The ProSeal™ laryngeal mask airway (PLMA; Laryngeal Mask Company, Henley-on-Thames, UK) is a new LMA designed for positive-pressure ventilation (PPV) (1). It achieves significantly larger airway seal pressures because of a new cuff design and also includes the safety feature of a drain tube for PPV (1). However, a distinguishing feature of this new LMA is that it can contribute to upper airway obstruction, even when it has been correctly positioned behind the cricoid cartilage. As previously described, the larger, more compliant outside cuffs can obstruct the airway by folding medially and approximating each other in front of the bowl (cuff infolding) (2). Also, the mask can compress the arytenoids, false cords, and true cords and significantly narrow the airway at both supraglottic and glottic levels (3,4). Therefore, despite the significant advantages of this new LMA for PPV, every anesthesiologist must be aware of its potential for upper airway obstruction.
We sought alternative methods for diagnosing airway obstruction with the PLMA. When we had encountered upper airway obstruction with the PLMA, we had often felt an abnormal feel of the anesthesia bag. Although difficulty was often noted while squeezing the bag, the abnormal feel was particularly evident during the filling phase; the bag simply refilled slowly. We believed that this represented a tactile diagnosis of diminished expiratory flows, such as a diagnosis of a low forced expiratory volume in 1 second (FEV1). We were thus struck by comments in a pulmonary function textbook that “…some lesions of the major airway cause the maximum voluntary ventilation to be reduced out of proportion to the FEV1… If a significant decrease in the maximum voluntary ventilation occurs in association with a normal FEV1, a major airway obstruction should be strongly suspected (5).” In fact, maximum voluntary ventilation is one of the most useful pulmonary function tests for detection of fixed airway obstruction, variable extrathoracic airway obstruction, or variable intrathoracic airway obstruction (6,7). Perhaps a hyperventilation test, analogous to the maximum voluntary ventilation test, might be a more appropriate method for detection of upper airway obstruction with the PLMA compared with subjective assessment alone of expiratory flow by feel of the anesthesia bag.
We have adopted the pulmonary function laboratory maximum voluntary ventilation test to anesthetized and paralyzed patients with the PLMA. The maximum minute ventilation test (MMV test) is performed by maximally hyperventilating a patient with a PLMA for 15 s:MATH
It represents the total exhaled volume of breaths achieved in 15 s and extrapolated to 1 min. Below, we present data collected for 6 mo pertaining to normal and abnormal MMV values in adult women and men with PLMAs. The MMV test is an easy and surprisingly reliable method appropriate for rapid identification and quantification of airway obstruction after insertion of the PLMA.
Hospital IRB approval and patient informed consent were obtained. The study took place for 6 mo beginning mid-October 2001 through mid-April 2002, representing clinical use of the PLMA in the authors’ practice. Patients with known or anticipated difficult airways did not receive a PLMA. All patients were fasted for solids for a minimum of 6 h before the induction of anesthesia. Patients were not excluded from PLMA use because of gastroesophageal reflux, hiatal hernia, diabetes, obesity, positioning, or American Society of Anesthesiologists classification status. To every extent possible, the authors attempted to have a fiberscope immediately available in the operating room (OR) at the time of the anesthetic induction and PLMA insertion.
General anesthesia was induced IV, lungs were ventilated with oxygen 100% via a facemask, and a muscle relaxant was administered. The PLMA was then inserted using the finger technique, and the cuff was inflated to 60 cm H20. Women received size 4 PLMAs and men size 5 PLMAs. Next we assessed for correct location of the PLMA behind the cricoid cartilage excluding both glottic insertion or backwards folding of the tip using soap bubble solution (such as used by children to blow bubbles) placed on the drain tube port, as previously described (8,9). We determined satisfactory depth of insertion behind the cricoid cartilage with a combination of (a) positive suprasternal notch test and (b) complete absence of drain tube leak during maximal lung inflation. If a drain tube leak was detected (bubble formation) during maximal lung inflation, we advanced the mask further into the hypopharynx. On the anesthetic records, we documented the successful placement of the PLMA behind the cricoid cartilage with positive suprasternal notch test and zero drain tube leak. We also recorded the maximum seal pressure (maximal achievable airway pressure squeezing the circuit bag) and noted that it was caused by a leak from the pharynx and not the drain tube.
