Bull. Org. mond. Santd 11973, 49, 433-442Bull. Wld Hlth Org. J

Rickettsiae and rickettsial diseasesR. BREZINA,1 E. S. MURRAY,2 M. L. TARIZZO,3 & K. BOGEL 4

This paper summarizes present knowledge on rickettsiae and rickettsial diseases, and ontheir epidemiological characteristics, control, and public health significance. There are manynatural foci of rickettsial diseases, from where the disease may spread to other areas inthe world under changing socioeconomic conditions. Because of rapid long-distance travel,sporadic cases ofserious rickettsial diseases may today appear farfrom endemic areas wherethe infection occurred. Even in endemic areas the disease may be misdiagnosed and deathsmay occur as a result of inadequate treatment. Rapid treatment of rickettsial infections(preferably with tetracyclines) is therefore most important. Epidemic louse-borne typhus,though no longer subject to the International Health Regulations, remains one of the diseasesin the WHO epidemiological surveillance programme. This disease continues to be a majorcause ofmorbidity and mortality in some parts ofAfrica and it is present also in parts of theAmericas and ofAsia. Scrub typhus remains a continuing and serious public health problemin areas ofSouth-East Asia and in the Western Pacific. The annual number ofreported casesof Rocky Mountain spotted fever in the USA showed an increase during the last twodecades, which may be due to improved recognition as well as to increased outdoor activitiesand migration ofpeople from the city centres to the suburbs. Related forms of tick-bornetyphus occur in South America, the Mediterranean region, Africa, South-East Asia, the FarEast, and the Western Pacific. Increasing in number, though still sporadic, are reports ofserious illness from chronic Q fever infection in many parts of the world.

After a rather long period of stagnation, rickettsialdiseases are now again the subject of interest and newapproaches to their study are being developed, inpart because of the recognition of rickettsial diseaseas a persisting and important cause of morbidity andmortality in certain countries. This paper is a synopsisof present knowledge on the epidemiological charac-teristics, control, and public health significance ofrickettsiae and rickettsial diseases.

HOST RANGE AND PATHOGENICITY

Rickettsiae are generally transmitted to man byarthropod vectors, the only exception being Coxiellaburneti, the agent of Q fever.

Infected ticks and mites transmit rickettsiae pri-

1 Head, Department of Rickettsiae, Institute of Virology,Bratislava, Czechoslovakia (WHO Collaborating Centre forRickettsial Reference and Research).

' Professor of Microbiology, Department of Micro-biology, Harvard School of Public Health, Boston, Mass.,USA.

' Medical Officer, Virus Diseases, WHO, Geneva, Switzer-land.

' Veterinary Public Health, WHO, Geneva, Switzerland.

marily through their infected salivary glands directlyinto the bite. In lice and fleas, rickettsiae do notinfect the salivary glands but multiply in the cells ofthe gut wall and are excreted in the faeces in largeamounts, contaminating the skin area around thebite. Rickettsiae may gain entry into the bodythrough bites or other breaks in the skin, or driedaerial suspensions of them may be inhaled or mayinfect the conjunctival mucosa. Dried tissue fluids orthe excreta of mites and ticks may also be a source ofinfection.Once rickettsiae have penetrated the skin, they

multiply inside cells, reach the blood stream, andeventually become located in the endothelial cells ofsmall veins, arteries, and capillaries. These infectedendothelial cells enlarge, degenerate, and causethrombus formation with partial or complete occlu-sion of the vascular lumen. Mononuclear leucocytesand plasma cells infiltrate the adventitia. Focalnecrosis of the walls of small vessels leads to in-creased permeability, rupture, petechiae, and hae-morrhages. The rupture of vessels and thrombosisinterrupt the blood supply, causing macro- andmicro-infarcts and necrosis of the tissues.

3138 - 433

R. BREZINA ET AL.

Localized cellular responses lead to the formationof nodules, commonly in the spleen, liver, kidneys,brain, heart, and skeletal muscles. Acute nonsup-purative myocarditis, both diffuse and local, is acommon occurrence. Apart from vascular damagecaused by the organism or its toxin, other lesions arerare. Rickettsiaemia is present throughout most ofthe acute febrile period. The overall pathologicalfeatures are similar in most rickettsial diseases;however, the extent and degree of involvement varyand in general are indicated by the severity of theillness.

Fatalities occurring early in the acute phase of theillness are generally due to vascular collapse; whenthey occur later, they are related to chronic vasculardamage producing thrombosis and necrosis.The long survival of rickettsiae in various organs

and lymphatic tissues after infection of both man andanimals is a general characteristic in the pathogenesisof rickettsial disease.

Classical typhus group

This group includes Rickettsia prowazeki andR. mooseri (R. typhi), respectively the agents oflouse-borne (epidemic) typhus and of flea-borne ormurine (endemic) typhus. R. canada, antigenicallyrelated to this group, has recently been incriminatedas a possible cause of infection in man (1).

