Journal of Internal Medicine 2000; 247: 674±678

Ehrlichiosis associated vasculitis 1

1

2

2

N. PICK , I. POTASMAN , C. STRENGER , A. KEYSARY & I. SCHWARTZ 1

3 2

From the Infectious Diseases Unit, Bnai Zion Medical Center, and Technion Faculty of Medicine, Haifa, Israel; Department of Infectious Diseases, Israel Institute for Biological Research, Ness-Ziona, Israel; Department of Biochemistry and Molecular Biology, New York Medical College, Valhalla, New York, USA 3

Abstract. Pick N, Potasman I, Strenger C, Keysary A, Schwartz I (Bnai Zion Medical Center, and Technion Faculty of Medicine, Haifa; Israel Institute for Biological Research, Ness-Ziona, Israel; New York Medical College, USA). Ehrlichiosis associated vasculitis. J Intern Med 2000; 247: 674±678. Objective. To test the hypothesis that some cases of primary vasculitis are caused by ehrlichiosis. Design. A retrospective case study and serological analysis of stored sera. Setting. University hospital. Subjects. Fifty-five patients discharged with any type of vasculitis over a 6-year period. Main outcome measures. Serology for human monocytic ehrlichiosis, and the human granulocytic

Introduction Vasculitis is a clinicopathologic process characterized by inflammation of and damage to blood vessels. A broad and heterogeneous group of syndromes may result from this process. Vasculitis and its consequences may be either primary or secondary to other diseases. The actual antigen contained in the immune complex or triggering the immune response has only rarely been identified [1]. However, in a minority of vasculitides an association between an infectious agent and the vasculitis has been found. Some of the well-known examples are the hepatitis B virus [2] and polyarteritis nodosa [3], the hepatitis C virus and cryglobulinemia [4] and lately parvovirus B19 and Wegener's granulomatosis [5]. Ehrlichiosis is an emerging infectious disease that has been linked to several clinical presentations. Most common are acute febrile illness, headache and myalgias [6]. Occasionally patients present with a predomi674

ehrlichiosis agent, and polymerase chain reaction (PCR) analysis of biopsy specimens. Results. Three patients (5.5%) had titres of 1 : 128 or higher against E. chaffeensis; none was positive for the human granulocytic ehrlichiosis agent. Skin biopsies of these patients showed lesions compatible with polyarteritis nodosa, allergic purpura and unspecified vasculitis. PCR analysis of the biopsies was unrevealing. Conclusions. Infection with human monocytic ehrlichiosis may underlie some forms of vasculitis. If confirmed, these findings may help identify patients with vasculitis who would benefit from antibiotic treatment. Keywords: E. chaffeensis, human granulocytic ehrlichiosis agent, PCR, serology.

nant single organ system involvement such as pneumonia, meningitis, or gastroenteritis [7] or prolonged fever [8]. Vasculitis has only rarely been associated with ehrlichiosis. We were triggered to carry out this study by a patient who presented with fever, arthralgia, a vasculitic type rash and a skin biopsy suggestive of polyarteritis nodosa; eventually, the patient was found to have ehrlichiosis. The association between this pathogen and vasculitis is the subject of this study.

Methods Patient selection The Bnai Zion Medical Center is a 380-bed university hospital with 28 000 hospitalizations per year. The diagnoses of all charts are computerized and coded. # 2000 Blackwell Science Ltd

EHRLICHIOSIS ASSOCIATED VASCULITIS The hospital archives were searched retrospectively (years 1993±98) for any chart having fever and vasculitis as the main discharge diagnosis (Table 1); 85 patients were found to meet these criteria. A search of sera of these patients carried out at the serology and immunology laboratories yielded 55 specimens; this constitutes our study group. Serology The sera were sent to the national reference laboratories for rickettsioses and ehrlichioses. The sera were tested for the prevalence of IgG antibodies to Ehrlichia chaffeensis and the human granulocytic ehrlichiosis agent using the immunofluorescence antibody methods (IFA) [9]. Suspensions of DH82 cells infected with E. chaffeensis, Arkansas strain (infection rate .80%) and HL60 cells infected with human granulocytic ehrlichiosis agent, nch-1 (infection rate .33%) were applied to glass slides, dried and fixed in acetone for 15 min at room temperature. Sera were diluted 1 : 64 in phosphate buffered saline (the minimal significant titre recommended by the CDC, Atlanta) and 5 mL were applied on the antigens spots. The slides were incubated for 30 min at 37 8C in a wet chamber and washed. Next 5 mL of anti human IgG:FITC conjugate solution (Sigma, St. Louis, USA; 1 : 30 in phosphate buffered saline) were applied on the spots, incubated and washed

