New Ultrathin-walled Endotracheal Tube with a Novel Laryngeal Seal Design: Long-term Evaluation in Sheep

The manufacturing process of the new line of ETTs has been previously described. [13]Briefly, a steel mandrel was fabricated to the dimensions of the desired internal configuration of the finished ETT. The mandrel was made of two diameters: a smaller diameter section to accommodate the trachea (first stage) and an oropharyngeal section 1–2 mm larger (second stage). The mandrel was coated with a thin layer of Teflon, after which it was placed in a rotating lathe, and a 15% dispersion of polyurethane (Surethane, segmented polyether urethane, Cardiac Control Systems, Palm Coast, FL) in dimethylacetamide was metered onto the mandrel at a constant thickness along its entire length. After air drying at 70 degrees C, the thickness of the polyurethane layer was 0.05 mm. After two such coatings, stainless steel 304 flat wire (0.10 mm x 0.250–0.500 mm) was wound around the mandrel so that approximately one-half of the area was covered with wire. In the second stage (oropharyngeal section), the stainless steel 304 flat wire was replaced by a superelastic nickel-titanium alloy (Nitinol, Fort Wayne Metals, Fort Wayne, IN). This alloy rendered the oropharyngeal section of the ETT kink- and crush-resistant, i.e., it recovered its shape after forceful manual occlusion. The mandrel was coated with an additional layer of polyurethane solution, air-dried at 70 degrees C, and cured at 80 degrees Celsius. The ETT was removed from the mandrel and placed in a press, which transformed a 4–5 cm-long laryngeal portion of the tube into an oval shape (i.e., the anteroposterior axis was larger than the transverse axis). Finally, we vacuum-molded 8–20 gills, ring-shaped disks made of thin polyurethane film (0.025–0.075 mm thick), which were solvent-cemented to the laryngeal portion of the tube. The diameter of those gills was approximately 25 mm for the adult range of ETTs. The gills were designed to conform to, rather than to deform, the opening of the glottis.

All animal studies were approved by the Animal Care and Use Committee of the National Heart, Lung, and Blood Institute of the National Institutes of Health.

Only one adult-size ETT was evaluated in 22 female sheep of a weight range (body weight 24.4+/-2.8 kg, range 19–28 kg) appropriate for a no. 8 standard ETT. The tracheal portion of the new ETT had the same external diameter as the standard 8-mm ETTs.

Anesthesia was induced with sodium pentobarbital (25 mg *symbol* kg sup -1 loading dose, followed by a maintenance dose of 2 mg *symbol* kg sup -1 *symbol* h sup -1). Paralysis was sustained with pancuronium bromide (0.15 mg *symbol* kg sup -1 loading dose, followed by a maintenance dose of 0.06 mg *symbol* kg sup -1 *symbol* h sup -1). The right external jugular vein was cannulated with a 16-G catheter to monitor central venous pressure, and for fluid replacement, normal saline was infused at 4–8 ml *symbol* kg sup -1 *symbol* h sup -1. The animals were positioned prone throughout the whole experiment. A Foley catheter was inserted to monitor urinary output.

Sheep were randomly assigned (with the method of cards in sealed envelopes) to four experimental groups as follows.

Group 1: New ETT, 1-Day Study (n = 6). Animals were tracheally intubated with the new ETT, connected to a Servo ventilator 900 C (Siemens, Elema) and their lungs mechanically ventilated on continuous positive-pressure ventilation (CPPV) for 1 day. The positions of the oval portion of the ETT and of the gills were checked by direct laryngoscopy and roentgenographic films. The tube was secured to a bite block. Tidal volume was set at 10–15 ml *symbol* kg sup -1, respiratory rate was set at 10–12 breaths/min, the peak inspiratory pressure (PIP) usually ranged between 10 and 15 cmH²O, the inspiratory/expiratory ratio was set at 1:2, and positive end-expiratory pressure was set at 5 cmH²O. Throughout the procedure, the lungs were ventilated with humidified air (Conchatherm III, Hudson Respiratory Care, Tennecula, CA) at an inspired fraction of oxygen (FI^O2) of 0.40. Every 4 h, we analyzed venous blood for blood gases, electrolytes, and glucose, and we recorded urinary output.

