Acute volume loading exacerbates direct ventricular interaction in a model of COPD

Volume loading may exacerbate adverse cardiopulmonary interaction in COPD however, the mechanisms remain unclear. We found that when negative intrathoracic pressure is increased, acute volume loading paradoxically reduces stroke volume. This reduction in stroke volume is considerably greater in a model of COPD, owing to the effects of lung hyperinflation.

Abstract

Volume loading increases left ventricular (LV) stroke volume (LVSV) through series interaction, but may paradoxically reduce LVSV in the presence of large increases in right ventricular (RV) afterload because of direct ventricular interaction (DVI). RV afterload is often increased in chronic obstructive pulmonary disease (COPD) as a result of pathological changes to respiratory mechanics, namely increased negative intrathoracic pressure (nITP), dynamic lung hyperinflation (DH), and increased pulmonary vascular resistance (PVR). These hallmarks of COPD negatively impact LV hemodynamics in normovolemia. However, it is unknown how these heart-lung interactions are impacted by acute volume loading. Twenty healthy subjects (23 ± 2 yr) completed the study protocol, involving acute volume loading via 20° head-down tilt (HDT) in isolation and with 1) inspiratory resistance of −20 cmH₂O (HDT+nITP) and 2) nITP, expiratory resistance to induce DH and hypoxic-mediated increases in PVR (HDT+COPD model). LV volumes and geometry were assessed using triplane echocardiography. HDT significantly increased LVSV by 10 ± 10% through an 8 ± 6% increase in LV end-diastolic volume (LVEDV). HDT+nITP paradoxically decreased LVSV by 11 ± 12% and LVEDV by 6 ± 9% from supine baseline, or −14 ± 10% LVSV and −15 ± 13% LVEDV from HDT (P < 0.001). HDT+COPD model decreased LVSV (21 ± 10% and 28 ± 11%) and LVEDV (16 ± 10% and 22 ± 10%) from both supine and HDT, respectively (P < 0.001). Under all conditions, significant septal flattening (increased radius of septal curvature) occurred, indicating DVI. Thus, when RV afterload is increased and/or an external constraint to ventricular filling exists, acute volume loading appears to paradoxically reduce LVSV. These findings have important implications for understanding how volume status impacts cardiopulmonary interactions in COPD.

NEW & NOTEWORTHY Volume loading may exacerbate adverse cardiopulmonary interaction in COPD; however, the mechanisms remain unclear. We found that when negative intrathoracic pressure is increased, acute volume loading paradoxically reduces stroke volume. This reduction in stroke volume is considerably greater in a model of COPD, owing to the effects of lung hyperinflation.

increased negative intrathoracic pressure (nITP), dynamic lung hyperinflation (DH), and elevated pulmonary vascular resistance (PVR) are hallmarks of chronic obstructive pulmonary disease (COPD) and interact to impair left ventricular (LV) function through both series and direct ventricular interaction (DVI) (12, 57). These pathological changes to pulmonary function are known to significantly impact LV hemodynamics in normovolemia (13). However, the influence of volume status on mediating these cardiopulmonary interactions remains unclear. Previous evidence supports that acutely increasing right ventricular (RV) preload when RV afterload is increased, or an external constraint to the heart exists (i.e., DH), may paradoxically reduce LV end-diastolic volumes (LVEDVs) via leftward septal shift and exacerbation of DVI (29). This paradox has been demonstrated in various animal models under conditions of acute volume loading and severe increases in RV afterload or external constraint (6, 22, 38, 42). However, there is only limited evidence in humans, consisting of one previous investigation in individuals with COPD following an acute exacerbation (29). Given the inherent difficulty in isolating the hemodynamic effects of pleural pressure and lung volume in patients with COPD, it remains unclear whether this interaction is fundamentally influenced by nITP (51) or by the interaction of nITP with changes in lung volume and/or PVR (13).

