Effects of midline thoracotomy on pulse pressure variations during pressure-control ventilation

Objective

During mechanical ventilation, the heart-lung interaction induces a cyclic oscillation of the arterial pulse pressure. The measure of this respiratory changes, called pulse pressure variation (PPV), is one of the most reliable index of fluid responsiveness. It is used as a functional hemodynamic monitoring in several conditions (general anesthesia, sepsis, ALI/ARDS). During cardiac surgery, midline thoracotomy significantly alters heart-lung interaction and, consequently, PPV. The aim of the study was the evaluation of the effects of sternotomy on PPV during pressure-control ventilation (PCV).

Methods

Nineteen patients (age 62 ± 10 years) undergoing elective CABG in a tertiary university hospital were enrolled. A Swan-Ganz catheter, an arterial catheter and a central venous catheter was inserted in order to collect pressure waveforms. After the induction of general anesthesia all the patients were mechanically ventilated (PCV), setting pressure values in order to obtain a tidal volume of 8 ml/kg. Hemodynamic data were collected 5 min before and after sternotomy. PPV was calculated offline from the collected waveforms, according to the formula reported by Michard and colleagues [1].

Results

The PPV, cardiac index, stroke volume, mean arterial pressure, airway pressure and tidal volume did not change after sternotomy. We subsequently differentiated patients according to PPV values (Fig. 1). In the subgroup of patients with PPV > 13% (7/19 patients), we found a good correlation between PPV and Paw (Pearson correlation 0.861 P = 0.03; R² = 0.74 P = 0.049); after sternotomy, PPV was significantly reduced (15.4 ± 2.8% vs 8.2 ± 1.6%, P = 0.043) and it was no more correlated with Paw. In the subgroup of patients with PPV < 13% (12/19 patients), we did not find any correlation between PPV and Paw, and sternotomy had no effects on hemodynamic data.

Conclusion

During PCV, airway pressure affects PPV only when patients are in a 'fluid responsive' status (PPV > 13%); similarly, sternotomy reduces PPV only when baseline is above the threshold value of 13%. It may thus be possible that midline thoraco-tomy makes a 'fluid responsive' patient unresponsive to a fluid challenge by leading his heart to work on the plateau portion of the Frank-Starling curve. This hypothesis would be confirmed by the lack of correlation between airway pressure and PPV after opening the thorax.