Multifocal atrial tachycardia (MAT) is defined as a tachycardia with at least 3 morphologically distinct P-waves. It is also characterized by irregular P-P, R-R, and P-R intervals, and an isoelectric baseline between the P-waves.1) MAT episodes typically last for several minutes to hours. Although the rhythm typically returns to normal between episodes, it can also coexist with atrial fibrillation (AF), atrial flutter (AFL), or both. During the tachycardia period, aberrant ventricular conductions are frequently observed, which must be differentiated from ventricular tachycardia (VT) (Figure 1).

MAT is prevalent in the elderly, particularly those with pulmonary diseases including chronic obstructive pulmonary disease. However, it is relatively rare in the pediatric population. Therefore, the clinical features of MAT in children is not well established.2)3) Several investigators have described MAT as having a mild clinical course with few sequelae and rare mortality,4)5) while others reported adverse outcomes including cardiogenic death or persistent arrhythmia.6)7) The purpose of this report was to ascertain the clinical course and prognosis of pediatric patients with MAT, and identify its potential prognostic factors.

Thirty-three patients who were diagnosed with MAT between 1997 and 2015 were identified by searching the medical records of Seoul National University Children's Hospital. The clinical histories, vital signs, and laboratory and imaging findings, including ECG, Holter monitor, and echocardiography were reviewed.

MAT with a P-wave interval <200 milliseconds or with coexisting AF or AFL was considered to have chaotic features. We defined patients as having “incessant-type MAT” if the arrhythmia persists for >30% of the day on Holter monitoring. In contrast, it was defined as “paroxysmal-type MAT” if the arrhythmia persists briefly. Echocardiographic data, including left ventricular (LV) ejection fraction (EF), ventricular dimension, and atrioventricular valve regurgitation were also reviewed. The ECG and echocardiographic data, as well as the patients' age at onset and accompanied diseases, were investigated as potential prognostic factors. Patients were grouped into infant and non-infant groups based on the age of MAT onset. We additionally placed patients in the idiopathic infant group if the patient had no underlying disease and had onset of MAT at <1 year of age.

With regard to the treatment response, MAT was considered to be controlled if the patient became asymptomatic and had evidence of sinus rhythm either on Holter monitoring or on 2 consecutive ECGs (performed at an interval of >3 months). Cardiopulmonary mortality, or persistence of the arrhythmia for >1 year despite treatment was considered an unfavorable outcome.

Continuous variables were recorded as mean±standard deviation or as median and range. Continuous variables were compared between the groups using the Mann-Whitney U test. Fisher's exact test was used to determine statistical significance between categorical variables. Multivariate logistic regression analysis was conducted to decide the major determinant of the nominal outcome. A p value < 0.05 was considered statistically significant. Statistical analyses were conducted using SPSS 20.0 (IBM, Armonk, NY, USA). This study was approved by the Institutional Review Board at Seoul National University Hospital (No. 1701-087-824, approved, 2017).

Among 33 patients with identified MAT, 19 were boys and 27 (82%) were younger than 1 year old at the time of diagnosis (infant group). The median age at diagnosis of MAT was 1.7 months (range, 0 days–14.2 years). Ten patients (30%) had fetal arrhythmias/tachycardia on fetal monitoring, while nine were first detected incidentally on physical examination (27%). Otherwise, the initial manifestation of MAT included fast/irregular heart beat on monitoring (n=8, 24%), tachypnea/dyspnea/feeding difficulty (n=4, 12%), and lethargy (n=1, 3%).