Louse-borne typhus is essentially a human disease,and flea-borne typhus is a disease of rats occasionallyinvolving man. Domestic cats and the cat flea,Ctenocephalides felis, have also been incriminated inthe transmission of murine typhus.

Contradictory evidence has been put forward bysome authors with regard to the possible role ofdomestic animals in maintaining the chain of infec-tion with R. prowazeki.

Several animal species have been shown to besusceptible to infection with R. prowazeki andR. mooseri but only a few have practical use aslaboratory models. The chick embryo is by far themost important of these and yolk sac cultures areroutinely used for producing antigens and vaccines.White mice are useful in testing for and in titratingthe toxin content of yolk sac preparations.

Guinea-pigs are somewhat less sensitive and lesspredictable than chick embryos and mice. However,they can be used for primary isolation. Intraperi-toneal injection of infected human blood into guinea-pigs leads to rickettsiaemia and a characteristic feverpattern. R. prowazeki can be isolated from the blood,

brain, suprarenal gland, or spleen of infected guinea-pigs.

Cotton rats and various species of African gerbilsare highly sensitive to R. prowazeki and are ideal forisolation or research studies.

Spottedfever groupThis group of rickettsiae, the agents of various

forms of infection in man, has been divided into4 subgroups on the basis of antigenic characteris-tics (8). Subgroup A consists of R. rickettsi andR. siberica, the agents of Rocky Mountain spottedfever (RMSF) and of north Asia tick typhus; sub-group B is formed by R. conori, the agent of fievreboutonneuse, and by R. parkeri, which causes infec-tion in guinea-pigs but has not been identified as apathogen for man; subgroup C consists of R. akariand of R. australis, the agents of rickettsialpox(transmitted by mites, not by ticks) and of Queens-land tick typhus; subgroup D includes only R. mon-tana, a new species isolated from Dermacentor ticksin eastern Montana (USA) and not yet incriminatedas a pathogen for man. All these rickettsiae multiplyin the cytoplasm and in the nucleus of infected cells.

Rickettsiae are transmitted transovarially by ticks,which therefore represent both the reservoir and thevector of the agent. The infection may be transmittedto the vertebrate host by any of the larval stages andby the adult tick. Ticks remain infected for life, arenot harmed by the rickettsiae, and pass them in theirfaeces. Transmission to man, however, is believed tooccur primarily through the bite since rickettsiae alsoinvade the salivary glands of the ticks. All therickettsiae of this group pass through natural cyclesin wild or domestic animals.

In man, RMSF may show very acute symptomswith intense vascular lesions. In contrast with otherforms of tick typhus, eschars rarely develop at thesite of the tick bite. The virulence of RMSF strainsmay vary considerably for man and for laboratoryanimals.

Rickettsialpox, the mildest of the rickettsial dis-eases of man, is caused by R. akari and transmittedby mites instead of ticks, and is mainly a disease ofrodents. It is characterized in man by the presence ofvesicular lesions superimposed on the initial papules.

Scrub typhusMite-borne scrub typhus, or tsutsugamushi, is

caused by R. tsutsugamushi (R. orientalis), and istransmitted by a number of trombiculid mites thatact both as vector and as reservoir, thus playing a

434

RICKETTSIAE AND RICKETTSIAL DISEASES

role similar to that of ticks in the transmission of theetiological agent; transovarial transmission has alsobeen demonstrated. While the role of rodents andother animals is yet to be clarified, the variation inthe virulence and antigenicity of strains is consider-able.

Q feverQ fever, caused by Coxiella burneti, is primarily a

zoonosis solidly established in domestic livestock theworld over; it is a widely disseminated chronicinfection involving many organs of the infectedhosts. Of primary importance to transmission isinfection of the uterus and udder tissue in femaleanimals. The animal's milk is thus infected, but ofmajor importance is the fact that at parturition theinfected products of conception contaminate theenvironment and remain for months or years as asource of infection for man, animals, and birds.Man contracts Q fever usually by inhaling infec-

tious dust generated from infected animals or theirproducts. Clinically, Q fever in man is characterizedby an interstitial pneumonia regardless of the site ofentry of the infecting organism. C. burneti can spreadthrough almost all the organs of the body formingsmall lesions of varied character showing oedema,round cell infiltration, and patchy necrosis. Occa-sionally the liver or heart may become intenselyinvolved leading to a frank hepatitis or a subacuteendocarditis with vegetations. Although pathologiclesions may be widespread both in man and animals,fatalities are rare.

Ticks are an important element in the maintenanceof the Q fever reservoir in nature and transovarialtransmission of C. burneti has been demonstrated inthese arthropods. A wide variety of ticks have beenshown to be infected and in view of the extensivehost range of C. burneti many different host-vectorcycles probably occur.Man to man transmission is rare. Recrudescence

of latent infections in man may occasionally occurand Q fever can remain latent in the tissues for2-3 years.

Strains of the Q fever agent with different degreesof virulence have been isolated.

Trench feverR. quintana, the agent of trench fever, has a very

limited host range. Besides man and the louse,several species of monkeys only have been shown tobe infectable. The attempts to infect common labora-tory animals have met with uniform failure.

