Beyond First-Degree Heart Block in the Diagnosis of Acute Rheumatic Fever

Cardiology in the Young Beyond first-degree heart block in the diagnosis

cambridge.org/cty of acute rheumatic fever

Joshua Agnew1, Nigel Wilson2, Jonathan Skinner2 and Ross Nicholson3

1 2 Original Article Bay of Plenty District Health Board, Tauranga, New Zealand; Green Lane Paediatric and Congenital Cardiac Services, Starship Children’s Hospital, Auckland, New Zealand and 3KidzFirst Children’s Hospital, South Auckland, Cite this article: Agnew J, Wilson N, Skinner J, New Zealand and Nicholson R (2019) Beyond first-degree heart block in the diagnosis of acute rheumatic Abstract fever. Cardiology in the Young 29:744–748. doi: 10.1017/S104795111900026X Objectives: First-degree heart block is a minor manifestation of acute rheumatic fever. Second and third degree heart block and junctional rhythms occur less commonly. We report patients Received: 27 September 2018 presenting with these latter three electrocardiographic abnormalities and investigate their Revised: 23 December 2018 Accepted: 21 January 2019 diagnostic utility. Design: Patients admitted to our centre meeting the 2014 New Zealand Rheumatic Fever Guideline Diagnostic Criteria for rheumatic fever over a 5-year period from Key words: January 2010 to December 2014 were identified. Clinical, haematologic, electrocardiographic, Rheumatic fever; junctional rhythm; and echocardiographic records were reviewed. Electrocardiograms (ECG) were considered heart block abnormal if there was second- or third-degree atrioventricular block or junctional rhythms. Author for correspondence: Comparative data from patients with advanced conduction abnormalities without a diagnosis Dr Ross Nicholson, Consultant Paediatrician, of rheumatic fever during the same time period were reviewed. Results: A total of 201 patients KidzFirst Children’s Hospital, Counties met inclusion criteria for rheumatic fever. Of these, 17 (8.5%) had transient abnormalities of Manukau Health, Private Bag 93311, Otahuhu, atrioventricular conduction, 5 (2.5%) with second or third-degree atrioventricular block, and Auckland 1640, New Zealand. Tel: þ64 9 276 0044 ext 57003.; 12 (6%) junctional rhythms. The remaining 173 (86%) patients had evidence of rheumatic Fax: þ6492760192; E-mail: valvulitis at presentation. Only one patient without rheumatic fever was found to have [email protected] advanced conduction abnormalities over the study period, from a total of 3702 ECG. Conclusions: This large contemporary cohort of acute rheumatic fever shows that 8.5% of cases had either advanced atrioventricular block or junctional rhythms both highly suggestive of the diagnosis in our population.

Rheumatic fever remains the most common form of heart disease in the world due to its preva- lence in the developing world and subsets of the population in the developed world, such as the indigenous Māori and Pacific Islanders in New Zealand. The diagnosis of acute rheumatic fever remains a clinical diagnosis based on major and minor criteria.1–4 Various conduction and rhythm disturbances have been associated with rheumatic fever. Abnormalities in atrioventricular conduction including first-degree heart block and more advanced degrees of atrioventricular block, junctional acceleration, supraventricular and ventricular tachycardia, bundle branch block, ectopic atrial and ventricular complexes, prolongation of the QT interval, and torsade de pointe have all been reported.5–14 We present a 5-year retrospective review of confirmed acute rheumatic fever cases from our institution, focusing on presentations with advanced atroventricular block and junctional rhythms.

Methods Patients under 15 years of age, satisfying the New Zealand Heart Foundation 2014 definition of rheumatic fever,1 presenting to KidzFirst Children’s Hospital in South Auckland between January 2010 and December 2014 were retrospectively identified from a regional Rheumatic Fever Registry. Patients were excluded if they had incomplete data sets or if they were registered for chronic or indolent rheumatic heart disease. Patient age, gender, NZ Heart Foundation 2014 major criteria, initial erythrocyte sedimen- tation rate (ESR), C-reactive protein, and streptococcal titres (anti-streptolysin and anti-DNAse B antibodies) were collected from clinical and laboratory records. Electrocardiography data were analysed as described below.

