79-Year-Old Man With Shortness of Breath and Fevers

Abstract

A 79-year old man with a history of myocardial infarction and coronary artery bypass grafting in 2000, biventricular systolic heart failure with an ejection fraction of 27%, hypertension, persistent atrial fibrillation for which he was taking high-dose aspirin given a previous history of hemorrhagic stroke, stage 3 chronic kidney disease, left cerebellar ischemic stroke, and implantable cardioverter-defibrillator (ICD) placement 5 years previously presented to his local medical facility with a 2-day history of breathlessness, fevers, and chills. His temperature was 37.6°C. Physical examination revealed normal peripheral pulses, a quiet precordium with a murmur consistent with aortic sclerosis, clear vesicular breath sounds throughout the chest, and no tenderness, mass lesion, or organomegaly in the abdomen. Examination of his mouth revealed dental caries. He had no new skin or nail changes. There was no jugular venous distention appreciated on examination. Blood was drawn for cultures, and 4 culture bottles out of 4 grew β-hemolytic group D Streptococcus (GDS).1.Which one of the following diagnostic tests should be performed next?a.Transthoracic echocardiographyb.Magnetic resonance imaging of the thoracolumbar spinec.Computed tomography (CT) of the abdomen and pelvisd.Panoramic radiographye.Urine culture When searching for a source of a bloodstream infection, clinicians should be guided by the likely source of the identified pathogen. Given the most likely source of this patient's bloodstream infection, transthoracic echocardiography would be the most appropriate next diagnostic test. β-Hemolytic GDS species are gram-positive cocci that represent an important cause of bloodstream infections and infective endocarditis. Streptococcus bovis biotype 1 subspecies have been most commonly associated with infective endocarditis.1Boleij A. van Gelder M.M. Swinkels D.W. Tjalsma H. Clinical importance of Streptococcus gallolyticus infection among colorectal cancer patients: systematic review and meta-analysis.Clin Infect Dis. 2011; 53: 870-878Crossref PubMed Scopus (241) Google Scholar Group D streptococcal species are not usually related to infections originating from the spine, and although secondary seeding of a bloodstream infection is an important complication to consider, dedicated imaging of the thoracolumbar spine would not be indicated in the absence of relevant clinical features such as back pain or tenderness. Indeed, group D streptococcal species are strongly associated with colonic neoplasia and other gastrointestinal tract lesions,1Boleij A. van Gelder M.M. Swinkels D.W. Tjalsma H. Clinical importance of Streptococcus gallolyticus infection among colorectal cancer patients: systematic review and meta-analysis.Clin Infect Dis. 2011; 53: 870-878Crossref PubMed Scopus (241) Google Scholar but identifying them is not the most important next step in the management of this patient given the potential mortality and morbidity associated with endocarditis,2Ballet M. Gevigney G. Garé J.P. Delahaye F. Etienne J. Delahaye J.P. Infective endocarditis due to Streptococcus bovis: a report of 53 cases.Eur Heart J. 1995; 16: 1975-1980Crossref PubMed Scopus (107) Google Scholar nor would they be best investigated with CT. In such patients, however, it is important to consider colonoscopy in the near future given the strong association between S bovis and colonic malignancy. Similarly, enteroccal endocarditis should also prompt the clinician to consider endoscopy. Nevertheless, it is important to first determine the extent of valvular involvement, and thus transthoracic echocardiography would be the best next diagnostic step. Viridans streptococci are typically associated with dental infections, in which case panoramic radiography may be a useful test. However, it would not be appropriate for this patient who presents with a group D streptococcal infection. Although urine culture is a simple and commonly used test to identify the source of bloodstream infections, it would not be helpful in this case because GDS does not usually cause urinary tract infections. In light of the positive blood culture results, our patient was empirically started on vancomycin and was then transitioned to ceftriaxone, 2 g once daily, on transfer to our facility.Transthoracic echocardiography yielded findings consistent with vegetations involving the right ventricular lead of the pacemaker and the right atrial wall.2.Which one of the following represents the most appropriate next step in management?a.Continue antibiotics and watchful waitingb.Transesophageal echocardiographyc.Repeated blood culturesd.Complete ICD generator and lead extractione.Cardiothoracic surgery consultation Cases of infected implanted cardiac devices and device leads represent a special and more complex subset of infective endocarditis. Continuing antibiotics is crucial to the management of these patients but is insufficient without source control and is in fact linked to a higher mortality.3Sohail M.R. Uslan D.Z. Khan A.H. et al.Management and outcome of permanent pacemaker and implantable cardioverter-defibrillator infections.J Am Coll Cardiol. 