CASE REPORT

A 40 year old woman was found to be anaemic at a routine blood donor session. Shortly after she developed joint pains and pruritus, and autoimmune haemolytic anaemia was diagnosed. A mediastinal mass was observed on her chest radiograph which was shown to be a thymoma on needle biopsy. The autoimmune haemolytic anaemia was complicated by red cell aplasia. Following prednisolone therapy the haemoglobin rose from 6.2 to 12.8 g/dl. There were no symptoms of myasthenia such as diplopia, limb weakness or dyspnoea, and the FEV₁/FVC was 2.39/2.98 litres. The patient underwent a thoracotomy during which the thymic tumour was found to be adherent to the right hilum, right phrenic nerve, and pericardium. The tumour was resected with strips of the right lung at the hilum. Histologically, the tumour was a cortical thymoma. Initially the patient made a good recovery on the ICU and was rapidly extubated. However, over the following 2 weeks she became progressively more breathless and orthopnoeic. The chest radiograph showed elevation of the right hemidiaphragm and bibasal consolidation (fig 1). A ventilation-perfusion scan confirmed matched defects at both lung bases. Little response was seen to several courses of antibiotics and physiotherapy, and the patient reported continued orthopnoea and fragmented sleep. She was therefore referred to the weaning programme at the Royal Brompton Hospital for further assessment.

On arrival, breathing spontaneously she was unable to lie flat and the FEV₁/FVC while sitting was 1.05/1.2 litres. Arterial blood gas tensions on air were Po₂ 10.6 kPa, Pco₂ 5.9 kPa, HCO₃ 30.3 mmol/l. Overnight monitoring during which the patient slept very lightly showed a rise in transcutaneous (Tc) CO₂ to 12 kPa with dips in arterial oxygen saturation (Sao₂) to 60% breathing air (fig 2A). A CT scan showed patchy subsegmental atelectasis affecting the right and left lower lobes with marked elevation of the right hemidiaphragm. There was no evidence of thromboembolism. Baseline respiratory muscle test results are shown in table 1 and indicate a marked reduction in inspiratory muscle strength with a modest reduction in expiratory muscle strength. Stimulation of the phrenic nerves generated no transdiaphragmatic pressure spike bilaterally. A tensilon test using the diaphragm as the test muscle was negative (fig 3).

Thyroid function tests demonstrated hypothyroidism (free thyroxine 7.6 pmol/l (NR 9–23), thyroid stimulating hormone 17.7 mU/l (NR 0.32–5)).

Diagnosis

The patient had developed bilateral basal atelectasis and consolidation after thoracotomy for thymic resection because of marked diaphragmatic weakness. Ventilatory decompensation occurred during sleep due to loss of intercostal muscle tone in REM sleep leaving ventilation dependent on the already weak diaphragm. Ventilation is further compromised during sleep by a reduction in central drive and basal ventilation-perfusion mismatch. In this patient the differential diagnosis of respiratory muscle weakness lies between:

• myasthenia gravis;

• phrenic nerve injury caused by thymic tumour involvement and/or surgery;

• a combination of myasthenia and phrenic palsy;

• respiratory muscle weakness associated with hypothyroidism.¹

The tensilon test was negative and respiratory muscle tests showed absent conduction down both phrenic nerves indicating that the main problem was phrenic nerve injury. The situation was probably exacerbated by hypothyroidism induced myopathy. Acetylcholinesterase (Ach) antibody was subsequently shown to be positive in this patient, but the combination of red cell aplasia, thymoma and Ach antibodies without clinical features of myasthenia has been reported previously. In a Japanese series of 17 cases of red cell aplasia and thymoma, only two patients had myasthenia.²

Management

The patient was started on nocturnal NIV via a nasal mask and thyroxine replacement therapy. She also received physiotherapy during NIV to facilitate coughing and secretion clearance. Theoretically, in this situation bilevel non-invasive positive pressure support may be more beneficial in addressing atelectasis than volume preset ventilation or inspiratory pressure support alone.³ The patient’s sleep quality improved, and overnight monitoring on NIV showed improvements in Sao₂ and Tcco₂ (fig 2B). Persistent anaemia was treated by transfusion to raise the haemoglobin above 8 g/dl. Basal atelectasis resolved over 1 week but the right hemidiaphragm remained elevated. The patient was discharged home after 16 days and continued to use nocturnal NIV.

Two months later the sniff inspiratory pressure was 27.6 cm H₂O compared with a value of 16.2 cm H₂O on arrival, and the patient had returned to work having completed a course of radiotherapy. Respiratory muscle test results obtained 6 months after surgery are given in table 1. These show a further improvement in inspiratory muscle strength with some recovery in conduction down the left phrenic nerve (fig 4), indicating that it was probably traumatised during the difficult surgical resection. There was no recovery in right phrenic nerve function, presumably because of partial resection. A sleep study with the patient breathing spontaneously showed normal Tcco₂ and Sao₂ values, so the patient was weaned from nocturnal ventilatory support.

DISCUSSION

Phrenic nerve injury occurs variably after cardiothoracic surgery⁴ and ranges from complete nerve ablation to the “frostbitten phrenic” seen after procedures involving topical cooling to produce cardioplegia.⁵ If the nerve is not irretrievably damaged, recovery is usually seen over a number of months.⁶ Patients with bilateral phrenic nerve injury often present with bibasal atelectasis. In those with underlying respiratory compromise overt ventilatory failure is precipitated.⁷ In this non-smoker, ventilatory decompensation only occurred at night due to the effects of sleep on respiration, but sleep fragmentation exacerbated her daytime symptoms. Respiratory muscle tests⁸ were used to help differentiate a true myasthenic syndrome from phrenic nerve damage.

NIV is a useful weaning tool. A randomised trial comparing rapid extubation on to NIV with continued intubation and pressure support ventilation in COPD patients showed more rapid weaning with fewer complications in the NIV group.⁹ In patients with restrictive ventilatory defects due to neuromuscular or chest wall disease, case series data¹⁰ suggest that NIV can shorten weaning and reduce the time spent on the ICU. NIV can also be used to prevent the need for reintubation if ventilatory failure recurs after extubation.¹¹ In the case presented here, reintroduction of invasive ventilation following extubation was not indicated but NIV is likely to have facilitated recovery from basal atelectasis. NIV was successfully carried out on a high dependency unit and subsequently on a general respiratory ward, thereby eliminating the need for continued ICU bed occupancy.