The MMV tests were performed using an adult circle system with regular length disposable circuit (SIMS Portex Inc, Ref C3710647, volume 900 mL, Fort Meyers, FL). Spirometer location was immediately adjacent to the expiratory unidirectional valve with exhaled tidal volumes displayed electronically on the machine panel (North American Drager, Narkomed 2B and GS models, Telford, PA).
Before the test, fresh gas was set to oxygen 2 L/min and the circuit pop-off valve completely closed off. Observing the large second hand on the OR clock, we then steadily hyperventilated the patient while counting breaths during a 15-s episode. The sequence was begun with a full 3-L circuit bag. Optimum MMV was achieved by squeezing the circuit bag as hard as possible but also insuring that the bag would remain mostly full for the subsequent breath (e.g., avoiding excessive oropharyngeal leak and collapse of the bag). We glanced once at the anesthesia machine to note a value for exhaled tidal. By documenting an exhaled tidal volume near the end of the breath sequence, we chose a single representative value. Neither the initial breath nor the final breath yields a reliable value. The final breath, in particular, is too large and represents complete exhalation to functional residual capacity that does not typically occur in the midst of the hyperventilation sequence. Usually the representative volume was larger than the expected setting for intraoperative ventilator management. We also deliberately ignored the anesthesia machine display of minute volume, a time-averaged value not appropriate for the 15-s MMV test.
The patient was then managed according to the judgment of the anesthesiologist, including fiberoptic examination via the airway tube when airway obstruction was suspected based on physical signs or small measured MMV value. Fiberoptic observations and pertinent physical findings were recorded. Any change in airway management modality was noted, and all perioperative complications were documented.
To check the significance of our measurements of exhaled tidal volume, we measured the ventilator circuit compression volume for our anesthesia machines and disposable circuits (10). This volume represents the component of measured exhaled tidal volume that does not actually emerge from the patient airway but is solely due to gas compression within and expansion of the circuit. Fresh gas flow was set to zero, the pop-off valve was completely closed, the patent end of the disposable circuit (15-mm connector) was occluded, and the ventilator bellows was turned on with a series of plateau pressures. Exhaled tidal volume measured by the spirometer was then plotted against plateau pressure to obtain the compression volume coefficient (mL/cm H2O).
Although 318 patients were in the study, one man was excluded because a fiberscope was not available to diagnose the cause of critical MMV before the PLMA was removed. Data are expressed as range (mean ± sd). There were 169 women aged 18–91 yr (56 ± 16 yr), height 145–183 cm (163 ± 7 cm), weight 42–212 kg (76 ± 21 kg), and body mass index (BMI) 18–67 kg/m2 (29 ± 7 kg/m2). There were 148 men aged 19–86 yr (57 ± 16 yr), height 153–193 cm (177 ± 7 cm), weight 44–168 kg (87 ± 19 kg), and BMI 17–52 kg/m2 (28 ± 5 kg/m2). A total of 43 patients (29 women and 14 men) were obese with a BMI >35.
We found that the MMV test yielded a clinically useful value that did not change dramatically when a different respiratory rate was used to repeat the test. Figure 1 illustrates variation of MMV for four patients when different respiratory rates were used for hyperventilation. Slow respiratory rates generated larger tidal volumes, whereas more rapid rates yielded smaller tidal volumes. Because MMV is the product of respiratory rate and tidal volume, the two opposite tendencies tend to cancel each other. Rates between 5 and 10 breaths/15 s usually yielded the optimum MMV in normal patients. A rate between 5 and 10 breaths/15 s was used in 91% of all patients in the study. When MMV was severely compromised, it was required to use a slower rate. In five of the six cases where the PLMA was removed, the MMV measurement used the slower rate of 4 breaths/15 s.
