Long-term Doppler echocardiographic evaluation of the right heart after major lung resections Free

Abstract

1 Introduction

Right ventricular morphology and function have gained increasing clinical importance [1–5], but only limited information on adaptation after major lung resections has been reported, focusing on the immediate postoperative period and the first 6 months after surgery [6–12]. Invasive hemodynamic measurements have been performed before, during, and after pulmonary resections; however, catheterization is certainly not comfortable for the patient and may carry complications. Doppler echocardiography has progressively gained popularity in the last two decades to evaluate right ventricle (RV) morphology and function; this technique allows also us to estimate pulmonary artery systolic pressure (PASP) [13,14]. It has also been validated in patients with chronic obstructive pulmonary disease (COPD) [15–19] and successfully employed to study early- and medium-term effects of pneumonectomy and lobectomy on RV function [8,12]. However, no data are currently available on the late effects of major lung resections on the right heart.

We hereby describe the results of a prospective study designed to assess if right heart echocardiographic parameters change more over time after lobectomy or after pneumonectomy and when these modifications occur.

2 Patients and methods

Patients suitable for lobectomy or pneumonectomy for lung cancer were enrolled in this prospective study. Exclusion criteria were: previous history of myocardial infarction, angina, valvular disease, atrial fibrillation or other major arrhythmias, and heart surgery; also patients with a FEV₁ lower than 60% were excluded to avoid right heart modifications related to severe chronic obstructive pulmonary disease. All patients underwent preoperative chest X-ray, ECG, spirometry, and blood gas analysis. They also underwent two-dimensional echocardiography preoperatively and 1 week, 3 months, 6 months, and 1 year after the operation. At 6 months and 1 year after surgery, spirometry and blood gas analyses were repeated; 4 years after surgery only blood gas analysis was monitored. All patients were informed about the project and agreed to respect the postoperative follow-up timetable. Echocardiographic studies were performed using a Toshiba Powervision 8000 (Japan) with a 3.7 Mz harmonic imaging/continuous-wave Doppler and color Doppler transducer. We have studied the RV diastolic diameter (RVDD), the RV free wall thickness (RVT), the tricuspid valve morphology (TVI), the regurgitation jet speed (TRJ), and the pulmonary artery systolic pressure. The apical four-chamber view and the parasternal short axis view were used to study the RV size and the tricuspid valve was studied through the subcostal and apical view. However, after pneumonectomy the echocardiographic windows had to be searched on a case-by-case base due to mediastinal shifting. The TRJ was localized with color flow Doppler and then interrogated with image-directed continuous-wave Doppler. An accurate search was performed to identify and record the greatest maximal velocity corresponding to the most homogeneous flow profile. The systolic transtricuspid pressure gradient (P) was calculated with the modified Bernoulli’s equation P = 4

, where V_max is the maximal tricuspid regurgitation speed and is measured in ms. The V_max was calculated as the average peak velocity among five consecutive measurements. The pulmonary artery systolic pressure was calculated as the sum of the systolic transtricuspid gradient and the estimated right atrial pressure (RAP). RAP was introduced in the equation on the base of the caliber and the degree of inspiratory collapse of the inferior vena cava (IVC) (collapse > 45% RAP = 5 mmHg; collapse 35–45% RAP = 9 mmHg; collapse < 35% RAP = 16 mmHg; IVC uniformly dilated without inspiratory collapse RAP = 20 mmHg). PASP is equal to the right ventricle systolic pressure (RVSP) in the absence of pulmonary outflow obstruction; thus, PASP = RVSP = 4

 + RAP. The morphology of the tricuspid valve was analyzed by our cardiologists (S.S. and F.F.) assigning a subjective score: 0 = valve normal and competent; 1 (+) = low-grade insufficiency; 2 (++) = moderate-grade insufficiency; 3 (+++) = high-grade insufficiency.

Data are presented as a mean ± standard deviation. Paired two-tailed Student’s t-test was used to analyze quantitative continuous variables comparing the means. Repeated measures analysis of variance was performed using the Friedman ANOVA and Kendall coefficient of concordance. A p-value < 0.05 was considered to indicate statistical significance.

