Introduction Electrical storm (ES) is usually defined as the occurrence of three or more episodes of sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) within 24 hours requiring either anti-tachycardia pacing (ATP) or cardioversion/defibrillation. The occurrence of ventricular tachyarrhythmias appears to follow circadian, daily, and seasonal distributions. Unfortunately, the temporal patterns described so far vary widely in patients with ischemic and non-ischemic heart disease, hypertrophic cardiomyopathy, and Brugada syndrome. Although some studies have shown potential candidates as ES predictors, there is currently no evidence regarding the association of this condition with environmental and external factors such as temperature, time of the day and time of the year. The aim of the present study is to describe the incidence of ES over time and test the potential association between ES incidence and time of the year, time of the week, time of the day and short-term temperature variations. Methods Study selection The Circannual pattern and TEMPerature-related incidence of Electrical Storm (TEMPEST) study was conducted following the current guidelines, and registered in the PROSPERO International prospective register of systematic reviews of the University of York, United Kingdom (registration # CRD42013003744). Two medical databases (MEDLINE and Embase) were systematically searched in order to include all available papers. MEDLINE was searched using the following query: “electrical storm” [Mesh] OR “arrhythmic storm” [Mesh] OR “recurrent ventricular arrhythmias” [Mesh] OR “ventricular tachycardia clusters” [Mesh] OR “electrical instability” [Mesh]. ISI Web of Science was searched using the following query: title contains “electrical storm” OR “arrhythmic storm” OR “recurrent ventricular arrhythmia” OR “ventricular tachycardia clusters” OR “electrical instability”. The search was performed up until November 1st, 2015 and was limited to English-language literature. Two authors independently screened all the records reviewed full-texts articles and determined their eligibility. In order to be selected, a study has to meet all the following criteria: a) diagnosis of ES as the occurrence of three or more episodes of VT/VF within 24 hours (each episode at least 5 minutes apart) or VT for more than 12 hours; b) theabsenceofacutecoronarysyndromeasthearrhythmictrigger; c) tenormorepatientsincluded; d) selection of the most recent publication when the same group reported on the same patients in separate publications. Inter-observer concordance was optimal during the whole selection process (k = 0.97). At the end of the selection process, 35 papers were included in the data collection process. Data collection Corresponding authors of aforementioned papers were contacted by e-mail and asked to participate. Of those, 19 agreed and the study protocol was provided to them. All authors were asked for additional information on a patient-level basis regarding clinical characteristics, lab exams, ES characteristics, therapies delivered by the ICD, and pharmacological and non-pharmacological management. Nine authors replied back with a complete dataset. In order to ensure optimal data quality, datasets with no data available regarding ES time and geographical location of each patient were excluded, as were datasets collecting non-consecutive patients. At the end of the selection and collection processes, five centres had the datasets from their papers (9–13) merged into the present registry. Temperature data collection Temperature data were retrieved from the “historical weather” section of Weather Underground (www.wunderground.com), the largest weather database available online. For each patient, average, maximum, and minimum temperatures during the event day were collected, as were average mean, maximum and minimum temperatures during the week and the month of the ES, and the week and the month before. If the weather almanac was not available for the exact geographical location at that time, the nearest forecast station with available data was used instead. Weekly and monthly temperature ranges were defined as the average of maximum temperatures minus the average of minimum temperatures during a week or a month, respectively. The same database was used in order to retrieve sunrise and sunset hours. If ES started after sunrise and before sunset it was considered as happened during daytime; if it started after sunset and before sunrise it was considered as happened during night-time. Results Population characteristics Total population included 246 patients presenting with ES (221 males, age 66±9 years). While mean serum potassium levels were within the normal ranges (4.3±0.6 mEq/l), 23 patients (9.3%) had hypokalemia and 6 patients (2.4%) had hyperkalemia on presentation. Mean creatinine levels were 1.3±0.5 mg/dl and 22 patients (8.9%) were admitted with creatinine serum levels ≥2 mg/dl. Each ES episode was made up by a median