Alcohol septal ablation (ASA) is a well-established procedure for
patients with obstructive hypertrophic cardiomyopathy (oHCM). It is safe and
less invasive compared with surgery, and the long-term survival rates after ASA
seem similar to those after myectomy at 5-year follow-up at least [1, 2]. But
the higher reoperation rates and the higher frequency of residual obstruction
comparing with myectomy are still concerning [3, 4]. However, the residual LVOT
gradients may lead to adverse outcomes [5]. Understanding of ASA effects seems
crucial for better patient selection and improvement of outcomes.
This study is
focusing on ASA efficiency in different groups of patients with oHCM depending
on the severity of the left ventricle outflow tract (LVOT) gradient and the
severity of basal septal hypertrophy.
Material
and methods
Since 2001,
212 patients with oHCM underwent ASA in Sverdlovsk regional hospital No.1 and
Tyumen Regional clinical hospital No.1. The diagnosis was based on typical
clinical findings, electrocardiograms and comprehensive imagine [transthoracic
(TTE), transesophageal echocardiography (TEE) and/or cardiac magnetic resonance
(CMR)]. Indications for the septal reduction were as follows: the presence of
severe left ventricle outflow obstruction (≥50 mmHg) at rest or after
provocation in cases when initial maximal medical therapy was not effective
enough to reduce heart failure symptoms and/or to reduce LVOT gradient. The
mean age was 52±15. Female patients were 50.7% (107 patients). All relevant
clinical characteristics are presented in table 1. Clinical follow-up was
performed by office visit, phone contact, or structured follow-up.
Intervention
All patients
underwent a classic ASA technique with echo-contrast guidance [6]. Procedures
were done under local anesthesia with analgosedation during ethanol infusion.
One or two arterial accesses were used to perform ASA and hemodynamics
evaluation. The electrode for the temporary pacing was routinely placed in the
right ventricle apex for 3-7 days depending on the AV conduction. Desiccated
alcohol 96% (1-3 ml) was used. Intraoperative TTE was used to evaluate the
perfusion zone of the target septal artery.
Comparison: study design
The study was
performed in compliance with the Helsinki II declaration. We used 2x2 factorial
design to define the ASA performance in different patient’s subsets. All
patients were divided in 4 groups (table 2):
• group 0 -
patients without severe LVOT gradient (<50 mmHg at rest) or severe basal
hypertrophy (<20 mm);
• group 1 -
patients with severe LVOT gradient (>50
mmHg at rest), but without severe basal hypertrophy (<20 mm)
• group 2 -
patients without severe LVOT gradient (<50 mmHg at rest), but with severe
basal hypertrophy (>20 mm)
• group 3 -
patients with severe LVOT gradient (>50
mmHg at rest) and with severe basal hypertrophy (>20
mm).
Clinical,
echocardiographic data were collected at baseline, in the mid-term (<1 year)
and in the long-term follow-up (>1
year).
Definitions
A criterion of
residual obstruction was maximal PG >50
mmHg after provocation. The term “severe obstruction” in this study was defined
as LVOT obstruction with PG >50
mmHg at rest. “Severe hypertrophy” in this study meant the basal thickness of
IVS >20 mm. The combined endpoint
was defined as death, stroke, myocardial infarction, new-onset atrial
fibrillation. Reoperation was defined as repeated ASA or mitral replacement
and/or surgical myectomy.
Outcomes and study endpoints
In this study
dealing with different groups (with or without severe hypertrophy or gradient)
we aimed to determine (1) frequency of residual obstruction after ASA in each
subset of patients; (2) freedom from reoperation in the long-term; (3) the
difference in the long-term survival rates; (4) freedom from combined endpoint
including all-cause death, stroke, myocardial infarction, and new-onset atrial
fibrillation.
Mortality data
were collected from State Insurance Fund and via direct calls to relatives. The
local regional database was used to assess the frequency of residual
obstruction, reoperations, hospital deaths, stroke, myocardial infarction, and
new-onset atrial fibrillation.
Statistical analysis
Data were analyzed
using IBM SPSS v26.0. Graphs were created in R 4.0.4 package using the
“survminer” application. The Shapiro-Wilk test was used to test continuous
variables for normality. Continuous data are presented as a mean ± standard
deviation for variables with normal distribution and median (25th-75th
percentile) for variables with non-normal distribution. Categorical data are
described as absolute numbers and relative frequencies. Non-categorical
variables are summarized using means and were compared using ANOVA or
Kruskal-Wallis tests according to normality of distributions. The Kaplan-Meier
method was applied to construct survival curves. The survival rate was
presented with a 95% confidence interval. A p-value <0.05 was
considered statistically significant.
