Prevalence and Mechanisms of Subendocardial Septal Impact Scar in Patients with Hypertrophic Cardiomyopathy

Background: In hypertrophic cardiomyopathy (HCM), left ventricular outflow tract (LVOT) obstruction is associated with systolic anterior motion (SAM) of the mitral valve leaflet. Conventional grading distinguishes SAM grades II and III, which involve physical contact of the mitral valve (MV) with the basal septum, from grade I, which describes MV motion toward the septum without contact1. Pathology studies suggest that MV contact can lead to impact lesions in the basal septum, presenting as subendocardial fibrosis2, however, it is unclear if physical contact is necessary or accelerated/turbulent blood-flow alone can cause scar. On CMR, subendocardial scarring may be attributed to “silent” or “unrecognized” MI even in patients without history of coronary artery disease (CAD). Flow-independent dark-blood delayed enhancement CMR (FIDDLE) is superior to conventional bright-blood DE-CMR in the detection of subendocardial scars3. In this study of HCM patients without history of CAD, the aim was to define the prevalence and mechanisms of subendocardial septal impact scarring–which could be misinterpreted as scarring of ischemic origin–accounting for different grades of SAM and LVOT shape and blood flow characteristics.

Methods: 120 patients with documented HCM were studied. Patients with CAD or other non-HCM cardiomyopathy were excluded. Septal subendocardial scarring was evaluated on the two most basal short axis (SAX) slices showing the complete circumference of the LV myocardium, and a corresponding long-axis view, on both conventional DE-CMR and FIDDLE. SAM grades including chordal SAM (involving only chordae tendineae) were as previously defined. Additional parameters assessed were: peak resting LVOT flow velocity, blood flow acceleration at the septal boundary, and basal septal overhang into the LVOT (Fig. 1).

Results: Mean age was 57 years (50% male). Basal anteroseptal wall thickness was 16±5mm. Basal septal scar was detected by FIDDLE in 38 pts (32%). Fig. 2 shows a progressive, increasing relationship (p < 0.001) between scar prevalence and SAM grade. 100% of patients with physical MV–septal contact (grade II or III) showed septal scarring. Notably, 37% of patients without physical contact (grade I or chordal) nevertheless demonstrated scarring. In this group without contact, septal scar was associated most strongly with flow acceleration at the basal septal boundary (p < 0.001). A moderate relationship with peak resting LVOT flow velocity (p=0.05) and no relationship with basal septal overhang (p=0.57) was found (Fig. 3). There was also a progressive, increasing relationship (p < 0.001) between DE-CMR scar prevalence and SAM grade, albeit prevalence was far lower (Fig 2).

Conclusion: Physical contact of the MV with the basal septum reliably results in subendocardial scarring as detected with FIDDLE. In the absence of physical MV contact, scarring may be related to impact of accelerated LVOT blood flow with the basal septum.

Figure 1. Panel A shows an example of SAM with physical MV–septal contact (left) as well as examples of SAM without physical MV–septal contact, but with flow acceleration occurring along the basal septal boundary (middle, yellow arrows) or with septal overhang into the LVOT (right). Panel B: Patient 1 shows septal basal scar (blue arrows) on the short axis flow-independent dark-blood delayed enhancement (FIDDLE) image, but not on the corresponding conventional bright-blood DE-CMR image. Patient 2 shows septal basal scar on both FIDDLE and conventional bright-blood DE-CMR.
Figure 2. Prevalence of basal septal scar for different grades of SAM. Panel A shows the prevalence of basal septal scar detected using flow-independent dark-blood delayed enhancement imaging (FIDDLE). Panel B shows the prevalence of septal scar detected by regular bright-blood delayed enhancement (DE) CMR. There is a progressive, increasing relationship (p < 0.001) between scar prevalence and SAM grade for both techniques.
Figure 3. LVOT blood flow characteristics and shape in patients with chordal SAM and SAM grade I. Panel A shows a strong relationship between flow acceleration at the basal septal boundary and basal septal scar (p < 0.001). Panel B shows a moderate association between peak resting LVOT flow velocity and basal septal scar (p=0.05), while Panel C shows no association of basal septal scar presence and basal septal overhang (p=0.57).