Man is the only known reservoir and the bodylouse the only known natural vector. The faeces oflice become infectious 5-10 days after an infectiousmeal; after that the lice remain infectious for lifealthough they are themselves unharmed. There is notransovarial passage of R. quintana in lice.The disease in man leads frequently to a chronic

rickettsiaemia, R. quintana having been isolated fromthe blood of patients up to 2 years after an acuteattack. True relapses have occurred as long as 20years after a primary attack. The chronic rickett-siaemia and late relapses are obviously importantfactors that help give rise to trench fever epidemicsover long periods of time.

ANTIGENIC STRUCTURE

Classical typhus groupThree species belong to this group, R. prowazeki,

R. mooseri, and R. canada. R. canada can be differen-tiated from R. prowazeki and R. mooseri by the toxinneutralization test -and by the complement fixationtest with mouse antiserum (9). The relationshipbetween R. canada and the spotted fever group ofrickettsiae, however, has not yet been clarified.The nature of the soluble antigen and the antigenic

properties of various rickettsial fractions have beenreviewed (2). A thermolabile antigenic fraction capa-ble of inducing protection after a single vaccinationof animals was recently obtained from the classicalsoluble antigen. Vaccination of human beings withthis antigen provoked no local or general reactionsand was followed by the development of comple-ment-fixing, haemagglutinating, and toxin-neutraliz-ing antibodies (6).The antigenic activities of intact R. prowazeki, and

of its cell wall, protoplasm, and soluble antigens incomplement fixation (CF) and mouse protection testsdiffer in several points. The protoplasm antigensshow the greatest specificity and heat lability in theCF test. The cell wall and soluble antigens and intactrickettsiae contain more heat-stable CF antigens andtend to be more active than the protoplasm antigensin the mouse protection test. In vaccine potency tests,the soluble antigens released by ether treatment arethe most active.

Spotted fever groupThe division of this group into four subgroups has

already been mentioned in the preceding section.This differentiation was made on the basis of toxin-neutralization and complement fixation tests withmouse species-specific sera.

435

R. BRESINA ET AL.

Scrub typhus groupExcept for the mouse cross-immunity tests, which

show a close relationship among the strains of thisgroup, other procedures, such as serum protection,complement fixation, toxin neutralization, and vac-cination-challenge, reveal a considerable hetero-geneity within the species. The impossibility of de-tecting antibody induced by a variety of antigenicallydissimilar strains of R. tsutsugamushi represents aserious difficulty in the serologic diagnosis of scrubtyphus. Serotypes Gilliam, Kato, and Karp wereidentified by complement fixation tests. The strainsof R. tsutsugamushi recovered from patients can be amixture of two or more antigenically distinctivetypes. Moreover, among the strains of R. tsutsuga-mushi in Thailand, 5 distinctive antigenic types havebeen recognized thus far: Gilliam, Karp, Kato, and2 other types not previously described.

Coxiella burnetiCoxiella burneti is the only rickettsia showing

phase variation. This phenomenon is similar in manyways to the rough-smooth variation of Diplococcuspneumoniae. In nature, C. burneti has been foundonly in phase I; in the laboratory, the adaptation togrowth in chick embryos produces phase II, whichcan be converted back to phase I by subsequentgrowth in laboratory animals. The essence of thisphenomenon lies in the alternation of two main anti-gens. The strains in phase I possess both antigens anddo not show positive results in a variety of serologicreactions with sera containing only antibodies againstthe phase II antigen. The vaccines prepared from apurified suspension of phase I, when tested in guinea-pigs, have 100-300 times the potency of those derivedfrom phase Il, provided that the immunity is testedby the use of a strain in phase 1 (14). With trichloro-acetic acid it is possible to extract from coxiellae inphase I an antigen that resembles serologically thecorpuscular phase I antigen.

This trichloroacetic acid extract of coxiellae canbe used as a vaccine producing both phase II andphase I antibodies and it often produces allergic skinreactions. An antigenic component, with the serologicbehaviour of phase I, can also be extracted withphenol from purified phase I suspensions.The presence of antigen I in C. burneti is related to

its basic properties. The shift from the naturally occur-ring phase I to egg-grown phase II is accompanied bychanges in antigenic composition, immunogenicity,density, autoagglutinability, staining properties, re-sistance to phagocytosis, and virulence.

There are only a few data concerning the antigenicdifferences among phase II strains. Antigenic differ-ences among phase I strains have not yet beenobserved.

OTHER PROPERTIES OF RICKETTSIAE

MorphologyRickettsiae can usually be visualized with the

ordinary light microscope. They tend to be some-what pleomorphic, varying in size and in shape fromcoccoidal, to bacillary, to filamentous, depending onthe species and the conditions of growth.

Electron microscopy has greatly contributed toknowledge of the structure of rickettsiae. Commonto all studied so far are a trilamellar cell wall and atrilamellar cytoplasmic membrane. Both these showthe typical " unit membrane " structure of bacterialcells. Intracytoplasmic invaginations of the plasmamembrane, DNA strands, ribosomes, electrolucentspheres, vacuoles, and membranous organelles havebeen observed.

