Author + information
- Piergiuseppe Agostoni, MD, PhD∗ (, )
- Mauro Contini, MD,
- Carlo Vignati, MD,
- Alberico Del Torto, MD,
- Giorgio De Vecchi Lajolo, MD,
- Elisabetta Salvioni, PhD,
- Emanuele Spadafora,
- Carolina Lombardi, MD,
- Gino Gerosa, MD,
- Tomaso Bottio, MD,
- Marco Morosin, MD,
- Vincenzo Tarzia, MD,
- Silvia Scuri, PhD Eng,
- Gianfranco Parati, MD and
- Anna Apostolo, MD
- ↵∗Centro Cardiologico Monzino, IRCCS, Via Parea 4, 20138 Milan, Italy
Central sleep apnea (CSA) is reported in advanced heart failure (HF) (1). At least 3 factors have been suggested to play a pivotal role in CSA pathogenesis, specifically hyperventilation, which is the likely consequence of an enhanced stimulation of intrapulmonary receptors; low cardiac output leading to increased circulatory delay; and carbon dioxide cerebrovascular reactivity (1). The increased circulatory delay from the lungs to peripheral receptors imposes a temporary misalignment between chemical signals and ventilation (1). Although suggested, a role of low cardiac output in CSA pathogenesis has little supporting evidence. Indeed, no significant differences in cardiac output or in circulatory delay have been found between patients with HF with and without CSA (2).
Left ventricular assist devices (LVAD) are used to treat patients with severe HF. In some cases, LVAD blood flow can be increased simply by increasing LVAD pump speed. However, because pump flow is pre- and post-load dependent, a precise relationship between pump speed and blood flow changes cannot be anticipated (3).
We analyzed the effect of pump speed changes on blood flow in 17 patients (mean age 61.3 ± 7 years) who successfully underwent LVAD (Jarvik 2000; Jarvik Heart, New York, New York) implantation, as measured by inert gas rebreathing technology (Innocor; Innovision, Odense, Denmark) (4), on respiratory function and ventilation during sleep. The Jarvik 2000 has 5 different settings for rotational speeds, allowing progressively greater pump flow.
On the first day, patients (pump speed 3) underwent clinical, laboratory, echocardiographic, and spirometric (including alveolar-capillary diffusion [DL] for nitric oxide and carbon monoxide) assessments and cardiopulmonary exercise testing (SensorMedics, Yorba Linda, California). Thereafter, patients were studied with pump speed randomly set at 2 or 4 in the following sequence: clinical monitoring for a few hours to confirm hemodynamic stability, 5-channel portable polysomnography (Embletta X100; Embla Systems, Pleasanton, California), and, the following morning, cardiac output at rest, spirometry, DLCO, and DLNO. Thereafter, measurements were repeated with the same sequence after changing the pump speed. Spirometry and cardiopulmonary exercise testing at pump speed 3 showed a restrictive defect with moderate to severe DL reduction and relevant exercise impairment.
Changing pump speed induced trivial spirometric and DL changes. Cardiac output was 3.26 ± 0.94 l/min and 3.63 ± 1.08 l/min (p = 0.03) with pump speed set at 2 and 4, respectively. Polysomnography showed a time in bed of 475 ± 53 min and 504 ± 31 min (p = 0.07) and oxygen saturation signal integrity (time with evaluable oxygen saturation signal) of 78.8 ± 19.6% and 53.6 ± 28.1% of recording (p < 0.01) at speeds 2 and 4, the latter probably due to difficulty reading oxygen saturation in the absence of pulsatile flow. Increasing cardiac output (speed 4) decreased apnea-hypopnea index (AHI) (from 20.7 ± 16.3 to 11.85 ± 10.38, p = 0.008), hypopnea index (from 6.7 ± 4.4 to 2.7 ± 2.7, p < 0.001), and CSA, but not obstructive or mixed apnea (Figure 1). Patients with the highest AHI and CSA were those with the greater reduction at pump speed 4. In the 6 patients with greater CSA index reduction (>1/h), CSA decrease was related to cardiac output changes (R2 = 0.4949). Increasing pump speed caused reductions in Cheyne-Stokes respiration in all patients (from 44.5 ± 13.4% to 23.0 ± 23.0%, p < 0.05).
Circulatory delay and cardiac output did not correlate, nor did their changes, with the 2 pump speeds. Similarly, no correlation was detected between cardiac output and circulatory delay and AHI and CSA. A few reasons may explain this finding. First, cardiac output was by necessity measured during the waking state with patients in a sitting position, whereas circulatory delay was measured during sleep with patients lying in bed. Thus, the hemodynamic pattern was likely different. Moreover, patients with HF with elevated AHI and several CSAs have the same cardiac output and circulatory delay as patients who do not show these findings (1). However, patients with circulatory delay reductions >10% had greater AHI and CSA reductions and a trend toward greater cardiac output increases. Changing pump speed from 2 to 4 increased cardiac output by about 10%. The increased cardiac output did not influence lung mechanics and DL, but it significantly reduced sleep-related breathing disorders, with a >50% reduction of AHI, due to hypopnea and CSA reduction. Accordingly, changes in AHI are unlikely related to changes in pulmonary receptor stimulation. All reported observations confirm the dependence of CSA but not of obstructive and mixed apneas on cardiac output. However, it should be emphasized that both cardiac output and circulatory delay were not directly linked to AHI or CSA, showing that low cardiac output and high circulatory delay are among the triggers needed for the presence of AHI and CSA in patients with HF but, by themselves, not sufficient to trigger CSA (1). It is unknown if the presence of high AHI and CSA is associated with a worse prognosis in LVAD patients, as it is in patients with HF. However, this is a likely possibility, given that AHI and CSA are associated with worsening of HF and with sudden cardiac death. It is possible that regulating nighttime LVAD pump speed to promote better breathing during sleep is a worthwhile practice.
Please note: Dr. Scuri is a consultant for Artech Srl. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation