STUDIES ON CORTISONEAND ANTIBIOTICS FOR PROMPTTHERAPEUTICCONTROLOF TYPHOID FEVER

AND SCRUBTYPHUS

By C. L. WISSEMAN,JR., P. Y. PATERSON,J. E. SMADEL, F. H. DIERCKS,AND H. L. LEY, JR.

(From the Department of Virus and Rickettsial Diseases, Army Medical Service GraduateSchool, Washington 12, D. C.)

(Submitted for publication July 27, 1953; accepted October 21, 1953)

Prior to the use of chloramphenicol in the treat-ment of patients with typhoid fever the mor-tality rate in this disease was almost 10 per centand approximately one-fourth of the deaths wereattributable to the toxic-febrile state (1). Chlor-amphenicol has been shown to control the bac-teriologic manifestations of typhoid fever withina matter of hours but the toxemia has respondedmore slowly to the specific antibiotic (24). In-deed, fever and general toxemia generally havecontinued unabated or are even somewhat in-tensified (2, 3, 5) for two or three days afterantibiotic therapy has been instituted. The ob-servations of our group (6, 7) indicated thatcortisone, when used either alone or along withchloramphenicol, brought prompt relief of sub-jective and objective acute febrile manifestationsof typhoid fever. Moreover, others have madesimilar observations in such patients treated withACTHand chloramphenicol (8, 9).

The present report is concerned primarily withadditional studies on the use of a combined corti-sone-chloramphenicol regimen in the treatment oftyphoid fever patients, and, secondarily, with re-lated studies on the use of the same therapeuticcombination in the treatment of scrub typhus pa-tients. In both of the diseases, but particularly inthe former, the studies have been oriented towardattempts to understand the factors concerned withthe toxic-febrile reactions displayed by acutely illpatients and the physiological mechanisms bywhich these may be alleviated.

MATERIALS AND METHODS

Typhoid Fever

Selection of patients: All patients included in this studywere treated on the wards of the General Hospital, KualaLumpur, Federation of Malaya. Each exhibited pyrexiaand typical manifestations of severe typhoid fever; in

every instance a positive blood culture was obtained priorto treatment which was instituted between the ninth andsixteenth days of disease. The 18 patients mentioned inthis report included 6 males and 12 females who rangedin age from 7 to 35 years.

Treatment: All patients received both chloramphenicol 1and cortisone 2 orally. The regimens listed below werethose given to Asian adults who weighed about 45 Kg.;children received proportionately smaller doses.

In each instance the antibiotic was administered accord-ing to a fixed schedule in two courses as follows: A firstcourse consisting of 3.0 Gm. initially, followed by main-tenance doses of 1.5 Gm. at 12-hour intervals until thepatient had been afebrile for 48 hours; and a second courseconsisting of 1.5 Gm. every 12 hours for eight doses, be-ginning on the eighth day after completion of the firstcourse.

Cortisone was administered according to several sched-ules in order to learn more about its effect and optimaldosage.3 Thus, the patients were divided into five groupson the basis of the duration of cortisone treatment. Thefour patients in Group I received an average total dose of10.7 mg. cortisone per Kg. of body weight during a singleday of steroid therapy. For 45 Kg. adults this doseamounted to a total of 500 mg. cortisone of which 300 mg.were administered with the initial dose of chlorampheni-col, and 200 mg. 12 hours later. Groups II (five pa-tients), III (six patients), and IV (one patient) receivedcortisone twice daily over periods of 2, 3, and 4 days,respectively; here the amount of drug on a body weightbasis was somewhat smaller (8.1 to 6.4 mg. per Kg. dur-ing the first day) than in Group I. For an average Asianadult these dosages amounted to a total of 300 mg. on thefirst day, 200 mg. on the second, and 100 mg. on the thirdand fourth days. The regimen was so arranged that themorning dose was approximately twice that given on the

1 Supplied by Parke, Davis and Company.2 Supplied by Merck and Company, Incorporated.8 Oral administration of the 25 mg. tablets of cortisone

acetate, which were employed throughout the study, wasoriginally troublesome in stuporous patients and in chil-dren because the tablets quickly became soft and adheredto the palate or tongue. This difficulty was avoidedafter it was found that the tablets disintegrated readilyin water to form a suspension which could then be givenorally with a medicine dropper.

264

TYPHOID FEVER AND SCRUB TYPHUS

evenings of the first and second days, but the amountssupplied on subsequent days were divided equally betweenthe morning and evening. The two patients in Group V,who were cared for late in the study, received smalleramounts of this drug at more frequent intervals, i.e., 50mg. doses of cortisone at 6-hour intervals until the tem-perature returned to normal; subsequently the hormonewas given at 12-hour intervals. Full details of the treat-ment regimens employed with all patients are included inAppendix A.

Care of patients: Clinical and laboratory proceduresemployed in this study were similar to those used previ-ously by our group in Malaya (2, 7, 10). Members ofthe team administered the drugs personally and took thepatients' temperatures frequently during the initial re-sponse to treatment. Pyrexia in patients with typhoidfever was defined, in accordance with previous reports,as an oral temperature of 99° F. or above. Blood cul-tures were made during the first few days of therapy, onthe two days immediately prior to the second course ofchloramphenicol, and in any subsequent febrile period.Relapses of typhoid fever were treated with chloram-phenicol alone, and other complications by appropriatemeasures. Public health regulations for release of con-valescent typhoid fever patients in Malaya require thatthree consecutive stool and urine cultures taken at weeklyintervals be free of Salmonella typhosa. In order tominimize the possibility that negative cultures were theresult of continued antibiotic suppression instead of abacteriological cure, collection of specimens for culturefrom convalescent patients awaiting discharge was notbegun until five days had elapsed since the last dose ofchloramphenicol. An additional stool culture was ob-tained from those patients who reported for a follow-upexamination two weeks after discharge from the hospital.

Scrub TyphusSelection of patients: The patients in this study, with

one exception, were healthy young adult male volunteerswho had been inoculated intradermally with a few infec-tious units of the Karp strain of Rickettsia tsutsugamushiwhile serving as members of control groups in an in-vestigation dealing with immunization against scrub typhus(11, 12). Each of these volunteers bears the same iden-tifying code number here that was used in the otherreports. Since these reports contain many of the detailsof the course of the disease of each volunteer, includinglaboratory proof of scrub typhus infection, only informa-tion which is pertinent to the present study will be men-tioned here.