3 Results

There were 36 patients in the lobectomy group and 15 in the pneumonectomy group (M/F: 29/7 and 15/0, respectively; p = 0.008). The mean age was 66 ± 9 years and 69 ± 6 years, respectively (p = 0.14); the preoperative FEV₁ was 83 ± 15% and 83 ± 9% (p = 0.7); and PaO₂ was 85 ± 12 mmHg and 88 ± 0.6 mmHg (p = 0.6). Thirty patients underwent right (four bronchial sleeves) and six left lobectomy. Eleven patients received left pneumonectomy (two intrapericardial) and four right pneumonectomy. In the lobectomy group, 5 patients had stage IA lung cancer, 14 IB, 12 IIB, and 5 IIIA; in the pneumonectomy group 2 patients had IB lung cancer, 5 IIB, 6 IIIA, and 2 IIIB. Induction chemotherapy with a cysplatinum-based regimen was performed in four and five patients, respectively, in the lobectomy and pneumonectomy group (p = 0.02); in the two groups, respectively, nine (25%) and seven (47%) patients received adjuvant therapy. The mean duration of postoperative hospitalization was 7 ± 1 days and 12 ± 1.5 days for the two groups, respectively. Postoperative complications in the lobectomy group were: myocardial infarction (two patients), supraventricular tachycardia [4], pneumonia [2], prolonged air leak [3], and contralateral pneumothorax [1]; in the pneumonectomy group we observed myocardial infarction [1], supraventricular tachycardia [3], and pulmonary edema [1] in a patient undergoing left intrapericardial pneumonectomy. No mortality was observed during the postoperative course and the first year after surgery. All patients were available for 1-week, 3-month, 6-month, and 1-year follow-up; 4 years after surgery 29 patients undergoing lobectomy (1 lost to follow-up) and 11 receiving pneumonectomy (1 lost to follow-up) were available for follow-up.

Table 1 shows that in the group of patients undergoing lobectomy no statistically significant modification of any variable was observed at any time during the 4-year follow-up. On the other side, the RVDD significantly increased in patients receiving pneumonectomy (Table 2 ). Four years after surgery all patients undergoing pneumonectomy had moderate tricuspid valve insufficiency while only 50% of patients in the lobectomy group had low-grade insufficiency and 5% had moderate insufficiency of the valve. The PASP was significantly higher after pneumonectomy with differences starting to appear 1 week after the operation. The right ventricular free wall thickness showed a moderate increase 4 years after surgery in the group of patients undergoing pneumonectomy (lobectomy pre: 5.6 ± 0.8 mm, post: 5.7 ± 0.9 mm; pneumonectomy pre: 5.5 ± 0.9 mm, post: 5.9 ± 0.9 mm).

Table 1

Modifications of the echocardiographic variables during the 4 years of follow-up after lobectomy

Open in new tabDownload slide

Table 2

Modifications of the echocardiographic variables during the 4 years of follow-up after pneumonectomy

Open in new tabDownload slide

The PaO₂ was not statistically different in the two groups before and after surgery (lobectomy group pre: 85 ± 12 mmHg, post at 1 year: 84 ± 9 mmHg, post at 4 years: 82 ± 7 mmHg; pneumonectomy group pre: 88 ± 9 mmHg, post at 1 year: 80 ± 11 mmHg, post at 4 years: 78 ± 9 mmHg); spirometry showed a statistically significant decrease of FEV₁ in both groups (59 ± 9.6% after pneumonectomy and 76 ± 15% after lobectomy; p < 0.005 in both groups).

4 Discussion

The right ventricle has been the object of numerous studies in the past. Early reports postulated that RV could play only a minor role to sustain cardiac output [20]; it is now clear that RV performance shows an important impact on overall cardiac function, especially under stress [21]. In this setting, major RV dysfunction is correlated with poor outcome and it is implicated in the pathophysiology of the final event [1]. Non-invasive monitoring methods like echocardiography have met objective difficulties in the past; invasive approaches like right heart catheterization with thermodilution have been proposed [9–11] to investigate different clinical settings, including lung surgery; measurements performed immediately after pulmonary artery clamping and in the early postoperative period [9–11] documented an increase in PASP after major lung resections with RV overload and progressive dilatation. RV overload has been observed at rest but it is significantly increased under effort [11], demonstrating that in this situation ventricular dilatation may reach its limit. The use of flow-directed pulmonary artery catheters capable of determining indexes of right ventricle performance is certainly the ideal method to evaluate this section of the heart but it carries some discomfort to the patient and it is potentially related to complications.

