I wanted to share this fascinating and incredibly educational case – one where I learned an enormous amount about critical care medicine, pulmonary physiology, and ventilator management.
24 y.o. previously healthy female transferred from an OSH with ARDS. She developed nausea and vomiting three days after sinus surgery with subsequent shortness of breath. At presentation, had RLL infiltrate which progressed to diffuse bilateral infiltrates with some cavitary lesions, worse on the right. Sputum culture grew MRSA – she is on vancomycin. She also has developed a pneumothorax on the right and has a chest tube in place, although it was clamped during her medical flight to our hospital and she has developed lots of subcutaneous emphysema. The sq emphysema slowly resolves over a couple of days with chest tube on -20 cm of water suction, but she continues to have a persistent air leak of moderate volume (about 30% of her tidal volume) consistent with a bronchopleural fistula (BPF).
Initially, she was on 80% FiO2 and PEEP of 14 cm H2O. She remained this hypoxic for the first couple of days. Over the next week, her oxygenation slowly improved and she was weaned to 40% FiO2 and PEEP 5 cm H2O. She continued to have a significant air leak, now about half of her tidal volume, despite changing the pleurovac to water seal in an attempt to minimize the leak. Her ABG at this point was 7.33/62/70 on 40% FiO2. Despite having reasonable pulmonary function on minimal ventilator settings (Assist Control Volume Control ventilation, tidal volume 320 cc, PEEP 5, 40%, Plateau Pressure around 20 and peak airway pressure 30, RR 25/min), she rapidly failed any attempt at an SBT with both tachypnea and hypoxia. In addition, over the previous three days, she has had progressive collapse of her contralateral (i.e. left) lung (bronchopleural fistula is on the right).
What might be going on?
Given the inability to effectively wean her off the ventilator (repeated failure of SBTs), it was thought that functioning of her left lung was needed for further weaning. Bronchoscopy, looking for mucus plugging of the left mainstem bronchus, was performed but only minimal secretions were found. (Incidentally, during bronchoscopy, we also confirmed that her bronchopleural fistula on the right was isolated to a single subsegment of her right middle lobe). With mucus plugging not likely the cause of her left lung collapse, the possibility of her not receiving any (or at least markedly reduced) air flow to that left lung because of the extremely low resistance for air to flow through the rather large bronchopleural fistula, a sort of pulmonary steal phenomenon where the right lung (and specifically the BPF on the right) “steals” all the flow from the left lung, was entertained. In order to both diagnose, and treat this, we decided to exchange her endotracheal tube for a double lumen endotracheal tube and perform independent lung ventilation (left mainstem bronchial placement of double lumen ETT demonstrated by arrow).
After placing the double lumen endotracheal tube, we began independent lung ventilation (Cheatham ML, et al. Am Surg. 2006;72(6):530-533.) by applying half the previously delivered tidal volume to each lung separately. While this worked fine on the right lung with reasonable airway pressures, the left lung had markedly compromised compliance resulting in the ventilator on the left lung alarming for high peak airway pressures – when the alarm was adjusted to deliver the tidal volume (of 160 cc), the peak airway pressure in the left lung was 109 cm H2O. As this was not considered reasonable or safe to continue, ventilation of the left lung was changed to a slow, recruitment method using bivent (or Airway Pressure Release Ventilation (APRV)) ventilation. Initially 40 cm of driving pressure yielded a whopping 40 cc of tidal volume – fortunately, this gradually increased over about 40 hours to 200 cc of tidal volume (from the isolated left lung) as the lung slowly expanded.
Initially after transitioning to double lung ventilation, her left lung contributed little to her pulmonary function and she lived off her right lung (yes, the one with the bronchopleural fistula). Anytime her right lung lost ventilation (like with a mucus plug of that part of the double lumen endotracheal tube), she would quickly become hypoxic, hypercarbic, and hypotensive. On day 2, the ventilator ventilating her right lung rather quickly began alarming for high peak airway pressures, while the tidal volume on the left lung (still on Phigh of 30cm of H2O) had suddenly increased to almost 800cc. In addition, she became almost immediately hypotensive and the airleak into her pleurovac had entirely gone away.
Uh, oh, what happened?
In troubleshooting the situation, she was found to have significant mucus plug in the endotracheal tube lumen for the right lung, with subsequent collapse of the right lower and middle lobes. As a result of this, she developed hyperinflation (and herniation across midline) of her left lung (which was being ventilated with bivent, or a pressure control mechanism of ventilation). This was corrected by changing her left lung ventilator settings to volume control with a tidal volume of 200 cc and bronchoscopy with suctioning of the lumen of the endotracheal tube to the right lung.
Two days later, I was again called for her ventilator alarming. The ventilator for the right lung was again alarming high peak airway pressures, but this time her return tidal volume was 0 cc. In addition, the respiratory therapist reported to me that the patient had developed a cuff leak (and you could hear the airleaking with every breath from the ventilator for the right lung). The ventilator for the left lung, again had increased tidal volumes, this time to about 500 cc. The pleuravac continued to have an intermittent airleak, although interestingly, it now tidaled with the left lung ventilator than the right lung.
Now what has happened?
Troubleshooting this scenario revealed that the double lumen endotracheal tube had retracted a few centimeters so that the bronchial lumen (supposed to be in the left mainstem) had migrated back to the distal trachea. Because of this, the lumen was open to both the right and left mainstem bronchi and the left lung ventilator was now able to ventilate both lungs. Furthermore, the balloon for the bronchial lumen was now also in the distal trachea instead of in the left mainstem bronchus. With it in the distal trachea, it actually obstructed any volume of air that the right lung ventilator was attempting to deliver through the tracheal lumen, which was more proximal in the trachea. Because of this obstruction, pressure had built up between the bronchial and tracheal balloons, giving both high peak airway pressures on the right lung ventilator, and also an air leak back across the tracheal balloon (the pressure just distal to this balloon had increased to the point that the path of least resistance was back across the tracheal balloon and out the proximal trachea, oropharynx, through the mouth into the atmosphere (and this resulted in the airleak the respiratory therapist heard). Using bronchoscopy, the double lumen tube was repositioned so the bronchial lumen (and bronchial balloon) were once again in the left mainstem bronchus, resolving this problem.
She continued to have a significant bronchopleural fistula on the right and she was not thought to be a surgical candidate. As her pneumonia, septic shock, and ARDS had all improved at this point, it was clear that this BPF was preventing her from weaning from the ventilator and being able to be liberated from the ventilator. So what options are left for managing or treating her bronchopleural fistula?
After a few more days (now 8 days with independent double lung ventilation), she was trialed on the same settings for the ventilators for both her left and right lungs. She tolerated this well enough (and her pneumonia had been treated and resolved enough) that it was felt that an endobronchial valve could be placed in the subsegmental bronchus in the right middle lobe leading to her bronchopleural fistula. This type of subsegmental lobar bronchial block (Otruba Z, Oxorn D. Can J Anaesth. 1992;39(2):176-178.) using a one-way endobronchial valve was undertaken the next day with very good result. The subsegment of her right middle lobe collapsed over a couple of days and her airleak through her bronchopleural fistula essentially disappeared. Her double lumen endotracheal tube was change back to a single lumen tube and she tolerated this well. Both of her lungs remained inflated and functional and she began the process of ventilator weaning.
This was a fascinating case that I learned more critical care medicine and ventilator and pulmonary physiology than I had learned in the previous two to three years of caring for patients. Hopefully you have learned from this description of it.