2007; 49: 1851-1859Crossref PubMed Scopus (547) Google Scholar Transesophageal echocardiography is an invasive test not without potential complications and at this point would not add any diagnostic information because the likely source of infection has been illustrated adequately on transthoracic imaging. Given the absence of clinical signs of valvular disease and the absence of major valvular heart disease on transthoracic imaging, the guidelines would not support immediate transesophageal echocardiography. If transthoracic echocardiography in a patient with persistent bacteremia had revealed no abnormalities, then transesophageal echocardiography would be an important consideration. Repeated blood cultures are an important way of determining clearance of a bloodstream infection but would not be helpful until source control had been established and would therefore not be the most appropriate next step. In cases such as these, source control should almost always be obtained by extracting the infected device and leads and is the critical next step in this patient's management. This procedure can be undertaken by heart rhythm disease specialists/electrophysiologists and does not require consultation with a cardiothoracic surgeon. A surgical consultation would be warranted if the patient experienced complications of endocarditis or medical therapy failed after source control had been obtained because the ultimate goal is cure and prevention of future recurrence of valvular endocarditis or complications such as mediastinits or septic emboli. The patient's pacemaker was subsequently interrogated to determine his pacing requirements in anticipation of new device implantation after the infected device had been extracted and confirmed that he was pacemaker dependent 98% of the time. Infectious disease specialists recommended continuing the current regimen of antibiotics and also reiterated the importance of extracting his pacing system. On the third day of hospital stay, the patient underwent an uncomplicated extraction of the right and left ventricular leads as well as the pacing device by the heart rhythm service, after which a temporary pacing device was implanted via the right internal jugular approach. Under the direction of the general surgery service, the patient also underwent an ICD pocket capsulectomy and had a subcutaneous vacuum-assisted closure dressing placed. Two days after the extraction procedure, the dressing was removed and his wound was debrided and closed over a drain, which was removed after an additional 2-day period during which drain output was minimal. Subsequent blood culture results were negative, and the patient underwent repeated transesophageal echocardiography, which revealed a new cigar-shaped, large (4-cm), mobile mass in the right atrium that attached to the medial aspect of the lower superior vena cava.3.Which one of the following is the most likely etiology of the right atrial mass?a.Thrombusb.Septic embolic.Infected fibrous material dislodged during lead extractiond.Segment of the old pacing leade.Atrial myxoma There were no clear echocardiographic features that distinguished the etiology of the lesion. Systemic infection is associated with a prothrombotic state, but an isolated thrombus in this location that has formed de novo in the time between the initial and current echocardiogram would be unusual. Indeed, endocarditis secondary to infections with GDS have been found to be associated with larger vegetations, more destructive and invasive disease, and potentially a higher propensity for septic emboli,2Ballet M. Gevigney G. Garé J.P. Delahaye F. Etienne J. Delahaye J.P. Infective endocarditis due to Streptococcus bovis: a report of 53 cases.Eur Heart J. 1995; 16: 1975-1980Crossref PubMed Scopus (107) Google Scholar, 4Pergola V. Di Salvo G. Habib G. et al.Comparison of clinical and echocardiographic characteristics of Streptococcus bovis endocarditis with that caused by other pathogens.Am J Cardiol. 