**3. Physiology of PVF**

78 Echocardiography – In Specific Diseases

settings such as valvular heart diseases, postoperative conditions, heart failure, hypertension, metabolic syndrome, thyrotoxicosis, and so on (**Fig. 1**). Since valvular heart diseases were historically a main etiology of AF, echocardiographic attention to AF patients was mainly rheumatic valvular lesions and detection of LA thrombi or spontaneous echo contrast which is based on the local hemostatic changes due to rheologic abnormalities (Kwaan et al, 2004; Topaloglu et al, 2007). In relation to thrombus formation, LA appendage function was highlighted in that impaired appendage function leads to thrombus formation and high risk of embolic event (Donal et al, 2005). According to an increased prevalence of coronary artery diseases, AF has been encountered in acute myocardial infarction, after coronary artery bypass grafting surgery and chronic phase of ischemic heart disease and subsequent heart failure. On the other hand, AF is often observed in patients with another kind of arrhythmias (e.g., preexcitation syndrome) or noncardiac disorders (e.g., thyrotoxicosis, chronic obstructive pulmonary disease). AF is often encountered in subjects without systemic or organic heart diseases (so-called 'lone' AF). According with relative decline of rheumatic valvular diseases, terminology of 'nonvalvular' or 'nonrheumatic' AF becomes familiar. Wide spectrum in clinical features of AF sometimes makes the therapeutic

AF is classified by the duration in which this arrhythmia sustains (e.g., paroxysmal, persistent and permanent). Paroxysmal AF is characterized as rare or repetitive paroxysms of short-lasting AF, which often undergoes spontaneous conversion to sinus rhythm, but rhythm-control treatment is required depending on symptom and hemodynamic deterioration. Persistent AF has the possibility of termination either by antiarrhythmic drugs or by electrical defibrillation. Permanent AF does not restore to sinus rhythm spontaneously, and hence conservative therapeutic option is the rate-control strategy. AF is not only responsible for substantial morbidity and mortality, but also impairs quality of life by limited capacities of physical activity and heart rate regulation. To date, the most effective treatment for drug-refractory AF is radiofrequency catheter ablation. Pulmonary vein (PV) isolation by circumferential ablation of PV-LA junction is a promising technique to terminate AF. Despite the introduction of novel and sophisticated ablation techniques such as irrigation catheters, pericardial approach and ganglionated plexi ablation, periprocedural

AF is characterized to date as an age-dependent, progressive disease, i.e., AF prevalence increases steeply from 0.5% at age 50 to 59 years to 9.0% at age 80 to 89 years (Kannel et al., 1998). Progressive nature of this arrhythmia is evident in that AF becomes refractory to pharmacologic treatment and electrical defibrillation in proportion with the duration of sustaining AF. This is the main feature that distinguishes AF from many other kinds of clinical arrhythmia (Wijffels et al., 1995). In a few decades, the mechanisms of such progression of AF have been clarified by many basic experiments using AF animal models and clinical studies of AF patients. Remodeling of LA plays an important role in the genesis, maintenance and perpetuation of AF. LA remodeling is a concept including electrical, contractile and structural aspects. Electrical remodeling induces abbreviated and dispersed electrical refractoriness and inhomogeneous slow conduction of electrical impulse. These are considered to be an arrhythmogenic substrate, prerequisite of AF development. Electrocardiograms (ECG) in patients with AF demonstrate characteristic fibrillation (f) waves that are evident in right precordial leads. According to the progression of AF,

decision-making difficult (Wyse & Gersh, 2004).

complications are not negligible.

PVF recording is feasible not only by transesophageal echocardiography (TEE) but also by transthoracic echocardiography (TTE). According with the prevalence of PVF estimation by TTE, there have been investigations comparing the PVF recorded by TTE with that recorded by TEE. To date, TTE estimation of PVF is reported to provide reliable quantitation of PVF recorded by TEE in patients with and without organic heart diseases (Masuyama et al., 1995). **Fig. 2** is an actual Doppler PVF pattern during an entire cardiac cycle recorded by TTE. When ultrasound probe is positioned at the apex of chest wall, a four-chamber apical view is obtained. Then color jet is visualized in the upper LA of real-time, B-mode image. This color image is forward blood flow signals in right superior PV (**Fig. 2**, upper). After overall color Doppler interrogation, Doppler velocimetry is obtained by positioning the sampling gate 2-3 cm distal from the orifice of right superior PV (**Fig. 2**, lower).