In the absence of increased afterload or systolic dysfunction, it is well established that volume loading increases left ventricular stroke volume (LVSV) through series interaction (15). Head-down tilt (HDT) has been used extensively as a noninvasive method of acutely volume loading the heart through increasing RV preload with minimal changes to afterload, contractility, or heart rate (HR) (45). Clinically, volume loading is a front-line therapy used to augment LVSV in the acute care setting for patients presenting with hemodynamic instability, including exacerbation of respiratory disease (38, 48). Additionally, systolic dysfunction is often seen to cause a compensatory increase in blood volume (hypervolemia) to maintain cardiac output (31, 42), a common occurrence in COPD with comorbid heart failure (12, 17). Thus the aim of this study was to better understand how volume status mediates cardiopulmonary interaction in the face of adverse respiratory mechanics in a spontaneously breathing model of the major perturbations to respiratory function that occur in COPD. We hypothesized that HDT during nITP would cause a paradoxical reduction in LVSV from the supine condition, and that this reduction in LVSV would be significantly greater in a model of COPD, owing to the interactive effects of DH and nITP.

MATERIALS AND METHODS

Subjects.

Twenty healthy subjects (10 men and 10 women, 23 ± 2 yr) volunteered for the study. Subjects were excluded if they were current smokers, had a history of cardiovascular or respiratory disease, a BMI > 30 kg/m², were hypertensive (BP > 140/90 mmHg), or had a poor echocardiographic window. This study received ethical approval from the University of British Columbia Clinical Research Ethics Board, and subjects signed an institutionally approved informed consent form before their participation.

Study design.

These data were collected as a separate and distinct study question during a broader experiment that investigated the hemodynamic responses to nITP, DH, and PVR alone and in combination (13). To address this unique aspect of the study, subjects had a small balloon-tipped esophageal catheter inserted through the nose and properly positioned in the lower third of the esophagus as previously described (13, 20). After catheter insertion and instrumentation, baseline variables were collected including echocardiographic measurements of LV function, ventilatory parameters, and measurements of ITP fluctuations during resting breathing. The study protocol involved three phases: phase 1) acute volume load via 20° HDT positioning, phase 2) 20° HDT with inspiratory resistive loading to generate −20 cmH₂O (HDT+nITP), and phase 3) 20° HDT with nITP, expiratory resistive loading to induce lung hyperinflation (DH) and hypoxic-mediated increases in pulmonary vascular resistance (hPVR), which we have previously referred to as a model of COPD (13) (HDT+COPD model). Subjects were returned to the supine position and given 5 min of rest between phases to allow cardiopulmonary parameters and blood pressure to return to baseline. During phases 2 and 3, HDT and respiratory perturbations were initiated simultaneously.

HDT procedure.

A 20° HDT was accomplished using a bed designed for 12° Trendelenberg and reverse Trendelenberg positioning, which was modified by elevating and supporting the foot end of the bed to achieve 20° HDT. Pilot data for this study, and previous work (32), demonstrated a peak effect of HDT on increased LVEDV within 1 min of initiating the procedure. Therefore, all echocardiography measurements were started 30 s after initiating HDT to capture peak effect.

Inspiratory loading and lung hyperinflation procedure.

Inspiratory pressures of −20 cmH₂O were generated through use of custom-built in-line resistances installed in the inspiratory port of a two-way valve (Hans Rudolph, Model 2700, Hans Rudolf, Shawnee, KS) that could be adjusted to achieve target inspiratory pressures without significantly altering ventilation. DH was induced through use of the same resistance devices inserted into the expiratory port of the two-way valve, while simultaneously pacing respiratory duty cycle such that expiratory time was restricted to 0.5 of inspiratory time.

Hypoxic exposure procedure.

Subjects received a normobaric hypoxic stimulus through a 45-min wash-in period of hypoxic gas, delivered from a 150-l nondiffusion reservoir bag connected to a hypoxic generator (HYP123, Hypoxico, New York) to increase PVR (19). Inspired O₂ fraction was individually titrated to achieve and maintain an S_pO₂ of 80% at the end of hypoxic exposure, and was maintained through the HDT+COPD model phase.

Ventilatory parameters and intrathoracic pressure measurement.

Ventilatory parameters were measured continuously via a two-way low dead-space, low-resistance valve (Hans Rudolph, Model 2700, Hans Rudolf) attached to inspiratory and expiratory pneumotachometers via large bore tubing. Signals from the pneumotachometer were converted to a digital signal by using a data acquisition system (Powerlab, ADI Instruments, Colorado Springs, CO). The flow signal was integrated over time to obtain volume. All data were sampled at 200 Hz and stored on a computer for analysis at a later date. Baseline PFT maneuvers were performed according to American Thoracic Society standards (39). Repeated inspiratory capacity (IC) maneuvers were used during phase 3, to quantify the relative change in operational lung volumes (59).

