Fulminant, acute bacterial endocarditis (ABE) or subacute endocarditis (SBE)? This case illustrates the confusion in the terminology discussed previously. A rapidly progressive disease leading to death within one week involving virulent organisms (pneumococcus and enterococcus), no obvious pre-existing valve pathology, extensive valvular and blood vessel destruction, and widespread systemic complications of embolization. It certainly fits with the features of ABE, yet he also had significant weight loss over several months, and glomerular inflammation (e.g. glomerulitis), which indicated an immune complication of a more indolent disease. Of course, with his history of continuing intravenous drug abuse, and homosexual contact, one could explain at least the weight loss as HIV/AIDS-related. It was only at autopsy, however, that his HIV status became known as negative. It was also evident that the glomerular inflammation resulted from microscopic emboli of infected vegetations, and not from the development of antigen-antibody complexes in the glomerular basement membranes. With the uncertainty, it is preferable to designate cases such as this as active IE. But as discussed previously, such designation should not be used to obscure the unique features of someone with a fulminant, aggressive infection superimposed on normal heart valves.
The other important lesson in this case, as in the previous discussion of endocarditis, is the failure of several clinicians to consider the diagnosis of active IE, even in the face of some very obvious clues. A local physician had seen this patient, probably for complaints of shortness of breath and exercise intolerance. He had an echocardiogram, which led to a diagnosis of CHF. Presumably, even though we do not have access to this echocardiogram, it most likely showed ventricular contractile dysfunction, and possibly mitral valvular insufficiency. As we have seen previously in patients with dilated cardiomyopathy, ventricular dysfunction and valvular insufficiency may result from ventricular remodeling and annular dilatation. If vegetations are not appreciated or looked for on the valves, the etiology may be missed, and ascribed to myocardial disease. Certainly, with his personal history and his weight loss, the assumption that he had HIV/AIDS cardiomyopathy, or cocaine-induced cardiomyopathy, is not unreasonable. On the other hand, with intravenous drug use and acute onset CHF, it is equally reasonable to include IE in the list of possible differential diagnoses. Regardless, this patient was diagnosed as having CHF, and he was started on diuretic and digitalis therapy. Shortly after the initial visit with a local physician, the patient presented to an affiliated hospital complaining of weakness. Unfortunately, the work-up did not demonstrate a cause, and he was discharged.
The failure to diagnose this patient with IE in the days before his final presentation to our hospital had devastating consequences. By the time he was admitted with fever, increasing dyspnea on exertion, and a left facial droop, his disease was already advanced. He had signs and symptoms of at least two compromised valves, even before confirmation by echocardiography. Specifically, he had aortic insufficiency, with a wide pulse pressure (140/60 mmHg), and a low diastolic pressure secondary to the rapid diastolic backflow into the left ventricle through an incompetent aortic valve. Normally, the diastolic pressure results at least partially from the recoil of the aorta as it are distended during systole. With stretching of the elastic fibers in the aortic media, the vessel normally 'snaps-back', providing the coronary arteries with blood flow during diastole. If the aortic valve is insufficient, the elastic recoil is attenuated, and the diastolic pressure falls. This also accounts for the II/VI diastolic murmur heard along the left sternal border, although one would expect such a murmur to be heard loudest at the heart base. As a correlate of the aortic insufficiency, he also had pistol-shot pulses (strong upstroke followed by a rapid downstroke). The upstroke is a sign of a vigorous ventricular contraction, while the rapid downstroke is a result of the valve being anatomically open during diastole, comparable to the low diastolic pressure. The pistol-shot sound, which can be auscultated, also occurs because of the diastolic collapse. It is typically a loud sound that can be heard over the femoral arteries. It is one of classical pulsation abnormalities reported in aortic insufficiency (these physical examination findings were all described by physicians in the 19th century, during a period in medicine when aortic insufficiency due to infectious endocarditis, chronic rheumatic heart disease, and syphilitic aortic valve disease were common). Other signs include: 1. Corrigan's pulse, a 'water-hammer' sound due to a rapid collapse of pressure late in systole and during diastole. This sound is the banging in pipes associated with rapid sudden flow, or gas bubbles in the fluid; 2. Duroziez's sign, a to-and-fro sound heard over the femoral arteries; and, 3. Quincke's pulse, which refers to the filling and emptying of nail-bed capillaries, seen when firm pressure is applied to the fingertip. All of these phenomena result from the movement of blood back through an insufficient valve, and its association with diastolic collapse.