Data are expressed as range (mean ± sd). For adult women and size 4 PLMAs, the MMV values obtained were 3.2–41.8 L/min (26 ± 6 L/min). For adult men and size 5 PLMAs, the MMV values obtained were 6.2–57.6 L/min (29 ± 9 L/min). (Table 1). We considered any MMV of 12 L/min or less to be critical (Table 2). In two women with known severe emphysema, the fiberoptic examination was normal. The remaining 15 patients had fiberoptically obvious airway obstructions caused either by bilateral cuff infolding (with or without downfolded epiglottis) or narrowing of the supraglottic and/or glottic airway termed laryngeal obstruction (Table 2). We emphasize that the PLMA was correctly positioned behind the cricoid cartilage in each of these 15 cases. The incidence of critical MMV value attributable to insertion of the PLMA was therefore 15 of 317 (4.7%) overall, five of 169 (3.0%) in women, and 10 of 148 (6.8%) in men. The incidence of PLMA removal was seven of 317 (2.2%) overall, three of 169 (1.8%) in women, and four of 148 (2.7%) in men. Different incidences for men and women were not statistically significant using the χ2 test and P < 0.05. Overall incidences represent our sample of 317 patients with 53% women and 47% men.
There were many complex processes resulting in severe narrowing of the airway at glottic and supraglottic levels including (a) symmetric inward displacements of supraglottic and/or glottic structures, (b) asymmetric displacements, (c) glottic airway narrowing characterized by arytenoid vocal processes abnormally contacting in the midline, (d) the true vocal cords, the anterior two thirds of the glottis, approximated in the midline, (e) PPV predominantly occurring through the posterior one third of the glottis, the triangular shaped posterior glottic chink surrounded by both vocal processes and base of the cricoid cartilage, (f) false cords compressed in the midline and failing to separate with PPV, (g) approximation of both arytenoid cartilage apices into a midline, supraglottic mass, (h) supraglottic obstruction of the posterior glottic chink, and (i) anterior-posterior compression of supraglottic and/or glottic anatomy.
Linear regression of MMV was performed with respect to age, height, weight, BMI, respiratory rate, exhaled tidal volume, and maximum airway seal pressure. Using a 0.05 significance level and two-tailed test, the following statistically significant correlations were obtained: adult women MMV was correlated with exhaled tidal volume, r = 0.66; height, r = 0.38; age, r = −0.29; and seal pressure, r = 0.22. Adult men MMV was correlated with exhaled tidal volume, r = 0.75; age, r = −0.37; and seal pressure, r = 0.24.
Compression volume measured in the OR using our anesthesia machines and circle systems with regular size disposable circuits yielded 1.6 mL/cm H2O. A tidal breath with plateau pressure of 25 cm H2O would, for example, produce a compression volume of 40 mL. Compression volume is therefore not a terribly significant proportion of exhaled tidal volume mea-surement using our circle system with spirometer adjacent to the expiratory unidirectional valve.
The PLMA can occasionally cause significant upper airway obstruction after it is inserted, even when it is correctly placed behind the cricoid cartilage (2–4). Here we have described our experience using a hyperventilation test to detect such upper airway obstruction. The MMV test consists of manually hyperventilating an anesthetized and paralyzed patient with a PLMA for 15 seconds and extrapolating total exhaled volume to one minute. The test is easy to perform and can be completed with equipment that is readily available to nearly every anesthesiologist. Also, our experience indicates that a clinically useful value for MMV can be obtained with just one or two performances of hyperventilation. Below, we discuss the incidence and sites of airway obstruction as well as management of such patients.
One of the important findings in this study is that upper airway obstruction with the PLMA is not necessarily due to improper insertion technique. Indeed, every patient with critical MMV (Table 2) was fiberoptically observed to have the PLMA correctly positioned behind the cricoid cartilage. Causes of airway obstruction fell into two broad categories: laryngeal obstruction and cuff infolding. Laryngeal obstruction referred to compression of normal supraglottic (false vocal cords, arytenoid apexes, and epiglottic tubercle) and glottic (vocal processes and true vocal cords) anatomy so that one or more levels of airway narrowing and obstruction were present. Cuff infolding referred to inward rotation of the large cuffs in front of the bowl so that they contacted each other in the midline and obstructed gas flow.