2001; 88: 871-875Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar but embolization to the atrial-caval junction would be against the normal flow of blood and would be an unusual site for an emboli to lodge. The temporal association between device and lead extraction and the new finding of the mobile mass lesion should prompt suspicion of a procedure-related complication. Because lead extraction involves simple manual traction, infected fibrous material that was previously adherent to the pacing leads can often be dislodged during removal, and that is the most likely etiology of the lesion in this case. The removed device and leads were all structurally intact after extraction, and so a fractured segment of pacing lead would not explain this finding. A primary cardiac tumor such as atrial myxoma is not typically found at the atrial-caval junction and is highly unlikely in this setting given the acuity and timing of the new mass lesion. Given the concern about infected material, the case was reviewed with the vascular radiology service, who performed removal of the mass from the right atrial-caval junction with an aspiration tool (AngioVac; AngioDynamics). Immediate histologic review revealed a large amount of fibrinous tissue, and the sample was subsequently sent for culture and sensitivity analysis. There is little data comparing clinical outcomes between patients with and without residual thrombus following extraction. Consequently, there are no specific guidelines informing decisions regarding repeated imaging with transesopheageal echocardiography or CT in patients following extraction or on the duration of antibiotic therapy, and thus these decisions are typically made on an individual case-by-case basis. In the interim, the patient’s initial blood cultures were further speciated as group D S bovis.4.Which one of the following is the most important next step in management?a.Colonoscopyb.Esophagogastroduodenoscopyc.Panoramic radiographyd.Device interrogatione.No further testing necessary At this point, the patient was receiving appropriate antibiotics under the guidance of the infectious diseases service, and source control had been obtained through removal of the infected system and the new mobile mass identified in the atrial-caval junction. Group D S bovis organisms have a strong association with gastrointestinal tract lesions, particularly colonic neoplasms, and one meta-analysis reported a pooled odds ratio of 7.26 (95% CI, 3.94-13.36)1Boleij A. van Gelder M.M. Swinkels D.W. Tjalsma H. Clinical importance of Streptococcus gallolyticus infection among colorectal cancer patients: systematic review and meta-analysis.Clin Infect Dis. 2011; 53: 870-878Crossref PubMed Scopus (241) Google Scholar for colorectal cancer in patients with a group D S bovis bloodstream infection. Also, given this patient's age, the next most appropriate step in management would be a colonoscopy. Group D S bovis infections are linked with pathology of the lower, as opposed to upper, gastrointestinal tract, so esophagogastroduodenoscopy would not be indicated in this scenario. Dental caries and poor dentition, which can be further assessed with panoramic radiography, are associated with viridians streptococci bloodstream infections and infective endocarditis, but these organisms have not been identified in this case. Ongoing interrogation of the patient's temporary device is important to ensure that the device is functioning and meeting the patient's pacing requirements but should occur concomitantly and not instead of a search for colonic neoplasia, which should be the most important next step in this case. Withholding further investigation at this point would also not be appropriate because the association between group D S bovis infection and colonic malignancy is strong and well established; colonic malignancy is associated with considerable morbidity and mortality, and screening for colorectal cancer using colonoscopy allows for early diagnosis and treatment. The patient underwent colonoscopy, which revealed a likely malignant tumor at the hepatic flexure in addition to multiple 10- to 25-mm polyps in the mid transverse colon, at the hepatic flexure, and in the ascending colon. Pathologic examination identified moderately differentiated adenocarcinoma, and CT of the chest, abdomen, and pelvis did not reveal any metastatic disease. The general surgery service was consulted, and the patient was scheduled for and underwent an uncomplicated extended laparoscopic hemicolectomy with anastomosis.5.Which one of the following