Chemical composition and metabolismComprehensive reviews of both the chemical com-

position and the metabolism of rickettsiae have beenpublished by Paretski (15), Wisseman (18), andOrmsbee (13).

Rickettsiae possess complex cell walls, containboth ribonucleic and deoxyribonucleic acids, syn-thesize their own structural and functional consti-tuents with their own enzymes, and generate high-energy phosphate bonds with energy derived fromthe oxidation of amino acids and carbohydrates.

Rickettsial cell walls closely resemble the cell wallsof Gram-negative bacteria in complexity of structureand in the presence of muramic and diaminopimelicacid components. The purified cell walls of R. mooseriand R. prowazeki contain protein, polysaccharide,and phosphorus in excess of that from nucleic acidcontamination. Further analyses show the presenceof 12-15 amino acids, including both lysine and thedistinctive diaminopimelic acid, and the presence ofthe amino sugars glucosamine and muramic acid.Teichoic acid has not been detected in R. mooseri.The DNA of C. burneti has been isolated and

characterized and found to be significantly differentfrom that of R. prowazeki. Several data suggest thatthe DNA of C. burneti possesses a double strandedstructure.Energy production is apparently accomplished

through the operation of part or all of the Krebs

436

RICKETESIAE AND RICKETISIAL DISEASES

cycle. The principal energy-yielding substrate is glu-tamate in the case of R. quintana, and pyruvate inC. burneti. Oxidation is accompanied by phosphory-lation. The electron transport system appears toutilize phosphopyridine nucleotides and cyto-chromes.There is convincing evidence that rickettsiae pos-

sess enzymatic mechanisms for the breakdown ofcarbohydrates, the formation of high-energy phos-phate bonds, and the synthesis of lipids and proteins.The host cell contribution probably includes primarysubstrates such as glutamate and pyruvate, factorssuch asATP,NAD,and coenzymeA, and amino acids.

Multiplication of rickettsiaeRickettsiae, except R. quintana, are obligate intra-

cellular parasites. It is generally accepted that rickett-siae maintain their morphological integrity duringgrowth and multiplication. Binary fission is consid-ered to be the only mode of their multiplication;Kordova and her associates (reviewed in (3)), how-ever, found " filterable particles " in C. burneti andsuggested some other form of replication. Koko-rin (7) studied various rickettsiae in tissue culturesand suggested the existence of two developmentalstages: vegetative, dividing, and moving forms on theone hand, and resting, immobile forms on the other.Resting forms can persist in the cell for a very longperiod of time, thus providing a possible explanationfor latency in rickettsial infections.

DIAGNOSTIC PROCEDURES

Laboratory methods for isolation of rickettsial strainsThe procedures for the isolation of rickettsiae are

well known. Guinea-pigs, mice, meriones, and cottonrats are the most useful animals for the isolation ofrickettsial strains from patients and animals, andfrom ectoparasites with the exception of R. quintana.Tissue or cell cultures have been used for thepropagation of various rickettsiae, but their value inprimary isolation has not been confirmed.

R. quintana can be propagated from the blood ofpatients directly on blood agar. Erythrocytes (hemin)are essential components of media; in a liquidmedium, fetal calf serum satisfied the requirement ofR. quintana for a red blood cell lysate.

Serologic methods for the study and diagnosis ofrickettsial infectionsComplement fixation test. The most frequently

used serologic method is the complement fixation

test with soluble or particulate antigens. Louse-borneand murine typhus can be differentiated by means ofwashed corpuscular suspensions of R. prowazeki andR. mooseri. R. canada, a new member of the typhusgroup (9), shares common antigens with R. prowa-zeki and especially with R. mooseri. However, thisnew agent has been shown to be different from bothR. prowazeki and R. mooseri when tested withantisera prepared in mice. The results of toxinneutralization tests confirmed that this was a newspecies of rickettsia. Several methods by means ofCF and agglutination reactions allow the differentia-tion of primary from secondary (Brill-Zinsser dis-ease) cases (10-12).

In the spotted fever group the soluble antigen ofeach strain is group-specific whereas washed antigensallow for species differentiation. However, it wasshown that antisera from white mice infected withvarious strains from the RMSF group containedspecies-specific complement-fixing antibody that didnot react with the soluble, cross-reacting CF groupantigen ofheterologous rickettsiae of the same group.

In the scrub typhus group there is a high degree ofantigenic heterogeneity among strains of R. tsutsuga-mushi. Purified or partly purified rickettsiae allow fordifferentiating between the antigenically representa-tive Gilliam, Karp, and Kato strains. Recently satis-factory antigens were prepared from infected tissuecultures.

In Q fever, complement fixation is the test mostfrequently used for identifying Q fever antibodiesalthough it is not the most sensitive method. Phase IIantibodies are of primary diagnostic value for thedetection of recent infection and phase II egg-adapted strains must therefore be used for thepreparation of antigens.

