In utero diagnosis of long QT syndrome by magnetocardiography

Abstract

Background: The electrophysiology of long QT syndrome (LQTS) in utero is virtually unstudied. Our goal here was to evaluate the efficacy of fetal magnetocardiography (fMCG) for diagnosis and prognosis of fetuses at risk of LQTS.

Methods and results: We reviewed the pre/postnatal medical records of 30 fetuses referred for fMCG because of a family history of LQTS (n=17); neonatal/childhood sudden cardiac death (n=3), or presentation of prenatal LQTS rhythms (n=12): 2° atrioventricular block, ventricular tachycardia, heart rate < 3(rd) percentile. We evaluated heart rate and reactivity, cardiac time intervals, T-wave characteristics, and initiation/termination of Torsade de Pointes, and compared these with neonatal ECG findings. After birth, subjects were tested for LQTS mutations. Based on accepted clinical criteria, 21 subjects (70%; 9 KCNQ1, 5 KCNH2, 2 SCN5A, 2 other, 3 untested) had LQTS. Using a threshold of corrected QT= 490 ms, fMCG accurately identified LQTS fetuses with 89% (24/27) sensitivity and 89% (8/9) specificity in 36 sessions. Four fetuses (2 KCNH2 and 2 SCN5A), all with corrected QT ≥ 620 ms, had frequent episodes of Torsade de Pointes, which were present 22-79% of the time. Although some episodes initiated with a long-short sequence, most initiations showed QRS aberrancy and a notable lack of pause dependency. T-wave alternans was strongly associated with severe LQTS phenotype.

Conclusions: Corrected QT prolongation (≥490 ms) assessed by fMCG accurately identified LQTS in utero; extreme corrected QT prolongation (≥620 ms) predicted Torsade de Pointes. FMCG can play a critical role in the diagnosis and management of fetuses at risk of LQTS.

Keywords: alternan; arrhythmias, cardiac; fetus; long QT syndrome; magnetocardiography; torsades de pointes.

Figures

Figure 1

Signal-averaged fMCG waveforms, showing QTc duration during sinus rhythm in the cohort of 30 fetuses. In each tracing, the signals from all channels are superimposed. Time is shown on the x-axis, and was adjusted to display a single cardiac cycle. (A): Group 1 fetuses. Rhythm was 2° AV block in subjects #14 and #15. Averaged waveforms were not computed for subject # 6 because the periods of sinus rhythm were brief. The LQTS type (1–3, CALM2, uncharacterized mutation (Un-C) or untested (Un-T)) are shown on the illustration. The T-waves exhibit a distinctive late-peaking morphology. T-wave humps near the end of the T-wave are evident in subjects #5, #10, and #17. (B): Group 2 fetuses. All had normal postnatal QTc and none had an LQTS mutation or phenotype.

Figure 1

Signal-averaged fMCG waveforms, showing QTc duration during sinus rhythm in the cohort of 30 fetuses. In each tracing, the signals from all channels are superimposed. Time is shown on the x-axis, and was adjusted to display a single cardiac cycle. (A): Group 1 fetuses. Rhythm was 2° AV block in subjects #14 and #15. Averaged waveforms were not computed for subject # 6 because the periods of sinus rhythm were brief. The LQTS type (1–3, CALM2, uncharacterized mutation (Un-C) or untested (Un-T)) are shown on the illustration. The T-waves exhibit a distinctive late-peaking morphology. T-wave humps near the end of the T-wave are evident in subjects #5, #10, and #17. (B): Group 2 fetuses. All had normal postnatal QTc and none had an LQTS mutation or phenotype.

Figure 2

Fetal heart rates of Group 1 and Group 2 subjects, compared with percentiles for normal fetuses. For reference, the 110 bpm FHR line across gestation is shown. For subjects with multiple visits, only the measurement from the first visit is shown.

Figure 3

Chronology of LQTS rhythms (sinus rhythm-grey, 2° AV block-white and TdP-black) during the fMCG session for the 4 fetuses with TdP. A) subject #15, B) subject #16, C) subject #18, D) subject # 6. The recording time in this subject was 120 s.