Group 2: Standard ETT, 1-Day Study (n = 5). Sheep were tracheally intubated with a standard 8-mm Mallinckrodt ETT with a hi-lo pressure cuff inflated to a pressure of 25–30 cmH²O, which was continuously monitored using a strain gauge transducer (Gould Statham P23 ID, Gould Instrument Systems, Valley View, OH). The lungs were ventilated using CPPV (as in group 1) for 24 h, and the sheep were monitored as in group 1.

Group 3: New ETT, 3-Day Study (n = 5). The sheep were tracheally intubated with a new ETT, and the lungs were ventilated using CPPV for 3 days. The lungs of two sheep were ventilated with the same ventilatory setting as in groups 1 and 2. The lungs of three other sheep were ventilated at a PIP of 30 cmH²O with a tidal volume of approximately 35 ml *symbol* kg sup -1, a respiratory rate of 5 breaths/min, an inspiratory/expiratory ratio of 1:2, positive end-expiratory pressure of 5 cmH²O, and FI^O2of 0.40. To avoid respiratory alkalosis, carbon dioxide was added to yield an inspired fraction of carbon dioxide (FI^CO2) between 0.03 and 0.04. The right common carotid artery was cannulated with a 16-G catheter for monitoring arterial blood pressure and blood gas analysis (Nova Biomedical, Stat Profile Plus 9). Every 4 h, we analyzed arterial blood for blood gases, electrolytes, and glucose, and we recorded urinary output. Blood pressure was continuously recorded using a strain gauge transducer (Gould Statham P23 ID).

Group 4: Standard ETT, 3-Day Study (n = 6). The sheep were tracheally intubated with a standard 8-mm Mallinckrodt ETT with a hi-lo pressure cuff inflated to a pressure of 25–30 cmH²O and were monitored as in group 2. The lungs of three sheep were ventilated on CPPV for 3 days at a PIP of 10–15 cmH²O, as in group 1. The lungs of the other three sheep were ventilated at a high PIP (30 cmH²O), as in group 3.

Air leak was computed as the ratio of the difference between inspiratory and expiratory tidal volume to delivered inspiratory tidal volume, as read from the Servo Ventilator, in percent. Data were collected at baseline and at 8 and 24 h during the 1-day studies (groups 1 and 2). In sheep tracheally intubated for 3 days (groups 3 and 4), air leak was recorded every 4 h. To assess possible air leak at a higher PIP, we transiently increased the PIP in increments of 5 cmH²O up to 50 cmH sub 2 O while keeping respiratory rate constant; after this, standard ventilation was resumed.

An air leak under 5% of the delivered volume was considered insignificant.* In the standard ETT group, the intracuff pressure was maintained constant with no adjustments, even if air leak exceeded 5%.

To assess liquid seepage across the balloon or the gills, we instilled 25 ml of 0.5% methylene blue dye in saline into the oropharynx of all sheep every 4 h. The operating table was tilted to a 30 degrees reverse Trendelenburg position to expedite distribution of the dye within the oropharynx. At the end of the experiment, the sheep were killed by intravenous injection of sodium pentobarbital and potassium chloride.

