Systolic anterior motion begins at low left ventricular outflow tract velocity in obstructive hypertrophic cardiomyopathy

Patients

We prospectively studied 25 consecutive patients with obstructive HCM who had resting LV outflow gradients greater than 36 mm Hg and technically adequate echocardiograms. Mean age was 63 years (27 to 86 years); 11 were women. Patients were referred for clinical indications; medications were continued at the time of the study. All patients had asymmetric LV hypertrophy with septal thickness >15 mm and had no apparent cause for the hypertrophy (12). None had hypertension or aortic valve disease. All had SAM and mitral–septal contact. The peak pressure gradient across the LVOT was determined using continuous wave (CW) Doppler in the apical five-chamber view (13). All patients were in sinus rhythm.

Echocardiogram acquisition, selection and measurement

Studies were performed at rest in the left lateral decubitus position using Hewlett-Packard Sonos 1000 echocardiographs.

M-mode echocardiogram

The M-mode echocardiogram recordings were made from the parasternal window during the same examination and within 5 min of the Doppler tracings. A modified electrocardiographic (ECG) lead I was continuously recorded. The mitral valve area of interest was magnified and recorded at 100 mm/s sweep speed. Views showing the mitral coaptation point and the most SAM of the mitral valve were recorded on videotape. The M-mode views were correlated with the two-dimensional (2-D) view to avoid mistakenly recording chordal SAM.

Tracings of SAM with both a clear mitral coaptation point and continuity with mitral–septal apposition were subsequently selected for measurement using a Nova Microsonics analysis computer. We measured the time interval from Q-wave onset till the first abrupt systolic anterior movement of the mitral valve. We measured the RR interval for each M-mode beat selected. Figure 1 illustrates the M-mode and Doppler measurements.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(Left) Continuous wave (CW) Doppler tracings were recorded in the five-chamber view through the LVOT as shown by the dashed line. Pulsed wave (PW) Doppler tracings were recorded in the LV at a point 2.5 cm apical of the mitral coaptation point and 1 cm from the septum, in LV outflow, as marked with the X. This location is the AMV point, apical of the mitral valve. (Right) The time interval Q-SAM is the interval from Q-wave onset to SAM onset. The velocity V-SAM is the CW velocity in the outflow tract at the time interval Q-SAM. A_r = atrial reflected wave.

Statistics

Continuous data are presented as mean ± SEM (range). The Student group t tests were calculated for comparisons of means. Relations between different variables were assessed by Pearson’s correlation coefficient. Significance level was based on a two-tailed test. A value of p < 0.05 was considered significant.

The RR intervals

In the 25 patients as a group, little difference was seen in the mean RR intervals of the analyzed M-mode, CW and PW Doppler tracings; the mean RR intervals were 868 ± 25, 872 ± 25, and 874 ± 25 ms, respectively. For individual patients, the mean difference between the average RR intervals of the M-mode and CW tracings was 24 ± 3 ms (range 2 to 48 ms). The mean difference between the average RR intervals of the M-mode and PW tracings was 20 ± 3 ms (range 0 to 48 ms).

Preejection SAM

In 17 patients (68%), the onset of SAM occurred before the onset of ejection flow recorded on both CW and PW Doppler. In these patients with preejection SAM, M-mode SAM began earlier than in the patients with SAM beginning after ejection onset, at 62 ± 4 compared with 92 ± 11 ms after Q-wave onset (p = 0.03). Examples of patients with early SAM are shown in Figures 2 and 3.

Velocity

It is the central premise of the Venturi theory of SAM that high velocity ejection flow causes a local underpressure in the LVOT, which pulls the protruding mitral leaflet toward the septum. In this study, we show that SAM actually begins very early in systole at a time when velocity in the LVOT is normal. Specifically, SAM begins at a time when mean outflow tract velocity is 89 cm/s (range 43 to 193), a velocity not unlike that routinely recorded in the LVOT of normal patients without SAM (20).