Care of patients: As the inoculated volunteers becameill they were placed alternately into two groups,4 i.e.,

4Group A includes volunteers C-10, C-li, C-13, C-14,C-17, C-19, and C-23. Group B includes volunteers C-9,C-12, C-15, C-16, C-18, C-20, C-21, and C-22. Volun-teers C-21, C-22, and C-23 were inoculated at one timewith 90 mouse minimal infectious doses while the re-mainder were inoculated on a different day with 60 mouseminimal infectious doses.

Group A (seyen patients) who received chloramphenicolalone and Group B (eight patients) who received cortisonein addition to the antibiotic. Each volunteer of bothgroups was given specific antibiotic therapy after approxi-mately 48 hours of sustained fever which was regarded asan oral temperature of 1000 F. or higher in the ambulantperson.5 The course of treatment with chloramphenicolconsisted of an initial 3.0 Gm. oral dose followed by 1.5Gm. amounts 12 and 24 hours later. Members of GroupA were given no additional medication, but those ofGroup B also received a total of 500 mg. of cortisonegiven orally in amounts of 300 mg. and 200 mg., respec-tively, along with the first two doses of antibiotic. Thesingle patient with naturally acquired scrub typhus, aBritish soldier hospitalized at the British Military Hos-pital, Kinrara, near Kuala Lumpur, was treated on theseventh day of disease with the same chloramphenicol-cortisone regimen used for the volunteers.

RESULTSTyphoid Fever

Results of the therapeutic studies are sum-marized in Table I and are presented in detailfor each patient in Appendix A. Since all patientsreceived essentially the same course of chloram-phenicol, which usually renders typhoid patientsafebrile in about four days, emphasis will be placedon the effect that different regimens of cortisonehad on the course of the disease during the firstfew days of treatment.

Prompt control of pyrexia and toxemia by com-bined therapy: Rapid and dramatic defervescencewas observed in all patients in the present studywho received the large dose of cortisone, i.e.,Groups I through IV in Table I. The averagetime required for these patients to become afebrileafter receiving the first oral dose of cortisone and

5Fever was defined as a sustained oral temperatureabove 990 F. in early studies of therapy of patients withnaturally acquired scrub typhus who were first seen aftersome days of high fever and who were content to remainin bed or on limited activity for a reasonable period aftertherapy. Later, however, it was found that this cri-terion did not suffice to distinguish abnormal temperaturein volunteers inoculated with Rickettsia tsutsugamushi.These persons remained ambulatory and quite active inthe tropical environment during the incubation period andfrequently exhibited oral temperatures above 990 F. evenduring the period prior to the test. Furthermore, sincethey were treated within 48 hours after the onset of clinicaldisease and, hence, experienced little or no debilitation,the volunteers refused to remain in bed for more than afew hours after therapy was instituted. Under these cir-cumstances, an oral temperature of 1000 F. or above moreaccurately indicates a real deviation from normal.

265

266 C. L. WISSEMAN,JR., P. Y. PATERSON,J. E. SMADEL, F. H. DIERCKS, ANDH. L. LEY, JR.

TABLE I

Summary &f responses of typhoid fever patients to treatment with chloramphenicol and cortisone:Comparison with patients treated with chioramphenicol alone

First Duration Duration Perm.Duration normal afebrile fever after afebrilecortisone temp. after period due escape from due

Rx Number of cortisone cortisone cortisone therapyGroup (days) patients (hours)* (hours)* (days)* (days)*

Present study

I 1 4 5.5 22.5 3.5 4.8II 2 5 6.0 49.6 1.9 3.8

III 3 6 6.0 45.Ot 1.8t 4.2IV 4 1 6.0 Perm. afebrile after cortisone 0.3

Other studies

it 4 4 15.5 Perm. afebrile after cortisone 0.72§ 0 45 No cortisone given 4.0311 0 500 No cortisone given 4.5

* Average value for group.t Average of the four patients who had recrudescence of fever; two patients were permanently afebrile after cortisone.t Data on intramuscular cortisone and chloramphenicol from Reference 7.

Combined data from References 2 and 10.It From data of Reference 4.

antibiotic was six hours. The three charts in Fig-ure 1 illustrate graphically the accelerated de-fervescence.

Signs and symptoms commonly attributed totoxemia decreased along with defervescence.Thus, a patient who was in a semi-stuporousfebrile state at noon when therapy was begun,might be sitting up in bed, smiling and showinginterest in his surroundings at 6:00 p.m. EvenPatient 18, who was markedly toxic with pro-nounced stupor, tachycardia and incontinence ofurine and feces, improved distinctly within a fewhours after initiation of therapy.

The defervescence, which began shortly aftertherapy was instituted, was precipitous in allpatients of the first four groups and in about halfof these the temperature fell to hypothermic levels,i.e., below 960 F. orally. During the period ofmarked hypothermia experienced by Patient 9,whose chart is shown in Figure 1, the rectal tem-perature fell to 940 F. Of the 10 patients whosetemperature fell below 960 F., 4 attained a mini-mumbetween 950 and 95.9° F., and 4 between940 and 94.9° F.; in 2 patients the minimum tem-perature was 93.80 F. The skin of patients withmarked hypothermia was cold and clammy (sweat-ing was profuse during defervescence). Thesepersons responded slowly to commands and ex-

hibited pronounced intention tremors. If asleep,they were awakened with difficulty. However,there was no evidence of cardio-vascular collapsesince the pulse was slow and strong and the bloodpressure normal or even slightly elevated. Aftera few hours the temperature rose and the signsassociated with hypothermia disappeared, leavingno residual deleterious effects. Hypothermia fol-lowing cortisone administration was more frequentamong children than adults of this series; it oc-curred in seven of the nine children between 7 and14 years of age and in three of the nine older pa-tients. This phenomenon is not peculiar to pa-tients treated with cortisone although it may bemore pronounced in them. It may be recalled thatsome degree of hypothermia, at times sufficientlysevere to warrant remedial measures, has beenknown to occur during early convalescence of un-treated typhoid fever patients (13) and, morerecently, in patients treated with chloramphenicolalone (3, 14).