Table 1 Distribution of various types of vasculitis in the study group

Entity

No. with diagnosis

Positive HME serology

PAN Allergic purpura SLE Wegener's granulom. Giant cell arteritis Erythema nodosum Behcet Cryglobulinemia Scleroderma Sarcoidosis Kawasaki Sicca syndrome Churg±Straus Vasculitis-unspecified Total

3 6 11 2 4 4 3 1 1 1 1 2 1 15 55

1 1

1 3

PAN = polyarteritis nodosa; SLE = systemic lupus erythematosus.

675

similarly. The slides were examined with a UV epifluorescence at 3400 magnification for specific ehrlichial fluorescence. Known positive human sera (Center for Disease Control: CDC) served as positive control. Positive sera were subsequently assayed at twofold dilutions. The first serum specimen was sent for reconfirmation at the CDC. PCR Analysis The archives of pathology were searched for biopsies of the seropositive patients. Two biopsy specimens of the index case were found and sent for PCR analysis at the New York Medical College, Valhalla, NY. In addition, three skin biopsies of patients with vasculitis but negative serology for ehrlichiosis were sent as controls. PCR for human monocytic ehrlichiosis/human granulocytic ehrlichiosis was carried out as follows. In serum. A 150 mL aliquot was processed using the Isoquick kit (ORCA Research Inc.) according to the manufacturer's instructions, as described elsewhere [10]. Formalin fixed tissues. Two 5±10 mm sections were processed for each specimen in separate tubes. Each section was placed in a 1.5-mL Eppendorf tube and 400 mL of xylene were added to remove the paraffin. The xylene extraction was repeated and after centrifugation in a microfuge for 5 min, excess xylene was removed by washing with 400 mL of 100% ethanol. This step was repeated. Tissues were dried in a speedvac and resuspended in 100 mL of lysis buffer containing 500 mcg mL21 proteinase K. Samples were incubated at 50 8C overnight, boiled for 15 min and then processed using the Isoquick kit. Final DNA pellets were resuspended in 50 mL of sterile water and a 10 mL aliquot was used for PCR. Human granulocytic ehrlichiosis specific PCR was carried out by a nested procedure using primers HS1/HS6 in the first round and HS43/HS45 in the second round as described by Sumner et al. [11]. In all PCR runs multiple positive and negative controls were included and in all cases control results were as expected.

Results Sera of 55 patients (40 female and 15 male) were available for analysis of ehrlichia antibodies. Three

# 2000 Blackwell Science Ltd Journal of Internal Medicine 247: 674±678

676

N . P I C K et al.

Table 2 Clinical and laboratory findings of patients with human monocytic ehrlichiosis and vasculitis Patient

1

2

3

Age/Sex Urban location Month Background Animal contact Skin biopsy Clinical findings Fever Rash type Arthritis/arthralgia Tonsillitis Headache Laboratory WBC Thrombocytes Alt/Ast (IU mL21) CRP Immunology Discharge diagnosis

52/male + July Healthy ‹ +

21/female + July Healthy None None

68/female + November Leukaemia None +

38.5 8C Nodules + ± +

39 8C Vasculitic ± + +

40 8C Purpura + ± +

13 800 203 000 42 152 Negative Polyarteritis nodosa 1 : 256 Cloxacillin, ciprofloxacin, NSAID

7 900 126 000 33 80 Not done Vasculitis, unspecified 1 : 128 Erythromycin, clarithromycin

9 300 80 000 45 ± Negative Allergic purpura 1 : 256 Cefazolin, amoxicillin

HME titres (IFA) Treatment

HME = human monocytic ehrlichiosis; Alt/Ast = Alanine and aspartate aminotransferase; IFA = immunofluorescent antibody; NSAID = nonsteroidal anti-inflammatory drugs; Immunology work-up included = antinuclear antibody, anti DNA, C3, C4 levels immunoelectrophoresis, rheumatoid factor, ANCA.