ITP was measured via esophageal balloon catheter (Ackrad Laboratories, Cranford, NJ) inserted and positioned as per the manufacturer’s recommendation. The balloon catheter was connected to a differential pressure transducer (MP45, Validyne, Northridge, CA), which was calibrated before each test with a water-filled manometer. Signals from the pressure transducer were converted to a digital signal by using a data acquisition system (Powerlab, ADI Instruments).

Echocardiography measurements.

Echocardiographic images were acquired by a trained sonographer and recorded on a commercially available ultrasound system (Vivid-E9, GE Medical, Horton, Norway) by using a M5S-D 1.5–4.6 MHz probe for 2D imaging and a 4V-D 1.5–4.0 MHz probe for 4D (triplane) imaging. All images were acquired with subjects in the left lateral supine position and saved for offline analysis (EchoPAC Ver. 113, GE Medical). Images were acquired for the assessment of LV structure, volumes, and function in accordance with current guidelines (37). LV parasternal short-axis images just below the level of the mitral valve were acquired and analyzed for determining LV geometry. Triplane imaging was used to acquire images in the apical four-, three-, and two-chamber views simultaneously, which were analyzed for LV end-systolic volumes (LVESV), end-diastolic volumes (LVEDV), and ejection fraction (LVEF) to American Society of Echocardiography standards by using the Simpson’s method, modified for triplane analysis (21, 25, 30, 37).

Systemic vascular resistance (SVR) was calculated as mean arterial pressure (MAP) divided by cardiac output (Q, determined from triplane volumes) and expressed as a percentage change from baseline. hPVR was quantified from the estimated increase in RV systolic pressure (RVSP) by using measures of right atrial pressure (RAP) (34) and tricuspid regurgitation velocity (TRV) (8, 60), as previously described (13). LV geometry, specifically the radius of septal curvature (RSC) and radius of LV free-wall curvature (RFWC), was calculated using a modified technique developed by Brinker et al. (10). In brief, this method involves identifying the septal and free-wall segment of the LV from the parasternal short-axis view, and calculating the radius of curvature of these segments by using chords to quantify the degree of septal flattening. A single trained and blinded observer performed all analysis. All echocardiographic data was analyzed and averaged over three cardiac cycles.

Statistical analysis.

All descriptive statistics are reported as means ± SD. A Shapiro-Wilks test was performed to determine normalcy, after which a one-way repeated measures ANOVA was performed on normally distributed data, and one-way repeated measures ANOVA on ranks for nonnormally distributed data. When a significant effect was found, a Holm-Sidak post hoc test (for normally distributed data) or Dunn’s post hoc test (for nonnormally distributed data) was used to determine the level of significance between a given phase, baseline measures, and the previous study phase. Data were also analyzed and reported as a change score (percent change from baseline), which was analyzed using a one-sample t-test. The alpha level for all analysis was set a priori at P < 0.05.

RESULTS

Subject characteristics.

Baseline cardiovascular and respiratory parameters are shown in Table 1. Subjects’ (10 men and 10 women) mean age was 23 ± 2 yr, height was 1.71 ± 0.10 m, and BMI was 23.5 ± 2.4 kg/m². All pulmonary function parameters were within normal ranges (52): FVC 4.53 ± 1.09 l (100 ± 11% predicted), FEV₁ 3.52 ± 0.81 l (90 ± 12% predicted), and FEV₁/FVC 78 ± 7%.

Table 1.

