Predictors of survival in patients treated with long-term oxygen therapy

Long-term oxygen therapy (LTOT) has been shown in two controlled trials to improve survival in patients with hypoxic chronic obstructive pulmonary disease. The cause of this improved survival is unknown and was best correlated in the Medical Research Council (MRC) trial with the sum of the PaCO2 and the red cell mass. However, the survival curves of treated and untreated patients in the MRC trial for males were not significant until greater than 500 days of treatment. This suggests that some patients derive little benefit from LTOT. Follow-up data from the National Institutes of Health (NIH) trial indicate that pulmonary haemodynamics, including pulmonary vascular resistance, predict survival in such patients. In a previous study in 115 randomly selected patients with chronic obstructive pulmonary disease (COPD) who were not treated with LTOT, we found that PaO2 and PaCO2, measured when breathing air, were more significant predictors of survival than the presence of cor pulmonale or right ventricular function. We have also compared a group of 154 patients treated with LTOT in two centres, in Warsaw and Edinburgh, where criteria for patient selection for LTOT and subsequent survival curves were very similar in both centres. A pulmonary arterial pressure of greater than 29 mm Hg and a PaO2 when breathing air of less than 7.0 kPa produced a significant decrease in survival using a Cox survival model. These results suggest that patients with more severe COPD may not benefit from LTOT. Other factors such as poor compliance with treatment, lack of adequate oxygenation and increasing age of patients prescribed LTOT may also contribute to the reduced benefit from this treatment.(ABSTRACT TRUNCATED AT 250 WORDS)