of these had high titres of antiehrlichial antibodies (Table 2). However, with the exception of the index case, ehrlichiosis was not suspected at any time point during admission or hospitalization. The index case was a 52-year-old patient hospitalized with fever, diffuse vasculitic rash, arthralgia, and epididymitis. Skin biopsy taken 2 days after admission disclosed necrotizing small vessel inflammation, luminal obstruction by fibrin and rare eosinophils ± consistent with polyarteritis nodosa. The patient had no history of tick bite, but used to stroll every evening in a public park. Treatment with cloxacillin and NSAIDS was ineffective, but the patient promptly defervesced following ciprofloxacin, which was given for the epididimytis. The clinical characteristics of the patients are summarized in Table 2. None of the three patients owned pets. Two of the cases occurred in July, and the third in November. All three had fever, in the range of 38.5±40 8C, with a mean duration of 4 days (range 3±6). None of the patients had chills. Epididymitis was present in the only man of the

group. One suffered from exudative pharyngotonsillitis, which heralded the rash by 2 days. None had hepatosplenomegaly, or mental changes. All patients were discharged with a diagnosis that did not hint to the infectious aetiology of their illness. The discharge diagnoses were: patient 1, nodular rash, fever, arthralgia and epididymitis; patient 2, tonsillitis and vasculitic rash; patient 3, fever and allergic purpura. The treatment consisted of macrolides and ciprofloxacin. Laboratory work-up in all three patients, including blood, urine and throat cultures were negative, as was serology for HIV, CMV (cytomegalo virus), EBV, Rickettsioses, typhoid, VDRL (syphilis), HAV, HBV and HCV (hepatitis A, B and C). Immunologic work-up, carried out in two patients, was negative. PCR was carried out on stored sera of the three patients and on two skin biopsies of one patient, but all were negative. Blinded negative biopsy specimens from two of the other 52 patients were also negative. All patients recovered from their acute febrile illness. During a follow up of 1 year no patient had

# 2000 Blackwell Science Ltd Journal of Internal Medicine 247: 674±678

EHRLICHIOSIS ASSOCIATED VASCULITIS symptoms of `collagen' disease or vasculitis, but one patient died of acute leukaemia 3 months after the acute ehrlichial infection.

Discussion We found that three out of 55 patients (5.5%) who were discharged with any diagnosis of vasculitis had positive serology for human monocytic ehrlichiosis. Ehrlichiosis is a relatively rare disease. Only about 1000 cases have been reported in the United States since the discovery of the organism in 1990 [12]. Ehrlichiosis has only recently been described in Israel in animals and man. A recent sero-survey carried out in Israel found ehrlichiosis in 30% of dogs and jackals in the northern part of the country [13]. As Haifa is situated in northern Israel, the chances of acquiring ehrlichiosis in this part of the country seem higher. The national reference laboratory has identified two cases of human monocytic ehrlichiosis and one case of human granulocytic ehrlichiosis amongst 1000 febrile patients (0.3%) [14]. In total, until today, eight cases of human monocytic ehrlichiosis and six cases of human granulocytic ehrlichiosis cases have been identified in Israel (personal communication, Avi Keysary, 1998). The rate of ehrlichiosis found in our selected group of patients with vasculitis is thus significantly higher than the rate amongst febrile patients (P , 0.001; Yates corrected x2 = 16.227), let alone the rate amongst the general population. The fact that our patients had, simultaneously, acute ehrlichiosis and a biopsy proven polyarteritis nodosa, as well as resolution of the two after antibiotic therapy suggests some association between these entities. Various infectious agents have been associated with vasculitis: the hepatitis viruses B and C [15], Cytomegalovirus [16], the HIV virus [17], Chlamydia [18] amongst others. The question remains whether ehrlichiae can interact with the vascular endothelium as well? Three reports indicate that ehrlichia can indeed cause vasculitis: (i) vasculitis was found in the postmortem of a patient who died of ehrlichiosis; (ii) ehrlichiosis was implicated as the cause of Cerebrovasculitis; (iii) ehrlichiosis was found to mimic thrombotic thrombocytopenic purpura (TTP) [19±21]. The tropism of ehrlichiae to blood vessels is not surprising since they are phylogenetically linked to the Rickettsiae, which are capable of causing vascular damage [22±24].

