Article Text

Young Research Workers' Prize
A Periodic breathing: novel experimental models to develop dynamic therapies
  1. R Baruah1,
  2. A Giannoni2,
  3. C Manisty1,
  4. K Willson1,
  5. Y Mebrate3,
  6. A Kyriacou1,
  7. S Sen1,
  8. M Shanmuganathan1,
  9. B Unsworth1,
  10. H Yadav1,
  11. J Davies1,
  12. R Sutton1,
  13. J Mayet1,
  14. A Hughes1,
  15. D P Francis1
  1. 1St Mary's Hospital, Paddington, London, UK
  2. 2Institute of Clinical Physiology, G. Monasterio Foundation, Pisa, Italy, Pisa, Italy
  3. 3Royal Brompton Hospital, London, UK


Background Periodic breathing (PB) is a common complication of heart failure (HF), characterised by oscillations in CO2 and ventilation and associated with increased morbidity and mortality. Studying potential interventions for PB is difficult, because it is highly variable. We propose two experimental models and test two new dynamic therapies that target CO2 oscillations directly.

Methods and Results First, we aimed to produce a human model of PB using pacemakers. Pacemakers were manipulated in twelve pacemaker recipients, six with HF (EF=23.7±7.3%) and six without, to induce oscillations experimentally. Second, we applied a real-time algorithm of pre-emptive dynamic exogenous CO2 administration, and tested different timings. We found that cardiac output alternation by pacemaker successfully induced CO2 and ventilatory oscillations. Dynamic CO2 therapy, when delivered coincident with hyperventilation, attenuated 64% of the experimentally-induced oscillations in end-tidal CO2: SD/mean 0.06±0.01 untreated vs 0.04±0.01 with treatment (p<0.0001) and 0.02±0.01 in stable state. This translated to a 53% reduction in induced ventilatory oscillations: SD/mean 0.19±0.09 untreated vs 0.14±0.06 with treatment (p=0.001) and 0.10±0.03 in stable state. Notably end-tidal CO2 did not significantly rise when dynamic CO2 was applied to the model (4.84±0.47 vs 4.91± 0.45 kPa, p=0.08). Furthermore, mean ventilation was also not significantly increased by dynamic CO2 compared with untreated (0.13±0.02 vs 0.14±0.02 l/s, p=0.17).

A second cohort of 14 patients with cardiac pacemakers (5 with HF (EF 27.6 ±13.9) and nine with normal systolic function) inhaled exogenous CO2 that cycled from 0 to 4.25±1.8% every 30 s again inducing PB. In this group, cardiac output was then oscillated every 30 s in an attempt to alter the induced ventilatory oscillations. When cardiac output was oscillated from low too high every 30 s to coincide with peak ventilation, the degree of induced oscillation in et-CO2 was reduced by 43% (SD/mean et-CO2 during induced oscillations alone=0.08±0.02 vs SD/mean et-CO2 with alternation of cardiac output =0.05±0.02, p<0.001). By reducing the degree of oscillation in et-CO2, the degree of induced ventilatory oscillation was reduced 55% (SD/mean ventilation during induced oscillations alone=0.13±0.06 vs SD/mean ventilation with alternation of cardiac output=0.09±0.05, p<0.01).

Conclusion Both dynamic CO2 administration and cardiac output modulation using a pacemaker, when correctly timed, successfully attenuated experimentally-induced ventilatory oscillations without increasing mean ventilation.

Abstract A Figure 1

Schematic of laboratory set-up. CO2 and air are allowed to mixed to any desired concentration using a computer-controlled motorised valve. This mixture is drawn in by the patient on inspiration and expired via a 3-way valve. Inspired and expired CO2 concentrations and repiratory flow were measured and fed back to an online computer program to establish the position within the ventilatory cycle and to display this and the end-tidal CO2.

  • Heart failure
  • periodic breathing
  • pacemaker

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.