1.

Introduction

Smoke inhalation injury is a major cause of morbidity and mortality among victims of fires. Pathophysiologic changes include hyperemia, edema, sloughing, and necrosis, which can cause airway compromise and acute lung injury. ^{1, 2, 3, 4, 5, 6, 7, 8, 9} Currently, there are no highly reliable diagnostic techniques in the clinical setting to predict or assess the degree of airway compromise following smoke inhalation. High-resolution flexible fiber optic optical coherence tomography (OCT) provides the potential to observe real-time changes in the airway mucosa and epithelium using minimally invasive imaging techniques. ^{10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25} Most importantly, some of the early pathologic changes following smoke inhalation that may be indicative of the extent of injury such as edema and hyperemia^{2, 26, 27, 28, 29} may not be apparent clinically, endoscopically, or following histologic preparations of excised specimens, but can be readily detected on OCT.³⁰

Previous OCT airway injury studies³⁰ in rabbit smoke inhalation models suggested to us that the upper trachea airway may react very differently to acute smoke inhalation compared to distal trachea and proximal bronchi. The purpose of this study was to demonstrate the potential role of OCT in quantitatively detecting early smoke inhalation injury and assessing differences in response of more proximal versus distal large airways to smoke inhalation injury.

2.

Materials and Methods

This protocol was approved by the University of California, Irvine (UCI) Academic Research Committee (2002-2397) and complied with all federal and state regulations for animal welfare assurance. General model and preparation methods have been described previously³⁰ and are briefly summarized here.

3.

Results

30 animals total were exposed to inhaled room-temperature smoke for this study (mean $127\pm 9$ breaths, range from 54 to 216 breaths total). 14 animals were in the lower trachea/bronchi exposure group, 16 in the upper tracheal exposure group, and 3 controls (no smoke exposure). As seen in our previous studies, the smoke-treated animal study groups had significant increases in airway thickness seen by in vivo OCT. Changes were evident in the thickness of the airway mucosa at the time of the first follow-up measurement $15\min$ following exposure with an average increase in airway thickness of $27\pm 5\%$ (SEM) for the overall smoke-treated animals $(\mathrm{p}<0.001)$ .²¹ No change was seen in the airway thickness of the control group animals (three upper and ten lower airway) that were not exposed to smoke during the first $60\min$ of anesthesia, intubation, and ventilator support administered in an analogous manner to the smoke-treatment animals. A slight increase in the mucosal thickness of the control group was observed during the experiments, although this increase was not significant. This increase was most likely due to the positive pressure of the ventilator or the anesthesia, as these are both known to cause effects similar to the ones we observed. Despite the marked changes in the subepithelial layers seen on OCT, only epithelial cell layer changes were seen on histologic examination (Fig. 1 ).

Fig. 1

(a) Histology showing upper rabbit trachea taken ( $10\times$ objective) $6\text{-}\mathrm{h}$ postsmoke exposure (the notable landmarks are: c-cartilage ring, e-epithelium, m-mucosa, sm-submucosa, and tm-muscularis). (b) Histology of upper rabbit trachea shown $6\text{-}\mathrm{h}$ postsmoke exposure ( $40\times$ objective). (c) Lower rabbit trachea $6\text{-}\mathrm{h}$ postsmoke exposure ( $10\times$ objective). (d) Lower rabbit trachea $6\text{-}\mathrm{h}$ postsmoke ( $40\times$ objective). Reference bars represent $0.25\mathrm{mm}$ .

Carboxyhemoglobin levels were obtained immediately postexposure in 26 of the 33 animals studied (7 lower trachea, 19 upper trachea). The range of carboxyhemoglobin levels attained in the smoke-treated animals was $21.7\pm 2.7\%$ overall (range 0.2 to 46.7%), while the control group averaged $1.0\pm 0.4\%$ percent (range 0.2 to 1.7%), $(\mathrm{p}<0.02)$ . There was no significant difference in the average carboxyhemoglobin levels in the upper trachea group (mean $22.7\pm 2.9\%$ , range 8.6 to 42.5%) versus lower trachea/bronchi (mean $19.8\pm 2.9\%$ , range 0.2 to 46.7%) $(\mathrm{p}=0.68)$ . There was no difference in the total number of smoke breaths administered to the upper tracheal group (mean $141\pm 11$ breaths, range 90 to 198) verus the lower trachea/bronchi (mean $110\pm 13$ breaths, range 54 to 216) $(\mathrm{p}=\mathrm{ns})$ .

