Faculty Bibliography

Clinical narratives are a valuable source of information for both patient care and biomedical research. Given the unstructured nature of medical reports, specific automatic techniques are required to extract relevant entities from such texts. In the natural language processing (NLP) community, this task is often addressed by using supervised methods. To develop such methods, both reliably-annotated corpora and elaborately designed features are needed. Despite the recent advances on corpora collection and annotation, research on multiple domains and languages is still limited. In addition, to compute the features required for supervised classification, suitable language- and domain-specific tools are needed. In this work, we propose a novel application of recurrent neural networks (RNNs) for event extraction from medical reports written in Italian. To train and evaluate the proposed approach, we annotated a corpus of 75 cardiology reports for a total of 4,365 mentions of relevant events and their attributes (e.g., the polarity). For the annotation task, we developed specific annotation guidelines, which are provided together with this paper. The RNN-based classifier was trained on a training set including 3,335 events (60 documents). The resulting model was integrated into an NLP pipeline that uses a dictionary lookup approach to search for relevant concepts inside the text. A test set of 1,030 events (15 documents) was used to evaluate and compare different pipeline configurations. As a main result, using the RNN-based classifier instead of the dictionary lookup approach allowed increasing recall from 52.4% to 88.9%, and precision from 81.1% to 88.2%. Further, using the two methods in combination, we obtained final recall, precision, and F1 score of 91.7%, 88.6%, and 90.1%, respectively. These experiments indicate that integrating a well-performing RNN-based classifier with a standard knowledge-based approach can be a good strategy to extract information from clinical text in non-English languages.

Dilated cardiomyopathy (DCM) is a clinical diagnosis characterized by left ventricular or biventricular dilation and impaired contraction that is not explained by abnormal loading conditions (for example, hypertension and valvular heart disease) or coronary artery disease. Mutations in several genes can cause DCM, including genes encoding structural components of the sarcomere and desmosome. Nongenetic forms of DCM can result from different aetiologies, including inflammation of the myocardium due to an infection (mostly viral); exposure to drugs, toxins or allergens; and systemic endocrine or autoimmune diseases. The heterogeneous aetiology and clinical presentation of DCM make a correct and timely diagnosis challenging. Echocardiography and other imaging techniques are required to assess ventricular dysfunction and adverse myocardial remodelling, and immunological and histological analyses of an endomyocardial biopsy sample are indicated when inflammation or infection is suspected. As DCM eventually leads to impaired contractility, standard approaches to prevent or treat heart failure are the first-line treatment for patients with DCM. Cardiac resynchronization therapy and implantable cardioverter-defibrillators may be required to prevent life-threatening arrhythmias. In addition, identifying the probable cause of DCM helps tailor specific therapies to improve prognosis. An improved aetiology-driven personalized approach to clinical care will benefit patients with DCM, as will new diagnostic tools, such as serum biomarkers, that enable early diagnosis and treatment.

Aims/UNASSIGNED:Data on predictors of time-to-first appropriate implantable cardioverter-defibrillator (ICD) therapy in patients with Brugada Syndrome (BrS) and prophylactically implanted ICD's are scarce. Methods and results/UNASSIGNED:SABRUS (Survey on Arrhythmic Events in BRUgada Syndrome) is an international survey on 678 BrS patients who experienced arrhythmic event (AE) including 252 patients in whom AE occurred after prophylactic ICD implantation. Analysis was performed on time-to-first appropriate ICD discharge regarding patients' characteristics. Multivariate logistic regression models were utilized to identify which parameters predicted time to arrhythmia ≤5 years. The median time-to-first appropriate ICD therapy was 24.8 ± 2.8 months. A shorter time was observed in patients from Asian ethnicity (P < 0.05), those with syncope (P = 0.001), and those with Class IIa indication for ICD (P = 0.001). A longer time was associated with a positive family history of sudden cardiac death (P < 0.05). Multivariate Cox regression revealed shorter time-to-ICD therapy in patients with syncope [odds ratio (OR) 1.65, P = 0.001]. In 193 patients (76.6%), therapy was delivered during the first 5 years. Factors associated with this time were syncope (OR 0.36, P = 0.001), spontaneous Type 1 Brugada electrocardiogram (ECG) (OR 0.5, P < 0.05), and Class IIa indication (OR 0.38, P < 0.01) as opposed to Class IIb (OR 2.41, P < 0.01). A near-significant trend for female gender was also noted (OR 0.13, P = 0.052). Two score models for prediction of <5 years to shock were built. Conclusion/UNASSIGNED:First appropriate therapy in BrS patients with prophylactic ICD's occurred during the first 5 years in 76.6% of patients. Syncope and spontaneous Type 1 Brugada ECG correlated with a shorter time to ICD therapy.