The two patients included in Group V in Ap-pendix A received smaller doses (50 mg.) of oralcortisone at intervals of six hours in an attemptto avoid the extremely rapid change in tempera-ture. Despite the fact that defervescence in Pa-tient 17 occurred somewhat more slowly than usu-ally observed in patients in Groups I through IV,

TYPHOID FEVER AND SCRUBTYPHUS

CHLORAM-1,00EPHENICOL °

mg/kg/day o

ORALCORTISONEtomg/kg/doy0 . ........0 f E1* is1a 1r

DAY OF DISEASE

FIG. 1. FEVER CHARTSIN CHLORAMPHENICOLTREATEDTYPHOID FEVER PATIENTS REcEIvING CORnSONEFOR ONE,TWOANDTHREEDAYS.

Note: Interrupted vertical line on charts 4 days afterinstitution of therapy indicates average duration of feverin patients treated only with chloramphenicol.

a minimum temperature of 95.60 F. was attained12 hours after the initial dose. Patient 18 re-

sponded more slowly and at no time was a tem-perature below 97.80 F. recorded.

Duration of antipyretic effect of cortisone: Inall nine of the patients in Groups I and II whoreceived cortisone for one or two days, and infour of the six in Group III who were given thehormone for three days, fever, headache and othermanifestations of toxemia reappeared within 18to 72 hours after combined therapy was instituted.

For convenience, this prompt return of the febrile-toxic state will be referred to hereafter as "escape"from cortisone effect in order to differentiate itfrom relapse in patients treated with chloram-phenicol; the latter has been associated with re-turn of demonstrable bacteremia and generally oc-curred some days after antibiotic therapy was dis-continued.

The escape phenomenon became manifest in 6to 24 hours after the last scheduled dose of corti-sone had been administered to the patients inGroups I and II; the average time was 16 hoursfor the former and 20 for the latter. Stated dif-ferently, the durations of the afebrile periods in-duced by cortisone were 22.5 and 49.6 hours, re-spectively, for the two groups, while the averagetimes of escape for the groups were 28 and 56hours after the beginning of therapy, see Table Iand Appendix A.

The patients in Group III presented no suchconsistency as regards escape. In the first place,two of the six members never displayed the phe-nomenon; they progressed to uneventful recovery.Among the four who suffered a return of fever andtoxemia, three demonstrated these findings somehours before the last dose of hormone (Patients10, 11, 15, see Appendix A) while the fourth(Patient 14) reacted in this manner 12 hours afterthe end of cortisone therapy. Several factors prob-ably contributed to the irregularity of the phe-nomenon in this group of patients who receivedthe hormone for three days. In the first place thepatients were approaching the period, i.e., thefourth day, when the antibiotic effect alone wouldhave rendered them afebrile, see Table I; hence,a proportion of the group might be expected toavoid the escape phenomenon. A potential factoris the possibility that a given dose of cortisone maybe less effective in controlling the febrile-toxicstate on the third day than on the first day oftreatment. Thus, in Patient 17, 100 mg." corti-

6 The temperatures of three patients (numbers 10, 11,and 15) in Group III, i.e., those who received the 3-daycortisone regimen, showed a distinct upward trend on thethird day of steroid therapy and experienced the onset ofthe escape phenomenon just before or shortly after thelast dose of cortisone. This suggested that the amountof hormone given the patients on the third day was in-adequate; hence, the two patients in Group V were givensmaller doses at 6-hour intervals in the initial phase oftherapy in the attempt to prevent the abrupt defervescence

267

268 C. L. WISSEMAN,JR., P. Y. PATERSON,J. E. SMADEL, F. H. DIERCKS, ANDH. L. LEY, JR.

TABLE II

Duration of bacteremia in typhoid fever patients afterinstitution of combined cortisone-

chloramphenicol therapy

Time of culture

2 hours before Rx begun24 hours after Rx begun48 hours after Rx begun72 hours after Rx begun

No. patientscultured

18141113

No. patientswith-positive

cultures

930

1*

* Not cultured at 48 hours.

sone produced a rapid defervescence from 103.20F. to 95.60 F. on the first day of therapy while on

the third day the same dose was followed insteadby a rise in temperature from 98.60 F. to 100.20 F.

The data presented in Table I and the charts inFigure 1 show that the patients in the presentstudy who escaped from the effects of cortisoneprogressed to recovery along courses essentially in-distinguishable from those of patients in earlierstudies who were treated with chloramphenicolalone. The former became permanently afebrilein about the same length of time from the start oftherapy as did the latter patients.

Laboratory studies: Table II summarizes the re-

sults of blood cultures taken during the first 72hours after institution of therapy. With one ex-

ception, blood cultures were negative after the first24 hours of therapy. The single culture whichyielded S. typhosa at 72 hours was from Patient9, the youngest of this series, who later suffereda relapse. None of the blood cultures taken im-mediately prior to the second course of chloram-phenicol was positive, a finding of considerableinterest since earlier studies (10) had shown thatthis time marked the beginning of the period inwhich relapses with recurrence of bacteremia were

most prone to occur.

Stool cultures in five patients were positive forS. typhosa on one or more occasions after therapywas begun. In two cases the pathogen was foundon the second and fourth days, respectively, of the

and 1.5 to 2 times as much steroid on the third day as on

the first in the attempt to prevent escape. It is of interestthat the smaller more frequent doses of the drug pro-duced, on the first day, rapid defervescence in one patientof Group V; but in the other patient, by far the mostsevere case in this series, the action was less rapid.Furthermore, the large dose on the third day was in-adequate to prevent escape.

initial course of chloramphenicol; and in another,on the day following the last dose of the firstcourse. One patient had three positive stool cul-tures immediately prior to the second course ofantibiotic but subsequent cultures were negative.The feces of two patients, one of whomhad a posi-tive stool culture on the fourth day of the initialcourse of therapy, were positive on a single occa-sion after completion of second courses of anti-biotic. The chronic carrier state did not developin any of the patients during the period of ob-servation, an average of 31 days after the last doseof chloramphenicol.