Regional differences were seen in the degree of airway thickness changes following smoke exposure in the lower trachea/bronchi in comparison to the upper tracheal exposure groups when the OCT images at baseline were compared to those images taken $360\min$ after smoke exposure (Fig. 2 ). Statistically significant increases from baseline were seen in the lower trachea/bronchi treated animal group at all time periods postexposure (Fig. 3 ). At $15\text{-}\min$ postexposure, the lower trachea/bronchi treated group showed a $29\pm 7\%$ increase in airway thickness compared to baseline $(\mathrm{p}<0.001)$ . Airway thickness peaked in at $240\text{to}300\text{-}\min$ postexposure with a mean increase of $85\pm 22\%$ increase $(\mathrm{p}<0.003)$ . The upper trachea smoke-exposed animals also showed early increases in mucosal thickness at $15\text{-}\min$ postexposure ( $25.\pm 8\%$ increase, $\mathrm{p}<.007$ ). However, no significant further increases occurred after $15\text{to}30\min$ (at $30\min$ $r=0.36$ , $r^2=0.13$ , $\mathrm{p}=0.23$ , and at $240\min$ $r=0.14$ , $r^2=0.02$ , and $\mathrm{p}=0.66$ ), with airway swelling peaking at $30\text{-}\min$ postexposure at a $29\pm 7\%$ increase from baseline $(\mathrm{p}<0.001)$ . The degree of airway swelling from baseline was statistically significantly higher in the lower trachea/bronchi groups compared to the upper tracheal group beginning at $180\text{-}\min$ postexposure; $21\pm 7\%$ for the upper group versus $76\pm 14\%$ increase for the lower trachea/bronchi $(\mathrm{p}<0.002)$ (Fig. 4 ). These differences continue to increase out to $300\text{-}\min$ postexposure ( $25\pm 8\%$ increase for the upper group versus $85\pm 22\%$ increase for the lower trachea/bronchi group, $\mathrm{p}<0.02$ at $300\min$ ).

Fig. 2

(a) Histology image of normal rabbit lower trachea ( $10\times$ objective). (b) Normal rabbit lower trachea ( $40\times$ objective). (c) Normal rabbit upper trachea ( $10\times$ objective). (d) Normal rabbit upper trachea ( $40\times$ objective).

Fig. 3

(a) Longitudinal in vivo OCT image showing baseline upper trachea (top) and the same upper trachea $360\min$ following smoke exposure (bottom) (the notable landmarks are: c-cartilage rings, e-epithelium, m-mucosa, sm-submucosa, and tm-muscularis). (b) Longitudinal OCT image showing baseline low trachea (top) and the same trachea $360\min$ after smoke exposure (bottom). Reference bar represents $1\mathrm{mm}$ .

Fig. 4

Percent change in airway thickness of trachea and bronchi from baseline following smoke exposure or controls over time as measured by in vivo OCT. Control animals showed a gradual, mild increase in airway thickness that became statistically significant after two hours, but leveled off at that time. Upper tracheal regions showed early increases in airway thickness that were statistically significant within $15\min$ following exposure, peaked at $30\min$ , and remained stable out to five hours postexposure. In contrast, lower trachea and proximal bronchi showed early and progressive increases in airway thickness over time to $4\text{to}5\mathrm{h}$ . The difference increase in airway thickness changes following smoke exposure between lower trachea and bronchi groups versus upper tracheal injury groups was statistically significant.