Complement-fixing phase I antibodies are of littleimportance for the diagnosis of Q fever in man asthey can only indicate past or present persistinginfection. They are also found in cases of chronicQ fever endocarditis.The CF test for trench fever provides a rapid

diagnostic procedure, which can be used in conjunc-tion with isolation of R. quintana from the patient'sblood. Both particulate and soluble antigens can beprepared from whole R. quintana propagated onblood agar.

Agglutination reaction. The microagglutination testfor typhus antibodies described by Fiset et al. (5) is avery reliable method for detecting typhus antibodies.The test appears to have the same limits of resolutionas the standard tube agglutination tests and is more

437

R. BREZINA ET AL.

sensitive than the CF test in distinguishing betweenmurine and louse-borne typhus infections.

Infections of the spotted fever group can bedetected by Giroud's slide microagglutination testwith semipurified suspensions of R. conori.

In Q fever, highly purified haematoxylin-stainedantigens are used in the method described by Fisetet al. (5). These are phase I antigens converted tophase II by treatment with trichloroacetic acid. Themethod has been further improved by Schrameket al. (17) producing phase II antigens by oxidizingthe phase I antigens with periodate. Natural phase Ifstrains cannot be used for this agglutination test asthey tend to agglutinate spontaneously and arestrongly agglutinated by normal serum.

Microagglutination on plastic plates with stainedantigens is more sensitive than CF with phase IIantigens. It is also suitable for demonstrating thephase I antibody with phase I antigen. During thepast two decades several agglutination methods havebeen elaborated for detecting Q fever antibodies: thecapillary agglutination test (CAT), the agglutination-resuspension test, and the radioisotope precipitationtest (RIPT).

Other tests. These are useful in particular instancesand include the following.1. The passive haemagglutination test (PHT).

This test, which requires sheep or human group 0erythrocytes sensitized with ESS (erythrocyte sensi-tizing substance), is sensitive, group specific, andappears to be particularly useful in detecting recentinfections in both typhus and RMSF group rickett-sioses.

In Q fever the passive haemagglutination test canbe used for detecting phase I antibodies. The erythro-cyte sensitizing factor consists of a phenol extract ofthe phase I antigen.2. The toxin neutralization test.A test of this type is used in cases of rickettsiae of

the typhus and RMSF group. It is valuable forevaluating vaccine potency in the typhus group; inthe spotted fever group this test is useful for studyingantigenic structure.3. The intradermal sensitivity test.

This involves a skin reaction of the delayed type. Apositive skin test against C. burneti is considered tobe a contraindication for vaccination against Q fever.4. Weil-Felix reaction.The Weil-Felix reaction is still a useful test in all

rickettsioses except for Q fever, trench fever, andrickettsialpox. However, it has several serious limita-

tions. It fails in approximately 75% of sporadic orrecrudescent cases of typhus. In scrub typhus it maybe positive in no more than 50% of cases.5. The haemolymph (haemocyte) test.

This is an economical, rapid, and simple test fordetecting rickettsiae in nature and in laboratoryexperiments (4, 16). The presence of rickettsiae inticks can be detected by examining the haemolymphobtained by amputating the distal portion of one ormore legs. The smears are fixed by heat; stained bythe Gimenez method, and examined microscopicallyfor the presence of rickettsiae or rickettsia-like orga-nisms. If positive, additional smears from the sameticks are examined by the direct or indirect fluores-cent antibody method with sera or immunoglobulinsagainst the various rickettsial groups or againstrickettsiae known or assumed to occur in the area.Adult ticks, both engorged and unengorged, may beused for the haemocyte test.

All tick-borne rickettsiae can be identified by thehaemocyte test. In scrub typhus the indirect fluores-cent antibody method has been reported to be group-specific, allowing the use of any strain for thediagnosis.Some other tests may be used for special purposes

such as cross-immunity tests (with live or killedrickettsiae for immunization), and the serum-protec-tion test.The formation of plaques by rickettsiae in some

types of tissue cultures may lead to the elaboration ofmore precise methods for the titration of rickettsiaeand their differentiation.

PUBLIC HEALTH SIGNIFICANCE

Louse-borne typhusEpidemic louse-borne typhus, historically one of

the major epidemic diseases, typically spreads duringtimes of wars and civil disturbances. Since the agentis transmitted by body lice, the disease is still presentwhere socioeconomic standards are low and wherelice are abundant.

In the present century, the main foci of the diseasehave been in the highlands of northern and easternAfrica and in eastern Europe. The disease is alsopresent in the mountainous regions of CentralAmerica, north-western South America, southernAsia, and of southern Africa.During and immediately after the Second World

War the disease spread again from its endemic foci inNorth Africa and eastern Europe through most ofEurope causing very severe epidemics among pris-

438

RICKETTSIAE AND RICKE1TSIAL DISEASES

oners of war and in concentration camps in Europeand in Asia.