A reliable non-invasive prognostic marker of RV dysfunction would be extremely useful. In recent years Doppler echocardiography has been repeatedly validated to evaluate cardiac performance and in particular to estimate RV morphology and function [8,16]. However, it is difficult to adequately visualize at echocardiography the right heart, since it does not approximate to any geometrical model. Most investigators attribute the difficulties in finding a reproducible method to assess RV function to its complex structural geometry. RV is composed of several anatomic segments that can be divided into two major components: the RV sinus and the infundibulum. Electrophysiological studies indicate that activation of the RV outflow tract occurs relatively late in systole [22], possibly leading to asynchronous contraction and relaxation of the sinus and infundibulum components of the RV. For this reason, some functional parameters may be difficult to evaluate echocardiographically; in particular, RV ejection fraction is usually underestimated [22–24], and this is the reason why it was not included among the variables investigated in our study. Cine magnetic resonance imaging [25] could certainly improve understanding and it should be included in future studies.

Doppler echocardiography has been extensively used to study the right heart in various situations including advanced COPD [15,16,19] and in patients undergoing major lung resections [8,12]. In the latter group of patients pneumonectomy, with the topographic modifications related to mediastinal shifting and rotation, contributes to increase the technical difficulty of the procedure. In this setting, all the reports were focused on the early postoperative period and medium-term follow-up (6 months) [8,12]. We have extended echocardiographic monitoring to the first 4 years after surgery. The early modifications secondary to major lung resections described by other authors [8,12] were confirmed by our study; pneumonectomy causes an important reduction of the vascular bed resulting in a progressive increase in PASP starting at the end of the first postoperative week and is able to induce modifications of RV morphology; these modifications appear later, and are certainly less impressive than in patients with other types of pulmonary hypertension with much higher PASP values. Clinical and experimental reports [25] suggested that this would be the trigger for right heart modifications along with an increased activity of the adrenergic system early after surgery. As a consequence, the right ventricle progressively enlarges, reaching significance after 6 months; our study documented that PASP and RV diameters continue to increase during the 4 years of follow-up after pneumonectomy; however, at this time point, even if the PASP increase is significant, it is still not enough to elicit hypertrophy of the RV free wall as documented by our measurements. These modifications are certainly more evident during the first 6 months of follow-up; during the following period of time the variation rate slows down. This may be interpreted as a progressive adaptation to RV overload. Although from our study it clearly appears what type of modification occurs within the right ventricle and the pulmonary circulation, a larger population of patients would be useful to evaluate the potential impact of several variables: induction and adjuvant therapy, site of pneumonectomy (right vs left), and intrapericardial dissection of the vessels.

If we compare our data with the study reported by Foroulis et al. [12], we come to very similar findings; however, there are some differences that deserve comments. Also those authors found some functional modifications in the lobectomy group, but the measured parameters were much more similar to the left pneumonectomy group than in our study. On the other hand, PASP rose much more in the pneumonectomy group analyzed by Foroulis et al. This may be explained by the higher number of right pneumonectomies reported by Foroulis et al. They also reported no differences between the mean measured PASP after standard and intrapericardial pneumonectomy and that RV enlargement was observed only in patients with postoperative respiratory distress. We also observed some RV enlargement but we had no patients in N.Y. class III or IV. Hypoxia is certainly one of the mechanisms that may favor pulmonary hypertension; however, in our study there were no differences between preoperative and postoperative blood gas analyses; pneumonectomy was functionally well tolerated by all patients even if after surgery there was a statistically significant decrease in FEV₁. This may also reflect preoperative selection since we intentionally excluded patients with relevant COPD.

Although there are certain limitations of the interpretation of data due to heterogeneousity, RV modifications are well evident after pneumonectomy; even if they do not show a clear clinical impact they should not be neglected. For this reason bronchovascular reconstructions should be encouraged as a viable and tested alternative to avoid pneumonectomy when oncologically feasible.

References