Case reports

A 20 year old man awaiting double lung transplantation was established on NIPPV because of deterioration in his spirometric values and blood gas tensions. He was infected withBurkholderia cepacia, had a forced expiratory volume in one second (FEV₁) of 0.6 l, forced vital capacity (FVC) of 1.2 l, required continuous intravenous antibiotic therapy and overnight gastrostomy feeding. Capillary blood gas analysis while breathing room air revealed oxygen partial pressure (Po ₂) of 51 mm Hg (6.8 kPa) and carbon dioxide partial pressure (Pco ₂) of 43 mm Hg (5.7 kPa). On 24% oxygen, capillary blood gas analysis showed a Po₂ of 59 mm Hg (7.9 kPa) and aPco₂ of 48 mm Hg (6.4 kPa). He was commenced on NIPPV using a pressure cycled ventilator (NIPPY, Friday Medical, London, UK) with the pressure set at 24 mm H₂O and the inspiratory and expiratory times matched to the patient's breathing pattern (0.9 and 2.0 seconds, respectively). The pressure was later increased to 28 mm H₂O as he developed CO₂retention. Once established on the ventilator for 8–12 hours at night with 2 l/min supplemental oxygen, early morning capillary blood gas tensions were Po ₂ 83.8 mm Hg (11.2 kPa) and Pco ₂ 44.7 mm Hg (5.6 kPa).

Four months after commencing NIPPV he developed a small right pneumothorax (less than 10%) following the insertion of a totally implantable venous access device. The patient was asymptomatic but, as it was thought that NIPPV could increase the size of the pneumothorax, ventilatory support was withdrawn. His gastrostomy feeding was also discontinued and he was oxygenated overnight using a 24% Venturi mask. The following morning capillary blood gas analysis showed a Po₂ of 71 mm Hg (9.5 kPa) andPco₂ of 51.2 mm Hg (6.8 kPa). Overnight feeding was reintroduced at 50% of the usual volume (500 ml). NIPPV continued to be withheld and five days after the first pneumothorax a chest radiograph showed a new left sided pneumothorax (less than 10%). The aetiology of the second pneumothorax was unclear and fortunately the patient remained in a stable condition. Both pneumothoraces slowly decreased in size and eight days after the initial pneumothorax capillary blood gas analysis on 24% oxygen showed aPo₂ of 62 mm Hg (8.3 kPa) and aPco₂ of 44.6 mm Hg (5.9 kPa). At this time his chest radiograph was deteriorating and he was experiencing difficulty in expectorating sputum despite adequate hydration, intravenous antibiotic therapy, and the concomitant use of DNase. NIPPV was therefore recommenced (at a reduced pressure of 18 cm H₂O) to facilitate sputum expectoration and to improve his blood gas tensions. After two hours a chest radiograph revealed that the pneumothoraces had not increased in size and his capillary blood gas tensions had improved (Po₂ 89 mm Hg (11.8 kPa), Pco₂ 43.4 mm Hg (5.8 kPa)). He continued to receive NIPPV overnight and the next morning he felt better and was more mobile. However, at midday he developed surgical emphysema on his chest wall. A chest radiograph showed that the pneumothoraces had not increased in size but the NIPPV was further withheld. The following day his surgical emphysema had resolved and NIPPV and full overnight feeding were recommenced without further problems. He was discharged home approximately one week later and capillary blood gas analysis at discharge using NIPPV with 2 l/min supplemental oxygen gave a Po₂ of 78 mm Hg (10.4 kPa) and Pco₂ of 42 mm Hg (5.6 kPa). Three months later he received a double lung transplant.

Two further transplant listed patients with cystic fibrosis dependent on NIPPV attending our unit have developed pneumothoraces. A 28 year old man with an FEV₁ of 0.8 l who had been established on NIPPV for 22 months developed a large (50%) left sided pneumothorax whilst using the ventilator. His NIPPV pressure was set at 19 mm H₂O. The pneumothorax was not associated with a respiratory exacerbation, but he was treated with intravenous antibiotic therapy and intravenous aminophylline to aid sputum expectoration while his NIPPV was discontinued. Capillary blood gas analysis while breathing room air gave a Po₂ of 53 mm Hg (7.1 kPa) and a Pco₂ of 41 mm Hg (5.5 kPa). The pneumothorax was drained using an intercostal tube and the lung re-expanded quickly, following which capillary blood analysis on 24% oxygen gave a Po₂ of 61 mm Hg (8.1 kPa) and a Pco₂ of 44.9 mm Hg (6.0 kPa). The patient maintained adequate blood gases for the duration of his admission and only required further NIPPV one month after his pneumothorax. He remains well, but NIPPV dependent, 27 months following his pneumothorax.

