Rickettsia and Orientia

TAXONOMY Back to top

The family Rickettsiaceae comprises two genera of small, obligately intracellular bacteria that

reside free within the host cell’s cytosol, namely, Rickettsia and Orientia. Although the

number and diversity of pathogenic strains in these genera are similar, the practices of

species designation differ remarkably. The second edition ofBergey’s Manual of Systematic

Bacteriology lists 20 validated names of Rickettsia species, and others have been proposed

(115). Strains of Orientia tsutsugamushi, the single species of the genus, have 0.8%

divergence of the rrs (16S rRNA) gene. Similarly, other obligately intracellular bacteria have

0.5% rrs divergence within a species (e.g., Ehrlichia chaffeensis, Coxiella

burnetii, and Chlamydia trachomatis). A proposal of criteria for the limits of divergence

of Rickettsia species based upon the unacceptable criterion of historical designations would

allow different species to be as closely related as 0.2% divergence for rrs, 0.8% for citrate

synthetase(gltA), 1.2% for outer membrane protein A (ompA), 0.8% for outer membrane

protein B (ompB), and 0.7% for the cytoplasmic antigen encoded by sca4 (21, 24). There

are no common or universal concepts that can be utilized to delineate prokaryotic species as

there are with eukaryotes. However, among prokaryotes 1% divergence of the rrs gene is

considered as a natural separation between species. Thus, the genus Rickettsiahas a

disproportionate number of designated species relative to the genetic divergence of the

bacteria.

The genus is divided by the phylogenetic clustering of species into the typhus group (TG) and

spotted fever group (SFG), defined originally by their distinctive lipopolysaccharide antigens,

and the transitional and other basal groups that are widely distributed in arthropods (111).

The TG consists of only two members, Rickettsiaprowazekii and R. typhi, whereas the SFG

contains bacteria that are generally recognized as human pathogens(R. rickettsii, R. conorii,

R. africae, R. sibirica, R. japonica, R. honei, R. parkeri, R. massiliae, R. monacensis, R.

slovaca, R. aeschlimannii, and R. helvetica) as well as others that have been identified only

in arthropods (6,32, 64, 76, 78, 111, 115). The transitional group, a clade between the TG

and SFG, contains the pathogens R. akari, R. australis, and R. felis (26).

Most Rickettsia species of undetermined pathogenicity, including R. montanensis, R. bellii, R.

peacockii, and R. rhipicephali, are much more prevalent in U.S. ticks than is pathogenicR.

rickettsii. SFG isolates from Israel and the Astrakhan region of Russia, which have wider

geographic distributions, are genetic variants of R. conorii. There are reports of SFG

rickettsioses in southern Asia for which the agents have not been determined (75, 85).

Orientia tsutsugamushi diverges from Rickettsia by approximately 10% in the rrs gene and

differs greatly in its cell wall structure, containing completely unrelated proteins and lacking

lipopolysaccharide (Table 1). Phylogeny based upon groEL, a member of the molecular

chaperone family, reveals similar genetic diversity for the unispecies genus Orientia as for

SFG Rickettsia, for which it may be that too many species have been named (Fig. 1)

(50, 99). O. tsutsugamushi, originally classified serologically, has subsequently been

analyzed genetically (40). In each geographic area there are several genetic variants.

Phylogeny reveals nine clusters that are not identical with serotypes. The genotypes do not

have strong evidence of geographic differentiation (40). Genetic variants of O.

tsutsugamushi appear to correspond to particular arthropod hosts in which divergence most

likely occurred.

DESCRIPTION OF THE GENERA Back to top

Species of Rickettsia are small (0.3 to 0.5 μm by 1 to 2 μm), obligately intracellular bacteria

of theAlphaproteobacteria with a gram-negative cell wall structure that contains

lipopolysaccharide, peptidoglycan, a major 135-kDa S-layer protein (OmpB), a 17-kDa

lipoprotein, and, for SFG rickettsiae, a surface-exposed protein (OmpA) containing different

numbers of nearly identical tandem repeat units (27, 98). Rickettsiaorganisms have small

(1.1 to 1.5 Mb), A+T-rich genomes resulting from reductive evolution with a high proportion

(19 to 24%) of noncoding sequence. Among SFG and TG rickettsiae the genomes have

remarkable synteny (57). The lack of genes for enzymes for sugar metabolism, lipid

biosynthesis, nucleotide synthesis, and amino acid synthesis and the presence of genes

encoding enzymes for the complete tricarboxylic acid cycle and several copies of ATP/ADP

translocase suggest both independent synthesis of ATP and acquisition of host ATP and

rickettsial utilization of host sources for nutrition and building blocks. Rickettsiae adhere to

the host cell receptor Ku70 by OmpB (and also by OmpA to an unknown receptor for SFG

rickettsiae), trigger signaling pathways leading to recruitment and activation of induced

phagocytosis, and escape from the phagosome by membranolytic activities of rickettsial

phospholipase D and TlyC (54, 55, 112). O. tsutsugamushi(0.3 to 0.5 μm by 0.8 to 1.5 μm)

has a 2.1-Mb genome with high proportions of genes encoding mobile genetic elements,

identical repeats, and fragmented genes and a low coding capacity (13). The organisms have

a major surface protein of 54 to 58 kDa as well as 110-, 80-, 47-, 42-, 35-, 28-, and 25-kDa

surface proteins but lack muramic acid, glucosamine, 2-keto-3-deoctulonic acid, and hydroxy

fatty acids, suggesting the absence of lipopolysaccharide and peptidoglycan. Orientia has a

more plastic gram-negative cell wall with a thicker outer leaflet and thinner inner leaflet of

the outer envelope than those of Rickettsia.

