Alternative (or additional) explanations for the observation that some COPD patients appear to hypoventilate (i.e. their PaCO2 rises) when given supplemental oxygen revolve around two pieces of basic physiology.
Firstly, (and possibly most importantly), pulmonary arteries respond to a hypoxic environment by contracting. This makes good physiological sense, as there is little point in sending deoxygenated blood to poorly ventilated regions of the lung. When lung architecture is severely damaged – as in emphysema – large regions of the lung are hypoxic and such hypoxic vasoconstriction can occur on a large scale. Indeed, this phenomenon leads to the development of pulmonary hypertension in such patients. So picture the lung of a COPD patient: some areas retain normal structure and function while others are useless as gas exchange surfaces. Both regions get a blood supply, but hypoxic vasoconstriction limits blood flow to the damaged regions, diverting blood to the functionally-intact zones. When the fraction of inhaled oxygen (FiO2) is increased, hypoxic vasoconstriction is lessened and blood is diverted away from the areas where gas exchange can take place and so CO2 accumulates rather than being blown off.
Secondly, the impression that hypoventilation is occurring could be a result of the Haldane effect. CO2 can be transported by haemoglobin (see CO2 transport), and in a reduced O2 environment blood can carry more CO2. This is the Haldane effect and accounts for venous blood being able to carry more CO2 than arterial blood. So, if you increase FiO2 in a patient with a low PaO2, you drive CO2 off haemoglobin and into the aqueous phase that is measured by blood gas analysers. Remember, an ABG reading tells you about dissolved gases only, not other transport mechanisms (which is why they can trick you in an anaemic patient).
These two mechanisms will make a contribution in addition to the blunting of hypoxic drive that may be occurring. How important each mechanism is in determining the response to increasing the FiO2 will vary enormously from patient to patient. Different mechanisms may make a contribution at different times, too. For example, the initial response to increased FiO2 may be blunting hypoxic drive, while other mechanisms may make a contribution later on. So, when a senior colleague announces that “hypoxic drive doesn’t exist!” nod enthusiastically, but remember that it’s been a long time since they were in medical school learning the basics.