Cardiopulmonary parameters across all study phases

		HDT		HDT+nITP			HDT+COPD Model
Parameter	Baseline	Condition	Δ BL	Condition	Δ BL	Δ HDT	Condition	Δ BL	Δ HDT
RR, bpm	14.5 ± 1.8	14.7 ± 1.9	0.2 ± 2.1	9.3 ± 2.6	−5.2 ± 2.9*†	−5.3 ± 3.2*	14.1 ± 0.9	−0.4 ± 2.2†	0.6 ± 2.1
VT, l	0.59 ± 0.25	0.55 ± 0.08	−0.05 ± 0.22	0.96 ± 0.39	0.35 ± 0.46*†	0.40 ± 0.39*	0.56 ± 0.10	0.03 ± 0.26†	0.01 ± 0.14
VE, l/min	8.5 ± 3.7	8.0 ± 1.2	−0.6 ± 3.3	7.9 ± 1.4	−0.6 ± 3.9	−0.02 ± 1.12	7.8 ± 1.4	−0.7 ± 1.7	−0.11 ± 1.98
TI, s	1.6 ± 0.4	1.7 ± 0.4	0.1 ± 0.3	4.7 ± 1.8	3.0 ± 1.8*†	2.9 ± 1.9*	2.3 ± 0.2	0.7 ± 0.4*†	0.6 ± 0.4*
TE, s	2.4 ± 0.4	2.4 ± 0.4	0.0 ± 0.4	2.4 ± 1.0	0.0 ± 1.1	0.0 ± 1.1	1.8 ± 0.1	−0.7 ± 0.4*†	−0.6 ± 0.4*
TI/TTOT	0.41 ± 0.07	0.42 ± 0.05	0.09 ± 0.05	0.66 ± 0.07	0.25 ± 0.09*†	0.24 ± 0.09*	0.56 ± 0.03	0.17 ± 0.07*†	0.15 ± 0.06*
Pes.insp, cmH2O	−3.5 ± 1.7	−3.5 ± 2.4	0.0 ± 1.5	−22.8 ± 2.2	−19.3 ± 2.4*†	−19.3 ± 2.7*	−24.0 ± 3.6	−20.5 ± 3.8*	−20.5 ± 4.0*
Pes.exp, cmH2O	1.9 ± 1.6	2.1 ± 1.8	0.2 ± 1.5	4.1 ± 2.0	2.2 ± 2.0*	2.0 ± 2.6	7.3 ± 6.3	5.4 ± 6.5*	5.2 ± 6.2*
IC, l	2.65 ± 0.71	—	—	—	—	—	0.66 ± 0.21	−1.89 ± 0.81*	—
LVEDV, ml	115 ± 21	125 ± 27	10 ± 8.0*	107 ± 20	−7 ± 10*†	−15 ± 13*	96 ± 22	−19 ± 13*†	−30 ± 13*
LVESV, ml	48 ± 10	50 ± 10	2 ± 4	47 ± 9	1 ± 5	−2 ± 6	42 ± 12	−5 ± 9*†	−8 ± 9*
LVEF, %	59 ± 2	60 ± 3	1 ± 3	55 ± 3	−4 ± 4*†	−5 ± 5*	57 ± 5	−3 ± 5	−4 ± 6*
SV, ml	68 ± 12	75 ± 17	7 ± 7*	59 ± 12	−8 ± 7*†	−14 ± 10*	54 ± 12	−15 ± 8*†	−22 ± 10*
Q, l/min	3.9 ± 0.5	4.4 ± 1.0	0.5 ± 0.6*	3.7 ± 0.7	−0.2 ± 0.5†	−0.56 ± 0.56*	3.7 ± 0.8	−0.2 ± 0.9	−0.64 ± 0.92*
HR, bpm	59 ± 12	60 ± 12	0 ± 4	64 ± 13	4 ± 7*	4 ± 6	70 ± 10	12 ± 10*†	11 ± 9*
MAP, mmHg	76 ± 8	76 ± 8	0 ± 8	76 ± 9	0 ± 8	0 ± 4	86 ± 9	10 ± 11*†	10 ± 8*
SpO2, %	99 ± 1	99 ± 1	0 ± 1	99 ± 1	0 ± 1	0 ± 2	80 ± 5	−21 ± 6*†	−21 ± 6*
RSC-ED, cm	3.2 ± 0.9	4.2 ± 1.4	0.9 ± 1.3	6.1 ± 2.9	2.8 ± 2.5*†	2.2 ± 2.3*	7.8 ± 3.2	4.3 ± 3.2*	0.7 ± 2.3*
RSC-ES, cm	1.9 ± 0.5	2.2 ± 1.0	0.3 ± 1.0	3.4 ± 2.9	1.5 ± 2.7*†	1.2 ± 2.7*	4.6 ± 3.7	2.5 ± 3.5*	0.8 ± 1.4*
RFWC-ED, cm	2.5 ± 0.2	2.7 ± 0.4	0.2 ± 0.3*	2.8 ± 0.3	0.3 ± 0.2*	0.1 ± 0.3	2.7 ± 0.4	0.1 ± 0.3	−0.2 ± 0.4
RFWC-ES, cm	1.6 ± 0.3	1.6 ± 0.3	0.0 ± 0.2	1.7 ± 0.3	1.7 ± 0.3	0.1 ± 0.2	1.6 ± 0.3	0.0 ± 0.4	−0.1 ± 0.4