At autopsy, the trachea and the larynx were opened through an anterior midline longitudinal incision up to the carina and inspected. We noted the presence of methylene blue dye below the cuff or the gills and the position of the ETT. Tracheal and laryngeal lesions were documented photographically and were scored grossly (Table 1) by the same observer using a modification of the scoring system of Hedden et al. [14]Lesions were scored by severity and area of involvement to arrive at a single score. The larynx and trachea were excised and fixed in toto in buffered 10% formalin. After fixation, each laryngotracheal specimen was divided into halves at the midline. The lesions were almost symmetric. Samples for microscopic analysis were taken from the left side, unless the lesions appeared more severe on the right side. Tissue blocks were taken from the vocal cords, the arytenoids, and the trachea in the 1-day study and also from the epiglottis in the 3-day study (Figure 2). The tracheal sample in the standard ETT groups included the region in contact with the cuff; the same approximate region was sampled in groups 1 and 3 (the new ETT groups). Tissue blocks were dehydrated and embedded in paraffin. Five micrometer-thick sections were cut and were stained with hematoxylin and eosin, the Movat pentachrome method, and the periodic acid-silver methenamine. Slides were scored by a pathologist not aware of the ETT used (Table 2).

To assess the epithelial injury, using the scoring system reported in the table, the pathologist applied one mean score to each field of view. The scores from other fields of view were added to arrive at a final mean score for the whole slide. Histologic lesions of epiglottis, vocal cords, trachea, and larynx were scored separately, with particular attention to the epithelial and submucosal layers.

The Mann-Whitney two-group nonparametric test was used to compare the severity of the gross and microscopic lesions between the new and the standard ETT group. A P value of 0.05 or less was considered significant.

There was no air leak in the trachea of any sheep in the new ETT or in the standard ETT group in which the lungs were ventilated continuously at a PIP of 15 30 cmH²O for 1 or 3 days. When the PIP was increased, the new ETT with gills and the standard ETT showed no air leak up to a PIP of 40 cmH²O. In some animals in each group, there was some air leak at a PIP of 45 and 50 cmH²O (Table 3, animals without air leak are not shown). In sheep with the new ETT, air leak at a higher PIP was found more commonly at the beginning of the study and then rarely throughout the rest of the study. Air leak at a higher PIP was limited in scope, and effective ventilation of the animals was never compromised.

We found no methylene blue on gills below the level of the vocal cords or in the trachea, in either short- or long-term studies using the new ETT (groups 1 and 3), at a normal and a high PIP. In sheep tracheally intubated with the standard ETT, we found methylene blue dye below the inflated cuff in one sheep ventilated for 1 day at a normal PIP and in one sheep ventilated for 3 days at a high PIP.

1-Day Intubation: Groups 1 and 2. The observed lesions were located as shown in the artist's drawing in Figure 3. Those sites were common to 1-day and 3-day experiments, although their severity was different. Grossly, the trachea appeared better in the new ETT group compared to the standard ETT group (P = 0.02, Table 4). In the new ETT group, the trachea appeared grossly normal in five of six animals. The one abnormal sheep in this group showed extensive inflammation of the entire tracheobronchial tree, most likely due to a preexisting acute inflammatory process rather than to the use of ETT. In sheep tracheally intubated with the standard ETT, the tracheal changes ranged from erythema to erosion of the mucosa in contact with the cuff, often extending beyond four tracheal rings.

In the larynx in both groups, we found symmetric lesions on the medial side of the arytenoids. The severity of the lesions appeared different between the two groups (P = 0.009;Table 4). In the standard ETT group, abnormal findings ranged from mucosal erythema and/or edema to erosion. In the new ETT group, abnormal findings were limited to mucosal erythema and/or edema, and there was no erosion.

In both groups, most of the subglottic area appeared normal. Grossly, the epiglottis appeared healthy. Edema of the vocal cords was more common in the control group than in the study group (P = 0.045;Table 4).

3-Day Intubation: Groups 3 and 4. The trachea appeared better preserved in the new ETT group, compared to the standard ETT group (P = 0.004;Table 4). Tracheal lesions in the standard ETT group ranged from erythema to erosion, and extended over two to seven tracheal rings in contact with the cuff.

In the larynx in both groups, symmetric lesions were found on the medial side of the arytenoids. In the standard ETT group, erosion was the predominant finding. In the new ETT group, three animals showed erythema, and two animals showed erosion (NS). In both groups, the subglottic area appeared spared. There was erythema of the epiglottis in two animals in the new ETT group and two animals in the standard ETT group. In one additional animal in the latter group, we found limited erosion on the epiglottis.