Electrocardiographic analysis Electrocardiograms (ECG) from the first week after presentation were analysed. The PR interval (lead II) and QTc (longest of lead II or V5) using Bazett’s formula15 were calculated. © Cambridge University Press 2019. The presenting rhythm was categorised into the following three groups:

1. Advanced atrioventricular conduction abnormality (AAVCA) was used to describe our group of interest. This term included patients with either acquired second- or third-degree atrioventricular block or a junctional rhythm or any type.

Downloaded from https://www.cambridge.org/core. Research4Life, on 17 Aug 2019 at 00:47:47, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S104795111900026X Cardiology in the Young 745

Table 1. Missing data. Table 2. Major diagnostic criteria.

Variable Normal PR Prolonged PR AAVCA Major criteria No. patients Percentage Age 0 0 0 Carditis (clinical þ subclinical) 173 86 ESR 0 0 0 Polyarthritis 52 26 QTc 1 1 1 Monoarthritis 45 22 CRP 6 3 1 Chorea 23 11 ASOT 1 0 0 Erythema marginatum 3 1 Anti-DNAseB 1 0 0 Subcutaneous nodules 0 0 Gender 0 0 0 Clinical carditis 0 0 0 Table 3. Comparison of AAVCA versus Prolonged versus Normal PR groups: Carditis on 1st echo 0 0 0 continuous variables. Carditis on last echo 0 0 0 Normal PR Prolonged PR AAVCA P* Major criteria 0 0 0 N = 91 N = 93 N = 17 Median (IQR) Median (IQR) Median (IQR) • Advanced atrioventricular block was defined as a second- or Age (years) 10 (8–12) 11 (9–13) 11 (9–12) 0.025 15 third-degree heart block. ESR 70 (29–103) 97 (68–117) 108 (90–116) <0.001 • Junctional rhythm was defined where in Lead II there is an absent – – – p wave or a p wave that follows a normal QRS, where the ven- CRP 56 (11 112) 81 (35 139) 123 (101 153) 0.0025 tricular rate was the same as or in excess of the atrial rate.15 ASOT 543 (268–894) 611 (374–782) 533 (384–660) 0.72 • Ventriculoatrial dissociation was recorded when in the presence AntiDNAse B 665 (354–900) 600 (361–900) 880 (300–1440) 0.59 of a junctional rhythm, the atria and ventricle operated inde- QTc 440 (425–455) 442 (430–459) 446 (435–51) 0.66 pendently for more than three beats.15 2. Prolonged PR: This included all cases with first-degree atrio- *P is p-value using the Kruskal–Wallis rank sum test to compare the three groups p < 0.05 is ventricular block. considered significant 3. Normal ECG: This includes all cases in sinus rhythm with normal PR interval. incomplete data sets (no electrocardiographic data þ/− a lack of adequate notes). The remaining 201 cases had an average age at Comparative data set presentation of 10.3 years (range 5–14 years), with 59% of patients Local hospital data over the study period were reviewed to identify being male. Missing data on included patents are recorded in patients presenting with AAVCAs who did not satisfy a diagnosis Table 1. The distribution of major criteria at presentation can of rheumatic fever using the International Classification of Disease be found in Table 2. In total, 173 (86%) patients had clinical or sub- Codes version 10. The total number of electrocardiographs per- clinical carditis at presentation: 122 (61%) patients presented with formed during the study period was also determined. clinical carditis and a further 51 (25%) cases had subclinical cardi- To assess the utility of an electrocardiographic diagnosis of tis. A total of 172 (86%) patients had two or more major criteria. AAVCA in diagnosing acute rheumatic fever, we reviewed all We identified only one patient meeting our diagnosis of advanced admissions to our centre during the study period using a broad conduction abnormality who did not have a diagnosis of rheumatic International Classification of Disease-10 Code search of conduc- fever. This patient was a 12-year-old child from Zimbabwe, who tion abnormalities. presented with an idiopathic complete heart block at age 12 years. His echocardiogram was normal and he had no history consistent with Acute Rheumatic Fever. Statistical analysis The Kruskal–Wallis rank sum test of non-parametric data was Rhythm analysis used for continuous variables and the Fisher t test was used for categorical variables. A total of 17 (8.5%) patients with acute rheumatic fever had All p-values were two-tailed and a p-value of <0.05 was consid- AAVCA on ECG with the confirmation of diagnosis by the study ered significant. paediatric electrophysiologist (JS). In total, 5 (2.5%) patients presented with advanced atrioventricular block; 3 patients with second-degree heart block (Mobitz type 1) and 2 with complete Results heart block. Junctional rhythm occurred in 12 (6%) patients. A There were a total of 16,135 admissions to our hospital over the total of four patients with junctional acceleration also had ventri- study period, with a total of 3702 electrocardiographs performed culoatrial dissociation. Three ECGs that were initially included in during this time on children under 15 years of age. A total of the junctional acceleration group were later excluded as they were 224 new cases of rheumatic fever were identified. Of these, 23 felt to demonstrate first-degree heart block with an extremely pro- patients were excluded: 9 patients were registered as chronic or longed PR interval. indolent rheumatic heart disease, 1 patient had an alternative diag- AAVCA presented with significantly higher ESR and C-reactive nosis of septic arthritis, and the remaining 13 patients had protein and significantly less carditis (Table 3 and Table 4).