Decreased velocity significantly decreases lift in two ways, and decreases its importance compared with drag forces. First, lift falls because it is roughly proportional to the square of velocity. Second, for any given shape, the ratio of lift to drag falls with decreasing velocity (10,11). At lower velocity, profile drag increases because increased contact of the fluid with the surface leads to increased friction with flow. Lower velocity allows viscosity to do its work. The difficulty of flying at low speed is an example of falling lift/drag ratio at lower velocity.

The modest outflow velocities we have measured indicate that, at the time of SAM onset, the magnitude of the Venturi force is much smaller than previously assumed; high velocities needed to generate significant Venturi forces are, in fact, not present. Therefore, the Venturi mechanism cannot be the main cause of SAM.

Flow drag: geometric shape and orientation relative to flow

Besides velocity, the other factors that determine the relative importance of lift and drag are shape and orientation relative to flow (10,11), in this case between the mitral valve and the LV intraventricular flow (6,8). Available data with regard to these geometric relations indicate that drag is the dominant hydrodynamic mechanism for SAM (6–9).

In obstructive HCM the mitral valve leaflets are anteriorly positioned, relatively large, and residual portions of leaflets extend past the coaptation point and protrude into the outflow tract (1,7,21–23). The anterior displacement puts the mitral valve into the edge of the flow stream of LV ejection, subjecting the mitral valve to the hemodynamic force of ejection flow (1).

This LVOT narrowing and anterior position of the mitral coaptation point can be cited as evidence for either Venturi or drag (pushing) mechanisms, as the narrowing could play a role in both mechanisms. On the one hand, flow velocity must increase as it enters the narrowed outflow tract producing Venturi (lift) forces (24,25). Such Venturi forces are necessarily present. On the other hand, narrowing of the LVOT and the anterior position of the coaptation point also places the protruding leaflet into the edge of the flow stream, subject to the pushing force of flow that strikes the undersurface of the leaflet (6–9,23). The flow drag mechanism is illustrated in Figure 4. Thus, the LVOT narrowing provides the substrate for, and is evidence to support, both theories. This has been a source of confusion in the debate.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

The pushing force of flow. Intraventricular flow relative to the mitral valve in the apical five-chamber view. In obstructive HCM the mitral leaflet coaptation point is closer to the septum than normal (1). The protruding leaflets extend into the edge of the flowstream and are swept by the pushing force of flow toward the septum. Flow pushes the underside of the leaflets (arrow)(6–9). Note that the midseptal bulge redirects flow so that it comes from a relatively lateral and posterior direction; on the five-chamber view, flow comes from “right field” or “one o’clock” direction. This contributes to the high angle of attack relative to the protruding leaflets. Also note that the posterior mitral leaflet is shielded and separated from outflow tract flow by the cowl of the anterior leaflet. Venturi flow in the outflow tract cannot be lifting the posterior leaflet because there is little or no area of this leaflet exposed to outflow tract flow. Venturi forces cannot be causing the anterior motion of the posterior leaflet.

Additional data show that drag forces are more important than Venturi for causing SAM.

Canine and in vitro experiments

Without septal hypertrophy, SAM and obstruction can be induced in normal dogs by anterior displacement of the papillary muscles with ligatures (7). In an in vitro model, SAM does not occur when the valve has its normal posterior coaptation point, regardless of how high velocities are elevated in the LVOT. Not until the valve is mechanically lifted into the outflow stream and chordal slack is introduced does SAM occur (9,26).