Total eosinophil counts were of no value inestimating the hormonal response since 12 of the18 patients had no demonstrable eosinophils priorto the first dose of cortisone. The counts of theremaining six patients were 22, 22, 37, 121, 275,and 300 eosinophils per cu. mm. at the time therapywas begun, and dropped sharply after the firstdose of cortisone. Since eosinophil counts are gen-erally low in typhoid fever (15) and since suchtests were employed here in an attempt to evaluatethe initial cortisone effect, they were not extendedbeyond the first few days of observation.

Complications and relapse: There was no in-stance of overt intestinal hemorrhage in this seriesof cases. Only in the case of Patient 18 did symp-toms and signs suggest a small intestinal perfora-tion. This occurred on the 45th day while the pa-tient was receiving chloramphenicol for a relapseof typhoid fever. The signs and symptoms disap-peared with continued administration of chloram-phenicol alone. Pneumonitis was present whentreatment was begun in two cases, both of whomreceived penicillin in addition to antityphoid ther-apy. One other patient exhibited a marked bron-chitis which improved without additional therapy.Acute cholecystitis complicated the convalescenceof Patient 2 with symptoms beginning on the 59thday of disease. A stool culture positive for S.typhosa just four days prior to onset of symptomssuggested the possibility of typhoidal etiology inthis complication.

Bradycardia was common during the first fewdays of convalescence. This can hardly be classedas a complication since the older literature on ty-phoid fever states that true bradycardia, in contrastto the relative bradycardia early in the disease, wasnot uncommon during early convalescence of un-

TYPHIOID FEVER AND SCRUBTYPHUS

treated cases (13). It is of interest that prematureconvalescence induced by chemotherapy may alsobe accompanied by the marked slowing of the heartrate.

Irregularity of the pulse was noted in four pa-tients one to three days after initiation of therapy.It was the clinician's impression that the irregu-larity in all four cases resulted from dropped beats.Indeed, an electrocardiographic tracing obtainedfrom a patient showing this sign demonstratedsecond degree heart block in the form of a typicalWenckebach's phenomenon. This condition dis-appeared spontaneously in the four patients in fromone to nine days after onset. Cardiac irregulari-ties were by no means rare in typhoid fever pa-tients before the era of specific therapy; irregu-larity and intermittency (16, 17), heart block andchanges in the ventricular complexes in the elec-trocardiogram (18), and a high incidence of pro-longation of the P-R interval (19, 20) were alldescribed. Furthermore, arrhythmias have beennoted recently in typhoid fever patients treatedwith chloramphenicol alone (21). Since the elec-trocardiographic changes encountered in cases be-fore the days of antibiotics were more frequent dur-ing the second and third weeks of disease, it wouldbe difficult to ascribe the cardiac irregularitiesnoted in the present series, which occurred at acorresponding period, to the effects of therapyalone. On the other hand, the possibility thatsome animal tissues may exhibit unusual responsesto bacterial products while under the influence ofthe cortisone must be admitted (22).

Relapses, with proven bacteremia in each in-stance, occurred in 3 of the 18 patients; these epi-sodes began 8, 9, and 21 days, respectively, aftercompletion of the second course of chlorampheni-col. Response to readministration of chloram-phenicol was prompt. This incidence of relapsedoes not represent a significant deviation from thecombined relapse rates observed by previous groupsusing chloramphenicol alone (23) and is inter-mediate between the relapse rates noted by Johnand Vinayagam (24) for continuous and for in-termittent schedules of chloramphenicol therapy.Marmion (3), on the other hand, recently ob-served relapses in 41 of 97 patients treated on avery similar schedule of interrupted antibiotictherapy. In contrast, there was only one relapseamong the 21 patients of Marmion who received a

10-day course of typhoid vaccine concurrent withthe interrupted antibiotic regimen. This authorcalls attention to the higher relapse rate amongpatients treated early in the disease as comparedwith those treated late but does not indicate theexact day of disease upon which therapy was be-gun in his large series of cases. One suspects thathis patients, military personnel, were hospitalizedand treated at an earlier stage of the disease thanwere the civilians of the present study in whomtherapy was begun, on the average, on the 12thday of disease. Observations on patients withscrub typhus clearly indicate that relapses are com-mon in those treated within the first few days ofillness but rare in those treated during the secondweek (25); it is possible that an analogous phe-nomenon may be found in typhoid fever.

Scrub Typhus

The response of patients with scrub typhus tocombined therapy with cortisone and chloram-phenicol is compared with the response to chlor-amphenicol alone in summary form in Table III.The febrile-toxic state terminated in the usualrapid manner in the volunteers treated only withantibiotic; but those who were given supplemen-tary cortisone experienced even more rapid allevia-tion of their illness. Thus, volunteers of Group A,who received chloramphenicol, were afebrile in anaverage of 21.7 hours while those of Group B,who received the combined cortisone-chloram-phenicol regimen, were afebrile in an average of6.7 hours. The charts of selected volunteers arepresented in Figure 2. Marked symptomatic im-provement accompanied the fall in temperatureand within a few hours after institution of therapyit was difficult to persuade most of these volun-teers to stay in bed. In a few patients, however,

TABLE III

Comparison of combined cortisone-chloramphenicol regimenwith chloramphenicol aone in the treatment of

volunteers infected with the Karpstrain of scrub typhus

Response to treatment:Treatment 1st normal tempera-

ture after RxNumber Corti- Chloram- (hours)

of sone phenicolGroup patients (mg.) (Gm.) Average Range

A 7 0 6 21.7 13-40B 8 500 6 6.7 4-14

Numberof

relapses

14

269

270 C. L. WISSEMAN,JR., P. Y. PATERSON,J. E. SMADEL, F. H. DIERCKS, ANDH. L. LEY, JR.

DAYOF DISEASE

FIG. 2. FEvE CHARTS OF PATIENTS WITH SCRUBTYI-HUS, SELECTED TO ILLUSTRATE (a) HYPOTHERMICRESPONSETO CHLORAMPHENICOLTHERAPY ALONE, (b)POSSBLEESCAPE PHENOMENON,AND (c) RELAPSE

regression of headache was not quite as rapid as

defervescence.None of these patients experienced such marked

hypothermia as did some of the typhoid fever pa-

tients. Almost half of the scrub typhus patientstreated with antibiotic alone exhibited an oraltemperature as low as 96° F. for a short time aftertherapy; this is illustrated by the temperaturecurve of volunteer C-19 in Figure 2. The chartof volunteer C-20 (Figure 2) was selected forpresentation partly because it demonstrated a

minimum temperature as low as that attained inany of the cortisone-treated group.