EPIDEMIOLOGY AND TRANSMISSION Back to top

Rickettsia spp. reside in an arthropod host (tick, mite, louse, flea, or other insect) for at least

a part of their life cycle, during which they are maintained by transovarian transmission

and/or cycles involving horizontal transmission to vertebrate hosts

(7, 17, 31, 46, 56, 69, 70, 92) (Table 2). Orientia resides free in the cytosol and is

maintained in nature by transovarian transmission in trombiculid mites, which transmit the

infection to humans during feeding at the larval stage (Table 2).

CLINICAL SIGNIFICANCE Back to top

In addition to Rocky Mountain spotted fever (RMSF), rickettsialpox, murine typhus, flying

squirrel-associated R. prowazekii infection, flea-borne spotted fever, and R. parkeri infection,

which are indigenous to the United States, the potential for imported cases is significant for

African tick bite fever, boutonneuse fever, murine typhus, and scrub typhus (Table 2)

(10, 18, 38, 39, 71, 77, 86, 103, 109). Other rickettsioses, either because of their

geographic distribution and the infrequency of travelers’ exposure to them or because of

their incidence, are unlikely to be imported. RMSF, louse-borne typhus, and scrub typhus are

life-threatening illnesses even for young, previously healthy persons. Murine typhus and

boutonneuse fever can have a fatal outcome in patients who are elderly or have underlying

diseases or other risk factors. A recent study of 140 patients infected with R. conorii who

were admitted to Portuguese hospitals indicated that alcoholism and infection with the Israeli

strain are risk factors for a fatal outcome. Fatal cases more frequently have acute renal

failure, hyperbilirubinemia, obtundation, tachypnea, petechial rash, gastrointestinal

symptoms, and coagulopathy (16).

An average of 7 days after tick bite inoculation of rickettsiae, patients with RMSF develop

fever, severe headache, malaise, and myalgia, frequently accompanied by nausea, vomiting,

and abdominal pain and sometimes cough (38). A rash typically appears only after 3 to 5

days of illness. Rickettsiae infect endothelial cells, frequently leading to increased vascular

permeability and focal hemorrhages. In severe cases, noncardiogenic pulmonary edema and

rickettsial encephalitis with coma and seizures are grave conditions that often presage death

(30, 102).

Rickettsia parkeri causes a milder illness with tick inoculation site eschar, fever, headache,

myalgia, usually a maculopapular or vesiculopapular rash, less frequently tender regional

lymphadenitis, and no reported deaths (66). Rickettsialpox has been recognized mainly as a

nonfatal urban disease with disseminated vesicular rash and an eschar at the location of

rickettsial inoculation by the feeding mite (39). The complete spectrum of clinical

manifestations of R. felis infections has yet to be determined. This disease suffers from

diagnostic neglect despite its widening recognized geographic distribution and the prevalence

of cat flea exposure (86,96, 116, 119, 120).

Murine typhus causes a rash in only slightly more than one-half of patients, cough and chest

radiographic infiltrates suggesting pneumonia in many patients, and in some patients severe

illness with seizures, coma, and renal and respiratory failure necessitating intensive care unit

admission in 10% of hospitalized cases (18).

Travelers who have returned from Africa and develop fever, one or more eschars, and, in

some cases, regional lymphadenopathy and a maculopapular or vesicular rash are very likely

infected with R. africae (77).

Rickettsia prowazekii, R. rickettsii, R. typhi, and R. conorii are bioterror threats via aerosol

exposure to organisms that are infectious at a low dose (100).

Scrub typhus caused by O. tsutsugamushi occurs in the geographic area that is bordered by

Japan, Korea, and Russia on the north, Australia and Indonesia on the south, Pakistan and

Afghanistan on the west, and the Philippines and Micronesia on the east (90). Clinical signs

and symptoms of the disease include fever, headache, maculopapular rash, eschar,

interstitial pneumonia, temporary deafness, lymphadenopathy, and central nervous system

involvement (11, 74, 88, 90, 106). Without treatment, mortality can reach up to 30%

(11, 51, 95, 106).