To further explain the individual animal and regional variability in airway response to smoke exposure, the degree of airway thickness changes was correlated with carboxyhemoglobin levels attained for the upper trachea group in comparison to the lower trachea/bronchi group. In the lower trachea/bronchi group, there was a very close correlation between the degree of airway thickening and the carboxyhemoglobin level beginning at the first follow-up measurement period, $15\text{-}\min$ postexposure ( $r=0.83$ , $r^2=0.70$ , $\mathrm{p}<0.01$ , $\mathrm{n}=7$ ), and this trend continued to be observed at $30\text{-}\min$ postexposure ( $r=0.86$ , $r^2=0.74$ , $\mathrm{p}=0.01$ , $\mathrm{n}=7$ ), and $240\text{-}\min$ postexposure ( $r=0.75$ , $r^2=0.56$ , $\mathrm{p}=0.05$ , $\mathrm{n}=7$ ) (Fig. 5 ). This close correlation was maintained throughout the postexposure measurement period as the airway swelling progressed. In contrast, in the upper trachea group, no correlation was seen at any time period postexposure between the degree of upper tracheal swelling and carboxyhemoglobin levels.

Fig. 5

Increases in airway thickness following smoke exposure correlated closely with the exposure dose as measured by arterial carboxyhemoglobin levels in animals in the lower trachea and proximal bronchi injury group. This close correlation was seen early as well as later following smoke exposure. (a) Graph shows the correlation between percent change in airway thickness and carboxyhemoglobin level at $30\text{-}\min$ postexposure in the lower trachea and bronchi injury group. (b) Graph shows the correlation in this group at $240\min$ (four hours) postexposure. (c) Graph shows absence of relationship between airway thickness and carboxyhemoglobin levels in the upper trachea exposed animals at $30\text{-}\min$ postexposure. (d) Graph shows similar lack of correlation at $240\text{-}\min$ postexposure.

4.

Discussion

This study demonstrated the ability of optical coherence tomography minimally invasive imaging to detect acute changes in airway thickening following inhalation smoke exposure at various levels within the trachea and proximal bronchi. Optical coherence tomography images used in this study were obtained in real time at resolutions of $10\mu \mathrm{m}$ using flexible fiber optic GRIN lens-based probes and were capable of providing a mean to quantitatively assess the degree of airway thickness through measurements of the images. The changes could be monitored continuously, as in this study, for more than five hours following smoke exposure. Previous studies using this model³⁰ have shown that substantial airway changes occur very early after room temperature smoke exposure. The current study was designed to use OCT to assess regional response differences in the degree of airway changes seen in the upper trachea compared to lower trachea/bronchi following smoke exposure. The ability to quantitatively measure such differences with OCT is important for understanding potential regional variability mechanisms in inhalation injury in vivo, and will be important in future studies designed to correlate acute injury with subsequent pathophysiologic events.

This study demonstrated substantial statistically and clinically significant differences in the degree of airway changes occurring in the upper trachea compared to proximal bronchi and distal trachea. Variable but dramatic increases were seen in the thickness of the lower trachea/bronchial airways, while only modest changes developed in the higher trachea. These changes began very early following smoke exposure, frequently being observed within five minutes following exposure.

We are uncertain why the upper trachea and lower large airway regions may have responded so differently. There may be inherent differences in sensitivity to smoke in upper trachea versus distal trachea and bronchi. It is possible that the differences result from differences in deposition of smoke particles in the distal trachea and bronchi following smoke exposure. The smoke exposure was performed with room-temperature smoke administered down the endotracheal tube of intubated animals. For the upper tracheal assessments, the endotracheal tube was cut short ( $14\mathrm{cm}$ length). For the lower trachea/bronchi assessments, the endotracheal tube was kept at full length. Thus, the tip of the endotracheal tube (where the smoke exits and exposure can begin) was equidistant from the OCT probe in both cases. It is possible that the diameter and curvature of the lower trachea and bronchi result in toxic particle impaction with resultant increase in airway injury. Prior studies have shown that tracheobronchial injury in smoke inhalation is primarily due to toxic chemicals contained in the smoke.² Future studies will be needed to further clarify the causes for these regional differences, and may be very important to understanding the nature of inhalation injury and improving treatment/outcomes.