would be the most appropriate duration of therapy for this patient?a.Two weeks of antibiotics from the time at which the first negative blood culture result was obtainedb.Four weeks of antibiotics from the time at which the first negative blood culture result was obtainedc.Six weeks of antibiotics from the time at which the first negative blood culture result was obtainedd.Four weeks of antibiotics from the time of the removal of the infected pacing devicee.Six weeks of antibiotics from the time of the removal of the infected pacing device When determining the optimal duration of antibiotic therapy for patients with bloodstream infections, clinicians should be clear on establishing an appropriate treatment start date. Ordinarily, this time frame would begin at the point at which the first negative blood culture result is obtained, but in patients who have an infected device, the start date should be modified to account for the point at which source control has been obtained. Therefore, the time at which the first negative blood culture result was obtained does not play a role in the determination of this patient's duration of antibiotic therapy. Source control through device and lead extraction was not obtained until almost 72 hours after starting appropriate antibiotic therapy, and without removal of the foci of sepsis, ongoing dissemination of infecting pathogens would prevent complete bacterial eradication even with appropriate antimicrobial therapy. Thus, the patient's start date was fixed at the point at which the infected pacing device and leads were removed. The atrial-caval mass lesion that had been removed did not grow any organisms on culture and so likely consisted of residual aseptic fibrous material that had encapsulated the pacing leads over time, and its removal date did not need to influence the patient's treatment start date. In keeping with recent guidelines outlining optimal treatment duration in infectious endocarditis without complications,5Baddour L.M. Wilson W.R. Bayer A.S. et al.American Heart Association Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and Stroke CouncilInfective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association.Circulation. 2015; 132 ([published corrections appear in Circulation. 2015;132(17):e215 and Circulation. 2016;134(8):e113]): 1435-1486Crossref PubMed Scopus (1572) Google Scholar the patient was recommended to complete 4 weeks of intravenous antibiotics from the time that the pacing device was removed. Treatment for a longer period has not been found to be necessary,5Baddour L.M. Wilson W.R. Bayer A.S. et al.American Heart Association Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and Stroke CouncilInfective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association.Circulation. 2015; 132 ([published corrections appear in Circulation. 2015;132(17):e215 and Circulation. 2016;134(8):e113]): 1435-1486Crossref PubMed Scopus (1572) Google Scholar and in keeping with appropriate antibiotic stewardship, antibiotics should not be continued for any longer than clinically indicated. Under the guidance of the infectious diseases service, the patient was advised to undergo reimplantation of his pacing system 14 days after the initial infected system had been removed, which he undertook as an outpatient without complication. Group D Streptococcus, of which S bovis represents a subspecies, is an important cause of infective endocarditis to maintain in the differential diagnosis, particularly due to its strong association with gastrointestinal tract polyps and malignancy.1Boleij A. van Gelder M.M. Swinkels D.W. Tjalsma H. Clinical importance of Streptococcus gallolyticus infection among colorectal cancer patients: systematic review and meta-analysis.Clin Infect Dis. 2011; 53: 870-878Crossref PubMed Scopus (241) Google Scholar Group D Streptococcus accounts for 2% to 6% of streptococcal bloodstream infections in hospitalized patients,6Pfaller M.A. Jones R.N. Marshall S.A. Edmond M.B. Wenzel R.P. Nosocomial streptococcal blood stream infections in the SCOPE Program: species occurrence and antimicrobial resistance.Diagn Microbiol Infect Dis. 1997; 29: 259-263Abstract Full Text PDF PubMed Scopus (47) Google Scholar with the gastrointestinal tract being the most likely portal of entry,7Murray H.W. Roberts R.B. Streptococcus bovis bacteremia and underlying gastrointestinal disease.Arch Intern Med. 1978; 138: 1097-1099Crossref PubMed Scopus (112) Google Scholar and up to 28% of these patients have concurrent infective endocarditis,8Zarkin B.A. Lillemoe K.D. Cameron J.L. Effron