Somepatients experienced a recurrence of fever,

toxemia and rickettsemia shortly after terminationof therapy. Such a sequence of events, which isinterpreted here as relapse, is illustrated by Volun-teer C-12 in Figure 2; he showed a characteristicstep-wise increase in fever over a period of severaldays. One of the seven patients treated with chlor-amphenicol and four of the eight volunteers whoreceived cortisone in addition to antibiotic showedrelapse of this type. The observed relapse rate inthe cortisone-treated group was numerically greaterthan that of the group treated only with chloram-phenicol. However, in this small series, the dif-ference was not statistically significant; nor was therate greater than that of another group of volun-teers similarly infected with a different strain ofscrub typhus rickettsiae the previous year andtreated with the short course of antibiotic, butwithout cortisone (26).

A picture exactly analogous to the escape fromcortisone exhibited by some of the typhoid feverpatients was not observed among the scrub typhuspatients, possibly because the rapid response toantibiotic therapy itself in scrub typhus leaves littletime for such occurrences. Volunteer C-20 (Fig-ure 2) and the patient with naturally acquiredscrub typhus (Figure 3) were the only instancesin which escape may have occurred. However, itshould be pointed out that the occasional patienttreated with the short course of antibiotic alonemay show a similar brief return of fever aftercessation of therapy.

Rickettsemia has been observed by previousgroups to persist for as long as 48 hours after in-stitution of specific antibiotic therapy; the intro-duction of cortisone to the therapeutic regimendid not alter this phenomenon.

The single case of naturally acquired scrub ty-phus treated with the combined cortisone-chlor-amphenicol regimen was a 20-year-old Britishsoldier with full-blown disease of seven days dura-tion. A graphic summary of this case appears inFigure 3. Defervescence occurred as promptlyin this individual as in the patients who had beenfebrile for only 48 hours when treated. More-over, the minimum temperature attained was nolower than in those treated early.

DISCUSSION

The practical need for control of toxemia intyphoid fever is emphasized by the experience of

271TYPHOID FEVER AND SCRUB TYPHUS

105

104

T 103E 102

e 101DE 100-lG.

9

F. 998

97CHLORAM6PHENICOL

gm/day o -

ORALCORTISONE600mg/doy

SCRUBTYPHUS:NATURALINFECTION20 YR. OLDWHITE a"

---W.-0'

1 la A a I t I I a

3 5 7 9 1 13 es 17 19 21 23

HGBGM% 153 K4 149

WBCx 1000 6.3 10.7 9.2ISOLATION OF + + + +RICKETTSIAOX-K AGGL. 0 320

FIG. 3. CLINICAL RESPONSEOF SCRUB TYPHUS PATIENT TREATED WITH CHLORAMPHENICOIAND CORTISONEIN SEVENTHDAY OF NATURALLYACQUIREDDISEASE

Stuart and Pullen (1), who found it a major causeof death in patients given only non-specific sup-portive treatment, and by the continued reports ofsome deaths associated with toxemia despite chlor-amphenicol therapy (3, 5, 27). The results ofthe present study confirm the earlier findings (6,7), i.e., that administration of cortisone in theinitial phases of chloramphenicol therapy induced,within a few hours, a rapid and dramatic regres-sion of the clinical manifestations of the febrile-toxic state in typhoid fever. Indeed, the oralpreparation of hormone employed in the currentinvestigation produced its effects with greaterrapidity than did the intramuscular preparationused in the earlier studies. The prompt clinicalimprovement induced by cortisone without inter-ference with the bacteriologic control afforded bychloramphenicol suggests that this hormone maybe administered profitably to those typhoid feverpatients who are so ill at the time antibiotic ther-apy is begun that the physician may anticipatedeath of the patient in the three or four days thatseparate initiation of antiobiotic therapy from thedefervescence and detoxification that ordinarilyfollow this treatment. The combined experiencesof our groups in over 30 patients treated witha cortisone-chloramphenicol regimen provide the

background for our opinion that it is desirable toshorten the toxic-febrile state in so far as possiblein the critically ill typhoid patient. These pa-tients stand in contrast to those with disease ofmild to moderate severity who respond satisfac-torily to antibiotic therapy alone.

The mechanisms by which cortisone producesits effects in typhoid fever remain essentially un-known. Of some interest, however, is the escapephenomenon noted in patients in whom cortisoneadministration was discontinued a day or so be-fore the anticipated occurrence of the defervescenceinduced by the antibiotic. The alleviatory effectof cortisone on typhoid pyrexia is transient andappears to dissipate itself within a day after thelast oral dose of hormone. The short durationof the cortisone effect is in general agreement withpublished studies (28) of the duration of cortisone-induced eosinopenia in normal subjects. The pres-ent observations indicate that cortisone does noteradicate the cause of the toxic state and suggestthat it exerts no direct neutralizing action on thetyphoidal toxins; the latter is in accord with re-sults of animal experiments (29). Furthermore,the persistence of bacteremia in patients treatedwith cortisone alone (6), despite clinical improve-ment, has shown that there is no suppression of

272 c. L. WISSEMAN,JR., P. Y. PATERSON,J. E. SMADEL, F. H. DIERCKS, ANDH. L. LEY, JR.

the growth of the organism by the hormone. Theobserved effects would appear, then, to be broughtabout by an action on the host through physio-logical mechanisms. It is worth re-emphasizingthat neither the antibiotic nor the hormone "cure"the patient; they merely buy time for him todevelop the immune mechanisms necessary forpermanent recovery and to repair the pathologicinjuries of disease.

The toxic typhoid patient does not display thesigns of a patient suffering from acute adrenalcortical insufficiency, except possibly for the rareoccurrence of a state resembling the Waterhouse-Friderichsen syndrome (30). Moreover, the de-fervescence observed in typhoid fever patientsin response to administration of ACTH (8, 9,31) suggests a reasonable degree of adrenal cor-tical reserve function even in these severely illpatients. It is, therefore, unlikely that adminis-tered cortisone produces the observed effects bysimple replacement therapy according to the cus-tomary connotation of this term.