of 7 VT/VF episodes (1st-3rd IQR 4-16). On average, each patient experienced a median of 8 anti-tachycardia pacing (ATP) therapies (1st-3rd IQR 3-10) and 5 shocks (1st-3rd IQR 3-6) before ES termination. Electrical storm incidence over time ES incidence over time followed a non-homogeneous distribution in all participating centres, with low-incidence periods alternating with clusters of many ES admitted within a few days from one another. Regarding ES distribution over the day, a significantly higher number of ES happened during daytime hours rather than during night-time hours (58.7% vs. 41.3%; p<0.001). Plotting ES incidence over time shows that 29% of all ES started from 8 to 10 a.m., with a significant incidence reduction during the following hours (p<0.05). Weekly incidence of ES was similar from Mondays to Fridays, while both Saturdays and Sundays showed a lower incidence of ES when compared to other days (all p<0.05). ES distribution over the year was homogeneous with no month significantly associated with a higher prevalence of ES. Similar incidences of ES have been found between winter, spring, summer and fall. Temperature and electrical storm No linear association was found between ES incidence and average, minimum and maximum daily, weekly or monthly temperatures. However, ES incidence was significantly associated with an increase in monthly temperature range when compared to the month before, with 68.9% of ES happening after an increase in monthly temperature range (p=0.003). Discussion Temperature-related variations From our data, ES presentation is not homogenous over time but seems to recognize a clustered or “hot-spot” pattern, alternating low-incidence with “high-intensity” periods, when many patients were admitted with ES within a few days. This pattern, while not superimposable for each center due to the different timeframes and geographical coordinates considered, was indeed quite evident for all of the five participating hospitals. The pattern described was not related to a specific circannual pattern neither to absolute temperatures, as centers varied widely in terms of latitude, longitude, and mean monthly averages (these latter ranging from 10°C in Warsaw to 20° C in Valencia and 21°C in Los Angeles). On the other hand, we found a significant association between ES incidence and variations in the temperature range, with nearly 70% of all ES happening after an increase in temperature ranges of their specific geographical locations. The mechanisms underlying the connection between ES and environmental triggers are unknown, but some hypothesis can be made. In favor of a direct causal relationship, we have many available data underlying an increased risk of ventricular arrhythmias in patients undergoing fast body temperature cooling and rewarming, as commonly experienced in therapeutic hypothermia. Animal models demonstrated that inducing hypothermia amplifies dispersion of repolarization and increases myocardial vulnerability to VF, thus potentially eliciting multiple arrhythmic events in predisposed patients. Of course, our patients experienced a variation in external rather than body temperature, which was not as big in magnitude neither so concentrated in time. Therefore, other hypotheses must be sought. Recently, it has been suggested that the increased mortality and morbidity caused by hot and cold temperatures could not be strictly related to hypothermia/hyperthermia but by other causes triggered by the human organism’s attempts to adapt to the external temperature. In this regard, increased mortality and hospitalization rates for cardiovascular diseases has been reported after both cold spells and heat waves in many different geographical locations. Hospitalization rates from a large population- based study showed a reverse J-shaped association between hospitalization for blood pressure, diabetes, and, with a minor extent, arrhythmias over a lag of 21 days, with higher risks at both temperatures extremes. Thermoregulation-mediated vasoreactivity, sympathetic nervous system activation and sodium and volume retention through the renin-angiotensin-aldosterone system could all be involved as underlying pathophysiological mechanisms. On a note, the changes in atmospheric pressure related to weather and temperature variations are also associated with the concentration of particulate and gaseous air pollutants, both of whose have been related to an increased risk of arrhythmia-related hospitalization or mortality. Weekly and daily variations The higher incidence of ES during working days in our population was not totally unexpected, as was already underlined on unclustered VT/VF. This behavior potentially highlights a predominant role of autonomic regulation in triggering the ES, similarly to the well-known circadian pattern in acute myocardial infarction. This hypothesis is supported by a recent sub-analysis of the SCD-HeFT trial, which demonstrated a significant weekly variation with