The PVF profile is characterized as forward flows during LV systolic (**S**) and early diastolic (**D**) phases, and as reversed flow during late diastole when LA contracts (**Ar**). There are strictly two components within the **S** wave of PVF, i.e., **S1** is caused by active LA relaxation

Pulmonary Venous Flow Pattern and Atrial Fibrillation: Fact and Controversy 81

P QRS

LA active contraction →LA emptying

LV relaxation

LA relaxation at the onset of LV contraction

LV

**A**

**B**

**C**

**D**

LV contraction

Reservoir functioning in mid-systolic phase

LV

Ao PV

Ao LA PV

Booster pump functioning in late diastolic phase

> PV back flow

LA active relaxation

LA

LV contraction toward apex

LA functioning as conduit during early diastolic phase

LV relaxation

LV

Ao LA PV

LA

LV

Ao PV PV forward flow

LA passive stretching

PV forward flow

PV forward flow

(reservoir)

PVS PVD

T

A E

v

T

PVS PVD

A E

v

T

PVS PVD

A E

v

T

PVS PVD

A E

v

y

y

y

y

PV flow

LA pressure

PV flow

LA pressure

PV flow

LA pressure

PV flow

LA pressure

LV inflow

LV inflow

ECG

LV inflow

ECG

LV inflow

ECG

PVA

P QRS

PVA

P QRS

PVA

PVA

Fig. 3. Illustration of normal blood flow pattern in PV and LA during an entire cardiac cycle.

<sup>a</sup> <sup>c</sup> x

<sup>a</sup> <sup>c</sup> x

LA as conduit ECG

P QRS

<sup>a</sup> <sup>c</sup> x

<sup>a</sup> <sup>c</sup> x

and **S2** is ascribed to passive LA wall stretching caused by vigorous LV contraction toward apical direction. Peak velocity and velocity-time integral of the **S** wave are usually greater than those of the **D** wave. Gentile et al (1997) investigated Doppler PVF parameters in 143 healthy individuals aged from 20 to 80 years by TTE. Age-dependent Doppler parameters are reported to be as follows; peak amplitude and time integral of both **S** and **D** waves, and **S**/**D** peak amplitude and integral ratios, whereas **Ar** wave is reported to be ageindependent. These findings indicate the possibility of **Ar** wave as a diagnostic tool of various hemodynamic abnormalities in a wide range of patients' age.

Fig. 2. Representative Doppler imaging of transthoracic echocardiography applied to a patient with hypertension. Upper image is an apical four chamber view with color flow indicating blood flow returning from right superior pulmonary vein (PV) into left atrium (LA). Middle is an ECG tracing (standard limb lead II). Lower is a continuous-wave Doppler PV flow velocimetry during an entire cardiac cycle. Upward direction indicates forward flow, whereas downward direction means reverse flow.

LA plays three different roles periodically in an entire cardiac cycle, i.e., LA acts as a 'booster pump' when LA contracts in late LV diastole, then as a 'reservoir' during LV systole, and finally as a 'conduit' during early LV diastole (**Fig. 3**). These three kinds of LA functions correspond with **Ar**, **S** and **D** waves respectively, and are estimated also by LA volume curve during an entire cardiac cycle using automatic boundary detection (Zhang et al, 1998), manual tracking (Ogawa et al, 2009), and speckle tracking techniques (Mori et al, 2011). The PVF pattern, especially **S** and **D** wave components, is influenced originally by many physiologic factors such as age, heart rate, respiration, LV function and loading conditions (Bollmann, 2007). These factors should be taken into account when evaluating PVF recording.

and **S2** is ascribed to passive LA wall stretching caused by vigorous LV contraction toward apical direction. Peak velocity and velocity-time integral of the **S** wave are usually greater than those of the **D** wave. Gentile et al (1997) investigated Doppler PVF parameters in 143 healthy individuals aged from 20 to 80 years by TTE. Age-dependent Doppler parameters are reported to be as follows; peak amplitude and time integral of both **S** and **D** waves, and **S**/**D** peak amplitude and integral ratios, whereas **Ar** wave is reported to be ageindependent. These findings indicate the possibility of **Ar** wave as a diagnostic tool of

various hemodynamic abnormalities in a wide range of patients' age.