It is not unreasonable that a properly placed LMA might cause obstruction of the laryngeal aperture. The fundamental role of the larynx is to act as a sphincter to protect the lower airways (11). Although the PLMA is not literally a bolus of food, its correct insertion does correspond to routes taken by food during deglutition, and therefore, its presence could by design produce a degree of laryngeal closure. Our fiberoptic observations reflect processes previously described for the laryngeal sphincter mechanism including (11) (a) shortening of the larynx in the anterior-posterior direction so that the true and false cords became shortened and flaccid, (b) inward rotation of the arytenoids resulting in opposition of the true cords, (c) sliding medially of the arytenoids on top of the cricoid cartilage to form a midline mass, and (d) tilting forward of the arytenoid apexes over the posterior glottic opening. We assume that the PLMA, with its large drain tube and cuffs, produces such effects by displacing the cricoid cartilage anteriorly as well as by pressing directly on the arytenoid bodies and muscular processes.
We can now summarize the sites of airway obstruction as follows. First, airway obstruction was never observed to be due to a downfolded epiglottis situated within the bowl of the mask. The drain tube always functioned to suspend the epiglottis off of the floor of the bowl, and critical airway obstruction was never caused by failure of this design feature. However, with cuff infolding (the two outside cuffs meet in the midline and the epiglottis cannot enter the bowl), a downfolded epiglottis becomes a risk factor for airway obstruction because it is now forced directly on the arytenoids. We recently observed complete airway obstruction when such a downfolded epiglottis completely covered both arytenoids. Cuff infolding, by itself, can cause obstruction. Finally, airway obstruction can occur in a normally positioned PLMA because of compression and obstruction of the laryngeal aperture, as described. This complex process can involve supraglottic and/or glottic level narrowing often resembling normal sphincter function of the larynx.
Our findings indicate that cuff infolding is a relatively uncommon cause of airway obstruction, occurring in only two of 317 patients. By comparison, laryngeal obstruction was the cause of critical MMV in 13 of 317 cases. Therefore, laryngeal obstruction is the most likely cause for upper airway obstruction after insertion of the PLMA. All together, 15 of 317 (4.7%) patients had significant upper airway obstruction due to insertion of the PLMA, occurring in roughly one of 20 cases. We removed the PLMA in seven of 317 (2.2%) patients, roughly one of 50 cases. Finally, critical MMV due to insertion of the PLMA occurred in 15 of 317 patients, whereas it was due to patient disease (emphysema) in just two of 317 patients.
Practically, it is worthwhile to compare the MMV with a patient’s basal minute ventilation requirement. The amount that the MMV exceeds this basal requirement constitutes an important margin of safety for minute ventilation. In adult men and women, basal minute ventilation is approximately 5–7 L/min, although it can vary based on CO2 production, physiologic dead space, and the desired level for arterial CO2 partial pressure (12). As a rough guideline, we usually consider 6 L/min to be the average basal minute ventilation requirement for an adult patient. The ability to achieve MMV in excess of 6 L/min therefore constitutes a rapid estimate of the margin of safety for minute ventilation. Clearly, the mean values of 26 L/min (women) and 29 L/min (men) provide ample margin of safety. With a compromised MMV however, the margin of safety is drastically reduced or may be nonexistent. When MMV is less than 6 L/min, the PLMA cannot even satisfy basal minute ventilation requirements.
When the study began, we were not familiar with normal and abnormal values of MMV. As the study progressed, we came to consider MMV values less than or equal to 12 L/min as critical for both adult women and men (Table 2). In every case with MMV less than 8 L/min, the PLMA was removed. However, both a prostatectomy and a hemicolectomy were completed in patients with a MMV of 10 and 12, respectively. Therefore, in our practice we have demonstrated a threshold for removal of the PLMA when the MMV reaches a level somewhere between 6 and 12 L/min. In all cases, the PLMA was removed during the start of anesthesia, before the initiation of surgery.