Results
In the
short-term one patient died from sepsis (0.5%). Technical success was achieved
in 99.5%. One patient required emergent coronary artery bypass grafting due to
left main dissection. A permanent pacemaker was implanted for 11.8% (25)
patients in 30 days. The mean of the heart failure functional class improved
from 2.5±0.6 to 1.5±0.6. The LVOT gradient decreased from 60 (40-89) mmHg to 15
(9-124) mmHg at rest and from 108 (80-135) mmHg to 26 (16-49) mmHg after
provocation. The mean IVS reduction and residual obstruction rates were
assessed at 1-year follow-up after ASA. The mean IVS reduction was 5±3 mm in
the last follow-up.
Residual
obstruction was observed in 42 (20%) cases in the whole cohort. Data about
residual obstruction rates in groups after ASA are summarized in table 2. The
highest rate of residual obstruction was observed in group 3 (gradient >50 mmHg, hypertrophy >20 mm) - 27% (25/69 patients).
Long-term results
The median of
follow-up was 74 (IQR 38-127) months. Total number of mortality cases in the
long-term was 41 patients. Long-term survival rates were as follows 98.1 (95%
CI 96.2-100.0)%, 92.3 (95% CI 88.5-96.3)%, 77.2 (95% CI 70.1-85.0)%, at 1-, 5-,
10- and follow-up, respectively. Causes of long-term death were as follows: stroke
(5 patients, 2.3%); myocardial infarction (3 cases, 1.4%); heart failure (4
patients, 1.8%); sudden death (6 cases, 2.8%), cancer (4 patients, 1.8%);
COVID-19 (2 cases, 0.9%); unknown causes (15 cases, 7%).
Freedom from
the combined endpoint in whole cohort was 95.2 (95% CI 92.3-98.1)%, 82.1 (95%
CI 76.6-87.9)%, 61.9 (95% CI 53.9-71.0) %, at 1-, 5-, 10- and follow-up,
respectively
Multivariable
analyses (Cox regression) identified 3 independent predictors of long-term
death: PG >50 mmHg at rest before ASA
(HR 2.09, 95% CI 1.04-4.20), stroke in the follow-up (HR 3.1, 95% CI
1.27-7.55), and age (HR 1.03, 95% CI 1.00-1.05). C-statistic was 0.64.
The
statistical difference in survival rates between groups was not reached in this
study (p=0.16) (fig. 1A), but comparing a group 3 (PG ≥50 mmHg and IVS
≥20 mm) versus the rest of patients (PG <50 mmHg and/or IVS <20 mm) the
statistically significant difference in survival rates was identified (p=0.029)
with a hazard ratio (HR) = 1.98 (95% CI 1.06-3.72) (fig. 1B).
No significant
difference was found between groups in terms of freedom from the combined
endpoint in the follow-up (p=0.5) (fig. 2)
Observed
reoperation rates in the entire cohort were as follows: 36 (17.1%) events. The
difference in reoperation rates (freedom from myectomy, mitral replacement, or
repeated ASA) between study groups in the follow-up was not identified (p=0.12)
(fig. 3A). Сomparing a group 3 (PG ≥50 mmHg and IVS ≥20 mm) versus the rest of
patients (PG <50 mmHg and/or IVS <20 mm) the statistically significant
difference in reoperation rates was not observed (p=0.2) (fig. 3B).
Comparing patients with PG ≥50 mmHg versus those with PG <50 mmHg the higher
reoperation rates were observed in patients with severe LVOT gradient at
baseline (p=0.046) (HR=2.12; 95% CI 1.00-4.49) (fig. 3C). In contrast,
patients with IVS >20 mm versus
those with IVS<20 mm had similar freedom from reoperations in the long-term
period (p=0.52) (fig. 3D).
Long-term
effects of ASA on common echocardiographic measures are presented below (fig.
4).
Discussion
Findings and its explanation
The essential
findings are as follows: (1) the mean effect size of alcohol ablation on septal
reduction is 5 mm in this study; (2) alcohol ablation works perfectly in
patients with severe hypertrophy as well as non-severe septal hypertrophy, but
it is less efficient in terms of the frequency of residual obstruction in
patients with high gradients in LVOT (≥50 mmHg) at baseline; (3) the higher
mortality rates were observed in the long-term after ASA in patients with oHCM
with both PG ≥50 mmHg and IVS ≥20 mm.