Table 4. Comparison of AAVCA versus prolonged versus normal PR: discrete variables.

Normal Prolonged Variable PR PR AAVCA P* N = 91 N = 93 N = 17 N (%) N (%) N (%) Gender F 38 (42) 36 (39) 9 (53) 0.54 Figure 1. Extreme PR prolongation,P waves merging with T waves M 53 (58) 57 (61) 8 (47) Clinical N 30 (33) 37 (40) 12 (71) 0.014 carditis Y 61 (67) 56 (60) 5 (29) Carditis N 15 (16) 15 (16) 8 (47) 0.008 (1st echo) Y 76 (84) 78 (84) 9 (53)

Carditis N 13 (14) 11 (12) 6 (35) 0.04 Figure 2. Accelerated Junctional rhythm at 71 bpm (FU echo) Y 78 (86) 82 (88) 11 (65) Major Nil 35 (38) 38 (41) 6 (35) 0.27** criteria Monoarthritis 18 (20) 23 (25) 4 (24) Polyarthritis 19 (21) 23 (25) 7 (41) Chorea 16 (18) 6 (6) 0 (0)

Erythema 0 (0) 1 (1) 0 (0) Figure 3. Junctional rhythm at 55 bpm marginatum

*P = p-value derived from the Fisher t test comparing the two groups **p-value relates to all non-cardiac major criteria combined Illustrative case p < 0.05 is considered significant Mesina, a previously well Samoan girl, aged 12 years presented acutely from her GP with a history of fever, lethargy, and chest pain. A throat swab taken 2 days prior isolated a moderate growth AAVCAs were a transient phenomenon in all patients, with a of group A streptococcus. normal electrocardiograph documented either during the initial On admission, we obtained a history of a painful left ankle, but admission or in early follow-up. AAVCA took a variable time to there was no objective evidence of arthritis. She was afebrile with resolve, usually days rather than weeks, and usually to sinus rhythm normal findings on cardiac, respiratory, and skin examination. with first-degree block. The exact time was not able to be deter- ESR was elevated at 90 mm/hour. Both anti-streptolysin and mined accurately in this retrospective review. No patients with anti-DNase B antibodies were moderately elevated. an AAVCA developed cardiorespiratory instability or required Her first ECG taken 1 hour after admission (Fig 1) at a heart rate treatment for their rhythm disturbance. We did not use corticoste- of 89 bpm shows extreme PR prolongation with PR greater than roids Intravenous Immunoglobulin or aspirin for the treatment of 300 msecond. The P-waves are seen to merge into the T-waves. rheumatic valvulitis or rhythm disturbance. The second ECG (Fig 2) obtained 45 hours after presenting The PR interval was prolonged for age and heart rate (first- demonstrates an accelerated junctional rhythm at a rate of degree heart block) in 93 (46%) and normal in 91 (45%) of patients 71 bpm. The P-waves are seen in the ST segment following the (Table 4). QRS complex. On day 6 after admission, the third ECG (Fig 3) shows a junctional rhythm at 55 bpm. This was initially thought to represent complete heart block. By day 8, her ECG had returned Comparative data set to sinus rhythm at 57 bpm with a PR interval of 190 msecond. (All From the total of 16,135 admissions, 92 admissions with a primary three ECGs are recordings from Lead II 10 mm/mV at 25 mm/ or secondary diagnosis of arrhythmia were identified from our seconds.) International classification of disease search. These were reviewed, and patients with advanced atrioventricular block and junctional Discussion rhythm were identified. Patients with congenital, rather than acquired atrioventricular block, were then excluded. This left one Rheumatic fever remains an important cause of cardiac morbidity patient with an acquired idiopathic complete heart block, who had for our New Zealand population, as well as other high-risk popu- a normal echocardiogram and did not meet criteria for acute rheu- lations worldwide.1–4 There remains limited data on both conduc- matic fever. tion and rhythm disturbances in patients with acute rheumatic During the study period, a total of 3702 electrocardiographs fever. We present a large contemporary cohort of acute rheumatic were performed in children under the age of 15 years, presenting fever specifically addressing presentation with advanced atrio- acutely to our children’s emergency department. ventricular block or junctional rhythms.