an increase in VT/VF incidence on Mondays only in patients not treated with β-blockers. Available analyses of VT/VF daily patterns report a higher incidence of ICD therapies in the early morning both in ischemic and non-ischemic cardiomyopathy, while patients with hypertrophic cardiomyopathy and Brugada syndrome experience ICD therapy more often during the afternoon and late-night, respectively. In our population, mainly composed of ischemic (63.0%) and idiopathic dilated cardiomyopathy (20.7%), ES starting hour distribution is similar to the circadian distribution of VT/VF described by Anand et al., although the late rise described during the afternoon hours is much less evident in our population. So, while it is feasible to hypothesize that the first VT/VF triggering the ES follows the same circadian pattern already described, some other factors must be held responsible for the coalescence of many VT/VF into an ES. Given the high prevalence of structural heart disease and severe left ventricular dysfunction in our population, it is feasible to identify heart failure (HF) as a major potential actor. Available data on a 5-year follow-up shows that ES patients share some striking similarities with patients admitted for HF exacerbation, and ES could be seen as a warning sign of impending pump failure rather than an independent event. In this setting, the many dysregulated biochemical pathways typical of HF could help in creating a substrate for arrhythmia recurrence, with ES as an epiphenomenon. Potentially overlapping as an arrhythmic risk factor in HF patients is sleep-disordered breathing (SDB). Recent observational data have shown that SDB is an independent predictor of appropriate therapy in ICD patients, and could explain, at least in part, the increased incidence of ES seen in the morning. Limitations The present paper shares all the limitation intrinsic to a retrospective, patient-level pooled analysis. Moreover, average temperatures were extracted from a web-based forecasting service, thus limiting the real assessment of external temperature to a wide geographical area. Regarding collected data, chronic pharmacological therapy was not routinely collected, as the papers selected focused on the acute treatment of ES. Therefore, is not cautious to extend the present findings to specific subpopulations (i.e. patients previously treated with β-blockers or other anti-arrhythmic drugs). Moreover, we were not able to provide data regarding ICD programming, which has been showed as an important risk factor for ES and therefore a potential confounder. Conclusions ES incidence is not homogenous over time but seems to recognize a clustered or “hot- spot” pattern. While ES does not seem to be associated with absolute temperature values, most ES happened in association with an increase in monthly temperature range. Although a higher incidence of ES can be demonstrated also during working hours and working days, the present findings are observational in nature, and new hypotheses must be tested in order to explain this peculiar behaviour.
Introduzione Electrical storm (ES) è definite come la presenza di 3 o più episodi di tachycardia ventricolare sostenuta (VT) o fibrillazione ventricolare (VF) in 24 ore che richiede o pacing anti tachicardico (ATP) o cardioversione/defibrillazione. Le tachiaritmie ventricolari sembrano seguire una distribuzione circadiana, giornaliera e stagionale. Sfortunatamente il pattern temporale descritto fin ora varia ampiamente in pazienti con cardiopatia ischemica e non, cardiomiopatia ipertrofica e sindrome di Brugada. Benché alcuni studi hanno mostrato alcuni predittori di storm aritmico, a tutt’oggi non esistono evidenze che associno questa condizione con fattori ambientali ed esterni come la temperatura, l’ora del giorno o il periodo dell’anno. Lo scopo di questo studio è di descrivere l’incidenza di ES nel tempo e valutare la potenziale associazione tra l’incidenza di ES ed il periodo dell’anno, della settimana, del giorno e le variazioni della temperatura a breve termine. Metodi Selezione dello studio Lo studio Circannual pattern and TEMPerature-related incidence of Electrical Storm (TEMPEST) è stato eseguito nel rispetto delle line guida correnti e registrato nel PROSPERO il registro prospettico internazionale delle revisioni sistematiche dell’Università di York United Kingdom (registrazione # CRD42013003744). Due database medici (MEDLINE and Embase) sono stati valutati per includere tutti gli articoli disponibili. MEDLINE è stata valutata utilizzando le seguenti parole chiave: “electrical storm” [Mesh] OR “arrhythmic storm” [Mesh] OR “recurrent ventricular arrhythmias” [Mesh] OR “ventricular tachycardia clusters” [Mesh] OR “electrical instability” [Mesh]. ISI Web of Science è stata valutata utilizzando le seguenti parole chiave: “electrical storm” OR “arrhythmic storm” OR “recurrent ventricular arrhythmia” OR “ventricular tachycardia clusters” OR “electrical instability”. La ricerca è