The eosinopenia noted in the majority of the18 patients in the present series, though suggestiveof adequate adrenal function, is difficult to assess,since bacterial pyrogens are known to produce leu-kopenia, eosinopenia and subsequent leukocytosis("alarm reaction" blood picture) in adrenalec-tomized dogs (32).

There is an increasing number of reports, re-cently reviewed by Thomas (33), of spontaneousinfections arising in the course of cortisone ther-apy and of the enhancement and dissemination ofexperimental infections under the influence of thissteroid. The exact nature and relative importanceof the factors involved in the cortisone-altered hostresponse to infection remain ill-defined. Thus,rapidly spreading and widely disseminated infec-tions, with bacteYremia, are produced in animals byorganisms which normally cause only localizedinfections (34, 35), yet the polymorphonuclearleukocytes present are still capable of phagocytosisand some of the blood-clearing mechanisms ap-parently continue to operate (34-36). In contrastto such bacterial infections, Aikawa and Harrell(37) observed that cortisone had a life-prolongingeffect in guinea pigs infected with R. rickettsii;however, the ultimate mortality rates were thesame in the treated and control groups. Contra-dictory reports have been forth-coming on anti-body production (33-35, 38), but the reaction of

antibodies with antigens is not prevented andpreviously immunized animals continue to showsome degree of resistance to infection while underthe influence of this steroid (33, 35). It is worthnoting that the enhancement of infection in ex-perimental animals, which are not given the benefitof specific antimicrobial measures, is usually great-est when cortisone is given over relatively longperiods of time and when administered prior to theintroduction of the infectious agent (33). Rela-tively little work has been reported which dealswith experimental infection in animals treated con-currently with cortisone and antibiotics. There-fore, the observations of Glaser and Loeb (39)are of particular interest in connection with thepresent study. These authors observed that ratswith streptococcal pneumonia had less severe dis-ease when given cortisone and penicillin than didinfected animals which received only penicillin.

In the typhoid fever patient adequately treatedwith chloramphenicol, multiplication and spreadof the invading micro-organism are quickly con-trolled (2, 10, 23). The experience gained in theearlier studies (6, 7), and in the present investi-gation indicates that the addition of cortisone tothe antibiotic regimen does not interfere with therapid disappearance of S. typhosa from the blood.

Protection against death from bacterial toxinswhich is afforded by cortisone is rather strikingin the case of experiments with adrenalectomizedanimals (40, 41). On the other hand, a similarprotective effect is not so readily observed in ani-mals with normal adrenals to whom extra cor-tisone is given (29). Although these animal ex-periments are of considerable interest, they do nothelp to clarify our understanding of the relief ofthe clinical manifestations of the toxic-febrile stateelicited in man by cortisone.

The febrile-toxic state of scrub typhus was re-lieved as rapidly by cortisone administered in con-junction with chloramphenicol as was the similarstate in typhoid fever. However, since the re-sponse of scrub typhus to specific chemotherapy isusually very rapid, the cortisone effect in this dis-ease is primarily of academic interest. Recently,similar findings have been reported (42) in pa-tients with another rickettsial disease, i.e., RockyMountain spotted fever, which generally respondsmore slowly to antibiotic therapy than does scrubtyphus.

TYPHOID FEVER AND SCRUB TYPHUS

SUMMARY

Eighteen typhoid fever patients were treatedbetween the 9th and 16th days of their disease withchloramphenicol and with different schedules oforal cortisone in order to learn more about theeffects of the steroid in this disease. In addition,the responses of eight scrub typhus patients treatedwith a uniform regimen of combined cortisone-chloramphenicol therapy were compared with thoseof a comparable group of seven patients who re-

ceived antibiotic therapy alone.The clinical manifestations of the febrile-toxic

state in typhoid fever and in scrub typhus dis-appear in about six hours after initiation of sucha combined therapeutic regimen. In both diseasesdefervescence is accompanied by a general sense

of well-being and an improved appetite which sim-plifies maintenance of an adequate liquid and ca-

loric intake during the acute phases of the diseases.Control of bacteremia was prompt in typhoid

fever treated with the combined regimen and was

comparable in all respects to cases treated withchloramphenicol alone. The incidences of relapseand of complications did not appear to be influ-enced by the inclusion of cortisone in the thera-peutic regimen.

The antipyretic effect of a single dose of oralcortisone was temporary and fever recurred intyphoid patients unless administration of this hor-mone was continued. Experience with the differ-ent dosage schedules of cortisone in patients withthis disease who were receiving uniform antibiotictherapy indicated that, in order to avoid recrude-scence of symptoms, administration of the steroidmust be continued for a period of time which cor-

responds to that required for patients to becomeafebrile with a regimen consisting solely of chlor-amphenicol, i.e., about four days.

Since the response of scrub typhus to antibiotictherapy is generally very prompt, the accelerateddefervescence afforded by cortisone in this diseaseis primarily of theoretical interest. On the otherhand, selected severe cases of typhoid fever may

benefit by the administration of cortisone duringthe period normally required for the patient to be-come afebrile from the antibiotic therapy. Theimportance of adequate control of infection bysimultaneous specific chemotherapy is emphasizedin patients receiving cortisone.

The mechanisms by which cortisone controls the

toxic-febrile state in these diseases remain un-known; however, the observations of others withanimals are discussed in the light of. our clinicalexperience with men.

ACKNOWLEDGMENTS

The authors are grateful to Dr. J. W. Field, Directorof the Institute for Medical Research, and members of thestaff of the Institute, and to the directors and medical andnursing staffs of the General Hospital, Kuala Lumpur, andthe British Military Hospital, Kinrara, for their unre-served cooperation and generous assistance in this study.

REFERENCES

1. Stuart, B. M., and Pullen, R. L., Typhoid. Clinicalanalysis of three hundred and sixty cases. Arch.Int. Med., 1946, 78, 629.

2. Woodward, T. E., Smadel, J. E., and Ley, H. L., Jr.,Chloramphenicol and other antibiotics in the treat-ment of typhoid fever and typhoid carriers. J. Clin.Invest., 1950, 29, 87.

