Cardiac pacing: A novel approach to right ventricle failure during pulmonary thromboendarterectomy

Abstract

Pulmonary thromboendarterectomy (PTE) reverses chronic thromboembolic pulmonary hypertension, decreasing pulmonary vascular resistance and pressure overload failure of the right ventricle (RV). Suboptimal cardiac function early after PTE¹ on cardiopulmonary bypass (CPB) may require inotropic or mechanical support. Our laboratory demonstrated increased cardiac output (CO) with temporary biventricular pacing (BiVP) in an RV pressure overload pig model.² In patients with left ventricular (LV) dysfunction, optimized temporary BiVP increases CO after CPB,³ reduces the vasoactiveinotrope score, and increases urine output.⁴ Our study tests the feasibility of BiVP reversal of RV failure after PTE.

MATERIALS AND METHODS

Four adults were recruited and studied with informed consent under an Institutional Review Board-approved protocol between June 1, 2011, and March 1, 2012. Inclusion criteria were PTE on CPB, sinus rhythm (NSR), and inotrope dependence after CPB. Exclusion criteria were atrial fibrillation, 2° or 3° atrioventricular block, congenital heart disease, intracardiac shunts, inability to pace or place an ultrasonic aortic flow probe (UFP), or heart rate (HR) > 120 bpm after CPB. Our surgeon performed PTE using standardized methods.⁵ There were no study-related complications.

This protocol resembles phase I of the Biventricular Pacing after Cardiac Surgery trial.³ Temporary epicardial pacing wires sewn to the appendage of the right atrium (RA), anterior RV, and LV obtuse margin were connected to an InSync III pacemaker (Medtronic Inc, Minneapolis, Minn) for BiVP. A UFP device (Transonic Systems, Ithaca, NY) was used to measure CO.

The protocol begins after weaning from CPB and hemodynamic stabi-lization. Pacing rate is 90 bpm or 10 bpm higher than intrinsic HR if >90 bpm. Seven values of atrioventricular delay (AVD) are tested (90 to 270 milliseconds; 30-millisecond increments). Nine values of interventricular delay (VVD) are also tested (−80 millisecond [LV first] to +80 millisecond [RV first]; 20~millisecond increments). Settings are tested twice for 10 seconds in random sequence. Optimum AVD and VVD are selected using a graphical display. Vasoactive medications and volume infusion are held constant during optimization. Finally, optimized BiVP, atrial pacing (AAI), and NSR are tested twice in randomized order over 30-second intervats. The UFP and LV wires are then removed. Paired t tests are used to compare NSR to optimized BiVP and AAI.

RESULTS

Representative AVD optimization is illustrated in Figure 1, A, CO peaking at 240 milliseconds. During VVD optimization, CO peaks at −40 milliseconds (LV pacing 40 milliseconds before RV pacing), as shown in Figure 1, B. The change in CO is <1% between VVDs of +20 and −40 milliseconds.

FIGURE 1.

A, Representative effect of atrioventricular delay (AVD) optimization on cardiac output (CO). B, Representative effect of interventricular delay (VVD) optimization on CO.

Table 1 illustrates clinical and optimization data for each patient. AVD optima were 120 to 270 milliseconds. VVD optima were −40, 0, 0, and 60 milliseconds. All patients were in RV failure preoperatively and were receiving inotropic support at the time of testing.

TABLE 1.

Clinical and optimization data for each patient (N = 4)

		PAP s/d/m	QRSd			CA	XC	CPB	BSA	AVD	VVD	NoP	AAI	BiV
Patient	A/G	mm Hg	msec	RBBB	RVH	h:m	h:m	h:m	m2	msec	msec	COΔ	COΔ	COΔ
PTE001	54m	68/34/54	168	No	Yes	57	1:21	4:10	2.1	270	0	0	3.1	−5.6
PTE002	31m	99/21/44	192	No	No	1:02	1:36	5:27	2.4	180	60	0	10.7	3
PTE003	66f	64/33/44	154	No	Yes	40	56	3:55	1.8	120	0	0	6.5	1.2
PTE004	45f	96/28/55	128	Yes	No	39	1:39	5:07	2.2	180	−40	0	12.6	11.5

Minimum temperature all cases was 18°C. A/G, Age/gender; PAP, pulmonary artery pressure; s/d/m, systolic/diastolic/mean; QRSd, QRS duration before CPB; RBBB, right bundie branch block; RVH, right ventricular hypertrophy; CA, circulatory arrest; XC, crossclamp; CPB, cardiopulmonary bypass; BSA, body surface area; AVD, atrioventricular delay; VVD, interventricular delay; NoP, no pacing; AAI, atrial pacing; BiV, biventricular pacing; h:m, hours: minutes; COΔ, percent change in cardiac output vs AAI; m, male;f, female.

Figure 2 compares mean CO (n = 4) during AAI, BiVP, and NSR. Mean ± standard deviation CO during AAI was 8.4% ± 3.8% higher than in NSR (P = .04). CO during BiVP was not significantly different versus NSR. Mean ± standard deviation HR during BiVP and AAI was 94.0 ± 4.7/minute versus 76.3 ± 6.9/minute during NSR (P < .01).

FIGURE 2.

Comparison of cardiac output (CO) in sinus rhythm (NSR) versus atrial pacing (AAI) or optimized biventricular pacing (BiVP) in 4 patients. Bars indicate standard error of the mean and P values are based on paired t test.

DISCUSSION

This is a novel pilot study of the hemodynamics of BiVP during PTE. As in the Biventricular Pacing after Cardiac Surgery trial,³ our results demonstrate that AVD and VVD optimization increase CO during BiVP. However, CO during AAI in our study was significantly higher than with both BiVP and NSR. CO after CPB in the Biventricular Pacing after Cardiac Surgery trial was 10% to 13% higher with optimized BiVP than with NSR or AAl.

Minimal efficacy of optimized BiVP in our study contrasts with recent catheter lab research in patients with chronic thromboembolic pulmonary hypertension.⁶ Additional research is needed to determine if our results are an anomaly, related to a small patient cohort. Dual-site RV pacing is a possible alternative approach, particularly in patients with right bundle branch block.

Ventricular mechanics analysis might explain the reduced efficacy of BiVP observed here. The time course of pulmonary and systemic arterial pressure might help, as suggested by laboratory studies utilizing RVand LV pressure and 2-dimensional echocardiography.² Heart block in those studies prevented comparison with AAI, but BiVP was effective at the intrinsic atrial rate²

We attribute BiVP efficacy in the Biventricular Pacing after Cardiac Surgery trial to reversal of ischemia-reperfusion injury. Such injury may have been minimized in our study by the long rewarming time. With no clear BiVP benefit, HR augmentation may be the only potential advantage of pacing after PTE. Clinical trials are needed to define benefits and problems with pacing in critically ill patients with PTE, although temporary HR augmentation is apparently recommended in some centers after CPB.

As in the Biventricular Pacing after Cardiac Surgery trial, the optimum AVD noted here for BiVP is often considerably longer than in “permanent” pacemakers or BiVP implants.

CONCLUSIONS

AAI improves CO, and temporary BiVP is feasible after PTE. However, BiVP was minimally effective in this series. Temporary pacing warrants further study to define the full spectrum of response after PTE.