Oxygenation pattern and compensatory responses to hypoxia and hypercapnia following bilateral carotid body resection in humans

Key points Five years after bilateral carotid body resection (bCBR) performed in four patients, the absence of the hypoxic ventilatory response persisted, suggesting no compensatory regrowth. Breathing hypoxic gas mixtures (15% and 12%) results in a lower (by similar to 10%) minimal blood oxygen saturation (SpO2) in bCBR patients compared to heart failure subjects (CHF) with intact peripheral chemoreceptors. After bCBR, patients were characterized by a greater short-term variability in SpO2 during mild hypoxia in comparison to the CHF group. The ventilatory response to hypercapnia was preserved following bCBR and was sufficient to maintain minimal SpO2 at levels comparable to controls when combined with hypoxia. Bilateral CBR - a novel treatment modality for sympathetically mediated diseases - should be used with caution due to the risk of significant desaturation even during mild hypoxia equivalent to that experienced during long-haul air travel and high altitude. Carotid body resection has been proposed as a novel treatment for sympathetically mediated diseases but the safety of bilateral carotid body resection (bCBR) for blood oxygenation during hypoxic stress (long-haul flights or high altitude) remains uncertain. Also unknown is whether central ventilatory drive is sufficient to maintain adequate oxygen saturation when exposed to hypercapnia with concomitant hypoxia. Thus, we administered: 15% O-2, 12% O-2, 5% CO2/12% O-2 and 5% CO2/95% O-2 to a group of four patients with congestive heart failure (65 +/- 2.9 years) in whom bCBR was performed 5 years earlier. Ventilatory, haemodynamic and blood oxygen saturation (SpO2) responses were recorded non-invasively and compared to control groups with intact peripheral chemoreceptors (both healthy and heart failure patients). First, we confirmed that the ventilatory response to hypoxia was eliminated in patients with bCBR, although the increase in cardiac output was preserved. Second, administration of hypoxic gas mixtures resulted in a larger decrease in SpO2 and greater short-term variability of the SpO2 leading to a lower minimal SpO2 for both hypoxia levels in the bCBR group compared to heart failure controls (82.5 +/- 1.2% vs. 91.6 +/- 2.3% for 15% O-2 and 73.8 +/- 4.0% vs. 83.7 +/- 3.1% for 12% O-2). Third, in bCBR patients the ventilatory response to hypercapnia was present and sufficient to maintain a minimal SpO2 at a level comparable to heart failure controls following administration of 5% CO2/12% O-2 (88.7 +/- 4.2% vs. 91.1 +/- 2.8%). We conclude that bCBR carries a risk of significant oxygen desaturation even during mild hypoxia. Despite preservation of central chemosensitivity, future studies should focus on unilateral CBR or on pharmacological modulation of peripheral chemosensitivity.