Strain imaging for risk stratification of atrial arrhythmias in d-transposition of great arteries post-anatomic repair

Neonatal anatomic repair is the preferred surgical approach for the management of d-transposition of great arteries (d-TGA) [1]. Unlike the atrial switch operation, patients that underwent anatomic repair do not typically have substrates for atrial arrhythmias (AA) such as atrial baffles and extensive atrial surgical scars [2]. As a result, we expect that these patients would have a low risk of AA, but data are currently lacking. Atrial strain imaging has been used for prognostication in patients with acquired heart disease because it can estimate the risk of incident or recurrent AA, independent of other clinical risk factors [3]. However, the role of atrial strain imaging for risk stratification has not been studied in adults with d-TGA. Our study objectives were to determine: (1) the prevalence of AA at baseline encounter, and the risk of new-onset and recurrent AA during follow-up. (2) the association between atrial strain and AA.

To attain these objectives, we studied consecutive adults with d-TGA and neonatal anatomic repair that had adequate echocardiographic images for offline assessment of atrial strain (2003–2021). Offline image analyses were performed in all patients as previously described [4]. Right atrial reservoir strain (RA_RS) and left atrial reservoir strain (LA_RS) were used to assess RA and LA function, respectively.

AA was defined as atrial flutter, atrial tachycardia, and atrial fibrillation documented on electrocardiogram, Holter monitor, event monitor, or rhythm strip. Atrial flutter and atrial tachycardia were grouped together because of difficulty to reliably differentiate between both arrhythmias on surface electrocardiogram. Prevalent AA was defined as AA present during (or prior to) the baseline encounter in the adult congenital heart disease clinic, while recurrent and new-onset AA were defined as arrhythmias occurring during follow-up in patients with versus without prior history of AA, respectively.

Of 166 patients, we identified 151 (91%) with atrial strain data. Of these 151 patients, 14 (9%) had prevalent AA at baseline encounter (atrial flutter/tachycardia n = 11; atrial fibrillation n = 3). The mean RA-RS and LA-RS were 36 ± 11% and 41 ± 13%, respectively. Compared to patients without AA at baseline, those with prevalent AA were older, had lower RA-RS and LA-RS, and more likely to have had a Rastelli operation (Table 1).

Table 1.

Baseline characteristics.

	All (n = 151)	Prevalent AA (n = 14, 9%)	No Prevalent AA (n = 137, 91%)	p
Age, years	21 (19–28)	29 (23–37)	20 (18–37)	<0.001
Male sex	93 (62%)	9 (64%)	84 (61%)	0.7
Associated structural lesions
Isolated d-TGA	52 (34%)	3 (21%)	49 (36%)	0.3
Complex d-TGA	99 (66%)	11 (79%)	88 (64%)	0.3
Anatomic repair
Arterial switch operation	89 (59%)	4 (29%)	85 (62%)	0.02
Rastelli operation	62 (41%)	10 (71%)	52 (38%)	0.02
Comorbidities
Hypertension	8 (5%)	2 (14%)	6 (4%)	0.7
Diabetes	6 (4%)	2 (14%)	4 (3%)	0.8
Antiarrhythmic drugs
Class II	24 (16%)	4 (29%)	20 (15%)	0.2
Class III	5 (3%)	5 (36%)	0	–
Class IV	2 (1%)	2 (1%)	0	–
Thromboprophylaxis
Antiplatelet therapy	42 (28%)	9 (64%)	33 (24%)	0.001
Vitamin K antagonist	9	7 (50%)	2 (2%)	0.002
Direct oral anticoagulant	2 (1%)	2 (14%)	0	–
Echocardiography
Right Heart Indices
RA reservoir strain, %	36 ± 11	26 ± 7	39 ± 9	0.005
RA volume index, ml/m2	26 ± 14	34 ± 9	25 ± 10	0.01
Estimated RA pressure, mmHg	5 (3–8)	6 (5–8)	4 (3–6)	0.04
Estimated RV systolic pressure, mmHg	41 (32–61)	52 (40–68)	38 (32–57)	<0.001
RV global longitudinal strain, %	−22 ± 6	−16 ± 4	−24 ± 5	<0.001
Left Heart Indices
LA reservoir strain, %	41 ± 13	32 ± 9	43 ± 11	0.01
LA volume index, ml/m2	28 ± 11	33 ± 8	27 ± 10	0.2
LV longitudinal strain, %	−21 ± 4	−20 ± 4	22 ± 4	0.7

AA: atrial arrhythmia; LA: left atrium; LV: left ventricle; RA: right atrium; RV: right ventricle; TGA: transposition of great arteries.

RA-RS (odds ratio [OR] 0.91, 95% confidence interval 0.87–0.95, p < 0.001), LA-RS (OR 0.94, 0.90–0.98, p = 0.008), prior Rastelli operation (OR 2.23, 1.04–4.17, p = 0.03) and right ventricular global longitudinal strain (RVGLS) (OR 0.93, 0.88–0.98, p = 0.004) were associated with prevalent AA, after adjustment for age, sex, and left ventricular global longitudinal strain (LVGLS). Noteworthy, that RVGLS and LVGLS were modeled as absolute (i.e., positive) values.

Of 137 patients without AA at baseline, 13 (10%) developed new-onset AA (atrial flutter/tachycardia n = 9, and atrial fibrillation n = 4) within 5.4 (3.2–8.6) years of follow-up. The annual incidence of AA was 1.7%/year, and 5-year cumulative incidence of AA was 9%. RA-RS (hazard ratio [HR] 0.94, 0.92–0.98, p < 0.001), LA-RS (HR 0.95, 0.91–0.99, p = 0.02) and RVGLS (HR 0.96, 0.93–0.99, p = 0.01) were associated with new-onset AA, while prior Rastelli operation (HR 2.08, 0.98–5.13, p = 0.09) did not reach statistical significance after adjustment for age, sex, and LVGLS.

Of 27 patients with AA (14 with prevalent and 13 with new-onset AA), 11 (41%) had recurrent AA (atrial flutter/tachycardia n = 6, and atrial fibrillation n = 5) within 4.7 (2.8–6.3) years of follow-up. The annual incidence of recurrent AA was 8.7%/year, and 5-year cumulative incidence of recurrent AA was 43%. Similarly, RA-RS (HR 0.92, 0.87–0.98, p = 0.01) and RVGLS (HR 0.95, 0.90–0.99, p = 0.04), but not LA-RS (HR 0.93, 0.84–1.02, p = 0.3), were associated with recurrent AA, after adjustment for age, sex, and LVGLS. Of note, 9 of the 27 patients with AA had prior Rastelli operation, and AA was more common among patients with Rastelli operation (29% [18/62]) as compared to arterial switch operation (10% [9/89]), p = 0.003.

In summary, AA was not uncommon in patients with anatomic repair, and atrial strain and RV strain were associated with risk of AA, and hence, can potentially be used for risk stratification. The higher prevalence of AA among patients with Rastelli operation suggests that right heart disease may be a major factor in the pathogenesis of AA in this population. The study was limited by a small sample size, and this did not allow for the assessment of the effect of antiarrhythmic therapy and cardiac interventions on the risk of AA. Therefore, there is a need for further studies with larger sample size to determine whether atrial strain can be used to proactively identify patients that would respond favorably to different antiarrhythmic therapies, thereby improving patient selection, and in turn, improving outcomes.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

As this was a retrospective analysis of data collected for routine clinical care, individual informed consent was not required. The study was registered and approved by the local Ethics Committee.