This patient also had evidence of mitral valve disease. He was noted to have a grade II/VI holosystolic murmur at the apex radiating to the left axilla. This is a typical location of a mitral valve murmur, and the holosystolic nature of the murmur is consistent with mitral regurgitation. The II/VI systolic murmur heard along the left sternal border is less easily characterized as to site of origin. It may represent turbulent flow across a distorted and damaged aortic valve, or ventricular cavity turbulence resulting from the aortic and mitral valve insufficiency. Ultimately, specific aortic valve and mitral valve disease was confirmed on echocardiography, with large vegetations on both valves, and prominent aortic insufficiency. It is noteworthy that the intensity of the murmurs was not as loud as one would expect from the degree of damage appreciated on echocardiography, or eventually found at autopsy. This is not an unusual observation, as the intensity of the murmur may vary with cardiac function, size of vegetation, and degree of valvular pathology. If, for instance, there is a perforation of the leaflet or cusp, this may intermittently be covered by vegetation with no flow through the hole. In the case of a torn cusp or ruptured group of chordae tendineae, the flail tissue may not move consistently with each cardiac cycle, thereby making the murmur variable in intensity. This anatomical variability is the explanation for the so-called 'changing murmur', which can occur in IE. Other manifestations of anatomic distortion of valvular tissues includes the musical or high-pitched cooing murmurs due to rapidly vibrating loose tissue or ruptured chordae; and the Austin Flint murmur (mid-diastolic rumbling), which is thought to result from the impingement of blood coming through an insufficient aortic valve, striking the ventricular surface of the mitral valve, and causing it to vibrate. At this point, you have probably realized that no other area in medicine is as rife with eponyms as valvular heart disease and endocarditis.
As noted above, the failure to diagnosis this patient earlier had major consequences that contributed to his death. By the time of presentation, he not only had fever and CHF, but also dysarthria, a left facial droop, and drooling. These findings were presumptive evidence of intracerebral emboli, and the presence of brain infarction or abscess. Confirmation came from the lumbar puncture, which demonstrated pleocytosis with polymorphonuclear predominance and decreased CSF glucose (CSF to blood ratio less than 2/3), suggesting meningitis. Cerebral abscess from embolized fragments of vegetation can often extend to the meningeal surface with the development of meningitis. Small fragments of vegetation preferentially embolize to the microcirculation at the cortico-medullary junction in the brain; then a micro-abscess develops and extends to the brain surface, where it secondarily infects the meninges. Alternatively, there may be direct seeding of meningeal blood vessels by the infective material. Cerebral abscess and meningitis are particularly common with pneumococcal endocarditis, which was one of the two organisms isolated from his blood culture (another eponym associated with endocarditis is the Austrian's triad: pneumococcal endocarditis, pneumonia, and meningitis). Often, by the time IE embolizes systemically, the process is too far advanced for any meaningful intervention. In this case, shortly after confirmation by echocardiography of the aortic and mitral valve involvement, the patient developed left hemiparesis, and mental status changes. Despite Ct scan findings of an old left temporal lobe infarction with no new lesions (it may take several days for new cerebral infarctions to become visible), clinically, he had obvious findings consistent with acute cerebral embolization. Although surgical intervention was considered, the clinicians caring for this patient considered his cardiac function stable, and before attempting valve replacement surgery, wanted to sterilize his valves with antibiotics.
Several days before his death he developed a new, and potentially ominous complication: a pericardial friction rub. This new physical finding should have led to additional diagnostic work-up, including a repeat echocardiogram and possibly a pericardiocentesis with culture of the aspirated material. No such studies were carried out. This finding should have been considered a serious new complication because it almost certainly represented spread of infection to the pericardial sac, with the development of fibrinopurulent exudate on the parietal and visceral pericardial surfaces. It is the contact of these surfaces during the cardiac cycle that leads to the auscultatory finding of friction rub (similar to the sound of two sheets of fine sandpaper rubbed together). If there is sufficient fluid or pus in the pericardial sac to separate the two layers of pericardium, the friction rub may not occur, or it may disappear if it was previously heard. There are multiple mechanisms leading to a pericardial friction rub in IE. None of them is benign:
1. There may be spread of infection from the valve surface to the fibrous annulus of the valve, where it produces an annular (or valve ring) abscess. This is a very difficult lesion to treat, because the annulus is virtually avascular, and bactericidal levels of antibiotic may not reach this tissue. Annular abscess is often a cause of a smoldering, persistent infection. It is also difficult to diagnose because blood cultures may not be positive. Additionally, it is common in patients with infection of prosthetic heart valves, particularly mechanical valves, which rarely become infected on the surface components of the valve, but are susceptible to bacterial growth in the area of the cloth-covered sewing ring (where it can spread to the annulus). The valve annulus is immediately adjacent to epicardial adipose tissue, and the epicardial surface. Thus, infection can easily spread to the pericardial sac.