P.N. Magnuson T.H. Pitt H.A. The triad of Streptococcus bovis bacteremia, colonic pathology, and liver disease.Ann Surg. 1990; 211: 786-791Crossref PubMed Scopus (110) Google Scholar These patients are also typically older and less likely to have predisposing risk factors such as intravenous drug use or structural heart disease compared to cases of infective endocarditis related to other microorganisms.9Moellering Jr., R.C. Watson B.K. Kunz L.J. Endocarditis due to group D streptococci: comparison of disease caused by Streptococcus bovis with that produced by the enterococci.Am J Med. 1974; 57: 239-250Abstract Full Text PDF PubMed Scopus (126) Google Scholar Infective endocarditis in the context of a likely infected pacing system is relatively uncommon with an incidence of less than 6% and is more typically caused by infection with coagulase-negative staphylococci such as Staphylococcus epidermidis.3Sohail M.R. Uslan D.Z. Khan A.H. et al.Management and outcome of permanent pacemaker and implantable cardioverter-defibrillator infections.J Am Coll Cardiol. 2007; 49: 1851-1859Crossref PubMed Scopus (547) Google Scholar Clinical manifestations of infective endocarditis related to GDS are similar to those related to other organisms.10Duval X. Papastamopoulos V. Longuet P. et al.Definite Streptococcus bovis endocarditis: characteristics in 20 patients.Clin Microbiol Infect. 2001; 7: 3-10Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar The presentation is often subacute, but occasionally acute onset is seen9Moellering Jr., R.C. Watson B.K. Kunz L.J. Endocarditis due to group D streptococci: comparison of disease caused by Streptococcus bovis with that produced by the enterococci.Am J Med. 1974; 57: 239-250Abstract Full Text PDF PubMed Scopus (126) Google Scholar; involvement of more than one valve is seen more frequently than with other streptococci species,11Bassetti M. Secchi G. Borziani S. et al.Successful treatment of four-valve native endocarditis caused by Streptococcus bovis.Int J Cardiol. 2004; 97: 159-160Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar and vegetations that are large (ie, >10 mm) are seen more frequently than with non-GDS species4Pergola V. Di Salvo G. Habib G. et al.Comparison of clinical and echocardiographic characteristics of Streptococcus bovis endocarditis with that caused by other pathogens.Am J Cardiol. 2001; 88: 871-875Abstract Full Text Full Text PDF PubMed Scopus (92) Google Scholar and may be highly destructive, leading to valve perforation and/or septal or valvular ring abscess formation and possibly a higher frequency of embolic events.2Ballet M. Gevigney G. Garé J.P. Delahaye F. Etienne J. Delahaye J.P. Infective endocarditis due to Streptococcus bovis: a report of 53 cases.Eur Heart J. 1995; 16: 1975-1980Crossref PubMed Scopus (107) Google Scholar Diagnosis of GDS bacteremia is made by isolating the organism from the blood, while diagnosis of GDS infective endocarditis requires additional evaluation based on clinical manifestations and auxiliary testing such as electrocardiography, chest radiography, and urinalysis.12Cahill T.J. Prendergast B.D. Infective endocarditis.Lancet. 2016; 387: 882-893Abstract Full Text Full Text PDF PubMed Scopus (476) Google Scholar Echocardiography is an important diagnostic tool, and in patients with a pacing device, it is important that the echocardiography includes a detailed evaluation of the pacing leads to rule out vegetations or valvular abnormalities.3Sohail M.R. Uslan D.Z. Khan A.H. et al.Management and outcome of permanent pacemaker and implantable cardioverter-defibrillator infections.J Am Coll Cardiol. 2007; 49: 1851-1859Crossref PubMed Scopus (547) Google Scholar However, it is important to be mindful that the absence of vegetations on echocardiography may not rule out device infection. With persistent valvular endocarditis and bacteremia, it may often be assumed that the device is infected despite absence of vegetations. The association between GDS and colonic neoplasia is well established, and in one study the proportion of patients with GDS infective endocarditis who had a colonic neoplasia detected on colonoscopy was 63% (compared to 18% among patients infected by viridans streptococci).1Boleij A. van Gelder M.M. Swinkels D.W. Tjalsma H. Clinical importance of Streptococcus gallolyticus infection among colorectal cancer patients: systematic review and meta-analysis.Clin Infect Dis. 2011; 53: 870-878Crossref PubMed Scopus (241) Google Scholar The mechanism for this association is incompletely understood but may relate to proteins