ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Mar. 1994, p. 558-562 Vol. 38, No.

3
0066-4804/94/$04.00+0
Copyright ) 1994, American Society for Microbiology

Randomized Comparison of Aztreonam and Chloramphenicol

in Treatment of Typhoid Fever
EDUARDO GOTUZZO,1' 2 JUAN ECHEVARRIA,1 2 CARLOS CARRILLO,1,2 JORGE SANCHEZ,
PABLO GRADOS,' CIRO MAGUINA,1 AND HERBERT L. DuPONT3*
Instituto de Medicina Tropical Alexander von Humboldt,' Universidad Peruana 'Cayetano Heredia,'1 and
Departamento de Enfermedades Transmisibles, Hospital Nacional Cayetano Heredia,2 Lima, Peru,
and University of Texas Medical School and School of Public Health, Houston, Texas 770303
Received 29 October 1993/Returned for modification 10 December 1993/Accepted 3 January 1994

Patients with clinical typhoid fever plus a blood, bone marrow, or bile culture positive for SalmoneUa typhi
or SalmoneUla paratyphi were included in an open clinical trial to compare the efficacy of aztreonam (6 g/day [2
g intravenously every 8 h]) given for 10 days with that of chloramphenicol (50 mg/kg of body weight per day
[intravenously or orally]) administered for 14 days. A total of 44 patients, 22 in each group, were included in
the study, and both groups were comparable in terms of baseline parameters. All patients randomized to
receive chloramphenicol completed the 14 days of treatment, while two patients randomized to receive
aztreonam developed an intestinal hemorrhage, and a third patient elected to withdraw from the trial.
Defervescence occurred more quickly in the subjects receiving chloramphenicol than in those receiving
aztreonam (P < 0.05). All patients in the chloramphenicol group were clinically cured during therapy, while
four patients (21%) in the group receiving aztreonam were declared clinical treatment failures. None of the 19
patients receiving aztreonam, compared with 7 of 22 (32%) patients receiving chloramphenicol, had a positive
blood culture after 24 h of therapy (P < 0.05). Adverse experiences were unusual and mild. In the study,
aztreonam was less effective than chloramphenicol with regard to clinical effectiveness and time of deferves-
cence but was more effective in the elimination of the infecting SalmoneUla organisms from the bloodstream.