COLLECTION, TRANSPORT, AND STORAGE OF

SPECIMENS Back to top

Blood should be collected as early as possible in the course of illness. For the isolation of

rickettsiae, blood should be obtained in a sterile heparin-containing vial prior to the

administration of antimicrobial agents that are active against rickettsiae (38, 48). For

isolation and immunocytologic diagnosis, blood may be stored temporarily at 4°C and should

be processed as promptly as possible. If inoculation of cell culture or animals must be

delayed for more than 24 h, plasma, buffy coat, whole blood, or biopsied tissue should be

frozen rapidly and stored at −70°C or in liquid nitrogen. EDTA- or sodium citrateanticoagulated

blood collected in the acute state has been used effectively for the diagnosis

of boutonneuse fever, murine typhus, epidemic typhus, Japanese spotted fever, scrub

typhus, and, with lower sensitivity, RMSF and African tick bite fever by PCR

(23, 77, 84, 86, 91, 118). PCR provides a higher diagnostic yield when applied to biopsy

specimens of rickettsia-infected lesions, particularly eschars (5, 16, 44, 61, 65). If whole

blood, plasma, buffy coat, or tissue cannot be processed for PCR within several days, it

should be stored at -20°C or lower. Serum has a lower sensitivity for PCR but is often

diagnostic in fatal cases (52, 62).

For serologic diagnosis, blood is collected as early in the course of disease as possible, a

second sample is collected after 1 or 2 weeks, and if a fourfold rise in antibodies has not

occurred, a third sample is collected 3 or 4 weeks after onset. The serum may be stored for

several days at 4°C but should be stored frozen at -20°C or lower for longer periods to avoid

degradation of the antibodies. However, blood samples collected by finger stick on

appropriate blotting paper in remote areas and sent by ordinary mail can be used for

serologic diagnosis (20, 72). Even after transport at ambient temperature, this collection

method yields a serologic sensitivity similar to that of testing of fresh serum for indirect

immunofluorescence assay (IFA) diagnosis of scrub typhus (72).

A 3-mm-diameter punch biopsy specimen of a skin lesion, preferably a maculopapule

containing a petechia or the margin of an eschar, should be collected as soon as possible

(59, 101). Although treatment should not be delayed, it is best to perform the biopsy prior to

the completion of 24 h of treatment with a tetracycline or chloramphenicol. For

immunohistologic detection of rickettsiae, the specimen can be snap-frozen for frozen

sectioning or fixed in formaldehyde for the preparation of paraffin-embedded sections

(27, 38, 39, 59,101–105). The former approach yields an answer more rapidly, but freezing

artifacts distort the architecture of the tissue, and fixed tissue is more convenient for

shipping to a reference laboratory. Aseptically collected autopsy tissues, e.g., spleen and

lung, are useful for rickettsial isolation, ideally inoculated fresh, or held for 24 h at 4°C or

stored frozen at −70°C for longer periods if the specimen must be shipped to a public health

or reference laboratory. Autopsy tissues can also be examined for rickettsiae by

immunohistochemistry or PCR.

Body lice (Pediculus humanus corporis) removed from the clothing of patients with suspected

epidemic typhus can be examined for the presence of rickettsiae. Body lice acquire

rickettsiae and remain infected for life, thus providing a useful specimen for PCR diagnosis

even after a prolonged period of shipping at ambient temperature and humidity, which do

not ensure survival of the lice (82).

DIRECT DETECTION Back to top

General Considerations

Molecular and immunohistochemical diagnostic tests, the most useful methods for

establishing a diagnosis during the acute stage of illness (when therapeutic decisions are

critical), are available at this time, to the best of our knowledge, in only a few reference

laboratories, including ours. Individual cases for immunohistochemistry may be referred to

the following laboratories after contacting the directors for consultation: David H. Walker,

Department of Pathology, University of Texas Medical Branch, 301 University Blvd., Keiller

Building, Room 1.116, Galveston, TX 77555-0609 [telephone, (409) 772-3989 ;

fax, (409) 772-1850; e-mail, dwalker@utmb.edu]; J. Stephen Dumler, Division of Medical

Microbiology, Department of Pathology, The Johns Hopkins Medical Institutions, Meyer B1-

193, 600 North Wolfe St., Baltimore, MD 21287 [telephone, (410) 955-8654 ;

fax, (410) 287-3665; e-mail, sdumler@jhmi.edu]; Sherif Zaki, Infectious Disease Pathology

Branch, National Center for Emerging and Zoonotic Infectious Diseases, Centers for Disease

Control and Prevention, 1600 Clifton Rd., Mail Stop G-32, Atlanta, GA 30333

[telephone, (404) 639-3133 ; e-mail, sxz1@cdc.gov]; and Robert Massung,

Rickettsial Zoonosis Branch, Centers for Disease Control and Prevention, 1600 Clifton Rd.,

Mail Stop G-13, Atlanta, GA 30333 [telephone, (404) 639-1082 ; e-mail,

rmassung@cdc.gov].