stata realizzata dal 01.11.2015 ed è stata limitata alla letteratura di lingua inglese. Due autori hanno controllato indipendentemente tutti gli articoli e controllato la loro appropriatezza per essere selezionati secondo i seguenti criteri: a) Diagnosi di ES come 3 o più episodi di VT/VF in 24 ore (ogni episodio almeno 5 minuti dal precedente) o VT per più di 12 ore; b) L’assenzadisindromecoronaricaacutacometriggeraritmico; c) 10opiùpazientiinclusi; d) laselezionedellepubblicazionipiùrecentiquandolostessogruppohariportato lo stesso paziente in pubblicazioni separate. La concordanza inter osservatore era ottimale durante l’intero processo di selezione (k = 0.97). Alla fine del processo di selezione, 35 articoli sono stati inclusi nel processo di raccolta dati. Raccolta dati Gli autori dei lavori selezionati sono stati contattati per email ed è stato chiesto loro di partecipare. Di questi, 19 hanno aderito ed è stato condiviso il protocollo dello studio. Sono state chieste informazioni cliniche aggiuntive su ciascun paziente, gli esami di laboratorio e le caratteristiche dell’ES, le terapie erogate dall’ICD, la gestione farmacologica e non. 9 autori hanno fornito un dataset completo. Per assicurare una buona qualità dei dati, i dataset senza dati disponibili sulla data dell’ES e sulla localizzazione geografica di ciascun paziente sono stati esclusi. Alla fine del processo di selezione e raccolta dati, 5 centri sono stati selezionati ed I dati sono stati inseriti nel presente registro. Raccolta dei dati sulla temperatura I dati sulla temperatura sono stati raccolti dalla sezione “historical weather” del Weather Underground (www.wunderground.com), il più ampio database sulle previsioni del tempo disponibile online. Per ciascun paziente, è stata registrata la temperatura media, massima e minima durante il giorno dell’evento, della settimana e del mese in cui è avvenuto l’ES e del mese precedente. Se la temperatura non era disponibile per l’esatta localizzazione geografica in quel momento, si utilizzavano i dati della stazione meteo più vicina. I range di temperatura settimanali e mensili sono stati definiti come la media delle temperature massime meno ma media delle temperature minime rispettivamente durante una settimana o un mese. Lo stesso database è stato utilizzato per valutare le ore diurne e notturne. Se l’ES iniziava dopo l’alba e prima del tramonto si considerava un evento accaduto di giorno; se iniziava dopo il tramonto e prima dell’alba si considerava accaduto di notte. Risultati Caratteristiche della popolazione La popolazione dello studio ha incluso 246 pazienti con ES (221 maschi, età 66±9 anni). I livelli medi di potassio sierico erano nella norma (4.3±0.6 mEq/l), ma 23 pazienti (9.3%) avevano ipokaliemia e 6 (2.4%) iperkaliemia all’arrivo in pronto soccorso. I livelli medi di creatinina erano 1.3±0.5 mg/dl e 22 pazienti (8.9%) sono stati ricoverati con una creatinina ≥2 mg/dl. Ogni episodio di ES era composto da una mediana di 7 episodi di VT/VF (1st-3rd IQR 4-16). In media ogni paziente era sottoposto ad una mediana di 8 ATP (1st-3rd IQR 3- 10) e 5 shock (1st-3rd IQR 3-6) prima di risolvere l’episodio di ES. Incidenza di ES nel tempo L’incidenza di ES nel tempo seguiva una distribuzione non omogenea in tutti i centri partecipanti, con periodi con bassa incidenza alternati con cluster di diversi ES ricoverati a pochi giorni l’uno dall’altro. Riguardo alla distribuzione dell’ES nella giornata, un più alto numero di ES sono avvenuti nelle ore diurne rispetto alle notturne (58.7% vs. 41.3%; p<0.001). Il plotting dell’incidenza di ES nel tempo mostra che il 29% di tutti gli ES iniziava tra le 8 e le 10 a.m., con una incidenza nettamente inferiore nelle ore successive (p<0.05). L’incidenza settimanale di ES era simile dal lunedì al venerdì, mentre i giorni festivi mostravano una minore incidenza di ES rispetto agli altri giorni (tutte p<0.05). La distribuzione di ES durante l’anno era omogenea, nessun mese era significativamente associato con una maggiore prevalenza di ES. Incidenze simili di ES sono state riscontrate nelle diverse stagioni dell’anno. Temperatura e electrical storm Non si è trovata un’associazione lineare tra l’incidenza di ES e la temperatura media, minima e massima giornaliera, settimanale o mensile. Comunque l’incidenza di ES era associate in modo significativo con un increment nel range della temperature mensile, paragonata al mese precedente, con il 68.9% degli ES che si verificava dopo un aumento del range di temperature mensile (p=0.003). Discussione Variazioni correlate alla temperatura Dai nostri dati si deduce che la presentazione dell’ES non è omogenea nel tempo, ma sembra riconoscere un andamento a cluster o “hot-spot”, con un’alternanza tra periodi a bassa incidenza e periodi ad alta intensità, con molti pazienti ricoverati in pochi giorni. Questo pattern, benché non sovrapponibile per ogni centro dato il fusorario e le diverse coordinate geografiche, era però evidente per