Most literature regarding pediatric MAT shows a predominance in infants.1)4)5)7)8) Children older than 1 year rarely have documented MAT.9)10)11) Although the patients in this study were mostly infants, the number of non-infant patients was also relatively large. Prior studies have found that MAT is can coexist with structural heart diseases, respiratory diseases, and RASopathy syndromes. Our study confirms these associations. A variety of specific diagnoses in each category was similar to those of previous studies. Consistent with previous studies, the majority of our patients had no definite presenting symptom. Instead, most arrhythmias were detected during routine physical examinations or during other monitoring. However, if tachycardia persists for several days without management, patients can present with heart failure symptoms such as poor feeding, irritability, vomiting, respiratory distress and lethargy due to tachycardia-induced cardiomyopathy. Therefore, if an infant or young child with tachycardia develops poor feeding, irritability, respiratory distress and lethargy, clinicians must consider the possibility of a long-standing tachyarrhythmia. In such cases, immediate management for tachyarrhythmias is necessary.

Atrial premature beats, AF, or AFL are known to accompany MAT in both adults and pediatric patients.1)4)12) The rate of accompanied AFL or AF in our study (49%) was also similar to that of adult reports (55%).13) However, the clinical association between MAT and AVRT was a unique finding in our study. There are 2 reported cases of MAT with Wolff-Parkinson-White syndrome.4)11) Moreover, we also described a phenotypical progression of MAT into CPVT (patients 29 and 30). Although MAT or AF may be found concomitantly with polymorphic VT in CPVT patients, our data emphasize that MAT in young children may be the initial manifestation of a potentially life-threatening arrhythmia of CPVT.

Based on prior studies, the prognosis of MAT varies widely from being clinically benign to refractory. However, the epidemiologic characteristics of MAT had not previously been well defined. Among 29 case reports and/or series identified using ‘MAT in pediatric patients’ as a search term, only 15 studies were available in full text and contained individualized information. These studies are summarized in Table 4. Bisset et al.12) reported 5 patients older than 1 year, 3 of whom were diagnosed with structural heart disease and one with cerebral gigantism. Bradley et al.1) and Dodo et al.7) reported 1 and 2 non-infant patients, respectively, with MAT and structural heart disease. In contrast to previous studies, non-infant patients in our study tended to have other arrhythmias (2 with AVRT and 2 with RyR2 (+) CPVT) with structurally normal hearts (except for a secundum atrial septal defect in patient 28). There are only 17 non-infant patients with MAT described in the prior literature and in our study; regardless, most on these patients (n=15) had MAT with other comorbidities, rather than MAT alone.1)12)14)15) Infant patients, in contrast, had a significantly lower rate of other comorbidities (52% of infant patients had underlying comorbidities, p<0.01). MAT in infant and non-infant patients may have distinct pathophysiologies, and therefore different clinical courses and prognoses. Furthermore, there are varying prognoses between studies, which might have been influenced by age differences and underlying comorbidities.

Considering these findings, we compared the clinical outcome of MAT between patient groups according to their age and presence of major comorbidities. The overall mortality (12%) in our study is similar to that of other studies (0%–22%).1)4)5)7)8)12) All 3 cases of cardiopulmonary mortality were associated with structural heart disease (with or without pulmonary disease). This finding corresponds to those of previous studies; 9 of 12 mortality cases from previous studies also had associated structural heart disease. When considering the prior findings with our results, the mortality rate was significantly higher in patients with structural heart disease than it was in patients without structural heart disease (21% vs. 5%, p=0.01).

Our patients demonstrated a control rate (87%) that is similar to that of patients from previous studies (85%). Individual disease categories failed to show statistically significant differences in the control rate. However, the idiopathic infants had a 100% control rate. When the previous studies were counted together, the idiopathic infants had a significantly higher control rate than did the remaining patients (95% vs. 77%, p=0.01).

Several previous studies have reported an association between the RyR2 mutation and AF and ectopic atrial tachycardia. Di Pino et al.16) reported the case of a novel RyR2 mutation in a 2-year-old patient with AF, AFL, and ectopic atrial tachycardia. Furthermore, Broendberg et al.17) reported 5 cases of MAT in 51 patients with known RyR2 mutations; the group suggested that the young patients with severe MAT may be RYR2 mutation carriers. However, to the best of our knowledge, we have reported the first 2 cases of known MAT confirmed to have RyR2 mutations with life-threatening events. In both patients, the first manifestation of this mutation was with MAT, rather than VT. Therefore, non-infant MAT patients with structurally normal hearts might need thorough evaluations (including exercise testing and genetic evaluation of arrhythmogenesis), and long follow-up.