Data presented as means ± SD; Δ BL = change from baseline; Δ HDT = change from volume-loaded state (HDT alone); n = 20, except for echocardiographic volume measurements n = 18, n = 17, and n = 16 for LV geometry; n = 20, n = 18, and n = 11 in HDT, HDT+nITP, and HDT+COPD model, respectively. RR, respiratory rate; V_T, tidal volume; T_I, inspiratory time; T_E, expiratory time; T_I/T_TOT, ratio of inspiratory time to total respiratory cycle time; P_es.insp, mean inspiratory esophageal pressure; P_es.exp, mean expiratory esophageal pressure; IC, inspiratory capacity; LVEDV, left ventricular end-diastolic volume; LVESV, left ventricular end-systolic volume; LVEF, left ventricular ejection fraction; SV, stroke volume; Q, cardiac output; HR, heart rate; MAP, mean arterial pressure; S_pO₂, oxygen saturation measured by pulse oximetry.

^*Significant difference from baseline (Δ BL or Δ HDT, P < 0.05);

^†significant difference from the previous phase (P < 0.05).

Cardiac effects of 20° HDT.

Compared with supine baseline, HDT increased LVEDV by 8 ± 6% (P < 0.001) and LVSV by 10 ± 10% (P < 0.001, Fig. 1), thus Q increased by 11 ± 15% (P = 0.005) as HR was unchanged. MAP and SVR did not change significantly. HDT increased the radius of septal curvature at end-diastole (RSC-ED) by 33 ± 47% (P = 0.006), while also increasing LV free-wall radius of curvature at end-diastole (RFWC-ED) by 7 ± 11% (P = 0.007). At end-systole, septal and LV dimensions were not altered from baseline (Fig. 2).

Fig. 1.

Mean LVEDV (●) and LVESV (○) defining LVSV (gray area) at baseline, during HDT, HDT+nITP, and HDT+COPD models. LVEDV, left ventricular end-diastolic volumes; LVESV, left ventricular end-systolic volumes; LVSV, left ventricular stroke volume; HDT, head-down tilt; nITP, negative intrathoracic pressure; COPD, chronic obstructive pulmonary disease. *Significant change from baseline (P < 0.05); †significant change from the previous phase (P < 0.05).

Fig. 2.

LV geometry at baseline, during HDT, HDT+nITP, and HDT+COPD models. RSC-ED, radius of septal curvature at end-diastole; RSC-ES, radius of septal curvature at end-systole; RFWC-ED, radius of LV free-wall curvature at end-diastole; RFWC-ES, radius of free-wall curvature at end-systole. Data shown as means ± SD; *significant change from baseline (P < 0.05); †significant change from the previous phase (P < 0.05).

Cardiac effects of 20° HDT with nITP.

Inspiratory loads of −20 cmH₂O in combination with HDT (HDT+nITP) reduced LVEDV by 6 ± 9% (P = 0.016), while LVESV did not change (P = 0.274) from supine baseline. As a result, both LVEF and LVSV were significantly reduced, by 4 ± 4% (P < 0.001) and 8 ± 7% (P < 0.001), respectively (Fig. 1). Q was maintained by a compensatory 8 ± 12% increase in HR (P = 0.021), while MAP and SVR were unchanged. Compared with HDT alone (Table 1), HDT+nITP significantly reduced LVEDV by 12 ± 11% (P < 0.001), LVSV by 18 ± 13% (P < 0.001, Fig. 1), LVEF by 5 ± 5% (P = 0.001), and Q by 13 ± 12% (P < 0.001). HDT+nITP increased RSC-ED by 87 ± 80% (P < 0.001) and RSC-ES by 73 ± 112% (P = 0.013) from baseline. LV free-wall geometry also changed, with an 11% increase in RFWC-ED (P < 0.001) and a 6 ± 10% increase in RFWC at end-systole (RFWC-ES, P = 0.020). This was also a significant increase in RSC-ED and RSC-ES from the previous phase (HDT alone, P < 0.05, Fig. 2).