2. Other mechanisms for purulent pericarditis include direct spread from the vegetation to epicardial blood vessels through the blood stream. There may also be embolization of coronary vessels leading to myocardial abscess. The abscesses that develop in the heart tend to be preferentially located in the subepicardium and mid-wall myocardium due to the anatomical organization of the microcirculation (similar to the anatomy of the cerebral microcirculation, which predominantly leads to the development of abscesses at the cortico-medullary junction). A subepicardial abscess can extend directly to the epicardium, where it causes infective pericarditis.
3. A final pathogenic mechanism is the one that affected the patient in this case. There was spread of infection from the aortic valve cusps to the aortic wall immediately superior to the valve, in the aortic root. Aggressive organisms can adhere to endothelial tissue triggering a local vascular infection (infective vasculitis). They can also spread from the annulus to the adventitia of the aorta where external invasion of the vessel may occur. In this case, the virulent infection resulted in the formation of a vascular abscess, and vessel wall destruction. A pseudo-aneurysm developed in the region, known as a mycotic aneurysm. With spread of infection through the vessel into the adventitia and beyond, pericarditis developed, since the pericardium envelops not just the heart but extends around the ascending aorta up to the aortic arch. Ultimately, the pseudo-aneurysm ruptured leading to hemopericardium. The blood in the pericardial sac (200 cc), together with the fibrinous pericarditis, was sufficient to cause pericardial tamponade, and his sudden death. Clinical recognition of fibrinous pericarditis in a patient with aortic and mitral valve endocarditis, might have led to urgent surgical intervention with valve replacement, and a vascular graft repair of the aortic root. On the other hand, at the time when patients have such widespread embolization, with cerebral and meningeal involvement, surgical cure is extremely unlikely.
Several closing remarks are worth mentioning. The renal findings were, in fact, infective glomerulitis due to micro-emboli, and not immune-complex glomerulonephritis. In addition, multiple micro-abscesses, secondary to embolization, were evident. Surprisingly, there were no infarctions or abscesses in the spleen, an organ that is commonly involved in patients with active IE. Moreover, neuropathology examination revealed infected thromboemboli associated with organizing right fronto-parietal and insular infarcts. The valvular involvement (e.g. more than one valve) is often observed in patients with fulminant endocarditis. Vegetations can spread relatively easily from the aortic valve to the anterior leaflet of the mitral valve. If there is extension into the aortic valve annulus, or to the sinuses of Valsalva above the valve, infection can cross to the tricuspid valve, which lies adjacent to these structures. In our case, the severe damage to the aortic and the mitral valves did spread secondarily to the tricuspid valve. However, as is usually the case, the pulmonic valve was not involved (the aortic, mitral, and tricuspid valve annuli are anatomically adjacent, and potentially at risk for infectious spread; the pulmonic valve is embryologically and anatomically separate). Furthermore, the pulmonic valve is relatively resistant to endocarditis because of the low-pressure flow across it, and the closure of the valve with pulmonary diastolic pressure. In contrast, the tricuspid valve, which is also affected by low pressure flow, closes with higher right ventricular systolic pressure.
The partial destruction of aortic and mitral valve tissues led to the valvular insufficiency, which is characteristic of IE. The bacterial organisms coupled with the inflammatory response, triggers the synthesis and release of local collagenases that destroy the collagen framework of the cusps and leaflets, along with the chordae tendineae. Even when endocarditis heals, it generally leaves evidence of tissue loss, with perforations, irregular tears, and flail leaflets associated with chordal rupture. Valve stenosis virtually never occurs, except when vegetations obstruct the function of a mechanical valve, or form very bulky vegetations that occlude the orifice of a native valve.
As an incidental finding, this patient had a moderately large, previously undiagnosed, adenocarcinoma of the right upper lobe of the lung. As it extended to the pleural surface, it was potentially a fatal neoplasm.
1. Alestig K, Hogevik H, Olaison L. Infective endocarditis: a diagnostic and therapeutic challenge for the new millennium. Scand J Infect Dis. 2000; 32:343-56.
2. Murphy JG, Steckelberg JM. New developments in infective endocarditis. Curr Opin Cardiol. 1995; 10:150-4.
3. Bayer AS, Bolger AF, Taubert KA, Wilson W, Steckelberg J, Karchmer AW, Levison M, Chambers HF, Dajani AS, Gewitz MH, Newburger JW, Gerber MA, Shulman ST, Pallasch TJ, Gage TW, Ferrieri P. Diagnosis and management of infective endocarditis and its complications. Circulation. 1998; 98:2936-48.
4. Pawsat DE. Inflammatory disorders of the heart. Pericarditis, myocarditis, and endocarditis. Emerg Med Clin North Am. 1998; 16: 665-81.
5. Taylor SN, Sanders CV. Unusual manifestations of invasive pneumococcal infection. Am J Med. 1999; 107(1 A):12S-27S.
This Page Intentionally Left Blank
Was this article helpful?