We agree with Dr. Jorde that the patients in our case series were less ill than those described in the midwestern United States [1]. Few of our patients had comorbid conditions, and our patients were significantly younger than those in the Midwest cohort (47 ± 16 years compared with 57 ± 20 years; P = 0.04). Other factors that might account for differences in illness severity, such as referral bias or infection with less virulent strains of Ehrlichia, cannot be excluded.

It is somewhat misleading, however, to regard our experience with human granulocytic ehrlichiosis as "benign." Five patients (28%) did require hospitalization, and one was admitted to intensive care. The latter was also the patient who reported chest pains. The cause of this patient's chest discomfort is unclear but was unlikely to have been of cardiac origin. According to electrocardiography and serial measurement of creatine phosphokinase-MB isoenzyme levels, the patient had no evidence of myocardial infarction. Cardiac catheterization showed normal coronary arteries. Two-dimensional echocardiography showed no evidence of myocarditis, endocarditis, or pericarditis. Cardiac evaluation was not considered necessary for the other patients in our series.

Few data support direct cardiac injury as a result of infection by Ehrlichia species. Of more than 200 patients with human granulocytic ehrlichiosis identified so far, 3 patients who died after severe disease had postmortem examinations that showed no evidence of cardiac inflammation or ehrlichial infection of coronary arteries [1, 2]. One patient with human granulocytic ehrlichiosis and possible systemic lupus erythematosus developed a pericardial effusion that led to cardiac tamponade, and ehrlichia nucleic acids were present in the effusion [3]. Because ehrlichiae infect leukocytes that are recruited to sites of inflammation, it was not unexpected that inflammatory processes would contain ehrlichia-infected cells. Another patient with myocarditis was found to have antibodies to E. chaffeensis, an association of uncertain significance [4].

Another possible explanation for cardiac involvement in cases of this disease would be concomitant infection with B. burgdorferi leading to carditis. We saw no clinical or laboratory evidence of Lyme disease (the patient had no erythema migrans rash, cranial nerve palsy, arthritis, or heart block, and results of acute- and convalescent-phase serologic testing were negative for antibodies to B. burgdorferi).

We share Dr. Jorde's interest in whether prophylactic antibiotic treatment of Ixodes scapularis tick bites will prevent human granulocytic ehrlichiosis. More studies are needed to address this concern and to answer the many other questions engendered by the emerging tick-borne diseases.