Prolongation of the PR interval is the most common finding generalisable, given that complete heart block is associated in rheumatic fever, occurring in about 50%.5–9,16,17 Despite its with other inflammatory conditions such as diphtheria, lyme dis- common occurrence, PR prolongation is not a specific sign of car- ease, and myocarditis20–22 which occur more commonly in other ditis; occurring in up to 2% of the normal population,18 in patients populations. following group A streptococcal throat infection who do not We chose to use all ECGs performed at our centre during the develop rheumatic fever,17 and with other febrile conditions. study period as a denominator for our analysis, but do not have Hence, it remains as a minor criterion in the diagnosis of acute data on the indications for why this test was ordered. Due to rheumatic fever. We thus choose to exclude first-degree atrio- the endemic nature of rheumatic fever in our population and strict ventricular block. local protocols, we assume that all patients who presented with Conduction abnormalities beyond first-degree heart block have symptoms that could be consistent with acute rheumatic fever been reported in case reports and small series. We choose to inves- would have had an ECG and been included in this analysis. tigate patients presenting with advanced degrees of AV block or However, as less than 23% of patients admitted had an ECG during junctional rhythms on electrocardiograph both of which have been the study period, this limited our ability to detect asymptomatic associated with acute rheumatic fever in small series. We identified patients admitted who may have shown an advanced conduction 17 (8.5%) patients with these transient electrocardiographic abnormality on ECG. abnormalities. Second- or third-degree atrioventricular block and Both types of advanced conduction abnormality have previ- junctional rhythm appear to be a temporary and self-limiting ously been described with minimal other evidence of rheumatic occurrence in acute rheumatic fever. Although reports of syncopal carditis. Cristal demonstrated that advanced atrioventricular block events and temporary pacing have been described complicating is not consistently associated with valvulitis.9 Karacan showed that complete heart block in rheumatic fever, most reports indicate junctional rhythms are not always associated with clinical carditis.5 that the majority of patients will resolve without intervention Our study confirms their observations. Although AAVCAs are being needed.19 Our data demonstrate all patients presenting manifestations of cardiac involvement in rheumatic fever, they with AAVCAs returning to sinus rhythm within days or weeks of appear to occur earlier in the course of the illness, prior to the diagnosis. development of valvulitis, given the significantly higher ESR and About 2.5% of our cohort demonstrated advanced atrio- C Reactive Protein we observed. ventricular block; three patients with second-degree heart block (Mobitz Type 1) and two with complete heart block. This is com- parable to other reported incidence ranging from 1.5 to 5.5%.5,7,8,10 Limitations to this study About 6% of patients presented with a junctional rhythm. This Despite collecting 5 years of data from our institution, we still only was defined by Karacan et al as narrow complex QRS rhythm, present small numbers of patients with both advanced conduction