Woman With Fever, Malaise, and Lesions on Her Hands and Feet

A 22-year-old woman presents with fever and malaise of1 month's duration. About 3 weeks earlier she went to theemergency department. Erythromycin was prescribed,and the patient was told to seek medical attention if hercondition did not improve. Since that time, her healthhas worsened, the fever has continued, and she has lostweight. She says she has had painful areas on her handsand feet but no rash.

HISTORY
The patient has no history of heart disease or othersignificant medical conditions. Since finishing the erythromycin2 weeks ago, she has taken no medications. Shesmokes 1 pack of cigarettes daily and drinks alcohol, primarilyon weekends. She is currently unemployed and liveswith friends, who appear to have no significant illnesses.

PHYSICAL EXAMINATION
The patient is thin and appears ill. Temperature is39.4^oC (103^oF); heart rate, 110 beats per minute; respirationrate, 18 breaths per minute; blood pressure, 100/60 mm Hg;and oxygen saturation on room air, 96%. Mucous membranesare pale. No neck vein distention or hepatojugular reflux.Chest is clear. Heart examination reveals tachycardiaand a grade 3/6 holosystolic blowing murmur at the apexthat radiates to the axillae. Normal abdomen with no hepatosplenomegaly.Several nontender hemorrhagic lesionsare noted on the left palm and the soles of both feet; scarificationlines are visible along the venous system of the leftarm. No edema in the extremities. Neurologic examinationis normal.

LABORATORY AND IMAGING RESULTS
Leukocyte count is 16,900/μL, with a predominanceof polymorphonuclear forms; hemoglobin level, 8.5 g/dL;mean corpuscular volume, normal; platelet count, 85,000/μL.A chemistry panel reveals mild hyponatremia and hypokalemia.A chest film shows no pulmonary infiltrates or grosscongestive heart failure (CHF). An ECG demonstratessinus tachycardia with a normal PR interval and no acuteinjury currents. Urinalysis is positive for hematuria. Preliminaryevaluation of one set of blood cultures showsgrowth of gram-positive cocci. A second set of culturesalso reveals growth; however, the organisms have notyet been identified.

Antibiotics have been started.

Which of the following is the most appropriate next step?

A. Order a bone marrow biopsy to evaluate the hematologic abnormalitiesthat may be the cause of the patient's illness.
B. Order a colonoscopy to investigate a possible colonic source for herbacteremia.
C. Evaluate the genitourinary tract for obstruction or other pathology thatmight be a source for her bacteremia.
D. Obtain a transthoracic echocardiogram.
E. Obtain a cardiothoracic surgery evaluation for urgent valve replacementor repair.

CORRECT ANSWER: D
A transthoracic echocardiogram is indicated to confirm the diagnosis of bacterialendocarditis.

Clinical and laboratory evidence. The findings here essentially fulfill themodified Duke criteria for diagnosis of endocarditis.¹ These guidelines include2 types of major criteria:

• Microbiologic criteria. These include typical organisms, such as Staphylococcusaureus, from 2 separate blood cultures without a primary focus; typical microorganismsisolated from persistently positive blood cultures; or a singleblood culture of Coxiella burnetii.
• Endocardial involvement criteria. These include new valvular regurgitation orechocardiographic evidence of vegetations.

The guidelines also include minor criteria, such as:

• Epidemiologic predisposition to endocarditis (eg, preexistingvalvular disease or injection drug use).
• Vascular phenomena, such as Janeway hemorrhagiclesions.
• Immunologic phenomena, such as Osler nodes orglomerulonephritis.
• Persistent fever.

The presence of 2 major criteria is diagnostic of endocarditis.The combination of 1 major criterion plus 3minor criteria is also diagnostic. This patient seems tohave 2 major criteria-2 separate blood cultures that willlikely demonstrate S aureus, and a new, obvious mitral regurgitationmurmur. Moreover, her persistent fever, vascularskin lesions, and active urine sediment probably represent3 minor criteria.