Trachea: 1-Day Intubation. Epithelial injury was significantly milder in the new ETT group than in the standard ETT group (P = 0.05;Table 5). In the new ETT group, most of the tracheal epithelium appeared intact, and there was no erosion (Figure 4). The submucosa was free of edema or hemorrhage. There was mild to moderate infiltration of neutrophils. The cartilage was not involved.

In the standard ETT group (Figure 4), the trachea showed widespread epithelial erosion, mild to moderate edema (in two sheep), mild hemorrhage (one sheep), and mild to severe infiltration of neutrophils. The cartilage was not involved.

Trachea: 3-Day Intubation. We found significantly milder changes in the trachea in sheep tracheally intubated with the new ETT compared to the standard ETT group (Table 5). There were statistically significant differences with respect to epithelial injury (P = 0.004), submucosal edema (P = 0.04), and hemorrhage (P = 0.04). The cartilage was never exposed or injured in either group. In the new ETT group, tracheal epithelium was well preserved in general. There was no erosion, but mild inflammation was seen in all sections, and mild edema and hemorrhage were seen in one case (Figure 4(C)). In the standard ETT group (Figure 4(D)), there was epithelial erosion, mild to severe submucosal edema, hemorrhage, and infiltration of neutrophils in five of six animals.

Larynx: 1-Day Intubation. No statistically significant difference between the two experimental groups was found in the arytenoids (Table 5). The epithelium lining the arytenoid cartilage showed mild to severe injury and mild to severe infiltration of neutrophils (Figure 5). In the new ETT group, mild edema and hemorrhage were found in one sheep; in the standard ETT group, there was mild to severe hemorrhage in four of five animals and mild edema in two. The vocal cords were normal in both groups.

Larynx: 3-Day Intubation. The epithelium lining the arytenoid cartilage was denuded in both groups (Table 5). Moderate edema, inflammation, and hemorrhage were present in both groups. The epithelium of the vocal cords was mostly intact in the new ETT group, whereas it showed erosion in the standard ETT group (Figure 5). Edema of the vocal cords was found more frequently in the latter group. Mild to moderate inflammation was observed in both groups. There was mild focal hemorrhage in one sheep in the new ETT group.

Changes in the epiglottis were not significantly different in the two groups. Both groups showed mild epithelial denudation, mild to moderate edema, inflammation, and hemorrhage.

In this controlled study using a sheep model, we assessed whether the new ETT could prevent the aspiration of liquids, reduce laryngotracheal damage, and sustain ventilation at a different PIP.

The interpretation of our results is subject to certain limitations:

- We used healthy animals.

- The duration of our studies (1 or 3 days) does not allow extrapolation beyond such time.

- The animals were paralyzed, hence any possible friction between the laryngeal and tracheal mucosa and the tubes (both standard and new) was minimized and may have led to reduced injury. The differences between the new ETT and the standard ETT groups, performed under identical conditions, remain highly significant.

In sheep, the oral route of intubation was the only one that was practical. We have not evaluated nasal intubation.

There was no air leak at a PIP less than 40 cmH²O. Air leak at a higher PIP was a rare event and tended to decrease throughout the experiment. We ascribe this improvement to the mucus entrapped over time in the gills. Hence, the new ETT may be suitable for use in the majority of patients managed in intensive care units, because it performs well at airway pressures up to 35 cmH²O, above which the risk of barotrauma becomes significant, and acceptably at pressures up to 50 cmH sub 2 O.**.