There are a number of limitations with these studies. It was not until the time of sacrifice that the position of the endotracheal tube and optical coherence tomography probe was determined with certainty. It was not possible to precisely place the OCT probe in the right main stem bronchus at specific locations prior to smoke treatment with this study design. Furthermore, the imaging direction of the OCT probe was generally pointed anteriorly (as evidenced by as the presence of cartilaginous rings and continuous airway wall). However, the exact angle of imaging was not clearly defined, and precise delineation of distal trachea from proximal bronchi, or right versus left mainstem bronchi, could not be determined until after sacrifice.

The exact pathophysiology of the airway injury seen on OCT is also uncertain. The very early appearance of the changes suggests that edema and/or hyperemia may be predominant components of the injury process. The findings are consistently present on OCT evaluation. However, a proportion of the airway thickening changes seen on OCT disappears immediately following sacrifice, six hours postexposure. This would suggest that previously described hyperemia^{1, 2, 26, 27} in the initial stages following smoke exposure may be a significant component of this process. Furthermore, histology of the tracheal specimens following sacrifice does not demonstrate significant changes from controls, even in animals that had dramatic changes seen on OCT; this is consistent with published literature and suggests edema or hyperemia as major components of the early injury process that are lost during histologic preparation.

Imaging with small diameter flexible fiber optic white-light bronchoscopy showed only mild evidence of early hyperemia following smoke exposure. Regional deposition of smoke particles could not be differentiated visibly, and as expected, the degree of airway swelling could not be determined by fiber optic bronchoscopy, despite significant changes seen on OCT.

Thermal effects are unlikely a significant component of these findings, because room-temperature smoke was used and thermal injury would be expected to be more severe closer to the tip of the endotracheal tube.²

Carboxyhemoglobin levels were obtained in the animals as a marker of the degree of exposure. A range of carbon monoxide levels was obtained in animals through variation in numbers of administered smoke-containing breaths, as well as intrinsic variability in the level of carboxyhemoglobin achieved for a given number of breaths. The average carboxyhemoglobin levels achieved and the total number of smoke breaths administered were not statistically significantly different between the upper trachea and distal trachea exposure groups, and were actually slightly higher in the upper trachea group where less effect was seen. Therefore, greater degrees of airway swelling in the lower trachea/bronchial injury group could not be explained by greater degrees of total smoke exposure.

Interestingly, the degree of airway swelling correlated very closely with the level of carboxyhemoglobin attained in the animals during the postexposure period in the lower trachea/bronchi group. This is evidence of a dose-response relationship between total smoke exposure and early airway injury effects in this lower tracheal and proximal bronchial region.³² In contrast, more upper trachea showed significantly less injury, and no correlation with carboxyhemoglobin levels at any time following exposure.

Future studies to further elucidate the causes behind the regional differences in response to smoke injury, association between airway changes and various pathologic mediators of tissue injury and repair, as well as correlation between early changes and later effects are needed.

Flexible fiber optic optical coherence tomography is minimally invasive, and can be performed with very narrow diameter flexible probes that can be left in place to image intermittently or continuously. This can potentially facilitate follow up evaluations of smoke inhalation injury, which progresses over time,^{1, 2, 26, 27, 33} and that may not be evident during bronchoscopic evaluation. This study was able to demonstrate that OCT has the capability of not only detecting changes in thickness in the layers of the mucosa, but also quantifying differences in the magnitude of changes in the upper and lower trachea. The ability of OCT to distinguish such changes in real time may someday assist clinically in determining the extent of injury, prognosis, and need for intervention. Quantitative determination of acute and longer term changes in the airway epithelium, mucosa, and submucosa with OCT may also provide a more sensitive tool for investigation of the effectiveness of various therapeutic interventions in smoke inhalation and other airway injuries.