Figure 4

fMCG tracings at initiation (panels A–D) and termination (panels E–I) of TdP. The x-axis tick marks are 1 s apart. First and last beats of TdP are marked by arrows. (A): Wide complex (QRS duration 85 ms) regular rhythm (cycle length (CL) 488 ms) leads directly to a sustained run of TdP (CL 247 ms) in subject #15. The two beats prior to TdP show marked aberrancy. (B): QRS alternans during wide complex (QRS duration 90 ms) regular rhythm (CL 418 ms) preceding spontaneous onset of a slow TdP (CL 351 ms) in subject #16). (C): An aberrantly conducted QRS complex at same coupling interval as the narrow beats (CL 462 ms) initiates TdP (CL 272 ms) in subject #18. (D): An aberrantly conducted beat in 2° AV block (arrow indicates conducted P-wave; ventricular CL 984, atrial CL 498 ms) leads to TdP (CL 238–281 ms) in subject #15. (E): Second-degree AV block (atrial CL 1064, ventricular CL 540 ms) followed by sinus rhythm with QRS alternans in subject # 15. (F): Transient 3°AV block (ventricular CL 1111 ms; atrial CL 751 ms) after an episode of prolonged TdP (4 min) in subject #16. (G): Pause (1.2 s) followed by narrow complex regular rhythm (CL 538 ms) and T-wave alternans in subject #15. (H). Narrow complex regular rhythm (CL 487 ms) following TdP (CL 244 ms) in subject #18. (I): Narrow complex regular rhythm (CL 487) with T-wave alternans. QRS complexes are shown by dashed arrows in subject #16.

Figure 4

fMCG tracings at initiation (panels A–D) and termination (panels E–I) of TdP. The x-axis tick marks are 1 s apart. First and last beats of TdP are marked by arrows. (A): Wide complex (QRS duration 85 ms) regular rhythm (cycle length (CL) 488 ms) leads directly to a sustained run of TdP (CL 247 ms) in subject #15. The two beats prior to TdP show marked aberrancy. (B): QRS alternans during wide complex (QRS duration 90 ms) regular rhythm (CL 418 ms) preceding spontaneous onset of a slow TdP (CL 351 ms) in subject #16). (C): An aberrantly conducted QRS complex at same coupling interval as the narrow beats (CL 462 ms) initiates TdP (CL 272 ms) in subject #18. (D): An aberrantly conducted beat in 2° AV block (arrow indicates conducted P-wave; ventricular CL 984, atrial CL 498 ms) leads to TdP (CL 238–281 ms) in subject #15. (E): Second-degree AV block (atrial CL 1064, ventricular CL 540 ms) followed by sinus rhythm with QRS alternans in subject # 15. (F): Transient 3°AV block (ventricular CL 1111 ms; atrial CL 751 ms) after an episode of prolonged TdP (4 min) in subject #16. (G): Pause (1.2 s) followed by narrow complex regular rhythm (CL 538 ms) and T-wave alternans in subject #15. (H). Narrow complex regular rhythm (CL 487 ms) following TdP (CL 244 ms) in subject #18. (I): Narrow complex regular rhythm (CL 487) with T-wave alternans. QRS complexes are shown by dashed arrows in subject #16.

Figure 5

Examples of ABAB pattern of T-wave alternans (TWA). (A) Subject #15 (SCN5A mutation and TdP). Notice that the widths of the QRS complexes alternate from beat to beat (QRS alternans). (B) Subject #16 (de novo SCN5A L409P mutation and TdP). QRS alternans is present. The T-waves are highly abnormal T-waves and are not well defined, but exhibit wide swings in amplitude and polarity. (C) Subject #10 (uncharacterized mutation, no TdP). D: Subject #11 (CALM 2 mutation and postnatal TdP ). This subject exhibits the most prominent beat-to-beat changes in T-wave morphology. The QRS complexes are biphasic and alternate polarity from beat to beat. E) Subject #17 (KCNH2 mutation, PVC, delivered at 34 weeks). This tracing shows a transition in the character of the TWA between the beginning and the end of the tracing. The second half of the tracing shows QRS alternans with alternating polarity from beat to beat and “R on T” phenomenon. F) Subject #18 (KCNH2 mutation, TdP). TWA and marked QRS aberrancy in alternating beats due to “R-on-T” phenomenon.

Figure 6

Relationship between T/QRS amplitude, fetal heart rate, and fetal movement in subject #9 (KCNQ1 mutation). The time scale is different for each panel. (A) 9 s rhythm tracing. The T/QRS amplitude shows a transient increase between 547–551 s. (B) 300 s fetal heart rate tracing. During the approximate period of 540–550 s (dashed line), which corresponds to the time of the T/QRS increase seen in (A), the fHR tracing is flat. (C). 300 s actogram tracing. There is vigorous fetal movement around time 550 s (dashed line), corresponding to the time of the T/QRS increase seen in (A), as well as at around time 360 s; however, as seen in (B), the fHR is nonreactive.