Cardiac effects of 20° HDT in a model of COPD.

In a model of COPD, end-expiratory lung volumes were increased by 1.89 ± 0.81 l, as indicated by a corresponding reduction in IC (Table 1), which resulted in an inspiratory reserve volume of 0.1 ± 0.2 l. HDT in our model of COPD (HDT+COPD model) significantly reduced LVEDV (−17 ± 11%), LVEF (−3 ± 5%), and LVSV (−21 ± 12%) from supine baseline. From HDT alone (Table 1), this translated into a 23 ± 10% reduction in LVEDV, 15 ± 18% reduction in LVESV, 28 ± 11% reduction in LVSV (Fig. 1), and 4 ± 6% reduction in LVEF. This was a significantly larger decrease in LVEDV, LVESV, and LVSV than during HDT+nITP (P < 0.05). HDT+COPD model significantly increased MAP (14 ± 14%) and SVR (26 ± 33%) from baseline. PVR increased 23 ± 30% from baseline, as measured at the end of the 45-min hypoxic wash-in period (n = 16 for TRV measurement) as previously reported (13). RSC-ED (135 ± 103%, P = 0.001) and RSC-ES (120 ± 130%, P = 0.012, Fig. 2) were also increased from baseline in HDT+COPD model, indicating significant septal flattening and DVI, although this was not significantly greater than the degree of septal flattening which occurred during HDT+nITP.

DISCUSSION

The novel findings of this study are threefold. First, while it is well established that HDT acutely increases preload and thus LVSV, we demonstrated septal flattening indicative of DVI with acute volume loading, despite an augmented LVSV. Second, despite increased preload, HDT+nITP paradoxically reduced LVSV from supine baseline, a response mediated through DVI. Finally, HDT+COPD model showed a significantly larger reduction in LVEDV and LVSV than HDT+nITP, likely due to the interaction of nITP and DH further exacerbating DVI.

Effects of acute volume loading on cardiovascular function.

The effectiveness of HDT as a noninvasive method of acutely increasing preload with no change to afterload or contractility is well documented (31, 45). Thus, as expected, we observed an increase in LVEDV mediating a 10% increase in LVSV and ultimately a similar increase in Q, as HR was unchanged. Interestingly, HDT alone increased RSC-ED by 33%, suggesting the presence of DVI despite no increases in RV afterload (Fig. 2). This finding supports previous work in animal models that suggest DVI may modulate LV filling through the restraining effects of the pericardium and compliant septal wall, even at unstressed ventricular volumes (4, 16, 22, 36). Subsequently, an increase in DVI may not be detrimental to LV function unless ventricular afterload is also increased or a pericardial/mediastinal constraint to ventricular filling exists (5–7, 42). In the present study, given that ventricular afterload is not increased with HDT (31, 44), and LV filling is not limited by an external constraint, LVSV is increased through series interaction. This mechanism has previously been reported in canine models; however, to our knowledge augmented DVI with acute volume loading has not previously been shown in spontaneously breathing humans.

Effects of volume loading with nITP on cardiovascular function.

Whereas the hemodynamic response to increased nITP is relatively well established (50), the role of DVI in mediating this response remains unclear, and limited work has been done utilizing a spontaneously breathing human model. Additionally, the importance of volume status in altering the magnitude of LV volume reduction for a given degree of nITP is not known. We have previously shown, during spontaneous breathing in the supine position, that −20 cmH₂O nITP during inspiration significantly reduces LVSV via a reduction in LVEDV (13). In the present study, the addition of nITP during HDT significantly reduced LVEDV, causing an 11% decrease in LVSV from supine baseline, or a 21% decrease from the volume-loaded state (HDT alone). This appears to be a larger reduction in LVSV than we have previously reported with nITP in non-volume-loaded subjects (−7 ± 7%) (13). Thus the hemodynamic impact of nITP may be volume dependent, having a larger deleterious effect on LVSV in the volume-loaded state. This finding may have important implications for individuals with COPD and comorbidities categorized by hypervolemia, including heart failure (47) and obesity (46).