In HCM patients: orientation, shape, temporal and Doppler velocity evidence

In patients, a high angle of attack of Doppler color flow relative to the protruding leaflet of the mitral valve has been found; this orientation precludes significant Venturi effects and implicates drag (6). In the apical five-chamber view, local flow direction comes from an angle lateral of the protruding leaflet. The mean angle at time of mitral coaptation was lateral by 21°; mean angle just before septal contact increased to 45° (6). At these high angles relative to flow, drag is the dominant hydrodynamic force on the leaflet. This is illustrated in Figure 5. Flow pushes on the underside of the protruding leaflet (6–9), as drawn in Figure 4 and imaged in a patient in Figure 6. Critical overlap exists between the inflow and outflow portions of the LV that allows flow drag to catch the leaflet (6,27).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

(Left) In the example of a wing, as it moves from a position parallel to flow to a more angled orientation, flow separates from the upper surface, lift decreases, and drag increases (10,11). (Top) Lift is the dominant force up to an angle of approximately 15° relative to air flow. (Middle and bottom) Above 15°, separation occurs and lift is greatly attenuated because the rapid flow and its local decrease in pressure are no longer in contact with the top surface. Lift is lost, drag forces are dominant, and the wing stalls (10,11). The angle between flow and a surface is referred to as the angle of attack. Streamlines are shown with continuous arrows; forces on the wing are shown with larger gray arrows. (Right) Similarly, in HCM the magnitude of flow drag on the protruding mitral leaflet is directly related to the angle between the local direction of LV flow and the protruding leaflet. (Middle) The angle between the direction of local flow and the protruding leaflet of the mitral valve is the angle of attack, α. The anterior position of the protruding leaflet contributes to the angle of attack. Also, the midseptal bulge redirects flow so that local flow comes from a relatively posterior direction as it curves around the septum. Color flow Doppler defines local flow direction (6). Measured angle of attack was high at the moment of coaptation, mean 12° in the parasternal view, and mean 21° in the apical five-chamber view. High angles preclude significant Venturi forces and implicate drag. (Bottom) As the leaflet is pushed toward the septum, both the angle and drag forces increase markedly.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Flow strikes the undersurface and lateral aspect of the mitral valve very early in systole, causing SAM. Two echocardiographic views of one patient with obstructive HCM are shown; resting gradient = 54 mm Hg. (Left) Apical five-chamber view. (Top left) The 2-D image shows the protruding mitral leaflet on the first frame in systole that showed mitral coaptation. Arrowhead points to mitral valve. O indicates outflow tract. (Bottom left) The same view showing the first systolic frame with color flow. Color flow is seen lateral to the leaflet tips (arrow). These images show the event graphically drawn inFigure 4.(Right) Apical three-chamber view. (Top right) The 2-D image again shows the protruding mitral leaflet on the first frame in systole that showed mitral coaptation. Arrowhead points to mitral valve. (Bottom right) This shows the same view of the first systolic frame with color flow. Color flow is seen posterior to the leaflet tip (arrow). Note that color flow velocity is low on both views. On 2-D imaging, the next systolic frames showed fully developed SAM on both views. On color flow, the next systolic frames showed aliasing in the outflow tract. The mitral leaflets are medially and anteriorly positioned into the edge of the flow stream. Low velocity flow strikes the undersurface of valve leaflets; they are swept toward the septum by the pushing force of flow.

The leaflet is swept by the pushing force of flow into the septum. As it moves toward the septum the angle relative to flow increases and constricting drag forces are correspondingly increased as the leaflet presents a greater surface relative to flow (Fig. 5)(6). An example of this phenomenon is a widely open door in a drafty corridor. The door starts moving slowly; it then accelerates as drag increases because the door presents a larger area, until it slams shut (28).

Shape of the mitral valve

The mitral valve was not designed for lift. To the contrary, the valve has a sharp anterior edge with no streamlining, and there is a concavity under the cowl of the protruding leaflet. It resembles other biologic structures with high drag coefficient (Fig. 5)(10).

The midseptal bulge

Besides the anterior position of the mitral valve, another contributor to the initial positive angle of attack is the midseptal bulge. In patients with obstructive HCM and resting gradients, hypertrophy of the midventricular septum is the rule, occurring in 92% of patients (29). This midseptal bulge is often associated with basilar or whole septal thickening (29,30). Resting gradients are uncommon in patients with only subaortic basilar septal thickening; in these patients resting gradients are found in just 12% (29). If it were Venturi forces from acceleration into a small outflow tract that caused resting obstruction, then isolated basilar hypertrophy and subaortic narrowing should commonly cause resting obstruction. But this is uncommon; patients with just basilar outflow tract hypertrophy usually just have provocable obstruction.