3. Marmion, D. E., The treatment of typhoid fever withchloramphenicol. A clinical study of 330 casesof enteric fever treated in Egypt. Tr. Roy. Soc.Trop. Med. & Hyg., i952, 46, 619.

4. Kraljevic, R., Perroni, J., Pearson, E., Sesnic, R.,Gonzalez, O., Borel, H., Rojas, M., Jimenez, L.,Cloromicetina y tifoidea. Experiencia sobre 500casos. Rev. Med. de Chile, 1952, 80, 521.

5. Mollaret, P., Reilly, J., Bastin, R., et Tournier, P.,Accidents du traitement des fievres typhoides etparatyphoides par le chloramphenicol (chloromy-c'tine). (Apropos d'une premiere statistique de100 cas.) Bull. et mem. Soc. med. d. h6p. deParis, 1950, 66, 85.

6. Woodward, T. E., Hall, H. E., Dias-Rivera, R., High-tower, J. A., Martinez, E., and Parker, R. T.,Treatment of typhoid fever. II. Control of clinicalmanifestations with cortisone. Ann. Int. Med., 1951,34, 10.

7. Smadel, J. E., Ley, H. L., Jr., and Diercks, F. H.,Treatment of typhoid fever. I. Combined therapywith cortisone and chloramphenicol. Ann. Int.Med., 1951, 34, 1.

8. Ruiz Sanchez, F., El tratamiento de la fiebre tifoidea;principios y normas. Medicina, Mexico, 1952, 32,580.

9. Burton, I. F., and McDermott, M., Observations ofthe efficacy of chloramphenicol (chloromycetin) andACTH in the treatment of typhoid fever in chil-dren. A report of two cases. J. Michigan M. Soc.,1952, 51, 865.

10. Smadel, J. E., Bailey, C. A., and Lewthwaite, R.,Synthetic and fermentation type chloramphenicol(chloromycetin) in typhoid fever: prevention ofrelapses by adequate treatment. Ann. Int. Med.,1950, 33, 1.

11. Smadel, J. E., Ley, H. L., Jr., Diercks, F. H., Pater-son, P. Y., Wisseman, C. L., Jr., and Traub, R.,

273

274 C. L. WISSEMAN, JR., P. Y. PATERSON,J. E. SMADEL, F. H. DIERCKS, ANDH. L. LEY, JR.

Immunization against scrub typhus: duration ofimmunity in volunteers following combined livingvaccine and chemoprophylaxis. Am. J. Trop. Med.Hyg., 1952, 1, 87.

12. Ley, H. L., Jr., Diercks, F. H., Paterson, P. Y.,Smadel, J. E., Wisseman, C. L., Jr., and Traub, R.,Immunization against scrub typhus. IV. LivingKarp vaccine and chemoprophylaxis in volunteers.Am. J. Hyg., 1952, 56, 303.

13. McCrae, T., Typhoid fever. Chapter II in Osler, W.,Modern Medicine, Its Theory and Practice, Phila-delphia and New York, Lea Bros. and Co., 1907,vol. 2, p. 70.

14. Beckermann, F., and Otto, G., tber eine Typhusepi-demie mit 156 Erkrankungen, vorwiegend behandeltmit Chloromycetin. Deutsche med. Wchnschr, 1951,76, 466.

15. Landor, J. V., Typhoid fever with special reference tothe value of new antisera in therapy and eosinopeniain diagnosis. Tr. Roy. Soc. Trop. Med. & Hyg.,1941, 35, 1.

16. Osler, W., The Principles and Practice of Medicine,New York, Appleton and Company, 1892, 31.

17. Landouzy, L., et Siredey, A., etudes des localisationsangio-cardiaques typhoidiques. Leurs consequencesimmediates, prochaines et eloignees. (Atude clini-que et anatomopathologique.) Rev. de med., Paris,1887, 7, 804.

18. Chagas, E., Contribution a l'etude des alterations ducoeur dans la fievre typhoide (etude electrocardio-graphique). Compt. rend. Soc. de biol., 1931, 106,505.

19. Brow, G. R., The heart in typhoid fever. Canad. M.A. J., 1929, 20, 606.

20. Porter, W. B., and Bloom, N., The heart in typhoidfever. A clinical study of 30 patients. Am. HeartJ., 1935, 10, 793.

21. Weller, E., and Kunze, H., Chloromycetinbehandlungdes Typhus abdominalis. Deutsch med. Wchnschr.,1951, 76, 673.

22. Thomas, L., and Good, R. A., Bilateral corticalnecrosis of kidneys in cortisone-treated rabbits fol-lowing injection of bacterial toxins. Proc. Soc.Exper. Biol. & Med., 1951, 76, 604.

23. Woodward, T. E., Smadel, J. E., Parker, R. T., andWisseman, C. L., Jr., Treatment of typhoid feverwith antibiotics. Ann. New York Acad. Sc., 1952,55, 1043.

24. John, A. T., and Vinayagam, V. S., Effect of inter-rupted courses of chloramphenicol on relapse-ratein typhoid fever. Lancet, 1952, 2, 757.

25. Smadel, J. E., Influence of antibiotics on immuno-logical responses in scrub typhus. Ricketts AwardLecture; University of Chicago, May 11, 1953.

26. Smadel, J. E., Ley, H. L., Jr., Diercks, F. H., andTraub, R., Immunity in scrub typhus: resistance toinduced reinfection. Arch. Path., 1950, 50, 847.

27. Lantin, P. T., Gamboa, E. L., and Sadili, F., Studies onchloramphenicol in the treatment of typhoid fever.Am. J. Med. Sc., 1951, 222, 285.

28. Thorn, G. W., Renold, A. E., Wilson, D. L., Frawley,T. F., Jenkins, D., Garcia-Reyes, J., and Forsham,P. H., Clinical studies on the activity of orally ad-ministered cortisone. New England J. Med., 1951,245, 549.

29. Jackson, E. B., and Smadel, J. E., The effects ofcortisone and ACTH on toxins of rickettsiae andSalmonella typhosa. J. Immunol., 1951, 66, 621.