In excess of 12.5 million cases of typhoid fever occur peared to be effective for the therapy of typhoid disease (7, 18).
annually throughout the world (5). The expected mortality rate The present study was carried out to examine its comparative
without therapy of 15 to 30% (3, 13) can be reduced to very low therapeutic efficacy in a clinical trial employing chloramphen-
levels with antimicrobial therapy. The conventional drug of icol as standard therapy.
choice worldwide has been chloramphenicol because of its
excellent bioavailability by the oral route, its low cost, and its
clinical efficacy (11, 17, 19). Unfortunately, therapy with chlor- MATERUILS AND METHODS
amphenicol has little effect on the recurrence of disease Patients 14 years of age or older with clinical typhoid fever
relapse or chronic intestinal carriage of Salmonella typhi (temperature higher than 38.3°C, systemic toxicity, headache,
posttherapy (9, 10). Because of the drug's low cost, these anorexia, diarrhea, constipation, or presence of rose spots) and
limitations had been accepted by many public health officials a positive bone marrow, bile (obtained by Entero-Test), or
and clinicians in the developing world until drug resistance
began to occur. In some areas of the world, resistance of S. blood culture were admitted to Hospital Nacional Cayetano
typhi to each of the three standard antityphoid drugs (chlor- Heredia in Lima, Peru, and were entered into the trial.
amphenicol, ampicillin, and trimethoprim-sulfamethoxazole) Exclusion criteria included the following: complicated typhoid
has become common (1, 8). fever (bowel perforation, gastrointestinal bleeding, endocardi-
Between 2 and 4% of patients with typhoid fever become tis, meningoencephalitis, etc.), previous hypersensitivity to
chronic intestinal carriers of S. typhi for more than 1 year (10). beta-lactam antimicrobial agents, and receipt of antityphoid
Typhoid carriers are important to the epidemiology of this antimicrobial agents, including ampicillin, amoxicillin, tri-
infection, in which humans represent the exclusive reservoir of methoprim-sulfamethoxazole, or a fluoroquinolone or broad-
the organism. New chemotherapeutic drugs are needed that spectrum cephalosporin. Lactating or pregnant females were
would be more predictably effective in areas where resistance also excluded. Patients meeting the enrollment criteria were
is reported and that, hopefully, would decrease the occurrence randomized to receive aztreonam (2 g intravenously every 8 h)
of posttreatment typhoid relapses and chronic intestinal car- or chloramphenicol (50 mg/kg of body weight per day in four
riage of the organism. divided doses, with a maximum daily dose of 3 g/day intrave-
Because of its in vitro and in vivo activity against bacterial nously or orally). The durations of therapy were 10 days for the
enteropathogens (8), aztreonam was examined as an antity- aztreonam group and 14 days for the chloramphenicol group.
phoid drug in clinical studies with patients with typhoid fever. Patients were kept in the hospital for the duration of the trial.
In two preliminary uncontrolled clinical trials, the drug ap- Immediately before initiation of therapy, blood, stool, bile
(through an Entero-Test small bowel study), and bone marrow
samples were obtained for culturing S. typhi. During therapy,
*
Corresponding author. Mailing address: University of Texas- blood cultures were obtained on days 1, 2, 3, and 4 1 h before
Houston, Center for Infectious Diseases, Medical School/School of the first dose of the following day of treatment (5 to 10 ml of
Public Health, 6431 Fannin, 1.729 JFB, Houston, TX 77030. Phone: blood was placed in 30 ml of culture broth), and stools were
(713) 794-4254. Fax: (713) 792-4937. cultured on days 1, 2, 3, 4, and 5. After the end of treatment,
558
VOL. 38? 1994 AZTREONAM VERSUS CHLORAMPHENICOL AGAINST TYPHOID FEVER 559
blood cultures were obtained from outpatients on days 1, 2, 7, TABLE 1. Baseline microbiologic and serologic findings from
and 14 posttreatment, and stools were cultured on days 1, 2, 7, comparison of aztreonam and chloramphenicol
and 21 posttherapy. Patients then returned for clinical assess- typhoid fever therapy
ment 8 weeks after completing therapy. No. positive/no. tested (LX) in:
The trial was a two-center study which was to include a total
of 134 patients. A randomization table (randomization of 1:1)
Test Aztreonam Chloramphenicol
group group
was provided by the sponsor. Investigators were not able to see
the assignments for treatment prior to enrollment of the Blood culture 19/22 (86) 12/22 (54)"
subjects. The study was stopped at both sites when each found Bone marrow culture 22/22 (100) 21/22 (95)b
an inordinate number of failures in the aztreonam group. The Bile (Entero-Test) culture 11/16 (69) 12/21 (57)"
sample size for the study was statistically adequate on the basis Stool culture 7/22 (32) 8/21 (38)"
of the expected clinical cure rates of 100% for chlorampheni- Typhoid agglutination 15/21 (71) 17/22 (77)"
col and between 75 and 80% for aztreonam. With an overall (O agglutinin > 1/80)
statistical significance limit set at P = 0.05 (two sided) and a Salmonella sp. identified
power of 80%, the numbers of subjects necessary to detect the S. typhi 17/22 (77) 17/22 (77)"
differences are between 15 and 20 in each group. S. paratyphi 5/22 (23) 5/22 (23)"
To be included in bacteriologic analysis, patients needed to
have had at least two blood cultures performed during therapy {Pp < O.t)5.
'P, not significant.
and immediately posttreatment. The age, sex, weight, height,
duration (days) of fever, and clinical severity of the illness as
determined by the admitting physician were assessed and
compared between groups. The clinical parameters compared trial and serve as the basis of this study. Three subjects
between groups were hours until defervescence, clinical cure randomized to the aztreonam group were removed from the
(defervescence and resolution of preenrollment symptoms), clinical efficacy analysis because they failed to complete the
clinical failure (continued fever after 10 days of therapy), and required therapy. One of the subjects withdrew voluntarily
typhoid relapse (defined as clinical symptoms of typhoid fever from the study on day 4, and two were removed from the trial
with isolation of S. typhi or Salmonella paratyphi in blood (one on day 1 and the other on day 4 of therapy) because they
within 8 weeks after completing treatment). Bacteriologic developed gastrointestinal bleeding. These subjects were re-
response was defined as negative blood cultures during ther- moved from the efficacy analysis.
apy, and microbiologic relapse was defined as patients who had The groups (aztreonam versus chloramphenicol, respec-
a bacteriologic response but who had reappearance of the S. tively) were comparable at enrollment in terms of mean age
typhi in posttherapy blood samples. Convalescent carriage was (25.3 versus 21.9 years), mean body weight (52.3 versus 52.7
defined as reappearance of S. typhi in posttherapy stool kg), mean body height (156.4 versus 157.9 cm), complaints of
samples within the first 8 weeks after completion of therapy but malaise (82 versus 68%), headache (68 versus 77%), diarrhea
having negative cultures without further therapy by 3 months. (32 versus 45%), abdominal pain or cramps (36 versus 41%),
Patients were removed from the study if they had blood presence of fever (100% for both groups), hepatomegaly (86
cultures positive for the infecting organism after 5 days of versus 95%), splenomegaly (13 versus 4%), and the presence
therapy or if they experienced a severe, adverse event during of rose spots (23 versus 32%). Twelve subjects (55%) random-
therapy. ized to receive aztreonam were females, while 15 (68%) of the
To monitor for potential adverse drug effects, a complete chloramphenicol-treated subjects were females. The clinical
blood count and serum chemistry tests (including serum glu- severity of the patients as judged by the admitting physician
tamic oxalacetic transaminase, serum glutamic pyruvic differed slightly: 16 (73%) of the aztreonam-treated patients
transaminase, total bilirubin, alkaline phosphatase, creatinine, were felt to have mild illness, and 6 (27%) were felt to have
fasting blood glucose, total protein, albumin, sodium, and moderate illness; 11 (50%) of the chloramphenicol-treated
potassium) were performed pretreatment, on day 7, on the last patients were judged to have mild illness, and 11 (50%) were
day of therapy, and at the 7th day posttreatment follow-up. judged to have moderate illness.
Antimicrobial susceptibility testing with the Salmonella isolates By the enrollment criteria, each of the patients had a
was performed with agar dilution to determine MICs (14). positive isolate of S. typhi or S. paratyphi by blood, bone
Serum samples were drawn at admission for antityphoid marrow, or bile (Entero-Test) culture. The biologic fluid with
agglutinins (O and H), and a titer of anti-O agglutinin of -1:80 the highest yield was bone marrow aspirate, with bile and
was considered to be positive. blood cultures being positive with about equal frequencies
Analysis of variance for continuous variables and chi-square (Table 1). A significantly lower frequency of occurrence of
and Mantel-Haenszel tests for discrete variables were used in positive blood culture was seen in the patients randomized to
the analyses, employing the SPSS/PC+ program. receive chloramphenicol than in those randomized to receive
The study was reviewed and approved by the Ethical Insti- aztreonam (P < 0.05). Of the Salmonella organisms isolated
tutional Review Board of the Universidad Peruana 'Cayetano from the patients, 17 in both groups (77%) were S. typhi and 5
Heredia.' All patients included in the study gave written (23%) were S. paratyphi. All Salmonella isolates were found to
consent. be susceptible to aztreonam (MIC for 90% of the strains, 0.25
jLg/mI) and chloramphenicol (MIC for 90% of the strains, 8
RESULTS p.g/ml). Typhoid 0 agglutinins were detected in 71 and 77% of
the patients randomized to receive aztreonam and chloram-
A total of 52 patients were enrolled in the trial. Eight phenicol, respectively.
patients were later removed from the study, because five All patients randomized to receive chloramphenicol com-
patients had negative baseline cultures, and three patients pleted the 14 days of treatment, while two who received
elected not to enter the trial. A total of 44 patients, 22 in each aztreonam developed intestinal hemorrhages (one on day 1
group, with culture-proven typhoid fever were enrolled in the and the second on day 4 of therapy), and a third patient elected
560 GOTUZZO ET AL. ANTIMICROB. AGENTS CHEMOTHER.