Immunologic Detection

The diagnoses of RMSF, R. parkeri infection, boutonneuse fever, African tick bite fever,

murine typhus, louse-borne typhus, and rickettsialpox have been established by

immunohistochemical detection of rickettsiae in formalin-fixed, paraffin-embedded sections

of biopsy specimens of rash and eschar lesions (30, 38, 61, 101,103, 105) (Fig. 2, left).

Monoclonal antibodies that are specific for lipopolysaccharides of either SFG or TG rickettsiae

have been used to detect rickettsiae by immunohistochemical staining (Fig. 2, right). There

is no antibody commercially available (103, 104). The sensitivity and specificity of

immunohistochemical detection ofR. rickettsii in cutaneous biopsy specimens are 70 and

100%, respectively (38, 101). Eschar biopsies yield sensitive specimens for the diagnosis of

SFG rickettsioses that manifest that lesion and should be considered for diagnostic evaluation

for patients suspected to have rickettsialpox, R. parkeri infection, boutonneuse fever, or

African tick bite fever. Because histologic processing results in heat denaturation of speciesspecific

antigens, immunohistochemistry only distinguishes Rickettsia at the SFG or TG level.

Immunocytochemical detection of R. conorii in circulating endothelial cells has been

accomplished by capture of the endothelial cells from blood samples using magnetic beads

coated with a monoclonal antibody to a human endothelial cell surface antigen followed by

immunofluorescent staining of the intracellular rickettsiae (48). This method has a sensitivity

of 50% and a specificity of 94%. Rickettsiae are detected in 56% of untreated patients and

in 29% of patients receiving antirickettsial treatment. O. tsutsugamushi has been identified

using immunohistochemistry in human endothelial cells, macrophages, and cardiac myocytes

(60). The technique of in situ hybridization has been developed but has not been reported for

the detection of rickettsiae in clinical samples.

Molecular Detection

PCR has been applied to the amplification of the DNA of R. rickettsii, R. parkeri, R. conorii, R.

japonica, R. typhi, R. prowazekii, R. africae, R. sibirica, R. felis, R. akari, R. slovaca, and O.

tsutsugamushi, usually from peripheral blood, buffy coat, or plasma but occasionally from

fresh, frozen, or paraffin-embedded tissue or arthropod vectors from patients

(14, 23, 53, 65, 77, 82, 84, 86, 87, 89, 97). Nested PCR applied to skin biopsy specimens,

particularly of eschars prior to treatment, has a sensitivity of 78% (22). For all

pathogenicRickettsia spp., the 17-kDa lipoprotein gene is a target, employing the primers

CAT-TACTTGGTTCTCAATTCGGT and GTTTTATTAGTG-GTTACGTAACC, which amplify a 231-bp

DNA fragment (86). The gltA, rrs, groEL, ompA, andompB genes have also been amplified

diagnostically, with the Rickettsia being identified through either restriction fragment length

polymorphism analysis using AluI and XbaI or sequencing of the PCR product. The

availability of rickettsial genome sequences offers the possible design of an enormous

number of primer sets. The approach of using primer sets on a single occasion to reduce the

chances of amplicon contamination and false-positive results seems impractical. With batch

processing, the delay in laboratory results reduces the clinical value (22). The single use of

primers requires that their utility and sensitivity be unknown. The potential for amplicon

contamination originating from a positive patient sample remains even if there is no positive

control. Recent advances in technology, such as real-time PCR, allow for increased sensitivity

in the detection of rickettsiae (43, 47, 63, 67, 68, 73, 93, 97, 114). The advantage of realtime

PCR is detection of rickettsial organisms during the early or acute phase of disease

before the generation of antibody. The targets for primer design have ranged from

housekeeping genes (gltA) to antigen genes (ompA and ompB).The sensitivities for detection

vary among primer sets. For example, real-time assays utilizing primer set RR.190.547F

(CCTGCCGATAATTAT-ACAGGTTTA) and RR.190.701R (GTTCCGTTAATG-GCAGCAT), which

generates a product of 154 bp, can detect five copies of rickettsial DNA. The primer set CS-5

and CS-6 detects 1 copy ofR. rickettsii DNA and 10 copies of R. bellii DNA. Perhaps the best

potential demonstration of real-time PCR as a diagnostic assay was observed in a recent

study that compared real-time PCR evaluation with serology (114). In this study, primer set

PanRick_2_for (5′-ATAGGACAACCGTTTATTT-3′) and PanRick_2_rev (5′-

CAAACATCATATGCAGAAA-3′) and a probe, PanRick_3_taq (5′-FAMCCTGATAATTCGTTAGATTTTACCG-

TMR-3′), targeting a 70-bp region of the

rickettsial gltA gene, were utilized for diagnosis of a febrile returned traveler who also

presented with a macular rash and an eschar on his leg. DNA was extracted from a small

biopsied sample of the eschar and the leukocyte layer of the EDTA-treated blood samples

collected from the patient and analyzed by real-time PCR. The assay was able to detect

1,476 copies of PCR target per ml (1.4 copies per μl) from the initial patient sample. The

acute-phase serum of the patient had a positive IFA titer for SFG immunoglobulin M (IgM)

antibodies of 64 and no IgG antibodies. Intracellular bacterial growth was observed in

Gimenez-stained Vero cells on day 5 after inoculation. An IgG seroconversion was detected

using convalescent-phase sera obtained 4 weeks later. This rapid (total processing time is 3

to 4 h) molecular approach is currently being evaluated further for its effectiveness as a

rapid diagnostic tool (114).