present in GDS species that attach to ligands overexpressed in colonic neoplasms. Furthermore, there has also been an association between GDS bacteremia and other lesions of the gastrointestinal tract including adenomas, premalignant and benign polyps, lymphoma, and colitis irrespective of its cause.2Ballet M. Gevigney G. Garé J.P. Delahaye F. Etienne J. Delahaye J.P. Infective endocarditis due to Streptococcus bovis: a report of 53 cases.Eur Heart J. 1995; 16: 1975-1980Crossref PubMed Scopus (107) Google Scholar, 7Murray H.W. Roberts R.B. Streptococcus bovis bacteremia and underlying gastrointestinal disease.Arch Intern Med. 1978; 138: 1097-1099Crossref PubMed Scopus (112) Google Scholar, 8Zarkin B.A. Lillemoe K.D. Cameron J.L. Effron P.N. Magnuson T.H. Pitt H.A. The triad of Streptococcus bovis bacteremia, colonic pathology, and liver disease.Ann Surg. 1990; 211: 786-791Crossref PubMed Scopus (110) Google Scholar Consequently, detecting GDS bacteremia should trigger an evaluation with colonoscopy. The cornerstone of managing infective endocarditis is appropriate antibiotic therapy, although in the setting of an infected pacemaker system this approach must be combined with obtaining source control by way of device and lead extraction followed by appropriately timed implantation of a new pacing system.3Sohail M.R. Uslan D.Z. Khan A.H. et al.Management and outcome of permanent pacemaker and implantable cardioverter-defibrillator infections.J Am Coll Cardiol. 2007; 49: 1851-1859Crossref PubMed Scopus (547) Google Scholar Group D Streptococcus is more susceptible to penicillin and other β-lactam antibiotics than enterococci,9Moellering Jr., R.C. Watson B.K. Kunz L.J. Endocarditis due to group D streptococci: comparison of disease caused by Streptococcus bovis with that produced by the enterococci.Am J Med. 1974; 57: 239-250Abstract Full Text PDF PubMed Scopus (126) Google Scholar whereas resistance to tetracyclines, macrolides, clindamycin, and streptomycin is highly prevalent.13Leclercq R. Huet C. Picherot M. Trieu-Cuot P. Poyart C. Genetic basis of antibiotic resistance in clinical isolates of Streptococcus gallolyticus (Streptococcus bovis).Antimicrob Agents Chemother. 2005; 49: 1646-1648Crossref PubMed Scopus (33) Google Scholar Treatment of GDS infective endocarditis relies on susceptibility data. Choice of antibiotic and treatment duration may also be impacted by the presence of prosthetic heart valves, endovascular material, cardiac complications, evidence of embolic phenomenon, and other individualized patient factors.6Pfaller M.A. Jones R.N. Marshall S.A. Edmond M.B. Wenzel R.P. Nosocomial streptococcal blood stream infections in the SCOPE Program: species occurrence and antimicrobial resistance.Diagn Microbiol Infect Dis. 1997; 29: 259-263Abstract Full Text PDF PubMed Scopus (47) Google Scholar Treatment of choice usually consists of penicillin G or ceftriaxone,6Pfaller M.A. Jones R.N. Marshall S.A. Edmond M.B. Wenzel R.P. Nosocomial streptococcal blood stream infections in the SCOPE Program: species occurrence and antimicrobial resistance.Diagn Microbiol Infect Dis. 1997; 29: 259-263Abstract Full Text PDF PubMed Scopus (47) Google Scholar or in cases of serious penicillin allergy, vancomycin. Daptomycin is another alternative in those unable to tolerate a β-lactam or vancomycin.5Baddour L.M. Wilson W.R. Bayer A.S. et al.American Heart Association Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and Stroke CouncilInfective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association.Circulation. 2015; 132 ([published corrections appear in Circulation. 2015;132(17):e215 and Circulation. 2016;134(8):e113]): 1435-1486Crossref PubMed Scopus (1572) Google Scholar The ideal duration of therapy should be 4 weeks, but patients with uncomplicated native valve infective endocarditis due to a penicillin-susceptible strain who have a prompt response and the absence of preexisting renal disease could be treated with a shorter duration using the aforementioned therapy combined with gentamicin dosed daily.5Baddour L.M. Wilson W.R. Bayer A.S. et al.American Heart Association Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease of the Council on Cardiovascular Disease in the Young, Council on Clinical Cardiology, Council on Cardiovascular Surgery and Anesthesia, and Stroke CouncilInfective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association.Circulation. 2015; 132 ([published corrections appear in Circulation. 2015;132(17):e215 and Circulation. 2016;134(8):e113]): 1435-1486Crossref PubMed Scopus (1572) Google Scholar