677

The tropism is also evident in the typical skin rash of rickettsial diseases ± which may be petechial ± as well as in the damage found in internal organs [25]. This attachment to the endothelium may, in turn, trigger an immune response culminating in vasculitis. Ehrlichiosis usually presents as a nonspecific febrile disease. The diagnosis therefore rests on serology. The CDC guidelines for the diagnosis of ehrlichiosis rests on either a fourfold rise in titres, or a titre of 1 : 128 or higher in a patient with a compatible clinical picture [26]. All three patients had a titre of 1 : 128 or higher and had no recurrences of their `vasculitic' disease during a follow-up of at least 1 year (which would be unlikely with a rheumatic disease). All three patients had fever, headaches, rash, and thrombocytopenia, which are compatible with ehrlichiosis. False positive serology may occur with almost any serological test. However, two of our patients had negative rheumatoid factor and antinuclear antibody (ANA), so false positive results seem unlikely. An unexpected drawback of our findings was the negative result of the PCR; negative PCR results are not uncommon in ehrlichiosis [6]. This may be due to the fact that serum, as opposed to whole blood, is not the optimal medium for the detection of ehrlichia (an intracellular organism) by PCR. The reason for the negative results in the tissue specimens could be an imperfect sensitivity of the PCR technique in paraffin embedded specimens. A negative result could also have occurred because the specimen was obtained after the start of antibiotic treatment. Although ciprofloxacin is not the agent of choice for ehrlichiosis, it has demonstrated good in vitro activity against ehrlichial species [27]. Finally, the diagnosis of ehrlichiosis by blood smears, or culture was impossible due to the retrospective nature of the study. Indirect evidence to support the role of ehrlichiosis in vasculitis comes from the fact that tetracyclines, which are the drug of choice for ehrlichiosis have been reported to have anti inflammatory properties in skin diseases [28]. It can be postulated that some rheumatic enigmas that responded to tetracyclines were actually residual ehrlichial infections. This could be the missing link to the question: do the tetracyclines have a role in rheumatology [29]? In summary, vasculitis may be either primary or secondary to other diseases. In a minority of patients

# 2000 Blackwell Science Ltd Journal of Internal Medicine 247: 674±678

678

N . P I C K et al.

with vasculitis, especially those who present with an acute febrile disease, the diagnosis of ehrlichiosis should be entertained. Furthermore, treating an infectious disease with steroids, as is often the case in vasculitides may be damaging.

References 1 Fauci A. The Vasculitis Syndromes. In: Fauci A, Braunwald E, Isselbacher K, Wilson J, Martin J, Kasper D, Hauser S, Longo D, eds. Harrison's Principles of Internal Medicine 14 Ed, Vol 2 New York, USA: McGraw-Hill, 1998: 1910±22. 2 Sergent J, Lockshin M, Christian C, Gocke D. Vasculitis with hepatitis B antigenemia: long term observation in nine patients. Med 1976; 55: 1±18. 3 Guillevin L. Treatment of polyarteritis nodosa related to hepatitis B virus with interferon-alpha and plasma exchange. Ann Rheum Dis 1994; 53: 334. 4 Agnello V, Chung R, Kaplan L. A role for hepatitis C virus infection in type 2 cryoglobulinemia. N Eng J Med 1992; 327: 1490±5. 5 Somer T, Finegold S. Vasculitis associated with infections, immunization and antimicrobial drugs. Clin Infect Dis 1995; 20: 1010. 6 Everett E, Evans K, Henry R, McDonald G. Human ehrlichiosis in adults after tick exposure. ANN Intern Med 1994; 120: 730±5. 7 Dumler J, Bakken J. Ehrlichial diseases of humans: Emerging tick-borne infections. Clin Infect Dis 1995; 20: 1102±10. 8 Roland W, McDonald G, Caldwell C, Everett E. Ehrlichiosis ± a cause of prolonged fever. Clin Infect Dis 1995; 20: 821±5. 9 Nicholson W, Comer J, Sumner J, Gingrich-Baker C, Coughlin R, Magnarelli L, et al. An indirect immunofluorescent assay using a cell culture-derived antigen for detection of antibodies to the agent of human granulocytic ehrlichiosis. J Clin Microbiol 1997; 35: 1510±6. 10 Schwartz I, Varde S, Nadelman R, Wormer G, Fish D. Inhibition of efficient polymerase chain reaction amplifucation of Borrelia Burgdorferi DNA in blood-fed ticks. Am J Trop Med Hyg 1997; 56: 339±42. 11 Sumner J, Nicholson W, Massung R. PCR amplification and comparison of nucleotide sequences from the groESL heat shock operon of Ehrlichia species. J Clin Microbiol 1997; 35: 2087±92. 12 Walker D, Dumler J. Emerging and re-emerging rickettsial diseases. N Eng J Med 1994; 24: 1651±2. 13 Baneth G, Waner T, Koplah A, Weinstein S, Keysary A. Survey of Ehrlichia canis antibodies among dogs in Israel. Vet Rec 1996; 138: 257±9. 14 Keysary A, Amram L, Rzotkiewicz S, Strenger C, Waner T. The diagnosis of rickettsial diseases in Israel. In: 13th SesquiAnnual Meeting of the ASR Champion, PA, 1997. 15 Munke H, Stockmann F, Ramadori G. Possible association