For O. tsutsugamushi, the 56-kDa protein gene is the usual target of diagnostic PCR. The

nested 56-kDa gene PCR assay is useful for the diagnosis of O. tsutsugamushi infections

during the acute phase of the disease (84). PCR amplification of DNA from blood samples,

using primers p34 (TCAAGCTTATTGCTAGTGCAATGTCTGC) and p55 (AGGGAT

CCCTGCTGCTGTGCTTGCTGCG), which generate a 1,003-bp product, followed by nested PCR

with primers p10 (GATCAAGCTTCCTCAGCCTACTATAATGCC) and p11

(CTAGGGATCCCGACAGATGCACTATTAGGC), yields a 483-bp product. This assay detects 10

prototype strains of O. tsutsugamushi and does not amplify R. typhi or R. honei. Real-time

PCR assays utilizing the Orientia htrA gene have been applied to diagnosis with clinical

samples (35, 89), and the most sensitive levels of detection (two copies per μl of buffy coat

extracted DNA) have been observed when using the groEL gene as a target for analysis of

blood collected from 61 Thai patients at the time of clinic admission (67). In this assay, a set

of primers (forward primer, 5′-TTGCAACRAATCGTGAAAAG-3′, and reverse primer, 5′-

TCTCCGTCTACATCATCAGCA-3′) were used to amplify a 459-bp fragment of the gene. This

method shows promise as a diagnostic tool for the acute stage of scrub typhus. The

development of a loop-mediated isothermal PCR targeting groEL of O. tsutsugamushi offers

the possibility of a simple method that can be used in locations lacking costly infrastructure

(68).

ISOLATION PROCEDURES Back to top

Due to their high infectivity at a low dose, rickettsial isolation is performed in biosafety level

3 laboratories. Cumbersome historic methods, such as inoculation of adult male guinea pigs,

mice, or the yolk sac of embryonated chicken eggs, have been supplanted by cell culture

methods, except for isolation of O. tsutsugamushi, which is often achieved by intraperitoneal

inoculation of mice (38, 48). Vero, L-929, HEL, and MRC5 cells have been used in antibioticfree

media to isolate rickettsiae. The best results are achieved with heparin-anticoagulated

plasma, buffy coat, or skin lesion biopsy specimens collected prior to administration of

antirickettsial therapy.

Samples containing 0.5 ml of triturated clinical material mixed with 0.5 ml of tissue culture

medium are inoculated as promptly as possible onto 3.7-ml shell vials with 12-mm-diameter

round coverslips having a confluent layer of cells and centrifuged at 700 × g for 1 h at room

temperature to enhance attachment and entry of rickettsiae into host cells (3, 48). After

removal of the inoculum, the shell vials are washed with phosphate-buffered saline and

incubated with minimal essential medium containing 10% fetal calf serum in an atmosphere

containing 5% CO2 at 34°C. At 48 and 72 h, a coverslip is examined by Giemsa or Gimenez

stain or by immunofluorescence with antibodies against SFG and TG rickettsiae. Detection of

four or more organisms is interpreted as a positive result. This method has yielded a

diagnosis in 59% of samples from patients with boutonneuse fever who had neither been

treated nor developed antibodies to R. conorii prior to collection of the sample (48).

Rickettsiae were detected at 48 h of growth in 82% of the positive samples. Universal

precautions should be exercised, and work should be performed in a laminar-flow biosafety

hood with use of gloves, mask, and gown. Although the quantity of rickettsiae in the cell

culture is relatively low, care to avoid aerosol, internal, or contact exposure should be taken

as for mycobacteria, fungi, and viruses.

IDENTIFICATION

OF RICKETTSIA AND ORIENTIA ISOLATES Back to top

Rickettsiae isolated in cell culture can be identified by indirect immunofluorescence with

group-, species-, and strain-specific monoclonal antibodies. Rickettsial isolates are frequently

identified by molecular methods such as PCR amplification of genes that are genus specific

(17-kDa protein, citrate synthase [gltA], or ompB) or SFG specific (ompA) (80).

Determination of DNA sequences identifies unique isolates. O. tsutsugamushi, being more

distantly related to Rickettsia spp., lacks the above-mentioned cell wall genes but can be

identified by PCR of the gene encoding the major immunodominant 56-kDa surface

protein, groEL, or rrs (44, 68).