BACKGROUND:Information on young patients with Brugada syndrome (BrS) and arrhythmic events (AEs) is limited. OBJECTIVES/OBJECTIVE:The purpose of this study was to describe their characteristics and management as well as risk factors for AE recurrence. METHODS:A total of 57 patients (age ≤20 years), all with BrS and AEs, were divided into pediatric (age ≤12 years; n = 26) and adolescents (age 13 to 20 years; n = 31). RESULTS:Patients' median age at time of first AE was 14 years, with a majority of males (74%), Caucasians (70%), and probands (79%) who presented as aborted cardiac arrest (84%). A significant proportion of patients (28%) exhibited fever-related AE. Family history of sudden cardiac death (SCD), prior syncope, spontaneous type 1 Brugada electrocardiogram (ECG), inducible ventricular fibrillation at electrophysiological study, and SCN5A mutations were present in 26%, 49%, 65%, 28%, and 58% of patients, respectively. The pediatric group differed from the adolescents, with a greater proportion of females, Caucasians, fever-related AEs, and spontaneous type-1 ECG. During follow-up, 68% of pediatric and 64% of adolescents had recurrent AE, with median time of 9.9 and 27.0 months, respectively. Approximately one-third of recurrent AEs occurred on quinidine therapy, and among the pediatric group, 60% of recurrent AEs were fever-related. Risk factors for recurrent AE included sinus node dysfunction, atrial arrhythmias, intraventricular conduction delay, or large S-wave on ECG lead I in the pediatric group and the presence of SCN5A mutation among adolescents. CONCLUSIONS:Young BrS patients with AE represent a very arrhythmogenic group. Current management after first arrhythmia episode is associated with high recurrence rate. Alternative therapies, besides defibrillator implantation, should be considered.

RATIONALE/BACKGROUND:channel-blocking anesthetic lidocaine, is used to treat LQT3 patients. However, the patient response is variable, depending on the genetic mutation in SCN5A. OBJECTIVE:The goal of this study is to understand the molecular basis of patients' variable responses and build a predictive statistical model that can be used to personalize mexiletine treatment based on patient's genetic variant. METHODS AND RESULTS/RESULTS:channel voltage-sensing domain (VSD) conformational dynamics simultaneously with other gating properties for the LQT3 variants. To systematically identify the relationship between mexiletine block and channel biophysical properties, we used a system-based statistical modeling approach to connect the multivariate properties to patient phenotype. We found that mexiletine altered the conformation of the Domain III VSD, which is the same VSD that many tested LQT3 mutations affect. Analysis of 15 LQT3 variants showed a strong correlation between the activation of the Domain III-VSD and the strength of the inhibition of the channel by mexiletine. Based on this improved molecular-level understanding, we generated a systems-based model based on a dataset of 32 LQT3 patients, which then successfully predicted the response of 7 out of 8 patients to mexiletine in a blinded, retrospective trial. CONCLUSIONS:Our results imply that the modulated receptor theory of local anesthetic action, which confines local anesthetic binding effects to the channel pore, should be revised to include drug interaction with the Domain III-VSD. Using an algorithm that incorporates this mode of action, we can predict patient-specific responses to mexiletine, improving therapeutic decision making.