Despite the existence of conditions favouring thespread of the disease, the development of persistentinsecticides and their massive application to thecontrol of the vectors made it eventually possible tolimit somewhat the extent of epidemics. The develop-ment of specific vaccines prepared from rickettsiaegrown in the yolk sac of embryonate eggs or in thelungs of laboratory animals significantly contributedto the effective control of the disease; this was furtherfacilitated by the improvement in living conditionsthat occurred after the end of the Second WorldWar.

Since then the number of countries from whichcases are reported has considerably diminished. Thegreat majority of the cases now reported are from thehighlands of central and eastern Africa, namely fromBurundi, Rwanda, and Ethiopia. A large number ofcases of Brill-Zinsser disease (late relapses or recru-descent louse-borne typhus) are reported from someeastern European countries.

In Burundi, epidemic waves occurred in 1933-34,in 1939, and in 1945-46. During the latter period,3 752 cases and 429 deaths were recorded; in theendemic area the average number of lice on eachperson was approximately 40. A new epidemic startedin Burundi in 1967, with a peak of 17 200 casesreported in 1970. The number of cases since then hasdecreased, but there were still 2 817 cases in 1973.Although the reported number of deaths caused bytyphus ranged between approximately 1 % and 3% ofthe number of cases, the mortality rate in 1971 fromtyphus was estimated by physicians practising in thecountry at 5-20 %. Cases of louse-borne typhusstarted to be reported from neighbouring Rwanda in1971 with a peak of 5 509 cases and 78 deathsreported for 1973. In Ethiopia 2 000-3 000 cases areusually notified every year from all provinces, butonly a few deaths are reported. Most of the casesreported from other countries are from mountainousregions, where the climatic and socioeconomic fac-tors facilitate a close association between the humanpopulation and the louse vector.

Resistance of the louse vector to the commonlyused insecticides (DDT, gamma-HCH, and in a fewcases organophosphorus compounds) has occurredin some areas as a consequence of their use for vectorcontrol or for agricultural purposes.Although louse-borne typhus cannot be consid-

ered as a major risk from the point of view of itsspreading to other countries from the regions where

it is still present, it remains one of the most impor-tant causes of illness in some parts of Africa and aproblem in parts of the Americas and of Asia. Thedevelopment of tourism, rapid communications, andinternational travel in general make it possible forcases to occur far from endemic foci; if not recog-nized in time, it may present a considerable risk tothe individual, if not to the community. Also, it mustbe emphasized that the very reasons allowing for thepresence and dissemination of the infection and ofthe disease make it difficult to apply preventive andcontrol measures.

Patients with Brill-Zinsser disease are a potentialsource of infection in other countries, but spread ofthe disease still requires that the patient be alsoinfested with lice at the time of his relapse.

Other rickettsial infections

Murine typhus causes practically no public con-cern. It is in fact one of the relatively mild rickettsialdiseases and its cycle in nature is relatively easy tointerrupt, Xenopsylla cheopis, the normal vector ofR. mooseri, being easy to control.The number of reported cases of Rocky Mountain

spotted fever in the USA, which slowly decreasedfrom 1951 to 1960, has since risen from approxima-tely 200 to 305 in 1967, 298 in 1968, 498 in 1969, 380in 1970, 432 in 1971, and 528 in 1972. The majorityof the cases are now reported from states east of theMississippi river. This rise in reported cases may wellbe due to better recognition. However, mass migra-tions from the city centres to the suburbs havebrought people into closer contact with the tickreservoir of the disease and this is believed by manyto be responsible for the recent increases in thenumber of cases. Where the disease is relativelyfrequent, as for instance in Maryland, Virginia, andNorth and South Carolina, early recognition andtreatment are the rule. Elsewhere in North Americathe disease is so infrequent that it is often misdiag-nosed and a number of deaths occur each yearbecause of inadequate treatment. Sporadic cases ofthe same type of disease, caused by the same agent(R. rickettsi), are also reported from South America.Other forms of tick-borne typhus are of recognized

minor public health significance in localized geo-graphic areas: boutonneuse fever in the Mediterra-nean region, African and Indian tick typhus, Queens-land tick typhus in Australia, and Siberian ticktyphus in North-Eastern Asia. The incidence of thedisease is generally low and the clinical features so

439

R. BREZINA ET AL.

mild that there is very little public interest in controlmeasures.

Rickettsialpox is the mildest of all the rickettsialdiseases. In the USA at the time when the diseasewas first discovered in the mid-1940s, over 160 caseswere reported from New York. For several yearsthereafter about 180 cases were reported annuallyfrom New York City alone, but more recently onlyoccasional sporadic cases have been reported fromNorth America. Reports of the rickettsialpox orga-nism occurring in Korea and Russia as well as in theUSA indicate that it may have a worldwide distribu-tion. Rickettsialpox poses a minor public healthproblem only because it could appear in an epidemic.form if there should be a large buildup of infectedvectors and their hosts in close association with man.