The present data present robust evidence for the role of DVI for impairing LV function, substantiated by changes to septal geometry during HDT+nITP (Fig. 2). Significant septal flattening was observed to occur during HDT+nITP, resulting in an increased RSC-ED and RSC-ES, while RFWC-ED increased only modestly and RFWC-ES did not change. Although we cannot definitively rule out the role of series interaction, specifically IVC collapse attenuating venous return during increased nITP (43), previous authors have shown maintained RV output in the presence of nITP in normovolemic subjects (1), even with considerable IVC collapse (54). Additionally, nITP is seen to increase ventricular transmural pressure, and thus afterload (50, 51). Although afterload was likely increased in the present study, it does not appear to be the primary mechanism of LVSV reduction, since LVESV did not significantly increase as would be expected in an afterload-mediated reduction in LVSV. However, increased RV afterload may have contributed to DVI (especially in the presence of an augmented RV end-diastolic volume from HDT) by impeding RV outflow and increasing RV end-diastolic volume (1, 6), creating a greater degree of leftward septal displacement. Thus, in the current model, reduced LVSV appears to be a result of augmented DVI and restricted LV diastolic filling in response to elevated RV preload during HDT+nITP.

Effects of HDT in a spontaneously breathing model of COPD.

We have previously shown impaired LV function in a model of COPD, including the isolated and interactive hemodynamic effects of nITP and DH and the role of DVI in healthy individuals (13). In the present study, there are two novel findings from this phase. First, despite an increase in preload with HDT, LVSV was significantly decreased from supine baseline during HDT+COPD model, and appears to be reduced to a greater extent (−21 ± 10%) than our previously published model of COPD under normovolemic conditions (−14 ± 13%) (13). If series interaction predominated, we would expect to see an improvement in LVSV as compared with our previous model. In contrast, we observed the opposite, such that an increase in preload via HDT decreased LVSV, which suggests an increased contribution of DVI. Second, HDT+COPD model significantly worsened the paradoxical decrease in LVSV observed with HDT+nITP alone. This provides further evidence that the interaction of DH and nITP should be considered as an important hemodynamic stressor, rather than nITP alone being the primary mediator of cardiopulmonary interaction in obstructed breathing, as has been previously suggested (51).

We have previously shown that the isolated and interactive effect of increased PVR through hypoxic exposure does not considerably alter LVSV and is of relatively little hemodynamic significance in this model (13). Thus it would appear that the greatly reduced LVEDV and LVSV in HDT+COPD model was primarily driven by the interaction of DH and nITP. In this context, nITP increases venous return during diastole (50, 54) but also afterload during systole (50), while DH increases RV afterload via alveolar capillary compression (56) and acts as a mediastinal constraint to ventricular filling, which has previously been shown to amplify DVI (6 (11). Accordingly, with limited free-wall compliance, the LV in this scenario is highly sensitive to additional increases in RV volumes, which would increase RV end-diastolic pressure relative to LV end-diastolic pressure, further reducing LV compliance and filling (6). We suggest the interaction of these mechanisms explains the large reduction in LVSV seen with acute VL via HDT in this model of COPD.

The finding of a paradoxically reduced LVSV in response to acute volume loading has previously been reported in postexacerbation COPD patients in response to incremental volume loading, whereby LVSV was increased up to a critical right atrial pressure, after which point further volume loading caused a paradoxical decrease in LVSV, which the authors attributed to DVI (29). However, these patients were also characterized by LV hypertrophy, which would predispose them to impaired LV diastolic functioning even in normovolemia (29). Accordingly, the complexity of COPD with concomitant heart disease limited the extent to which the mechanism of LVSV reduction could be identified in this seminal study. Thus to our knowledge this is the first study to highlight the importance and influence of volume status on the hemodynamic response to nITP alone and in combination with DH and increased PVR.

Similar to the findings of Jardin and associates (29), we attribute the profound reduction in LVSV during acute VL in a model of COPD to DVI, as we observed the largest degree of septal flattening (increase in RSC-ED) during HDT+COPD model, which persisted through end-systole (increased RSC-ES, Fig. 3). Increased septal flattening through systole has been previously reported with pathological RV loading and may reflect abnormal ventricular systolic wall tensions (29), or an altered balance of forces at the right ventricular-septal junction (23). However, it should be noted that the addition of HDT to our model of COPD did not appear to further decrease LVSV, or further increase septal flattening from that which we have previously reported in a model of COPD without volume loading (13). In the present study, this may be explained by increased myocardial contractility of the RV, as has been documented in previous canine models investigating acute RV afterload and volume loading effects (7). Indeed, contractility is known to increase in response to increased afterload [Anrep effect (14)] and HR [Bowditch effect (26)], both of which are likely relevant to this model. Thus in the present scenario contractility likely increased to combat increased RV afterload and maintain RVSV, which may account for the lack of substantial difference between LV volumes and geometry in this model as compared with our previously published data (13). Accordingly, we postulate that the impact of volume status on cardiopulmonary interaction in COPD may be considerably more detrimental to LV function if contractile reserve were depleted (35).