16

17

18

19 20

21

22

23

24 25

26

27

28 29

between Behcet9s syndrome and chronic hepatitis C virus infection. N Engl J Med 1995; 332: 400±1(letter). Zhou Y, Leon M, Waclawiw M, Popma J, Yu Z, Finkel T, Epstein S. Association between prior cytomegalovirus infection and the risk of restenosis after coronary atherectomy. N Engl J Med 1996; 335: 624±30. Libman B, Quismorio F, Stimmler M. Polyarteritis nodosalike vasculitis in human immunodeficiency virus infection. J Rheumatol 1995; 22: 351±5. Genereau T, Wechsler B, Frances C, Godeau P. Cutaneous vasculitis during Chlamydia trachomatis seroconversion. Br J Dermatol 1995; 132: 1022±3(letter). Kiptoon J, Mugera G. Invasion of capillary endothelial cells by Ehrlichia ondiri. Vet Rec 1983; 112: 59±60. Grant A, Hunter S, Partin W. A case of acute monocytic ehrlichiosis with prominent neurologic signs. Neurol 1997; 48: 1619±23. Marty A. Ehrlichiosis mimicking thrombotic thrombocytopenic purpura. Case report and pathological correlation. Hum Pathol 1995; 26: 920±5. George F, Brouqui P, Boffa M, Mutin M, Drancourt M, Brisson C, et al. Demonstration of Rickettsia conorii-induced endothelial injury in vivo by measuring circulating endothelial cells, thrombomodulin, and von Willebrand factor in patients with Mediterranean spotted fever. Blood 1993; 82: 2109±16. Dumler J, Gage W, Pettis G, Azad A, Kuhadja F. Rapid immunoperoxidase demonstration of Rickettsia rickettsii in fixed cutaneous specimen from patients with Rocky Mountain spotted fever. Am J Clin Pathol 1990; 93: 410±4. Potasman I, Bassan H. Pulmonary embolism complicating murine typhus. J Roy Soc Med 1986; 79: 367±8. Linnemann J, Pretzman C, Peterson E. Acute febrile cerebrovasculitis. A non-spotted fever group rickettsial disease. Arch Intern Med 1989; 149: 1682±4. Fishbein D, Dawson J, Robinson. L. Human ehrlichiosis in the United States. 1985±90. Ann Inter Med 1994; 120: 736± 43. Klein M, Nelson C, Goodman J. Antibiotic susceptibility of the newly cultivated agent of human granulocytic ehrlichiosis: promising activity of quinolones and rifamycins. Antimicrob Agents Chemother 1997; 41: 76±9. Autio P, Keski O. Tetracyclines as anti inflammatory treatment in skin diseases. Nord Med 1996; 10: 348. Toussior E, Despaux J, Wendling D. Do minocycline and tetracycline have a place in rheumatology? Rev Rheum 1997; 64: 474±80.

Received 23 July 1999; revision received 16 November 1999; accepted 12 January 2000 Correspondence: Dr Israel Potasman, MD, Infectious Disease Unit, Bnai Zion Medical Center, PO Box 4940, Haifa 31048, Israel (fax: 972 4 8359755; e-mail: [email protected]).

# 2000 Blackwell Science Ltd Journal of Internal Medicine 247: 674±678