Identifying the species of rickettsial isolates by microimmunofluorescence serotyping requires

intravenous inoculation of mice with rickettsiae on days 0 and 7 and collection of sera on day

10. The high-titer antibodies react with conformational species-specific epitopes of OmpA

and OmpB. Antibodies against group-specific lipopolysaccharide develop later in the murine

immune response to high doses of Rickettsia. This rather cumbersome and expensive

method requires propagation of large quantities of the isolate and of the prototype strains for

immunofluorescence titration as well as for development of the typing sera. Genetic analysis

is currently favored for identification of isolates. An isolate that is known to be

a Rickettsia should be identified in a biosafety level 3 laboratory, and isolates of R.

rickettsii and R. prowazekii must be handled as required by U.S. federal regulations for select

agents.

SEROLOGIC TESTS Back to top

For most clinical microbiology laboratories, assays for antibodies to rickettsiae are the only

tests performed. This situation is unfortunate for the patient with a life-threatening, acutely

incapacitating rickettsial disease because these assays are useful principally for serologic

confirmation of the diagnosis in convalescence and usually do not provide information that is

helpful in making critical therapeutic decisions during the acute stage of illness. Patients who

die of rickettsioses usually have received many antibiotics, none of which have antirickettsial

activity owing in part to the lack of laboratory data providing clinical guidance for a rickettsial

diagnosis. The earlier a diagnosis is established, the shorter the course of rickettsial illness

after an appropriate antirickettsial antibiotic is administered.

Serologic assays for the diagnosis of rickettsial infections focus on the “gold standard,” the

IFA. Other approaches include indirect immunoperoxidase assay, latex agglutination, enzyme

immunoassay (EIA), Proteus vulgaris OX-19 and OX-2 and Proteus mirabilis OX-K

agglutination, line blot, Western immunoblotting, and rapid lateral flow assays

(8, 12, 15, 18, 19, 28, 33, 36–39, 41, 42, 77, 101, 110, 113). Only some of these assays

are available as commercial kits or in reference laboratories, and not for all rickettsial

diseases. Other serologic tests, such as indirect hemagglutination, microagglutination, and

complement fixation, are no longer in general use.

The IFA contains all the rickettsial heat-labile protein antigens and group-shared

lipopolysaccharide antigen and thus provides group-reactive serology. IFA reagents are

available commercially for SFG and TG rickettsiae from Scimedx Corp., Denville, NJ; Focus

Technologies, Cypress, CA; and Fuller Laboratories, Fullerton CA; they are also available

for O. tsutsugamushi from Scimedx Corp. In cases of RMSF, IFA detects antibodies at a titer

of ≥64, usually in the second week of illness. Effective antirickettsial treatment of RMSF

must be initiated by day 5 of illness to avoid a potentially fatal outcome. For boutonneuse

fever, an IFA titer of ≥40 occurs in 46% of patients between days 5 and 9 of illness, in 90%

of patients between 20 and 29 days, and in 100% of patients thereafter. In murine typhus,

diagnostic IFA titers are present in 50% of cases by the end of the first week of illness and in

nearly all cases by 15 days after onset (18). In areas where particular rickettsial diseases are

endemic, a higher diagnostic cutoff titer is required. For example, for the IFA diagnosis of

scrub typhus in patients residing in zones of endemicity, an IFA titer of antibody to O.

tsutsugamushi of ≥400 is 96% specific and 48% sensitive, with sensitivity rising from 29%

in the first week to 56% in the second week (4). Lowering the diagnostic cutoff titer to 100

raises the sensitivity only to 84% and reduces the specificity to 78%. These considerations

are not as important when testing patients who have visited regions of endemicity for only a

short period. Each laboratory performing the test should establish its own cutoff titers for the

patient population, the microscope and reagents used, and the laboratorian ’s judgment of

the minimal positive signal. Meta-analysis of IFA serology for the diagnosis of scrub typhus

led to recommendations that diagnosis be based on a fourfold or greater rise in titer and that

diagnosis not be based on a single antibody titer unless previous studies had determined the

seroprevalence in the local population justifying the cutoff titer (4).

Indirect immunoperoxidase assays for scrub typhus, murine typhus, boutonneuse fever, and

presumably other rickettsioses yield results similar to those of IFA when the IgG diagnostic

titer is set at 128 and that of IgM is set at 32 (42). Advantages include the use of a light

microscope rather than a fluorescence microscope and the production of a permanent slide

result. Latex agglutination test reagents are available commercially from Scimedx Corp., only

for R. rickettsii in the United States. Latex beads coated with an extracted rickettsial proteincarbohydrate

complex containing rickettsial lipopolysaccharide are agglutinated mainly by

IgM antibodies, with reports of a sensitivity of 71 to 94% and a specificity of 96 to 99% (28).

A diagnostic titer of 128 is often detected early in the second week of illness.

