusion. A peak TPG intensity value of greater than 20 % (during adenosine hyperemia) has been shown to strongly correlate with a significant fractional flow reserve (<0.8). 14 Supplemental Video Legend: Video S1. Cardiac Catheterization Laboratory Protocol Set Up.