tutti i centri partecipanti. Il pattern descritto non era correlato ad un pattern circannuale né alla temperatura assoluta, dato che i centri erano molto diversi in termini di latitudine, longitudine e temperatura media mensile (10°C a Varsavia, 20° C a Valencia e 21°C a Los Angeles). Inoltre, è stata riscontrata un’associazione significativa tra l’incidenza di ES e le variazioni del range di temperatura, con quasi il 70% degli ES che si verificava dopo un aumento del range di temperatura della specifica zone geografica. I meccanismi che mettono in relazione gli ES ei trigger ambientali non sono conosciuti a tutt’oggi, ma possono essere fatte delle ipotesi. Ci potrebbe essere una relazione causale diretta, per cui abbiamo molti dati disponibili che sottolineano il rischio maggiore di aritmie ventricolari nei pazienti che sono sottoposti ad un rapido raffreddamento e riscaldamento corporeo, come accade comunemente nell’ipotermia terapeutica. Modelli animali hanno mostrato che l’ipotermia indotta amplifica la dispersione della ripolarizzazione ed aumenta la vulnerabilità del miocardio ventricolare alla VF, quindi potenzialmente potrebbe favorire eventi aritmici multipli in pazienti predisposti. Certamente i pazienti presi in esame hanno subito variazioni della temperatura esterna, non corporea, che è di minore entità e meno concentrata nel tempo. Perciò altre ipotesi possono essere fatte. Di recente si è suggerito che l’aumento della mortalità e morbidità causato da temperature calde e fredde potrebbe non essere strettamente correlato all’ipo o ipertermia ma ad altri meccanismi triggerati dai tentativi di adattamento del corpo umano alla temperatura esterna. Infatti, un’aumentata mortalità e un incremento delle ospedalizzazioni per cause cardiovascolari sono stati documentati dopo un’ondata di freddo o di caldo in diverse localizzazioni geografiche. Il tasso di ospedalizzazioni valutato da un ampio studio di popolazione ha mostrato un’associazione inversa a J tra ospedalizzazioni per ipertensione, diabete e, in misura minore aritmie, in un lasso di tempo di 21 giorni, con un rischio più alto ad entrambi gli estremi di temperatura. La vasoreattività mediata dalla termoregolazione, l’attivazione del sistema nervoso simpatico e la ritenzione di sodio e acqua, attraverso il sistema renina-angiotensina-aldosterone potrebbe essere coinvolto come meccanismo fisiopatologico sottostante. Inoltre i cambiamenti nella pressione atmosferica correlati alle variazioni del meteo e della temperatura sono associate con la concentrazione di inquinanti particolati o gassosi, che sono stati messi in relazione con un aumentato rischio di ospedalizzazioni per aritmie e con un incremento della mortalità. Variazioni settimanali e giornaliere L’incidenza maggiore di ES durante i giorni lavorativi sottolinea un ruolo predominante del sistema nervoso autonomo, come trigger dell’ES, analogamente al ben noto pattern circadiano nell’infarto miocardico acuto. Questa ipotesi è supportata da una sottoanalisi recente del trial SCD-HeFT, che ha dimostrato una variazione significativa settimanale con un aumento dell’incidenza di VT/VF di lunedì solo nei pazienti non trattati con β-bloccanti. Le analisi disponibili del pattern giornaliero di VT/VF riportano una maggior incidenza di terapie dell’ICD nelle prime ore del mattino sia nei pazienti ischemici che non ischemici, mentre i pazienti con cardiomiopatia ipertrofica o sindrome di Brugada ricevevano più spesso le terapie dell’ICD rispettivamente nel pomeriggio o di notte. Nella nostra popolazione, principalmente composta da cardiomiopatia ischemica (63.0%) e dilatativa idiopatica (20.7%), la distribuzione dell’inizio dello ES è simile alla distribuzione circadiana di VT/VF descritto da Anand et al., benché l’aumento descritto nel pomeriggio è molto meno evidente nella nostra popolazione. Così, mentre si può ipotizzare che la prima VT/VF che triggera l’ES segue lo stesso pattern circadiano già descritto, alcuni altri fattori devono essere ritenuti responsabili dell’aggregazione di molte VT/VF in uno ES. data l’alta prevalenza di cardiopatie strutturali e severa disfunzione ventricolare sinistra nella nostra popolazione, si può identificare lo scompenso cardiaco (HF) come uno dei protagonisti principali. Dati disponibili con un follow-up a 5 anni mostrano che i pazienti con ES condividono alcuni elementi clinici con i pazienti ricoverati per riacutizzazione di HF, e lo ES potrebbe essere interpretato come un segno di scompenso cardiaco imminente piuttosto che un evento indipendente. Le alterazioni bioumorali tipiche dello HF potrebbero contribuire a creare un substrato per la ricorrenza di aritmie, con l’ES come un epifenomeno. Un fattore di rischio aritmico potenzialmente sovrapponibile nei pazienti con HF sono i disturbi respiratori del sonno (SDB). Recenti dati osservazionali hanno mostrato che SDB è un predittore indipendente di terapie appropriate nei pazienti portatori di ICD e potrebbe almeno