EIAs have been developed in various formats, including antigens coating microtiter wells or

immobilized on nitrocellulose or other sheets for use in the commercial reference laboratory

setting. Dot EIA kits are commercially available in the United States from Scimedx Corp. for

detecting antibodies against R. rickettsii, R. conorii, R. typhi, and O.

tsutsugamushi (41, 110). No peer-reviewed publications have described evaluation of the

use of the dot EIA for the diagnosis of RMSF. Compared with an IFA titer of ≥64 for the

diagnosis of murine typhus, the dot EIA showed a sensitivity of 88% and specificity of 91%

(41). The dot EIA for diagnosis of scrub typhus had sensitivities and specificities of only 80

and 77%, respectively, compared with an IFA cutoff titer of 64, and 89 and 66%,

respectively, at an IFA cutoff titer of 128 (110). These SFG rickettsia and R. typhi kits detect

cross-reactive antibodies, as demonstrated in an outbreak of African tick bite fever by clinical

and epidemiological data. The dot EIA of R. conorii antigen provided early diagnostic

evidence of an SFG rickettsiosis. Subsequent analysis revealed poor specificity with a high

rate of false-positive results. These assays are diagnostic tools that do not require expensive,

specialized equipment, but they suffer from apparent low specificity. Standard EIAs for

detecting IgG or IgM antibodies against SFG rickettsiae or O. tsutsugamushi are also

available from Panbio Diagnostics internationally but are not currently available for purchase

in the United States. The utilization of these tests for paired sera from populations with

clinical (fever, headache, or rash) and epidemiological (vector exposure) features consistent

with rickettsiosis would most likely yield useful information. A multitest dot EIA for scrub

typhus, murine typhus, and leptospirosis was not useful in Thailand, where nearly all patients

had antibodies to more than one agent and others had antibodies to an agent that was not

the cause of the illness (108). The diagnosis of scrub typhus by detection of antibodies in

clinical samples in an IgM capture enzyme-linked immunosorbent assay can be utilized for

single serum samples from early-stage infections (33). The performance values of this assay

are a sensitivity of 96.3% and a specificity of 99%. A comparison of serologic methods for

scrub typhus revealed that an EIA containing recombinant p56 of Karp, Kato, and Gilliam

strains was most sensitive (100%) and that rapid lateral-flow assay of antibodies against

recombinant Karp p56 had a sensitivity of 86% (33). The latter, which is especially useful in

situations with limited laboratory facilities and a low number of specimens, is available from

Panbio Ltd., Brisbane, Australia.

The assays that were historically most widely used for the diagnosis of rickettsial diseases

are agglutination of the OX-19 and OX-2 strains of Proteus vulgaris for TG and SFG

rickettsioses and the OX-K strain of Proteus mirabilis for O. tsutsugamushi infections. These

assays have poor sensitivity and specificity (37, 101). They should be replaced by more

accurate serologic methods such as IFA or EIA. However, there are situations in developing

countries where the choice is between the Proteus agglutination tests and none at all for the

detection of important public health problems such as outbreaks of louse-borne typhus (1).

In fact, the evidence leading to the discovery of Japanese spotted fever and Flinders Island

spotted fever includedProteus agglutinating antibodies.

Shared antigens of OmpA, OmpB, and group-specific lipopolysaccharide impede

establishment of a species-specific diagnosis by serologic methods. The criterion of a fourfold

or greater difference in IFA titers between the two suspected agents distinguished infections

by R. prowazekii and R. typhi in only 34% of cases and infections by R. africae from those

by R. conorii in only 26% (49, 77). Western immunoblotting detection of antibodies against

OmpA or OmpB of only one Rickettsia species has also been proposed as a criterion for

species-specific diagnosis. However, it was effective in distinguishing R. prowazekii and R.

typhi infections or R. africae and R. conorii infections in only one-half of the cases

(34, 49, 77). Cumbersome, expensive cross-absorption of sera prior to IFA or Western

immunoblotting is more effective in establishing a species-specific diagnosis. However,

interpretation of these results requires careful evaluation of the performance of valid

controls, the quality and quantity of each antigen preparation, and the potential for the

occurrence of infection by an untested, even as-yet-discovered, agent. In the past,

knowledge of the geographic origin of the case sufficed to designate the specific diagnosis.

However, the increasing number and geographic overlap of rickettsioses challenge the old

assumptions. The report of the reactivity of sera from patients with flea-borne spotted fever

but not patients with RMSF, rickettsialpox, or murine typhus with a recombinant fragment

of R. felis OmpA suggests that species-specific peptide antigens may be identified and

incorporated into assays that identify the disease more precisely (117).

