Case Report

Top Case Report Discussion Author & Article Info References

In July 1995, a 60-year-old woman who lives in an area of Westchester County, New York, that is endemic for Lyme disease and human granulocytic ehrlichiosis presented with a 5-day history of fever (body temperatures up to 40 °C). At presentation, the patient reported rigors, headache, fatigue, myalgia, and anorexia. She noted that she had been bitten by ticks twice within 2 weeks before the illness developed. In June 1994, she was treated with doxycycline for Lyme disease associated with erythema migrans. Her medical history did not feature other infectious problems that would suggest an immune deficiency, and she was taking no medications.

At presentation, the patient had a normal complete blood count, mildly elevated serum aminotransferase levels, and a normal serum globulin level. Evaluation under 500x and 1000x magnification of a Wright-stained buffy-coat smear did not show morulae within granulocytes. Polymerase chain reaction (PCR) of EDTA-anticoagulated whole blood with primers GE9f/GE10r [3] from the first visit revealed the presence of DNA from the human granulocytic ehrlichiosis agent. The patient returned to her normal state of health within 4 days of beginning a 14-day course of oral doxycycline therapy (100 mg twice daily). At the initial visit, antibodies to the human granulocytic ehrlichiosis agent were not detected by an indirect immunofluorescent assay. However, the antibody titer had increased to 1280 by 1 month after the acute illness, decreased to 640 by 3 months, and decreased to 80 by approximately 10 months (Table 1).

The patient remained well until July 1997, when she developed fever (peak body temperature, 38.5 °C), chills, myalgia, arthralgia, fatigue, headaches, and dizziness 18 days after a tick bite. At the time of her evaluation on 9 July 1997 (the second day of illness), her body temperature was 37.2 °C. A 16 x 8.5-cm erythema migrans rash was present in the left popliteal fossa. The leukocyte count was 4200 cells/mm³, and the platelet count was 190 000 cells/mm³. Results of serum chemistry studies (including globulin levels) were normal. Evaluation of a Wright-stained buffy-coat specimen showed morulae within granulocytes. Polymerase chain reaction of EDTA-anticoagulated whole blood done by using primers 521/747 [7] and GER3/GER4 done by using blood from 9 July 1997 [8] demonstrated DNA from the human granulocytic ehrlichiosis agent. Immunofluorescence immunoassay done on serum from this visit showed antibodies to the human granulocytic ehrlichiosis agent at a titer of 80. Eight days later, the titer had increased to at least 2560 [2].

HL-60 cells inoculated with the patient's blood grew the human granulocytic ehrlichiosis agent [8]. Identification of the cultured human granulocytic ehrlichiosis agent was confirmed by PCR and immunofluorescence assay by using another patient's serum. Punch biopsy of the erythema migrans lesion was done, and cultures of a 2-mm specimen in modified Barbour-Stoenner-Kelly media grew Borrelia burgdorferi [9]. Antibodies to Borrelia burgdorferi were detected by an IgG/IgM enzyme-linked immunosorbent assay (Borrelia burgdorferi IgG/IgM ELISA, Wampole Laboratories, Cranbury, New Jersey) and separate IgG and IgM Western blot analyses (positive by IgM and IgG criteria) (MarDx Diagnostics, Carlsbad, California) [9]. The cultured organism was confirmed as Borrelia burgdorferi by PCR that used primers IS1 and IS2 [9].

The patient rapidly responded to doxycycline, 100 mg twice daily for 14 days, and returned to her normal state of health after 8 days of treatment. The patient continues to be well 8 months after treatment.

Discussion

Top Case Report Discussion Author & Article Info References

Our case indicates that a single infection with the human granulocytic ehrlichiosis agent does not necessarily provide long-term immunity to reinfection. This may be due to a lack of long-lasting immunity (>2 years) or to strain differences among human granulocytic ehrlichiosis agents that do not share cross-reactive protective antigens [10]. Although reactivation of infection could be entertained as the cause of the patient's second illness, several factors argue against this scenario. First, the patient had a known tick bite and had erythema migrans before her illness and was likely to have been reexposed to human granulocytic ehrlichiosis at that time. Second, the patient had been treated for human granulocytic ehrlichiosis with a 14-day course of doxycycline 2 years earlier, with complete resolution of symptoms. Chronic or latent infection with human granulocytic ehrlichiosis has not been described and does not seem to occur in horses infected with E. equi after tetracycline treatment [11]. It is unlikely that the patient has an immune deficiency because she had no history of infectious problems, but laboratory studies for a specific immune deficiency were not done. We do not know whether Borrelia burgdorferi co-infection may have caused immune suppression [12], permitting the agent of human granulocytic ehrlichiosis to escape an otherwise successful immune response.

Antibodies develop in approximately 90% of humans infected with the human granulocytic ehrlichiosis agent [1, 2], and titers remain elevated in approximately 50% of patients for at least 1 year [1]. Of note, the titer of our patient's antibodies to the human granulocytic ehrlichiosis agent had decreased substantially before reinfection. In horses [11], sheep [13], and cows [14], short-term protection from rechallenge develops after acute E. equi or E. phagocytophila infection. Although protection can last as long as 20 months, immunity usually wanes by 1 year [11, 13]. Overall, little is known about the type of immunity, humoral or cellular, that protects against reinfection with the human granulocytic ehrlichiosis agent.

Ours is only the second reported case of culture-confirmed co-infection with the human granulocytic ehrlichiosis agent and Borrelia burgdorferi [15]. Another important aspect of our case is that the patient had a second episode of Lyme disease at presentation in 1997. Experimental studies in animals and long-term follow-up of patient cohorts in areas in which the human granulocytic ehrlichiosis agent is endemic are needed to 1) define better immune measures that lead to protection and 2) determine the durability of the immune response and the incidence of reinfection. Health care providers need to know that human granulocytic ehrlichiosis is another Ixodes tick-borne disease that can reoccur, as has been previously well documented for Lyme disease [16]. Furthermore, in geographic regions where infections with Borrelia burgdorferi and the human granulocytic ehrlichiosis agent are both endemic, doxycycline should be the antibiotic of choice for Lyme disease when co-infection with human granulocytic ehrlichiosis is a possibility.