in parte spiegare l’aumentata incidenza di ES dociumentata al mattino. Limitazioni Questo studio presenta tutte le limitazioni intrinseche ad uno studio retrospettivo. Inoltre, le temperature medie sono state estrapolate da un sito web, quindi limitano la reale valutaizone della temperatura esterna di un’ampia area geografica. Riguardo ai dati raccolti, la terapia farmacologica domiciliare non è stata raccolta routinariamente, dato che lo studio era focalizzato sul trattamento acuto dello ES. quindi, non si possono estendere questi risultati a sottopopolazioni specifiche (es. pazienti precedentemente trattati con antiaritmici o β-bloccanti). Inoltre, non possiamo fornire dati riguardo la programmazione dell’ICD, che è un fattore importante per l’ES ed anche un potenziale fattore di confondimento. Conclusioni L’ incidenza di ES non è omogenea nel tempo, ma sembra essere organizzata in cluster o in “hot-spot”. Anche se l’ES non sembra essere associato con i valori di temperatura assoluta, la maggior parte degli ES si verifica in associazione ad un aumento del range di temperatura mensile. Una maggiore incidenza di ES può essere dimaostrata anche durante i giorni lavorativi e le ore diurne. Questi dati sono osservazionali e nuove ipotesi potrebbero essere dimostrate per spiegare questo particolare comportamento dell’incidenza dell’ES.
Circannual pattern and Temperature-related incidence of Electrical Storm: the Tempest study / Cipolletta, Laura. - (2017 Mar 22).
Circannual pattern and Temperature-related incidence of Electrical Storm: the Tempest study
CIPOLLETTA, LAURA
2017-03-22
Abstract
Introduction Electrical storm (ES) is usually defined as the occurrence of three or more episodes of sustained ventricular tachycardia (VT) or ventricular fibrillation (VF) within 24 hours requiring either anti-tachycardia pacing (ATP) or cardioversion/defibrillation. The occurrence of ventricular tachyarrhythmias appears to follow circadian, daily, and seasonal distributions. Unfortunately, the temporal patterns described so far vary widely in patients with ischemic and non-ischemic heart disease, hypertrophic cardiomyopathy, and Brugada syndrome. Although some studies have shown potential candidates as ES predictors, there is currently no evidence regarding the association of this condition with environmental and external factors such as temperature, time of the day and time of the year. The aim of the present study is to describe the incidence of ES over time and test the potential association between ES incidence and time of the year, time of the week, time of the day and short-term temperature variations. Methods Study selection The Circannual pattern and TEMPerature-related incidence of Electrical Storm (TEMPEST) study was conducted following the current guidelines, and registered in the PROSPERO International prospective register of systematic reviews of the University of York, United Kingdom (registration # CRD42013003744). Two medical databases (MEDLINE and Embase) were systematically searched in order to include all available papers. MEDLINE was searched using the following query: “electrical storm” [Mesh] OR “arrhythmic storm” [Mesh] OR “recurrent ventricular arrhythmias” [Mesh] OR “ventricular tachycardia clusters” [Mesh] OR “electrical instability” [Mesh]. ISI Web of Science was searched using the following query: title contains “electrical storm” OR “arrhythmic storm” OR “recurrent ventricular arrhythmia” OR “ventricular tachycardia clusters” OR “electrical instability”. The search was performed up until November 1st, 2015 and was limited to English-language literature. Two authors independently screened all the records reviewed full-texts articles and determined their eligibility. In order to be selected, a study has to meet all the following criteria: a) diagnosis of ES as the occurrence of three or more episodes of VT/VF within 24 hours (each episode at least 5 minutes apart) or VT for more than 12 hours; b) theabsenceofacutecoronarysyndromeasthearrhythmictrigger; c) tenormorepatientsincluded; d) selection of the most recent publication when the same group reported on the same patients in separate publications. Inter-observer concordance was optimal during the whole selection process (k = 0.97). At the end of the selection process, 35 papers were included in the data collection process. Data collection Corresponding authors of aforementioned papers were contacted by e-mail and asked to participate. Of those, 19 agreed and the study protocol was provided to them. All authors were asked for additional information on a patient-level basis regarding clinical characteristics, lab exams, ES characteristics, therapies delivered by the ICD, and pharmacological and non-pharmacological management. Nine authors replied back with a complete dataset. In order to ensure optimal data quality, datasets with no data available regarding ES time and geographical location of each patient were excluded, as were datasets collecting