where the ventricular rate was the same as or in excess of the atrial abnormalities due to the low incidence of these rhythms in rheu- 5 rate. Junctional rhythm has an incidence in acute rheumatic matic fever. Whilst there are limitations in identifying patients – – fever of 5 15%, with the majority of patients (50 73%) showing retrospectively, especially with what may be a secondary diagnosis, 5,8 ventriculoatrial dissociation on electrocardiograph. We chose both advanced atrioventricular block and junctional rhythms are to include all patients with any form of junctional rhythm, who very uncommon in our population with the exception of children ’ met Karacan s definition, irrespective of rate as we surmised that subsequently diagnosed with acute rheumatic fever. Our ability to any junctional rhythm during an acute presentation of rheumatic collect a comparison population to assess their utility was also fever constituted an abnormality. limited by our retrospective design. Prospective multi-institutional Diagnosis of junctional arrhythmia proved more challenging reviews or larger single institutional reviews are needed to ascertain than the diagnosis of atrioventricular block, which was usually sensitivity specificity and diagnostic accuracy of these abnormal- self-evident. Those presenting with junctional acceleration on ECG ities of conduction. and little or no ventriculoatrial dissociation were often difficult to diagnose and required careful examination of the ECG. In our series, three ECGs that were initially thought to be junctional Conclusion acceleration were later excluded upon review by the study of electrophysiologist, as they represented extreme first-degree atrio- We present a large contemporary cohort of patients presenting ventricular block (see Fig. 1), resulting in P-waves “buried” in the with acute rheumatic fever in whom junctional rhythms and T-waves. advanced atrioventricular block occurring in 8.5% of cases. We found that an electrocardiograph showing advanced conduction abnormality, beyond first-degree atrioventricular block, is highly Utility of AAVCA in the diagnosis of ARF suggestive of acute rheumatic fever. Clinicians should therefore Ceviz et al compared patients with junctional acceleration to both a consider a diagnosis of acute rheumatic fever in children, from group of patients with confirmed group A streptococcal infection endemic regions, with an electrocardiograph showing advanced who did not develop acute rheumatic fever and a group of patients conduction abnormalities when echocardiography is negative. presenting with arthritis from rheumatic causes other than acute Further prospective data were warranted to establish if advanced rheumatic fever. They found no patients with junctional accelera- conduction defects could be used as a major criteria in the diagno- tion in either of their comparison groups, suggesting this was likely sis of acute rheumatic fever. a novel rhythm in acute rheumatic fever.6 We could not find any What is already known on this topic? data regarding advanced atrioventricular block and the diagnosis • Advanced conduction abnormalities occur in up to 15% of of acute rheumatic fever. rheumatic fever. Acute rheumatic fever is endemic in our population, with one • Conduction abnormalities can occur without evidence of new diagnosis every 10 days on average. Our findings may not be rheumatic carditis.