Instead, data indicate that midventricular septal thickening is generally a necessary condition for resting SAM with mitral–septal contact (29). We have observed that this occurs because the midseptal bulge forces the outflow to sweep from a relatively posterior and lateral direction in the LV, as shown in Figures 4 and 6. When viewed in the echocardiographic apical five-chamber view, flow comes from “right field” or “one o’clock” direction. This contributes to a high angle of attack relative to the protruding mitral leaflet (6). Klues and colleagues (30) reported that SAM was more common in patients with several hypertrophied segments rather than with one hypertrophied segment; SAM occurred more often when the proximal segments of the LV were thickened rather than the distal half.

Posterior leaflet SAM

The anterior motion of the posterior mitral leaflet is evidence for the drag mechanism. In 89% of patients with SAM and obstruction, the posterior leaflet moves anteriorly as well (31,32). However, the posterior leaflet is separated from the flow in the LVOT by the cowl of the anterior leaflet (Fig. 4). Venturi forces in the LVOT cannot be lifting the posterior leaflet because there is little or no area of the posterior leaflet that is exposed to LVOT flow. In light of this and the previously mentioned observations, it is concluded that the posterior leaflet is pushed anteriorly. This mechanism is shown in Figure 6. Is it likely that the anterior and posterior leaflets, which share a coaptation plane, have different causes for SAM? It is more reasonable that the anterior motion of the anterior leaflet is caused by the same force that triggers the abnormal posterior leaflet motion; both are caused by flow drag.

Finally, the last ingredient necessary for SAM is chordal slack. Without a reduction in chordal tension, no SAM would occur because the leaflets would be tethered (8,9,23). Systolic anterior motion can be described as anteriorly directed mitral valve prolapse. This analogy has merit; in both conditions, the mitral valve is often large and is pushed by flow from its normal systolic position, resulting in mitral regurgitation.

In summary, geometric data indicate that the necessary conditions for SAM are anterior position of the mitral coaptation point (which puts the valve into the edge of the outflow stream), positive angle of attack between outflow and the protruding leaflet, and chordal slack.

The geometric conditions for SAM are met in another condition where obstruction occurs, despite the absence of septal hypertrophy. Following mitral annuloplasty certain patients develop, as a complication, SAM. The mitral coaptation point has been shown to be anteriorly displaced into the outflow tract by the ring (33). Consequently, surgical techniques have been developed to ensure that the postoperative mitral coaptation point is posterior in the LV, explicitly out of the way of the outflow stream with its attendant drag (34,35).

In the present work, we have shown that rapid ejection flow velocity in the LVOT is not a necessary condition for SAM. In addition, we found that SAM begins before ejection onset in two-thirds of the patients. That SAM often occurs before ejection begins was first reported in 1987; in 10 patients on 2-D echocardiography SAM was seen to begin before the aortic valve opened (8). These temporal data, confirmed in the present work, support the drag hypothesis, rather than Venturi. In patients with preejection SAM, the early motion has been observed to be associated with a prominent LV preejection flow (36). Figure 7 summarizes the evidence in the debate between Venturi and drag forces as the cause of SAM.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7

Evidence in the debate between Venturi (lift) and drag (pushing) force as the dynamic cause for SAM. References are enumerated in the text. ASH = asymmetric septal hypertrophy.

Therapeutic implications

Recently, new methods to relieve obstruction have been developed: revised surgical techniques (37–40), dual chamber pacing (41–44) and percutaneous transluminal septal myocardial ablation (45,46). Interventions are not always successful (47), and the reason for heterogeneity in response is not clear (43). Despite active research in treatment, the hemodynamic mechanism of obstruction has hitherto remained controversial (2–5). By clarifying the circumstances and the nature of the dynamic force that causes obstruction, the present study should help explain how novel treatments work and also could permit their refinement. This understanding is of crucial importance for the development of future treatment strategies.