non-consecutive patients. At the end of the selection and collection processes, five centres had the datasets from their papers (9–13) merged into the present registry. Temperature data collection Temperature data were retrieved from the “historical weather” section of Weather Underground (www.wunderground.com), the largest weather database available online. For each patient, average, maximum, and minimum temperatures during the event day were collected, as were average mean, maximum and minimum temperatures during the week and the month of the ES, and the week and the month before. If the weather almanac was not available for the exact geographical location at that time, the nearest forecast station with available data was used instead. Weekly and monthly temperature ranges were defined as the average of maximum temperatures minus the average of minimum temperatures during a week or a month, respectively. The same database was used in order to retrieve sunrise and sunset hours. If ES started after sunrise and before sunset it was considered as happened during daytime; if it started after sunset and before sunrise it was considered as happened during night-time. Results Population characteristics Total population included 246 patients presenting with ES (221 males, age 66±9 years). While mean serum potassium levels were within the normal ranges (4.3±0.6 mEq/l), 23 patients (9.3%) had hypokalemia and 6 patients (2.4%) had hyperkalemia on presentation. Mean creatinine levels were 1.3±0.5 mg/dl and 22 patients (8.9%) were admitted with creatinine serum levels ≥2 mg/dl. Each ES episode was made up by a median of 7 VT/VF episodes (1st-3rd IQR 4-16). On average, each patient experienced a median of 8 anti-tachycardia pacing (ATP) therapies (1st-3rd IQR 3-10) and 5 shocks (1st-3rd IQR 3-6) before ES termination. Electrical storm incidence over time ES incidence over time followed a non-homogeneous distribution in all participating centres, with low-incidence periods alternating with clusters of many ES admitted within a few days from one another. Regarding ES distribution over the day, a significantly higher number of ES happened during daytime hours rather than during night-time hours (58.7% vs. 41.3%; p<0.001). Plotting ES incidence over time shows that 29% of all ES started from 8 to 10 a.m., with a significant incidence reduction during the following hours (p<0.05). Weekly incidence of ES was similar from Mondays to Fridays, while both Saturdays and Sundays showed a lower incidence of ES when compared to other days (all p<0.05). ES distribution over the year was homogeneous with no month significantly associated with a higher prevalence of ES. Similar incidences of ES have been found between winter, spring, summer and fall. Temperature and electrical storm No linear association was found between ES incidence and average, minimum and maximum daily, weekly or monthly temperatures. However, ES incidence was significantly associated with an increase in monthly temperature range when compared to the month before, with 68.9% of ES happening after an increase in monthly temperature range (p=0.003). Discussion Temperature-related variations From our data, ES presentation is not homogenous over time but seems to recognize a clustered or “hot-spot” pattern, alternating low-incidence with “high-intensity” periods, when many patients were admitted with ES within a few days. This pattern, while not superimposable for each center due to the different timeframes and geographical coordinates considered, was indeed quite evident for all of the five participating hospitals. The pattern described was not related to a specific circannual pattern neither to absolute temperatures, as centers varied widely in terms of latitude, longitude, and mean monthly averages (these latter ranging from 10°C in Warsaw to 20° C in Valencia and 21°C in Los Angeles). On the other hand, we found a significant association between ES incidence and variations in the temperature range, with nearly 70% of all ES happening after an increase in temperature ranges of their specific geographical locations. The mechanisms underlying the connection between ES and environmental triggers are unknown, but some hypothesis can be made. In favor of a direct causal relationship, we have many available data underlying an increased risk of ventricular arrhythmias in patients undergoing fast body temperature cooling and rewarming, as commonly experienced in therapeutic hypothermia. Animal models demonstrated that inducing hypothermia amplifies dispersion of repolarization and increases myocardial vulnerability to VF, thus potentially eliciting multiple arrhythmic events in predisposed patients. Of course, our patients experienced a variation in external rather than body temperature, which was not as big in magnitude neither so concentrated in time. Therefore, other hypotheses must be sought. Recently, it has been suggested that the increased mortality and morbidity caused