BACKGROUND:Triadin knockout syndrome (TKOS) is a rare, inherited arrhythmia syndrome caused by recessive null mutations in TRDN-encoded cardiac triadin. Based previously on 5 triadin null patients, TKOS has been characterized by extensive T-wave inversions, transient QT prolongation, and severe disease expression of exercise-induced cardiac arrest in early childhood refractory to conventional therapy. METHODS:We have established the International Triadin Knockout Syndrome Registry to include patients who have genetically proven homozygous/compound heterozygous TRDN null mutations. Clinical/genetic data were collected using an online survey generated through REDCap. RESULTS:Currently, the International Triadin Knockout Syndrome Registry includes 21 patients (11 males, average age of 18 years) from 16 families. Twenty patients (95%) presented with either cardiac arrest (15, 71%) or syncope (5, 24%) at an average age of 3 years. Mild skeletal myopathy/proximal muscle weakness was noted in 6 (29%) patients. Of the 19 surviving patients, 16 (84%) exhibit T-wave inversions, and 10 (53%) have transient QT prolongation > 480 ms. Eight of 9 patients had ventricular ectopy on exercise stress testing. Thirteen (68%) patients have received implantable defibrillators. Despite various treatment strategies, 14 (74%) patients have had recurrent breakthrough cardiac events. CONCLUSION/CONCLUSIONS:TKOS is a potentially lethal disease characterized by T-wave inversions in the precordial leads, transient QT prolongation in some, and recurrent ventricular arrhythmias at a young age despite aggressive treatment. Patients displaying this phenotype should undergo TRDN genetic testing as TKOS may be a cause for otherwise unexplained cardiac arrest in young children. As gene therapy advances, enrollment into the International Triadin Knockout Syndrome Registry is encouraged to better understand TKOS and to ready a well-characterized cohort for future TRDN gene therapy trials.

BACKGROUND:There is limited information on gender differences in patients with Brugada syndrome (BrS) who experienced arrhythmic events (AEs). OBJECTIVES/OBJECTIVE:To compare clinical, electrocardiographic (ECG), electrophysiologic (EP) and genetic characteristics between males and females in BrS-patients with their first AE. METHODS:The multicenter Survey on AE in BrS (SABRUS) collected data on first AE in 678 BrS-patients including 619 (91.3%) males and 59 (8.7%) females aged 0.27 to 84 (mean 42.5±14.1) years at the time of AE. RESULTS:After excluding pediatric patients, females were older than males (49.5±14.4 vs. 43±12.7 years, respectively, P=0.001). Higher proportions of females were observed in the pediatric and elderly populations. In Asians, male/female ratio of AE was ≈9-fold higher compared to Caucasians. Spontaneous type 1 BrS-ECG was associated with earlier onset of AE in pediatric females. A similar prevalence (≈65%) of spontaneous type 1 BrS-ECG was present in males and females above age of 60 years. Females less frequently showed a spontaneous type-1 BrS-ECG (31% vs. 59%, P<0.001) or arrhythmia-inducibility at EP study (34% vs. 64%, P<0.001). An SCN5A mutation was more frequently found in females (47.6% vs. 27.8% in males, P=0.007). CONCLUSIONS:This study confirms that female BrS-patients are much rarer, display less type 1 Brugada-ECG and exhibit lower inducibility rates than males. It shows for the first time that BrS females with AE have higher SCN5A mutation rates as well as the relationship between gender vs. age at onset of AE and ethnicity.

BACKGROUND:The literature on fever-related arrhythmic events (AE) in Brugada syndrome (BrS) is currently limited to few case reports and small series. OBJECTIVE:The present study aimed to describe the characteristics of fever-related AE in a large cohort of patients with BrS. METHODS:The Survey on Arrhythmic Events in Brugada Syndrome is a multicenter study on 678 patients with BrS with first AE documented at the time of aborted cardiac arrest (n = 426) or after prophylactic implantable cardioverter-defibrillator implantation (n = 252). RESULTS:In 35 of 588 patients (6%) with available information, the AE occurred during a febrile illness. Most of the 35 patients were male (80%), Caucasian (83%), and proband (70%). The mean age at the time of AE was 29 ± 24 years (range 0.3-76 years). Most patients (80%) presented with aborted cardiac arrest and 6 (17%) with arrhythmic storm. Family history of sudden death, history of syncope, and spontaneous type 1 Brugada electrocardiogram were noted in 17%, 40%, and 66% of patients, respectively. Ventricular fibrillation was induced at electrophysiology study in 9 of 19 patients (47%). An SCN5A mutation was found in 14 of 28 patients (50%). The highest proportion of fever-related AE was observed in the pediatric population (age <16 years), with a disproportionally higher event rate in the very young (age 0-5 years) (65%). Males were involved in all age groups and females only in the pediatric and elderly groups. Fever-related AE affected 17 Caucasians aged <24 years, but no Asians aged <24 years. CONCLUSION/CONCLUSIONS:The risk of fever-related AE in BrS markedly varies according to age group, sex, and ethnicity. Taking these factors into account could help the clinical management of patients with BrS with fever.