New surgical approaches address the problem of the large protruding mitral valve and its contribution to obstruction; the operation described by McIntosh reduces by plication the size of the anterior mitral leaflet, adding this to myomectomy (37). Another frees the bound papillary muscles from the surrounding LV muscle. This allows more normal posterior mitral coaptation, explicitly out of the outflow stream and its drag forces (38,39). As Messmer states: “This relieves the obstructive component of the mitral valve, which is rarely due to the often cited and never proved Venturi effect but has its origin rather in pathologic insertion of subvalvular apparatus” (38).

As discussed above, the impact of the midventricular septal bulge is to redirect LV flow so that it comes from a relatively postero-lateral direction. Consequently, an important goal of myomectomy must be to extend the myomectomy resection far enough down toward the apex to allow flow to track more anteriorly and medially along the surgically reduced septum and away from the mitral valve (23,38,39). Adequate extent of the myomectomy past the tip of the anterior mitral leaflets was stressed by Morrow (48) and others (23). A large decrease in the angle of attack of flow relative to the mitral valve has been shown after successful myomectomy; flow is made more parallel to the mitral valve (49). A recent modification extends resection to the deepest portion of the septum, to ensure complete resection of the midseptal bulge (38,39); 90% of patients had no postoperative SAM. Conversely, failure of myomectomy to alleviate SAM is often due to inadequate excision; in many such cases only the basal septum has been resected without resection farther down to the midventricle (47,50,51). Such considerations also apply to site selection for percutaneous alcohol ablation procedures. Targeting the basal septum alone will likely be inadequate; the midventricular bulge must also be addressed for complete relief of obstruction.

Dual chamber pacing with complete ventricular preexcitation through short atrioventricular delay significantly reduces outflow tract gradients (41–44). However, therapeutic effect is often incomplete; SAM persists with mean gradients of 30 to 55 mm Hg after three months of pacing (43,44). The mechanism by which pacing benefits SAM is unclear at this time (52). One hypothesis is that pacing might cause asynchronous or paradoxical septal motion, widening the outflow tract and decreasing Venturi forces (42). However, septal paradox is only rarely seen (41,52). Jeanrenaud (52) did find a modest decrease in regional septal wall motion but there was no uniform correlation between the magnitude of decreased septal motion and percent gradient reduction. Also, the present research indicates that a decrease in Venturi forces can only play a minor role in SAM improvement. Therefore, both direct observations and pathophysiology indicate that the mechanism by which pacing reduces SAM is more complex than just a widening of the outflow tract (52).

Study limitations

The M-mode tracings of SAM and Doppler tracings were not done simultaneously. Rather, they were done sequentially, within 5 min. This could have influenced results. However, tracings for M-mode, CW, and PW Doppler analyses were selected for matched RR intervals, and many tracings were selected and measured (4 M-mode and 7 CW Doppler beats). The mean RR intervals between modalities were quite close. For individual patients, the mean difference between the average RR intervals of the M-mode and CW tracings was 24 ± 3 ms (range 2 to 48 ms). No patients with atrial fibrillation were examined in this study. Interventricular pressures were not directly measured. In the parasternal M-mode view, as the mitral valve moves anteriorly, the point with SAM onset and the point of mitral–septal contact may not be on the same M-mode line. For this study, clear registration of SAM onset was deemed essential, sometimes to the detriment of the point of mitral–septal contact.

Conclusions

In patients with obstructive HCM we have shown that the onset of SAM of the mitral valve is a low velocity phenomenon. At SAM onset, though Venturi forces are necessarily present in the outflow tract, their magnitude is much smaller than previously assumed. This velocity data, combined with other experimental, geometrical and temporal observations, indicate that drag, the pushing force of flow, is the dominant hydrodynamic mechanism for SAM.