Scrub typhus poses a continuing serious and im-portant public health problem. While the diseasemay be mild, the severe clinical form is common,with a 20-60% mortality if not treated in time. Sincethe ubiquitous mite serves as both vector and reser-voir and has numerous small animal hosts, R. tsutsu-gamushi appears to be not eradicable at least with themite control measures now available. In the scrubtyphus endemic area of South-East Asia and thePacific Islands there is always the possibility ofsporadic cases or localized outbreaks, and no vaccineis available to help in control. As a result of travel,sporadic cases of the disease may appear in any partof the world.Q fever is a continuing and significant public

health problem. While there are not many humancases reported in the literature, pilot serologic sur-veys in certain high-risk groups and in animalsindicate that many undiagnosed cases occur. There iswidespread seeding of domestic livestock throughoutthe world with C. burneti and in many areas theincidence of infection in animals is rising. The diseasein man occurs in livestock handlers and processors ofanimal products, especially in abattoirs and textileplants. The milk of infected animals contains C. bur-neti, but there is conflicting evidence as to how muchinfection as well as how much overt disease istransmitted through milk.

There are an increasing number of reports, al-though still sporadic, of serious illnesses such asendocarditis, pericarditis, and hepatitis resulting fromchronic Q fever infection.The only known epidemics of trench fever oc-

curred during the First and Second World Wars.The virtual disappearance of the disease from publicnotice for nearly 50 years combined with the mild

character of the human illness when it does occuraccounts for the absence of any public health interestin this disease.

CONTROL MEASURES

Treatment ofhuman rickettsiosesThe broad-spectrum antibiotics, namely the tetra-

cyclines and chloramphenicol, are the drugs of choicefor the specific treatment of all rickettsial diseases.Tetracyclines are preferred to chloramphenicol be-cause of the higher risk of untoward reactions withthe latter. The treatment should be prolonged toavoid the occurrence of relapses upon withdrawal ofthe antibiotic.A new tetracycline, doxycycline, better absorbed

and more slowly excreted than other compounds ofthis group, has recently been used successfully for thetreatment of louse-borne typhus in endemic areas.The development of resistance to antibiotics com-

monly used in the treatment of rickettsial diseaseshas been induced under laboratory conditions, but asfar as is known it has not been observed in the field.

Chemoprophylaxis requires even more specialattention than chemotherapy with regard to thedosage and the duration of treatment in order toobtain suppression of overt symptoms and the devel-opment of active protection. A combination ofchemoprophylaxis and vaccination with killed vac-cine has been successfully used against scrub typhus.

Vector controlSince all rickettsial diseases (Q fever excepted) are

transmitted to man through the bite or excreta ofinfected arthropods, the obvious control measurewould be the elimination of the vectors. Althoughtheoretically the control of lice and fleas is relativelyeasy because of their close association with man andhis immediate environment, it is much more difficultto control transmission by ticks and mites. Thesearthropods in natural environments have life cyclesassociated with a multitude of wild animals that arethemselves difficult to eliminate or to control.

Rickettsialpox is the only mite-borne rickettsialdisease that appears to be controllable by attackingits vector, Allodermanyssus sanguineus, or its host,the house mouse, or both.

Resistance to pesticides may present a severeproblem. The resistance to chlorinated hydrocarbonsis stable and so high that it excludes its routine use inmany areas. Resistance to malathion is not a stablecharacteristic and depends on continual exposure.

440

RICKETTSIAE AND RICKE1TSIAL DISEASES

Some new organophosphorus compounds and otherrapidly degradable pesticides are therefore beingsought.

VaccinationDead and live vaccines have been prepared against

louse-borne typhus. A single dose of the dead vaccine(Cox type) may prepare for a booster effect at anytime within 15 years. However, there are significantproblems in the methods of assay for standardizingthis vaccine. Soluble antigen vaccines may offer someadvantages as the antibody response appears earlier.With the live vaccine (strain E) there is the problemof late reactions (more properly termed mild disease)that occurred in 0-14% of the vaccinees in differentstudies. A reversion to virulence after passage inlaboratory animals has been reported from onelaboratory, but was not fully confirmed by twoothers.A killed vaccine against Rocky Mountain spotted

fever is available and is indicated for persons at highrisk (such as laboratory workers).

Killed crude C. burneti vaccines are available butsevere and persistent local reactions have contraindi-cated their use. Purified phase I and live vaccines areunder experimental study. A trichloroacetic-acidextract from the phase I purified suspension promisesto be very effective in preventing human disease.

Other methods

In general, the only practical measure against tick-and mite-borne typhus is personal protection againstthe arthropod vector with repellents or by means ofadequate clothing to prevent the attachment of thevectors.

Health regulations

Louse-borne typhus was one of the six quaran-tinable diseases subject to the International SanitaryRegulations. These have been replaced as of 1 Janu-ary 1971 by the new International Health Regula-tions to which only cholera, plague, yellow fever, andsmallpox will henceforth be subject.