The third NIPPV dependent cystic fibrosis patient to develop a pneumothorax attended clinic with neck discomfort. She was not breathless. A chest radiograph demonstrated surgical emphysema in the neck and air in the pericardial space, but no evidence of a pneumothorax. NIPPV was withheld and she was maintained on 24% oxygen overnight. As she did not deteriorate it was decided to withhold NIPPV until the surgical emphysema had resolved, but two days after presentation she received a double lung transplant.

Discussion

At the Manchester Adult Cystic Fibrosis Unit we have treated 52 of 334 patients (16%) for a pneumothorax over the last 17 years and commenced 23 patients on NIPPV. Only three of our patients have developed a pneumothorax whilst dependent on NIPPV and one of these was iatrogenic. It is surprising that the reporting of pneumothoraces either as a consequence or coincident with NIPPV is not more common. However, our figures and those from another large adult cystic fibrosis unit indicate that this is a rare event.2 In other respiratory diseases the occurrence of a pneumothorax while patients are being treated with NIPPV is also rare. This probably reflects that, in conditions such as kyphoscoliosis and the neuromuscular disorders, the lungs are relatively normal. In contrast, cystic fibrosis is characterised by progressive pulmonary infection and apical subpleural bronchiectatic cysts that often rupture. In the future the incidence of NIPPV associated pneumothorax may increase as NIPPV is used more widely in patients with cystic fibrosis.

The ventilator described in these cases is pressure preset. The inspiratory and expiratory times are individualised to allow sufficient expiratory time to avoid autopositive end expiratory pressure and breath stacking.3 The pressure is then set to improve alveolar ventilation. It is important to balance adequate ventilation with peak inspiratory pressure, especially in those patients with low pulmonary compliance, to reduce the risk of barotrauma. Permissive hypercapnia may be acceptable in this situation as long as the patient remains asymptomatic. Pressure support ventilation may be preferable in patients with subpleural cysts who are at risk of barotrauma, as the peak airway pressure (including the swings associated with coughing on the ventilator) and the mean airway pressure may be lower. Conversely, it may be necessary to produce higher pressures in acutely sick patients. Some cystic fibrosis units use machines that generate higher pressures than the ventilator described in this paper. It is important to be aware of the maximum pressures generated and to balance the risk of barotrauma with patient benefit. We are not aware of any reports describing the relative risks of volume and pressure preset nasal ventilators in cystic fibrosis.

The development of a pneumothorax in a patient with cystic fibrosis receiving NIPPV can be a challenging clinical situation. As in the first case described here, pneumothorax may cause an abrupt deterioration in respiratory status and at the same time necessitate withdrawal of NIPPV. However, as demonstrated by the second case, other patients who are not in a respiratory exacerbation may remain relatively stable when NIPPV is withdrawn, reflecting that their “physiology” has been reset by NIPPV. If NIPPV is withheld or the inflation pressure is reduced and the Pco₂ rises, hypercapnia is acceptable in the short term as long as the patient remains asymptomatic. Alternatively, in a patient who requires continuing NIPPV following a pneumothorax, it can be delivered safely providing the pneumothorax is being drained through a patent intercostal tube.

An additional perceived benefit of NIPPV in patients with cystic fibrosis is that it may aid sputum expectoration. However, when NIPPV is withdrawn following pneumothorax without chest intubation, sputum clearance techniques are modified to avoid increases in intrathoracic pressure which may result in sputum trapping. Our experience shows that good hydration, intravenous aminophylline, intravenous antibiotic therapy, DNase, and adequate analgesia may be of benefit. The reintroduction of NIPPV can be associated with further air leaks and the length of time that a patient refrains from NIPPV following a pneumothorax is dictated by clinical circumstance. Reducing ventilator pressures may facilitate the reintroduction of NIPPV. However, it is important to appreciate that the pressure delivered by a non-invasive ventilator is measured from within the mask and is not necessarily the pressure achieved in the trachea and alveoli, as there is usually a significant fall in pressure across the nose and upper airway. This may be especially true in cystic fibrosis patients with rhinitis and nasal polyps. Overnight feeding can precipitate type 2 respiratory failure in patients with cystic fibrosis who have severe lung disease due to splinting of the diaphragm and increased CO₂ production. In NIPPV dependent cystic fibrosis patients with a pneumothorax it may therefore be prudent to discontinue feeding until NIPPV can be reinstituted.

This brief report highlights the dilemma of managing the rare complication of pneumothorax in NIPPV dependent adult patients with cystic fibrosis. Clearly, each patient requires meticulously supervised and individualised management with the risks of mechanical ventilation being balanced against the potential benefits.