**S D**

**Ar**

Fig. 2. Representative Doppler imaging of transthoracic echocardiography applied to a patient with hypertension. Upper image is an apical four chamber view with color flow indicating blood flow returning from right superior pulmonary vein (PV) into left atrium (LA). Middle is an ECG tracing (standard limb lead II). Lower is a continuous-wave Doppler PV flow velocimetry during an entire cardiac cycle. Upward direction indicates forward

LA plays three different roles periodically in an entire cardiac cycle, i.e., LA acts as a 'booster pump' when LA contracts in late LV diastole, then as a 'reservoir' during LV systole, and finally as a 'conduit' during early LV diastole (**Fig. 3**). These three kinds of LA functions correspond with **Ar**, **S** and **D** waves respectively, and are estimated also by LA volume curve during an entire cardiac cycle using automatic boundary detection (Zhang et al, 1998), manual tracking (Ogawa et al, 2009), and speckle tracking techniques (Mori et al, 2011). The PVF pattern, especially **S** and **D** wave components, is influenced originally by many physiologic factors such as age, heart rate, respiration, LV function and loading conditions (Bollmann, 2007). These factors should be taken into account when evaluating

flow, whereas downward direction means reverse flow.

PVF recording.

Fig. 3. Illustration of normal blood flow pattern in PV and LA during an entire cardiac cycle.

Pulmonary Venous Flow Pattern and Atrial Fibrillation: Fact and Controversy 83

Furthermore, both cholinergic and adrenergic nerve endings are found together within a single neural plexus of PV, and nerve density is highest in the PV antrum (Tan et al, 2006). The physiological meaning of these ganglionated plexi remains to be speculative. Considering the 'throttle' valve function of myocardial sleeves, ganglionated plexi located in PV-LA junctions may play a role of neural control of cardiac output by regulating proximal PV tonus. Elevated PV tonus associated with pathological condition such as heart failure (e.g., ganglionated plexi out of neural control) may lead to the occasion of acute pulmonary edema leading to severe dyspnea or orthopnea. AF *per se* also shows potential autonomic influence (**Fig. 1**). The correlations between the neural aspect of AF and the ganglionated

PVF is visualized by Doppler echocardiography not only in sinus rhythm but also during AF. AF is characterized by electrophysiological and mechanical properties such as rapid, irregular and fragmented electrical activities and absence of complete LA contraction and relaxation. Therefore, PVF during AF is known as loss of **Ar** wave, blunted **S** wave and relatively dominant **D** wave. Loss of synchronous LA contraction is reflected by disappearance of **Ar** wave. Similarly, loss of complete LA relaxation causes a delayed onset

plexi possibly influencing PV tonus or contraction are the subjects of future study.

LAA RAA

Fig. 4. Schematic illustration of human atria and adjacent great veins. Myocardial sleeves run in PV wall longitudinally, obliquely and cross-sectionally. Myocardial sleeve in superior

PV is usually longer than that in inferior PV. IVC, inferior vena cava; LAA, left atrial appendage; LIPV, left inferior PV; LSPV, left superior PV; RAA, right atrial appendage;

RIPV, right inferior PV; RSPV, right superior PV; SVC, superior vena cava.

**5. PVF during ongoing AF** 

of **S** wave.