We found no aspiration with the new ETT. Aspiration was found in some sheep tracheally intubated with the standard ETT in short- and long-term studies, despite meticulous attention to intracuff pressure. Because of the known correlation between aspiration of bacteria colonizing the oropharynx and nosocomial pneumonia, decreasing, if not eliminating, aspiration would be the main goal in the design of a new ETT. This would reduce the occurrence of nosocomial pneumonia, the leading cause of morbidity and mortality from nosocomial infections. [7]The gills prevent pooling of contaminated secretions within any portion of the trachea and restrict secretions to above the level of the glottis. It was shown previously that regular draining of subglottic secretions was associated with a 50% decrease of nosocomial pneumonia. [9]We can speculate that the glottic seal may further reduce nosocomial pneumonia. Our results on aspiration are encouraging. Those findings will have to be confirmed in a greater number of animals, under additional experimental conditions, using methods with greater sensitivity to detect aspiration.

We elected not to titrate tracheal cuff pressures to yield lowest possible cuff pressure without air leak ("minimal leak" method), because it would have complicated the interpretation of our results.

We hypothesized that substituting the cuff with the gills would substantially decrease tracheal lesions. The safe range of pressure to inflate the tracheal cuff is narrow. [5,15]No matter how carefully one monitors the inflatable cuff, tracheal damage remains predictable, particularly in patients ventilated at a high PIP, at which the intracuff pressure unavoidably exceeds this threshold. [16]The ideal solution to the problem is to eliminate the cuff.

Grossly, the trachea appeared better in the new ETT group compared to the standard ETT group, even though intracuff pressure in the standard ETT group was kept at all times within the recommended safe range. Our histologic findings confirmed our gross observations, in short- and long-term experiments. The tracheal epithelium remained mostly intact in sheep tracheally intubated with the new ETT at 1 and 3 days; mild hemorrhage and edema (not observable by gross inspection), if any, were found only after 3 days of intubation and were limited to the perivascular areas. In contrast, in the standard ETT group, histologic studies showed that tracheal epithelium was eroded at 1 and 3 days; hemorrhage and edema were focal in the 1-day study but were frequently observed and mostly diffuse at 3 days. After 3 days of intubation, the difference between the new ETT group and the standard ETT group became more pronounced and involved the submucosal layers.

Significant epithelial denudation can be found after even minimal contact. Dragging a gauze over the surface of the tracheal epithelium can remove the whole epithelial layers and injure the basement membrane. [17]In view of the above, the good preservation of the epithelium observed with the new ETT is even more striking.

The larynx appeared, grossly, less damaged at 1 day in the new ETT group, although the new sealing system was located within the glottic opening. Histologic evaluations of the larynx showed no significant difference between the two groups at 1 and 3 days. Hence, the new ETT with gills as a laryngeal seal does not appear to impose an additional burden on the integrity of the larynx compared to standard ETT with hi-lo inflatable cuffs. We ascribe the good performance of the new ETT to the oval shape of its laryngeal portion, which adapts well to the pentagonal shape of the glottic opening. The individual role of the two components, oval shape and gills, remains to be explored. It is possible that other contours of the laryngeal portion of the ETT and a better relationship between the size of the glottis and the size of the ETT would improve on our results. The optimal number, size, and thickness of gills are yet to be explored.

We conclude that it is possible to deviate successfully from the traditional method of sealing the trachea with an inflatable cuff and to replace it with a no-pressure system made of gills positioned at the level of the glottic opening. This design appears superior to the standard hi-lo cuff in preventing aspiration and tracheal injury. Those characteristics, with the greatly lower air resistance found in the new ETT, make this design attractive for further evaluation. The low resistance of the new ETT appears well suited for spontaneous forms of ventilation. Absent aspiration and the reduced tracheal injury make it attractive for prolonged intubation and for the newborn and pediatric patient population.

*American National Standards Institute: American national standard for anesthetic equipment-cuffed oral tracheal and nasal tracheal tubes for prolonged use, ANSI Z79, 16–1983. New York: American National Standards, 1983, pp 6–20.

**Slutsky AS: Consensus conference on mechanical ventilation, January 28–30, 1993, at Northbrook, Illinois. Intensive Care Med 1994; 20:64–79.