Louse-borne typhus, however, remains one of thediseases in the WHO epidemiological surveillanceprogramme, together with relapsing fever, influenza,paralytic poliomyelitis, and malaria. Under this pro-gramme, Member States are requested to inform theOrganization of the occurrence of any outbreak ofthe disease, providing pertinent details, and theOrganization is responsible for the dissemination toMember States of such information received. Inorder to facilitate the implementation of this pro-gramme a technical guide for typhus surveillance hasbeen prepared by WHO (19).

UMIt

RICKETITSIES ET RICKETTSIOSES

Dans la presente etude sont resumees les connaissancesactuelles concernant les rickettsies et les rickettsioses, etnotamment leurs caracteristiques epid6miologiques, leurimportance en sante publique et les moyens de lutte a leuropposer.

I1 existe de nombreux foyers naturels de rickettsioses, Apartir desquels elles peuvent se propager dans d'autresregions du monde A la faveur de changements desconditions soicio-6conomiques. La rapidite des voyages Alongue distance fait que l'on assiste aujourd'hui A l'appari-tion de cas sporadiques, parfois graves, dans des endroitstres 6loignes des zones d'endemicite. Meme dans cesdernieres, il arrive que ces affections ne soient pasdiagnostiquees et des deces peuvent survenir si un traite-ment correct, de preference par les tetracyclines, n'est pasappliqu6 rapidement.

Les rickettsies sont g6neralement transmises A l'hommepar des arthropodes (poux, puces, tiques et acariens); seuleCoxiella burneti, agent dela fievre Q, fait exception. Pourcertaines especes de rickettsies, le cycle naturel d'infection

fait intervenir comme reservoirs des animaux domestiquesou sauvages. La transmission transovarienne chez lesarthropodes, demontree pour les rickettsies du groupe dela fievre pourpree (chez des tiques) et du typhus de brousse(chez des acariens) joue un r6le important dans lapersistance des infections dans les foyers naturels. Pour lafievre des tranchees, l'homme est le seul reservoir connu;des infections persistantes chez l'homme et chez le pouassurent le maintien du cycle malgr6 l'absence de transmis-sion transovarienne de Rickettsia quintana chez le pou.Le typhus epidemique A poux, bien que raye de la liste

des maladies soumises au Reglement sanitaire internatio-nal, reste l'une des maladies figurant au programme desurveillance 6pidemiologique de l'OMS; cette affection estresponsable d'une morbidite et d'une mortalite impor-tantes dans certaines parties de l'Afrique, et des cas sontsignales localement dans les Ameriques et en Asie. Letyphus de brousse pose toujours un serieux probleme desante publique en certains endroits de l'Asie du Sud-Est etdu Pacifique Occidental. Aux Etats-Unis d'Amerique, le

441

442 R. BRESINA ET AL.

nombre annuel de cas signales de fievre pourpr&e desmontagnes Rocheuses a augment6 pendant les deuxdernieres decennies; les raisons possibles en sont demeilleurs diagnostics et une migration accrue de popula-tions du centre des villes vers les faubourgs. Des affectionsapparentees, transmises par des tiques, sont observees en

Am6rique du Sud, dans la region mediterraneenne, enAfrique, en Asie du Sud-Est, en Extreme-Orient et dans lePacifique Occidental. Bien que sporadiques, les cas d'affec-tions graves dues a une infection chronique par la fievre Qsont en augmentation dans beaucoup de regions duglobe.

REFERENCES

1. BOZEMAN, M. F. J. inf. Dis., 121: 367 (1970).2. BREZINA, R. Zbl. Bakt., I Abt. Orig., 206: 313 (1968).3. BREZINA, R. Curr. top. Microbiol. Immunol., 47: 20

(1969).4. BURGDORFER, W. Am. J. trop. Med., 19: 1010 (1970).5. FISET, P. ET AL. Acta Virol., 13: 60 (1969).6. GOLINEVICH, E. M. ET AL. Vestn. Akad. med. Nauk,

10: 77 (1969).7. KOKORIN, J. N. Acta virol., 12: 31 (1968).8. LACKMAN, D. B. ET AL. Hlth Lab. Sci., 2: 135 (1965).9. MCKIEL, J. A. ET AL. Canad. J. Microbiol., 13: 503

(1967).

10. MURRAY, E. S. ET AL. Proc. Soc. exp. Biol. Med.,119: 291 (1965).

11. MURRAY, E. S. ET AL. J. Immunol., 94: 723 (1965).12. MURRAY, E. S. ET AL. J. Immunol., 94: 734 (1965).13. ORMSBEE, R. A. Ann. Rev. Microbiol., 23: 275 (1969).14. ORMSBEE, R. A. ET AL. J. Immunol., 92: 404 (1964).15. PARETSKI, D. Zbl. Bakt., I Abt. Orig., 206: 283 (1968).16. AEHACEK, J. ET AL. Acta Virol., 15: 237 (1971).17. SCHRAMEK, S. ET AL. Acta Virol., 16: 487 (1970).18. WISSEMAN, Ch. L. Jr. Zbl. Bakt., I Abt. Orig., 206;

299 (1968).19. Wkly epidem. Rec., No. 28: 273-275 (1971).