### **4. Myocardial sleeves in PV**

In the PVF profile, **Ar** wave reflects physiologic PV regurgitation during LA contraction due to the absence of an anatomic valve at the PV-LA junction. Interestingly, PV wall contains myocardial sleeves instead of anatomic valve. **Fig. 4** is a schematic illustration of human atria and adjacent great veins. Posterior LA wall contains complicated myocardial layers for myocardial sleeves running longitudinally, cross-sectionally and obliquely within the PV walls. Histologically, myocardial sleeves exist in the mid-layer of PV walls (**Fig. 5A**). The myocardial sleeves are, therefore, considered to function as a 'sphincter', which minimizes the PV regurgitation caused by LA contraction. PV contraction is actually confirmed, and this phenomenon is mainly due to the presence of myocardial sleeves contracting synchronously with LA myocardium. This is validated by radiofrequency catheter ablation, i.e., perfect PV isolation (e.g., electrical disconnection of PV-LA junction) is reported to abolish the PV contraction (Atwater et al, 2011). These sleeves also function as 'throttle' valve that regulates cardiac output for systemic circulation (Burch & Romney, 1954). The myocardial sleeves show characteristic electrophysiological properties prone to yield spontaneous repetitive firings which propagate to LA and cause frequent ectopic beats. **Fig. 5B** is the microelectrode recording of the intracellular potentials of guinea-pig LA and myocardial sleeve in PV. Resting membrane potential in myocardial sleeve is less negative relative to that of LA. Moreover, myocardial sleeve in PV show the tendency of spontaneous electrical activity leading to the abnormal automaticity initiating AF. These arrhythmogenic foci act as a 'driver' to trigger and maintain paroxysms of AF. Highly compliant PV wall allows own cyclic stretching due to physiological PV regurgitation. This phenomenon is considered to accentuate intracellular Ca2+ dynamics mediated by stretch-activated ion channels, which is a prerequisite of repetitive electrical firing (de Bakker et al., 2002; Honjo et al., 2003; Chou et al., 2005, Takahara et al, 2011). Moreover, the myocardial sleeves within PV show a complicated anisotropic orientation of myocardial fibers separated by fibrotic tissues causing impaired electrotonic interactions, which accentuates intrinsic spontaneous firing and triggered activity (Nathan & Eliakim, 1966).

Since Haïssaguerre et al (1998) demonstrated the ectopic and spontaneous electrical activities in the myocardial sleeves located in PV responsible for triggering AF, main stream of the AF research has been changed over the past decade, in that recent AF study focused on many areas which had not been given much attention. Importance of the myocardial sleeves as arrhythmogenic foci in AF is confirmed also in the human postmortem studies. Tagawa et al (2001) investigated myocardial sleeve distribution in patients with AF or without AF. They showed that the significantly longer distance of sleeves extending to the peripheral end of PV in AF patients relative to the distance in control patients was confirmed in inferior but not superior PV. In addition, myocytes in PV of AF patients were not uniform and surrounded by fibrous tissues compared with those in controls. Moreover, Steiner et al (2006) reported that amyloid deposition and scarring in myocardial sleeves tended to be observed more frequently in AF patients relative to control patients. Interestingly, the incidence of atrial myocardium extending beyond the PV-LA junction up to the PV periphery in all the examined PV specimens is commonly reported to be 88 to 89% (Tagawa et al, 2001; Steiner et al, 2006).

According with an advance of immunohistochemical techniques, autonomic nervous innervation in PV has been elucidated. Ganglionated plexi are reported to be abundant around the great vessels of the human heart including PV (Armour et al, 1997). Furthermore, both cholinergic and adrenergic nerve endings are found together within a single neural plexus of PV, and nerve density is highest in the PV antrum (Tan et al, 2006). The physiological meaning of these ganglionated plexi remains to be speculative. Considering the 'throttle' valve function of myocardial sleeves, ganglionated plexi located in PV-LA junctions may play a role of neural control of cardiac output by regulating proximal PV tonus. Elevated PV tonus associated with pathological condition such as heart failure (e.g., ganglionated plexi out of neural control) may lead to the occasion of acute pulmonary edema leading to severe dyspnea or orthopnea. AF *per se* also shows potential autonomic influence (**Fig. 1**). The correlations between the neural aspect of AF and the ganglionated plexi possibly influencing PV tonus or contraction are the subjects of future study.