30. Cassoute, E., and Raybaud, A., Insuffisance surrenaleaigue par hemorragie capsulaire unilaterale aucours d'une fievre typhoide. Bull. Soc. pediat. deParis, 1928, 26, 193.

31. Roche, M., Clinical effects of ACTHin typhoid fever,Chapter 27 in Proceedings of the Second ClinicalACTH Conference, Vol. II, Therapeutics, J. R.Mote. ed.. New York, Blakiston, 1951, p. 373.

32. Soylemezoglu, B., and Wells, J. A., Studies on themechanism of the leucocyte response to bacterialpyrogen. J. Pharmacol. & Exper. Therap., 1951,101, 33.

33. Thomas, L., Infectious diseases. The effects of cor-tisone and adrenocorticotropic hormone on infec-tion. Ann. Rev. Med., 1952, 3, 1.

34. Germuth, F. G., Jr., Ottinger, B., and Oyama, J.,The influence of cortisone on the evolution of acuteinfection and the development of immunity. Bull.Johns Hopkins Hosp., 1952, 91, 22.

35. Mogabgab, W. J., and Thomas, L., The effects ofcortisone on bacterial infection; Group A hemolyticstreptococcal infection in rabbits. J. Lab. & Clin.Med., 1952, 39, 271.

36. Martin, S. P., and Kerby, G. P., Effect of adrenalhormone overdosage on bacterial removal by thesplanchnic viscera. Proc. Soc. Exper. Biol. & Med.,1952, 81, 73.

37. Aikawa, J. K., and Harrell, G. T., Effect of cortisoneacetate on experimental Rocky Mountain spottedfever in the guinea pig. Proc. Soc. Exper. Biol.& Med., 1953, 82, 698.

38. Bjorneboe, M., Fischel, E. E., and Stoerk, H. C., Theeffect of cortisone and adrenocorticotrophic hormoneon the concentration of circulating antibody. J.Exper. Med., 1951, 93, 37.

39. Glaser, R. J., and Loeb, L. H., Effect of combinedcortisone and penicillin therapy on experimentalstreptococcal pneumonia. J. Clin. Invest., 1952, 31,631.

40. Lewis, L. A., and Page, I. H., Method of assayingsteroids and adrenal extracts for protective actionagainst toxic material (typhoid vaccine). J. Lab.& Clin. Med., 1946, 31, 1325.

41. Chedid, L., et Boyer, F., Association de la cortisoneau chloramphenicol dans la typhoide experimentaledu Rat. Compt. rend Acad. d. sc., 1952, 234, 2108.

42. Workman, J. B., Hightower, J. A., Borges, F. J.,Furman, J. E., and Parker, R. T., Cortisone as anadjunct to chloramphenicol in the treatment ofRocky Mountain spotted fever. New England J.Med., 1952, 246, 962.

275TYPHOID FEVER AND SCRUB TYPHUS

APPENDIX A.

CHLORAMPHENICOLCOMBINEDWITH DIFFERENT SCHEDULES

OF ADMINISTRATION OF ORAL CORTISONEIN TYPHOID FEVER

G _PATIENT TREATMENT RESPONSETO TREATMENT SEROLOGYR NO0. AGE SEX WT. DAY OF CORTISONE CHILORAMPHENICOL CORTISONE EFFECT LAST DAY COMPLICATIONS TYPHIOIDVbKG.- DISEASE FERtILE AGGLUTININ0 BEGUN MGJ/ KG.

FIS ORETTLBGN-I NE,H.OST TITER+u ON DAY ~~~~~~~~DRUG AFTER AFTER DAYS AFTER TYPE DAY OF

I- GM!o FIRST P FIRST P FIRST R- DISEASE

P lIt ~~~~~~~34 DAYS GM. ACUTE CONy.I 55 F 33 10 10.11 7 55 36 4 I8 5 R 49 40 802 56 F 54 119 16 49.5 61.5 4 28 6 B, C 17,59 1280 64013 32 M41 13512.2 6 19.5 35.5 6 34 4 40 2043 F45 14 11.11 7 22.5 34.5 8 32 4 H _ _ (50 320AVERAGLW 52 50.7 5.5 28 4.8 ___5 10 F 124 52 8.3 6.3 7 171251 6 54 4 JH 256015606j25 M ~56 53 5. 3.6 7 22.5 34.5 4 60 4 IH 15601560

70 M 22 1 6 9.513A4 4 7 13~ 4 60 3 jI9H 19 16018081I50 Ft23 508.76.5 17572518148141 _ _18015609 7M 57. 54 8.8.5.9. 1 7 55.3 p22.5 8 56 4 RH 39 801320

1AVERAGE 53 8.5 5.5 6.0 56 3.8 __ _____50 28, F 33 52 6.7 4.5 2.3. 7 22.5 34.5 8 32 6 80 56055 56 F 45 50 4.4 22 1.7j 7 22.5 34.5 8 48 4 H 560 32052 58 F 40. 9 7.5 &L02.5j 4 53.5 35.5 6 - (so8 80MIK13 551 M 29 1556.9 4.32.6 5 53.0 25.0 4 - (I H 20 8054 551 F 26 551 7.74.8 2.9~ 6 55.0 23.0 4 72 3 H -2560 52801559 F 57 it1418215.911.51 1

7 18.8 54.8 8 156 4 1is H 55 560 180AVERAGE 55.8 7.04.4 23 6.3 52* 2.9

26121 M 1471 11 16.A414312.112.11 4 1 13.5 1 25.5 1 6 1 - I< Ii I 13 I40 I80

57514 F 25 10 6.0j4.0 8.0 7 55.8 57.8 8 j60 4 BIH 110 1 60 320r158 54 F 33 152 4.S 6.5 9.5 8 57 38.5 48 72 4 jBIRPI11,14,39,45 so8 15604KEY TO COMPLICATIONS: R- RELAPSE; B. BRONCHOPN4EUMONIA;C-CHOLECYSTITIS; I's CARDIAC IRREGULARITY;P-INTESTINAL PERFORATION; He-HYPOTHERMIA.*SUM OF DRUG, GIVEN IN FIRST AND SECONDCOURSEAND) IN RELAPSE.

't RECIPROCAL DILUTION.* AVERAGEOF THE FOUR PATIENTS WHOESCAPED FROMCORTISONE.