ANTIMICROBIAL SUSCEPTIBILITY TESTING Back to top

Data supporting the use of doxycycline or another tetracycline antibiotic as the drug of

choice for the treatment of infections caused by Rickettsia spp. and O. tsutsugamushi and

the use of chloramphenicol as an alternative drug have been derived principally by empirical

experience, retrospective case studies, and a few prospective studies

(2, 9, 25, 29, 45, 58, 79, 81, 83, 94). In addition to historic studies of the activity of

antimicrobial agents against these obligately intracellular bacteria in infected animals and

embryonated eggs, studies of the effects of antimicrobial agents in cell culture have

supported the consideration of alternative drugs such as fluoroquinolones, josamycin,

azithromycin, and clarithromycin. Indeed, several fluoroquinolones, josamycin, and

azithromycin have been used successfully for the treatment of boutonneuse fever under

certain circumstances but cannot be recommended for more pathogenic rickettsioses

(2, 9, 58, 79). Mediterranean spotted fever has also been treated effectively in clinical trials

with fluoroquinolones such as ciprofloxacin and macrolides such as azithromycin or

clarithromycin. A retrospective study of patients with murine typhus demonstrated that

ciprofloxacin is an effective drug. Except for cases of scrub typhus in Thailand which have not

responded to doxycycline or chloramphenicol but for which azithromycin has been reported

to be effective, there is little concern regarding rickettsial development of antimicrobial

resistance (45, 94, 107). During pregnancy, chloramphenicol has been used to treat RMSF

and josamycin for boutonneuse fever. Antimicrobial susceptibility studies of rickettsiae are

not routinely performed clinical laboratory tests.

INTERPRETATION AND REPORTING OF RESULTS Back to top

When reporting the results of an assay for antibodies in a single serum sample, the

laboratorian seldom knows the duration of illness and whether the serum sample is from the

acute or the convalescent phase of the disease. For sera that are nonreactive by dot EIA, by

IFA at a dilution of 1:64, by indirect immunoperoxidase assay at a dilution of 1:128, by latex

agglutination at a dilution of 1:64, or by Weil-FelixProteus agglutination at a titer of 1:160,

the laboratory report should state that no antibodies were detected at the particular cutoff

dilution, which may differ among some laboratories and some patient populations, that

negative results are expected in the acute stage of rickettsial illness, and that a second

sample should be submitted to evaluate the possibility of seroconversion if no alternative

diagnosis has been established. If paired acute- and convalescent-phase sera separated by

an appropriate interval are available, they should be tested simultaneously. It is wise to test

for all the rickettsial and ehrlichial agents to which the patient is likely to have been exposed

in the United States. SFG rickettsiae, Ehrlichia chaffeensis, Anaplasma

phagocytophilum,and R. typhi are the likely agents unless travel to an area where scrub

typhus is endemic has occurred. If the paired sera are negative, the report should state that

the results do not support the diagnosis of rickettsial infection but that occasionally antibody

synthesis is delayed, particularly in cases with early antirickettsial therapy. If a single serum

sample contains an IFA antibody titer of ≥64, an IgM IFA titer of ≥32, an indirect

immunoperoxidase antibody titer of ≥128, a latex agglutination titer of ≥64, or a Weil-Felix

titer of ≥320, the laboratory report should state that antibodies reactive with the particular

rickettsial antigen were detected at the measured titer, that the result provides supportive

evidence for the diagnosis of the rickettsial disease, and that a convalescent-phase sample

should be submitted to assess the possibility of a diagnostic rise in titer. If paired sera

measured simultaneously show a fourfold or greater rise in titer, the interpretation is stated

that the results strongly support the rickettsial diagnosis indicated by the tested antigen. If a

significant titer was detected in the acute-phase sample, but no rise or only a single doubling

dilution rise was measured, it should be stated that an additional later sample should be

tested to evaluate a fourfold rise or fall in titer. The concept that recrudescent typhus could

be distinguished from primary louse-borne typhus by the absence of IgM antibodies to R.

prowazekii and of Proteus OX-19 agglutinating antibodies has been challenged (19). The

manufacturers of the dot EIA have recommended the interpretation that strongly reactive

samples (three or four dots) may indicate the presence of a specific antibody response and

that weakly reactive samples (one or two dots) are infrequent but possible in normal

populations. Retesting 2 to 3 weeks later would establish the diagnosis if three or four dots

develop in the convalescent-phase serology and should always be performed.

Isolation of a rickettsia from blood or tissue may be interpreted as indicating an etiologic

role. The level of identification of the isolate should be stated, whether identified only as to a

group containing particular organisms or to the species level.

Immunohistologic and immunocytologic diagnostic interpretation states the method,

reactivity of the method (e.g., antibody reactive with SFG rickettsiae), and location of the

antigen (e.g., in vascular endothelium and frequently adjacent vascular smooth muscle for R.

rickettsii). Detection of three or more rickettsiae in vascular endothelium in biopsy specimens

or four or more rickettsiae in captured circulating endothelial cells is diagnostic of rickettsial

infection.

Interpretation of PCR results should state the target gene, the organisms that would be

detected, and the presence or absence of a DNA product. If a specific oligonucleotide probe

or DNA sequencing confirmed the specificity of the identification, this result should be stated.

For negative immunohistologic, immunocytologic, and PCR results, it should always be stated

that the failure to detect the agent does not exclude the diagnosis, along with data regarding

the sensitivity and specificity of the assay in the particular laboratory and the effects of

antirickettsial treatment on the sensitivity.

Special efforts should be made to establish the diagnosis of fatal cases, including rickettsial

isolation, immunohistology, PCR, and serology on samples collected at necropsy.