What this study adds 6. Ceviz N, Celik V, Olgun H, et al. Accelerated junctional rhythm in children • An advanced conduction abnormality is highly suggestive of with acute rheumatic fever: is it specific to the disease? Cardiol Young 2014; acute rheumatic fever in our population. 24: 464–468. • Advanced conduction abnormalities are commoner than eryth- 7. Balli S, Oflaz MB, Kibar AE, et al. Rhythm and conduction analysis of – ema marginatum and subcutaneous nodules established major patients with acute rheumatic fever. Pediatr Cardiol 2013; 34: 383 389. diagnostic criteria. 8. Clarke M, Keith JD. Atrioventricular conduction in acute rheumatic fever. Br Heart J 1972; 34: 472–479. 9. Cristal N, Stern J, Gueron M. Atrioventricular dissociation in acute rheu- Author ORCIDs. William Nicholson 0000-0003-4190-0876 matic fever. Br Heart J 1971; 33: 12–15. 10. Zalzstein E, Maor R, Zucker N, et al. Advanced atrioventricular conduction Acknowledgement. We thank Charlene Nell for preparing the manuscript block in acute rheumatic fever. Cardiol Young 2003; 13: 506–508. and Pernille Christensen for statistical analysis. 11. Freed MS, Sacks P, Ellman MH. Ventricular tachycardia in acute rheumatic fever. Arch Intern Med 1985; 145: 1904–1905. Financial support. This research received no specific grant from any funding 12. Tan KS, Lau YS, Teo WS. T wave alternans and acute rheumatic myocar- agency, commercial, or not-for-profit sectors. ditis: a case report. Ann Acad Med Singapore 1999; 28: 455–458. 13. Liberman L, Hordof AJ, Alfayyadh M, et al. Torsade de pointes in a child Conflicts of interest. None. with acute rheumatic fever. J Pediatr 2001; 138: 280–282. 14. Yahalom M, Jerushalmi J, Roguin N. Adult acute rheumatic fever: a rare Contribution statement. RN – Substantial contributions to the conception or case presenting with left bundle branch block. Pacing Clin Electrophysiol design of the work, or the acquisition, analysis, or interpretation of data. RN, JA, 1990; 13: 123–127. NW, and JS – Drafting the work or revising it critically for important intellec- 15. Park MK. Park’s Pediatric Cardiology for Practitioners. In: Park MK (ed), tual content. RN – Final approval of the version published. RN – Agreement to 6th edn. Elsevier Saunders, Philadelphia, PA, 2014. be accountable for all aspects of the work in ensuring that questions related to 16. Parkinson J, Gosse AH, Gunson EB. The heart and its rhythm in acute the accuracy or integrity of any part of the work are appropriately investigated rheumatism. Q J Med 1920; 13: 363–379. and resolved. 17. Tani LY. Rheumatic fever and rheumatic heart disease. In: Allen HD, Driscoll DJ, Shaddy RE, Feltes TF (eds). Moss and Adams Heart Disease in References Infants, Children, and Adolescents, Including the Fetus and Young Adult, 7th edn. Lippincott Williams & Wilkins, Philadelphia, PA, 2008: 1256–1280. 1. New Zealand Heart Foundation. New Zealand Guidelines for Rheu- 18. Ziegler RF. Electrocardiographic Studies in Normal Infants and Children. matic Fever: Diagnosis, Management and Secondary Prevention of Charles C Thomas Publishers, Springfield, IL, 1951: 3–9. Acute Rheumatic Fever and Rheumatic Heart Disease: 2014 Update. 19. Carano N, Bo I, Tchana B, et al. Adams-Stokes attack as the first symptom Auckland, 2014. of acute rheumatic fever: report of an adolescent case and review of the lit- 2. Jaine R, Baker M, Venugopal K. Epidemiology of acute rheumatic fever in erature. Ital J Pediatr 2012; 38: 61. New Zealand 1996–2005. J Paediatr Child Health 2008; 44: 564–571. 20. Phornphutkul C, Damrongsak D, Silpisornkosol S. Steroid therapy in car- 3. Wilson N. Rheumatic heart disease in indigenous populations – diac conduction disturbances in children with diptheria. Mod Med Asia New Zealand experience. Heart Lung Circ 2010; 19: 282–288. 1978; 14: 38, 43–44. 4. Neutze JM. Rheumatic fever and rheumatic heart disease in the western 21. Seferovic PM, Ristic AD, Maksimovic R, et al. Cardiac arrhythmias and Pacific region. N Z Med J 1988; 101: 404–406. conduction disturbances in autoimmune rheumatic diseases. Rheumatology 5. Karacan M, Isikay S, Olgun H, et al. Asymptomatic rhythm and conduction (Oxford) 2006; 45: iv39–iv42. abnormalities in children with acute rheumatic fever: 24-hour electrocar- 22. Sigal LH. Early disseminated Lyme disease: cardiac manifestations. Am J diography study. Cardiol Young 2010; 20: 620–630. Med 1995; 98: 25S–28S; discussion 28S–29S.