First,
selecting patients for ablation, we should understand its effects and which
patients will get benefit from this kind of intervention. Data from large ASA
registry (Euro-ASA) showed exactly the same mean value of septal reduction (5
mm) as we found [7]. What does it mean for practice? In reports from
experienced surgical teams the average depth of IVS resection after myectomy
ranges from 5 to 10 mm [8, 9]. However, the septal reduction after ASA seems
limited comparing with surgery. It means that we should clearly understand for
which patients this “moderate” effect will be sufficient enough in terms of the
LVOT gradient abolishment and the long-term prognosis.
Second, our
data showed that ASA is seemed very effective in cases of small resting
gradients (<50 mmHg) and its efficiency less depends on septal thickness as
it was considered previously [10]. Anyway, our finding goes in line (are
consistent) with data from J. Veselka et al. [11]. They showed the similar
reoperation rates after ASA in HCM patients with septum ≥30 mm and <30 mm.
Third, the
group 3 (PG ≥50 mmHg and IVS ≥20 mm) demonstrated the highest mortality rates
in the long-term comparing to the rest patients undergoing ASA. Comparing
patients with PG ≥50 mmHg versus those with PG <50 mmHg the statistical
difference in the long-term survival rates did not reach, but seemed borderline
(p=0.054). In contrast, patients with IVS ≥20 mm versus those with IVS
<20 mm showed no impact on the long-term survival (p=0.229).
However, in
our cohort the IVS thickness alone was less impactful on the long-term
prognosis than the LVOT gradient or than a combination of two variables: the
basal hypertrophy thickness and the LVOT gradient.
Guidelines and previous studies
The last HCM
guidelines 2020 recommended surgical myectomy as a preferred option especially
in cases of associated disorders (associated anomalous papillary muscle,
markedly elongated anterior mitral leaflet, intrinsic mitral valve disease,
multivessel CAD, valvular aortic stenosis), but there are no recommendations
regarding the ASA performance in different anatomical settings [15]. However,
just a few studies were addressed to these issues.
С. Van der Lee
et al. showed that myectomy combined with mitral valve plasty results in better
outcomes than ASA in patients with oHCM and elongated anterior mitral leaflet
[12].
To the best of
our knowledge, the only one previous study demonstrated the impact of baseline
gradients on patients’ outcomes after ASA. The research from Р. Sorajja et al.
showed that PG >100 mmHg at baseline is an independent predictor of
long-term death in patients undergoing ASA [14].
Groups reasoning
Our
multivariable analysis identified 3 predictors of the long-term all-cause
death: PG ≥50 mmHg at baseline, age, stroke during follow-up. C-statistic was
0.64. However, it is reasonable to study this Cox model in terms of its
validity in different patients’ cohorts or to build the survival curves
depending on a presence of 1 or 2 or all 3 variables as it was done in a
research form Р. Sorajja et al [14]. But this study was designed to assess the
ASA performance in different groups of patients in a factorial way. We chose
two factors (PG ≥50 mmHg, and IVS ≥20 mm) to divide patients into 4 groups. The
factor of “IVS ≥20 mm” was chosen empirically because the mean of the basal
septal thickness in this study was 21 mm.
Value of findings for practice
This study
adds a better understanding in terms of patient’s selection and which of them
will receive a benefit from this kind of intervention.
So-called
mitral abnormalities (septal elongations, hypertrophy of papillary muscles,
displacement of papillary muscles, etc.) in HCM and its impact on the LVOT
obstruction were already studied previously [12, 16]. However, the obstruction
is a consequence of 2 anatomical structures - septum and mitral valve, which
lead to the occurrence of mitral-septal contact. To the best of our knowledge,
the longevity in the time of this contact is linearly related to the LVOT
obstruction severity. In general, the longer contact means the higher gradient.
The continuous in-time contact may be created predominantly by the thick septum
or the long mitral leaflet or even both. However, even in cases of the thin
basal septum, long mitral leaflet may cause severe or long mitral-septal
contact and the severe gradient as a consequence. Mitral valve in such cases is
looking like a predominant cause of the LVOT obstruction and, however, septal
reduction alone may not be so efficient option.
It was shown
that the mitral valve surgery or mitral chords cutting may be an efficient
enough to abolish the LVOT gradient in patients with the thin septum [13].
This
retrospective, observational study has its own inherent limitations that should
be considered prior to generalization of the results. This study was not
initially designed or powered to assess the survival among the different ASA
groups. Future studies are needed to assess this matter.
Conclusion
Alcohol septal
ablation showed higher reintervention rates in cases with severe resting left
ventricle outflow gradient at baseline. Patients with severe resting gradient
and severe basal hypertrophy at baseline had higher long-term mortality rates
despite of the septal reduction.