by hot and cold temperatures could not be strictly related to hypothermia/hyperthermia but by other causes triggered by the human organism’s attempts to adapt to the external temperature. In this regard, increased mortality and hospitalization rates for cardiovascular diseases has been reported after both cold spells and heat waves in many different geographical locations. Hospitalization rates from a large population- based study showed a reverse J-shaped association between hospitalization for blood pressure, diabetes, and, with a minor extent, arrhythmias over a lag of 21 days, with higher risks at both temperatures extremes. Thermoregulation-mediated vasoreactivity, sympathetic nervous system activation and sodium and volume retention through the renin-angiotensin-aldosterone system could all be involved as underlying pathophysiological mechanisms. On a note, the changes in atmospheric pressure related to weather and temperature variations are also associated with the concentration of particulate and gaseous air pollutants, both of whose have been related to an increased risk of arrhythmia-related hospitalization or mortality. Weekly and daily variations The higher incidence of ES during working days in our population was not totally unexpected, as was already underlined on unclustered VT/VF. This behavior potentially highlights a predominant role of autonomic regulation in triggering the ES, similarly to the well-known circadian pattern in acute myocardial infarction. This hypothesis is supported by a recent sub-analysis of the SCD-HeFT trial, which demonstrated a significant weekly variation with an increase in VT/VF incidence on Mondays only in patients not treated with β-blockers. Available analyses of VT/VF daily patterns report a higher incidence of ICD therapies in the early morning both in ischemic and non-ischemic cardiomyopathy, while patients with hypertrophic cardiomyopathy and Brugada syndrome experience ICD therapy more often during the afternoon and late-night, respectively. In our population, mainly composed of ischemic (63.0%) and idiopathic dilated cardiomyopathy (20.7%), ES starting hour distribution is similar to the circadian distribution of VT/VF described by Anand et al., although the late rise described during the afternoon hours is much less evident in our population. So, while it is feasible to hypothesize that the first VT/VF triggering the ES follows the same circadian pattern already described, some other factors must be held responsible for the coalescence of many VT/VF into an ES. Given the high prevalence of structural heart disease and severe left ventricular dysfunction in our population, it is feasible to identify heart failure (HF) as a major potential actor. Available data on a 5-year follow-up shows that ES patients share some striking similarities with patients admitted for HF exacerbation, and ES could be seen as a warning sign of impending pump failure rather than an independent event. In this setting, the many dysregulated biochemical pathways typical of HF could help in creating a substrate for arrhythmia recurrence, with ES as an epiphenomenon. Potentially overlapping as an arrhythmic risk factor in HF patients is sleep-disordered breathing (SDB). Recent observational data have shown that SDB is an independent predictor of appropriate therapy in ICD patients, and could explain, at least in part, the increased incidence of ES seen in the morning. Limitations The present paper shares all the limitation intrinsic to a retrospective, patient-level pooled analysis. Moreover, average temperatures were extracted from a web-based forecasting service, thus limiting the real assessment of external temperature to a wide geographical area. Regarding collected data, chronic pharmacological therapy was not routinely collected, as the papers selected focused on the acute treatment of ES. Therefore, is not cautious to extend the present findings to specific subpopulations (i.e. patients previously treated with β-blockers or other anti-arrhythmic drugs). Moreover, we were not able to provide data regarding ICD programming, which has been showed as an important risk factor for ES and therefore a potential confounder. Conclusions ES incidence is not homogenous over time but seems to recognize a clustered or “hot- spot” pattern. While ES does not seem to be associated with absolute temperature values, most ES happened in association with an increase in monthly temperature range. Although a higher incidence of ES can be demonstrated also during working hours and working days, the present findings are observational in nature, and new hypotheses must be tested in order to explain this peculiar behaviour.File | Dimensione | Formato | |
---|---|---|---|
Tesi_Cipolletta.pdf
accesso aperto
Descrizione: Tesi_Cipolletta.pdf
Tipologia:
Tesi di dottorato
Licenza d'uso:
Tutti i diritti riservati
Dimensione
923.92 kB
Formato
Adobe PDF
|
923.92 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.