Click here to view next page of this article

Endocarditis and Bacteremia

There is good data to describe the epidemiology of endocarditis - how common this infection this. This data is from nationwide surveys. This is not a reportable disease, so it is not absolute data on heart valve infection. However, there have been several large studies that have used basically hospital-based discharge diagnosis reporting to look at this. The incidence ranges from two to six cases per 100,000 persons per year; this is United States data.

The reason why gram positives are more likely to cause disease is because they are sticky. They produce factors that allow them to stick to damaged endothelial surfaces. The pathogenesis of endocarditis begins with damage to the valve and then it exposes endothelium that can allow these organisms to stick to.

Who gets this disease? What are the risk factors for it? The most common group are those with underlying heart disease. That underlying heart disease can be of variable etiologies, including rheumatic valvular disease, which becomes less likely as we get less and less people who have had rheumatic fever when they were younger, but this is still something that you need to be aware of; certain kinds of congenital heart disease and in general, it's those that lead to turbulent flow.

Prosthetic heart valves - where the heart disease is iatrogenically induced, so to speak - is where we put a foreign body in that region, so that puts people at risk. Previous endocarditis is looked at as a different group, but think of them in this way - they have underlying heart disease because of previous endocarditis.

In terms of clinical presentation, I would like to break this down into acute versus subacute. We can use this breakdown to help determine what the likely etiologies are in each of these settings.

Acute endocarditis presents with signs of acute infection, not surprisingly. These patients have high fever and they may present with a sepsis-like syndrome. They are acutely ill and toxic. Usually, the illness has been present for less than seven days, so people can really date a time of onset for this presentation.

When we look at criteria - how we determine that someone has endocarditis - I just wanted to go briefly over these criteria that came from Duke University, published about six or seven years ago, and was really a major advance in our ability to classify people with endocarditis so that when we look at groups of people we could determine the right way to make a diagnosis and the right way to treat people. The basis of these criteria are that definite endocarditis requires both.

The major criteria would include blood cultures and echocardiography. So you can see where the workup of this disease has evolved to and these are the major things that we are doing from a laboratory standpoint to make this diagnosis. So you want to look for (a) a positive blood culture for either a typical organism from two separate blood cultures or persistently positive blood cultures. That is why I went through the likely flora in endocarditis initially.

Clearly, the transesophageal echocardiogram is more sensitive at detecting these things, so the lack of demonstration on a transthoracic echo doesn't mean that someone doesn't meet these criteria Viridans streptococci are often reported as alpha hemolytic streptococcus and what some laboratories do is not to further identify them. This is part of a lost of cost-cutting measures that a lot of laboratories are doing. They will then put a little caveat that says if you need to know more, you need to call us and tell us. If you have a patient in whom you are suspecting endocarditis and you are doing those blood cultures, as soon as you get that report back, you need to call the laboratory and say that they need to go further with this. Really, what you want to know are the susceptibilities. The name of the organism of a particular alpha hemolytic strep is actually not so important. What is more important is the susceptibility. As I'll talk about, some of them are intermediately resistant to penicillin.

The minor criteria that we use as part of this definition are for the most part clinical criteria, but they include defaults for those other things. This would include evidence of a predisposition - in other words, is the person in one of those risk groups that I talked about, is there fever, are there vascular phenomena (and this is where the emboli come in), immunologic phenomena such as kidney involvement, glomerulonephritis being the most common one and skin involvement, were you see immune complex deposition in the skin. If there is an abnormal echocardiogram, but it doesn't meet the major criteria.

The major complications of the disease are mechanical and heart failure due to valvular dysfunction is the one that is the most serious complication. When we talk about indications for surgery, I want you to keep that in mind. This is clearly something that you need to watch for. It is why many people believe that everybody with endocarditis still should be admitted to the hospital, although there is a movement for out-patient parenteral therapy in certain settings. I think it is dangerous.

The best way to detect that is with a transesophageal echocardiogram. You see it more commonly with aortic than mitral valve because of location and anatomy. Another major complication of endocarditis is embolization. This can be most serious when it is big vessel embolization. By far the most serious is central nervous system involvement - as I said, endocarditis can present as stroke. You can also see it to the abdominal organs, to the kidneys, to the spleen, rarely to the liver and rarely to the gut.