Lin et al.18) analyzed the cardiovascular abnormalities of RASopathy in their 61 patients with Costello syndrome. In their study population, 15 (25%) of 61 patients with Costello syndrome had MAT or ectopic atrial tachycardia. In addition, 13 (87%) patients had a HRAS mutation in p.Gly12Ser, which was also the most frequent mutation in the entire study population.18) Although less frequent, the mutation foci in our 2 patients with gene-confirmed Costello syndrome (p.Gly12Ala in patient 12 and p.Gly12Cys in patient 17) were also reported to be associated with atrial tachycardia.

According to the review by Lin et al.,18) RASopathies other than Costello syndrome, such as cardiofaciocutaneous syndrome, Noonan syndrome, and neurofibromatosis, were infrequently associated with supraventricular tachycardia. In contrast, Noonan syndrome with the RAF1 mutation is associated with supraventricular tachycardia. There had been 4 cases of Noonan syndrome in the MAT reports—2 from our study and 2 from the reports of Bradley et al.1) and Dodo et al.7) The specific type of gene mutation was only assessed in one patient (patient14), who also had the RAF1 gene mutation. Therefore, the specific type of gene mutation in RASopathy with MAT in our study corresponds well to the report of Lin et al.18) However, since MAT and/or atrial tachycardia have recently been reported in neurofibromatosis or Noonan syndrome with mutations other than RAF1, MAT may not be a differential feature distinguishing Costello syndrome from other RASopathies.19)20)

Two seemingly sporadic cases of rhabdomyosarcoma—patient 31 and a patient from the report of Liberthson and Colan5)—further suggests an association between the RAS gene and MAT.

The prevalence of pulmonary disease in MAT has been well established in adult MAT patients. Right atrial enlargement, hypercapnia, hypoxia, or adrenergic stimulation have been suggested as possible mechanisms by which pulmonary disease causes MAT.3) The majority of MAT reports in pediatric patients have also reported >20% of coexisting pulmonary disease, even though the rate is much higher in adults (up to 60%). However, alveolar growth and decline in pulmonary vascular resistance continues for at least 2 years; therefore, infant patients might be susceptible to MAT even in the absence of overt pulmonary disease. Furthermore, the immaturity of the atrium may also contribute to the infant-predominant age distribution of MAT. Pickoff et al.21) have reported that there is atrial vulnerability in neonatal puppies, showing an atrial refractory period that increases as the dog ages. Therefore, immaturity of both the lungs and the heart might explain the infant-predominant age distribution of MAT and its prevalence in idiopathic infant cases. The benign clinical course in the idiopathic infant group might also be explained by improvement of the immaturity of the lungs and heart as patients’ age.

Despite its valuable findings, this study has several limitations. Since this was an observational study, it has an intrinsic vulnerability to confounding. For example, the treatment strategy changed over different periods: β-blockers and digoxin were the 2 mainstay antiarrhythmic drugs in the 1990s and 2000s, while amiodarone was frequently combined with others in the 2010s. The follow-up interval also varied between patients, which could have influenced the duration of the arrhythmia and/or treatment. The statistical analyses were also limited given the study's small sample size. Given this small sample size, we were unable to identify a significant risk factor for unfavorable outcomes on the multivariate analysis.

Pediatric MAT mostly affects infants, who tend to have a favorable prognosis, especially in the absence of other comorbidities. However, MAT in patients with other underlying comorbidities may have a varying clinical course, ranging from mild and self-limiting to fatal. The prognosis of MAT in these patients may follow the prognosis of their underlying comorbidities. Therefore, it is worthwhile to evaluate for potential underlying causes in order to predict the prognosis of MAT, and potentially correct some of its reversible factors. Non-infant patients with MAT should undergo a thorough survey of underlying conditions, and prolonged follow-up to monitor for the development of other arrhythmias (such as CPVT).