Injection drug use and endocarditis. The patient's"needle tracks" are the likely source of the infection.Among persons who use injection drugs, the median ageof those in whom endocarditis develops is lower than thatseen in the general population. Also, the incidence of theinfection is far higher-estimated at 150 to 2000 cases ofacquired native valve endocarditis per 100,000 personyears,compared with 1.7 to 6.2 cases in those who do notuse injection drugs.² This patient is thus at risk for endocarditis,which-together with more than sufficient clinicaland laboratory evidence to support the diagnosis-justifiesimmediate initiation of antibiotic therapy.

Confirmatory and follow-up tests. Despite this strong evidence, most authoritieswould recommend a transthoracic echocardiogram (TTE) (choice D).The test is rapid and noninvasive; in appropriate patients, such as this womanwho is thin and has no chronic obstructive pulmonary disease, the specificityfor vegetations is 98%. In patients who have at least an intermediate probabilityof endocarditis-which is the case here-TTE is the initial procedure of choice.

Although a transesophageal echocardiogram (TEE) provides more dataand has greater sensitivity than a TTE with equal specificity, it is far more invasiveand costly. Reserve TEE for patients with prosthetic valves and/or problemsthat interfere with TTE (eg, obesity).³ In addition to evidence of valvularvegetations, TTE provides other useful data, including valve gradients and anestimated ejection fraction.

Colonoscopy and genitourinary evaluation (choices B and C, respectively)are appropriate for specific types of endocarditis. Streptococcus bovis infection ismost commonly seen in elderly patients and is strongly associated with coloniclesions, such as carcinoma. When this organism is isolated in a patient withsuspected endocarditis, order a careful examination of the colon. Enterococciare also seen in endocarditis in elderly patients, as well as patients with nosocomialinfections that are chronic or that stem from manipulation of the genitourinarytract. In this patient, the epidemiologic profile is not consistent witheither of these infections.

Infective endocarditis can produce a variety of interesting systemic phenomena.¹ These include anemia of inflammation and myriad immune complexphenomena, such as skin lesions and thrombocytopenia, that resolve as bacteremiais cleared. This patient's leukocytosis, anemia, and thrombocytopeniaare secondary to her profound infection and the inflammatory state it has produced.Chronic myeloid leukemia or a similar marrow disease can cause cellcounts similar to those seen here; however, marrow disease is very unlikely inthis patient, in whom an obvious underlying cause for hematologic abnormalityis present. Thus, bone marrow biopsy (choice A) is inappropriate.

Treatment options. Although this patient has a life-threatening infection,she does not have obvious CHF. Antibiotics, the initial therapy of choice,should be curative. In certain settings-including CHF, perivalvular abscess,and infection that is refractory to antibiotics-a combination of medical andsurgical therapy decreases endocarditis-related mortality. Surgery may also berequired for infections with certain organisms (eg, pseudomonads, fungi) thathave been associated with poor outcomes when treated with antibiotics alone.Because surgery is difficult and has its own associated mortality, it is best reservedfor patients whose illness has these features.

This patient has none of the indications for surgery (although echocardiographymay subsequently reveal them). Thus, initial antibiotic therapy andmonitoring are sufficient. Urgent valve replacement or repair (choice E) is tooaggressive and not appropriate at this time.

Outcome of this case. TTE revealed vegetations on the mitral valve butno evidence of CHF. It could not be determined whether there was preexistingprolapse. The patient received an initial 5-day course of intravenous vancomycinand gentamicin. Microbiologic analysis revealed the staphylococci to bemethicillin-susceptible; nafcillin was subsequently administered for 6 weeks,with apparent microbiologic cure.

References:

REFERENCES:1. Li JS, Sexton DJ, Mick N, et al. Proposed modifications to the Duke criteria for the diagnosis of infectiveendocarditis. Clin Infect Dis. 2000;30:633-638.
2. Mylonakis E, Calderwood SB. Infective endocarditis in adults. N Engl J Med. 2001;345:1318-1330.
3. Lindner JR, Case A, Dent J, et al. Diagnostic value of echocardiography in suspected endocarditis. Circulation.1996;93:730-736.