Fig. 3.

Representative parasternal short-axis image from one subject demonstrating septal flattening during HDT+COPD model as compared with baseline. Baseline end-diastole (A), HDT+COPD model end-diastole (B), baseline end-systole (C), and HDT+COPD model end-systole (D). Blue line indicates ventricular septum segment; red line indicates LV free-wall segment.

Methodological considerations.

It must first be acknowledged that the pathology of COPD is far more complex than the interplay of altered respiratory mechanics and airflow limitation that we report in this and a previous study (13) as a “model of COPD.” Although we do not want to oversimplify the complexity of COPD pathophysiology, the focus of this study was to evaluate the hemodynamic effects of the “mechanical” aspects that commonly accompany the disease (i.e., airflow limitation, hyperinflation, and increased PVR). We have previously addressed a number of methodological considerations associated with extrapolating our model of COPD to the disease state (13). Briefly, the use of an extrathoracic resistance does not accurately model the dynamic airway resistance seen in COPD (58). Furthermore, the likely homogenous lung hyperinflation that occurs during DH in healthy individuals may have different cardiac effects as compared with a likely more heterogenous lung hyperinflation pattern in COPD (9). This may influence the extent of mediastinal constraint dependent on regional differences in lung volume and compliance (53).

Additionally, owing to the inherent difficulty of acquiring echocardiographic images during inspiration, especially at high lung volumes, the majority of our data was acquired and analyzed at end-expiration. Within-breathe hemodynamic variation is well documented and worsens with respiratory perturbation, i.e., pulsus paradoxus (24, 27). As the majority of optimal echo images in the present study were collected during expiration, the data reported in this manuscript may be underestimating true LV volume changes during inspiration. Nevertheless, given that the intent of this study was to evaluate the global response (i.e., across consecutive cardiac and respiratory cycles) to these perturbations, we feel this does not limit the interpretation of our data.

As a final and clinically relevant point, this investigation of mechanical heart-lung interaction with acute volume loading does not take into account the potential for acute volume loading to improve ventilation-perfusion matching and/or reduce PVR, depending on the etiology of PVR increase and lung zone conditions (40, 55). Thus extrapolation of this data to the clinical setting, and determining whether volume loading will have beneficial or adverse effects, requires an understanding of and investigation into this complex interplay of mechanisms.

Conclusions.

During spontaneous breathing with nITP, acute VL paradoxically reduces LVSV through exacerbation of DVI. This reduction in LVSV is significantly larger during acute VL in a model of COPD, owing to the interaction of DH (10, 13) and nITP (13), and subsequent amplification of DVI. These findings provide further evidence for the importance of lung hyperinflation in influencing hemodynamics in COPD, and may have important implications for understanding and managing hemodynamics in COPD patients, specifically COPD with comorbid hypervolemia (i.e., obesity, heart failure) (46), or in the critical care setting (18, 41).

GRANTS

Grants held by N. Eves from the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Canadian Foundation of Innovation provided operating and infrastructure funding for the study. N. Eves was also supported by a Clinical Scholar Award from the Michael Smith Foundation for Health.

DISCLOSURES

The funding source had no influence on study design, writing of the manuscript, or the decision to submit for publication.

AUTHOR CONTRIBUTIONS

W.S.C. and N.D.E. conceived and designed research; W.S.C., A.M.W., and M.I.H. performed experiments; W.S.C. analyzed data; W.S.C. and N.D.E. interpreted results of experiments; W.S.C. prepared figures; W.S.C. drafted manuscript; W.S.C., A.M.W., M.I.H., and N.D.E. edited and revised manuscript; W.S.C., A.M.W., M.I.H., and N.D.E. approved final version of manuscript.