Anesthesiologists should be alerted to the potential for significant airway obstruction in any patient with a MMV less than 12 L/min. Oxygen saturation may remain normal, and we emphasize that normal oxygen saturation does not guarantee the ability to satisfactorily eliminate CO2(12). Instead, the decision to remove the PLMA must usually be based on patient and surgical factors. For example, in one case, the PLMA was removed because of critical MMV and patient obesity (136 kg). In other cases, the PLMA was removed in light of surgical duration and complexity, including its intended use for patients undergoing sigmoid resection, colovesical fistula repair, and prostatectomy. Finally, when the MMV is dramatically reduced, it is common to ventilate using the maximum achievable airway pressures. This in turn taxes the PLMA to guarantee such a seal for the entire surgery and is yet another reason to opt for removal of the device.
Despite the criticisms expressed in this study, we believe that the PLMA has the potential to alter anesthesiology practice like its Classic™ predecessor (Laryngeal Mask Company). It is usually a very capable LMA for provision of PPV. All the patients in this study received PPV and muscle relaxation, something we only occasionally do with the Classic™ LMA. Of the 317 patients, roughly 40% involved surgeries where previously the authors had strictly used endotracheal intubation including 26 prostatectomies, 37 hysterectomies, 16 major colorectal resections, 16 laparoscopic cholecystectomies, 12 laparotomies, and 24 other varied cases requiring muscle relaxation. In addition, the PLMA was used successfully in 42 of 43 obese patients with a BMI >35.
In summary, anesthesiologists should not have the expectation that the PLMA will work ideally with PPV in every patient. The MMV test provides a powerful technique to quantitatively assess the quality of airway patency after PLMA insertion. When the MMV is critical, consideration of the margin of safety of minute ventilation, patient factors, and surgical duration and complexity should be made with regard to suitability of the PLMA as the airway for the operation. In the current study, approximately one of 20 patients encountered critical MMV, and we removed the PLMA at the start of the anesthetic in approximately one of 50 patients. Usually oxygen saturation remains normal, and it is possible to suction the stomach with an orogastric tube before the exchange of the airway if that decision is made. We strongly believe that such decisions are best made at the outset of the anesthetic before the operation.
1. Brain AIJ, Verghese C, Strube PJ. The LMA ‘ProSeal’: a laryngeal mask with an oesophageal vent. Br J Anaesth 2000; 84: 650–4.
2. Stix MS, Rodriguez-Sallaberry FE, Cameron EM, et al. Esophageal aspiration of air through the drain tube of the ProSeal laryngeal mask. Anesth Analg 2001; 93: 1354–7.
3. O’Connor CJ Jr, Davies SR, Stix MS, Dolan RW. Gastric distention in a spontaneously ventilating patient with a ProSeal laryngeal mask airway. Anesth Analg 2002; 94: 1656–8.
4. Brimacombe J, Richardson C, Keller C, Donald S. Mechanical closure of the vocal cords with the laryngeal mask airway ProSeal. Br J Anaesth 2002; 88: 296–7.
5. Hyatt RE, Scanlon PD, Nakamura M, eds. Interpretation of pulmonary function tests. Philadelphia, PA: Lippincott-Raven, 1997:18–22.
6. Miller RD, Hyatt RE. Obstructing lesions of the larynx and trachea: clinical and physiologic characteristics. Mayo Clin Proc 1969; 44: 145–61.
7. Miller RD, Hyatt RE. Evaluation of obstructing lesions of the trachea and larynx by flow-volume loops. Am Rev Respir Dis 1973; 108: 475–81.
8. O’Connor CJ Jr, Borromeo CJ, Stix MS. Assessing ProSeal laryngeal mask positioning: the suprasternal notch test. Anesth Analg 2002; 94: 1374–5.
9. O’Connor CJ Jr, Stix MS. Bubble solution diagnoses ProSeal insertion into the glottis. Anesth Analg 2002; 94: 1671–2.
10. Fisher DM. Anesthesia equipment for pediatrics. In: Gregory GA, ed. Pediatric anesthesia. 3rd ed. New York: Churchill Livingstone, 1994: 197–225.
11. Pressman JJ. Sphincter action of the larynx. Arch Otolaryngol 1941; 33: 351–77.
© 2002 International Anesthesia Research Society
12. Lumb AB. Nunn’s applied respiratory physiology, 5th ed. Oxford, UK: Butterworth Heinemann, 2000: 222–48.