TABLE 2. Distribution of febrile patients by day of treatment' TABLE 4. Microbiologic response to aztreonam and
chloramphenicol typhoid fever therapy
No. with fever/no. treated (%) in:
Treatment No. positive/no. treated (%) in:
day Aztreonam Chloramphenicol Chloramphenicol
group group Test Aztreonam
group group
1 22/22 (100) 22/22 (100)
2 21/21 (100) 20/22 (91) Blood culture
3 20/21 (95) 18/22 (82) Day of treatment
4 18/19 (95) 14/22 (64) 1 0/19 (0) 7/22 (32)-
5 15/19 (79) 10/22 (45) 2 0/19 (0) 3/22 (13)
6 12/19 (63) 4/22 (18? 3 0/19 (0) 2/22 (9)
7 11/19 (58) 2/22 (9) 4 0/19 (0) 1/22 (5)
8 8/19 (42) 0/22 (0) Day posttreatment
9 4/19 (21) 0/22 (0) 1 0/15 (0) 0/22 (0)
7 0/15 (0) 1/19 (5)
a Temperatures were taken each 6 h during therapy. Fever was defined as an 21 0/15 (0) 0/19 (0)
oral temperature of >37.5°C. To be considered afebrile, the patient had to have
a temperature of '37.5°C and be without fever thereafter during treatment.
"P < 0.001. Stool culture
Day of treatment
1 1/19 (5) 9/22 (41)
2 7/19 (37) 3/22 (14)
to withdraw from the trial without explanation on day 4. Each 3 4/19 (21) 3/22 (14)
of these three patients was withdrawn from the trial and 4 2/19 (11) 1/22 (5)
excluded from the efficacy analysis. Defervescence occurred 5 1/19 (5) 1/22 (5)
more quickly in the subjects receiving chloramphenicol than in Day posttreatment
those receiving aztreonam. The average time from initiation of 1 2/15 (13) 1/15 (7)
7 1/15 (7) 1/19 (5)
therapy until defervescence was 6.6 ± 3.58 days in the aztreo- 21 0/15 (0) 1/19 (5)
nam group, with the corresponding time in the chloramphenicol
a p < 0.01.
group being 4.5 ± 2.3 days (P < 0.03). By analysis of variance,
adjusting for sex, day of treatment, and culture positivity, there
was a statistical difference between the durations of fever
between the two treatment groups (P < 0.001). The number of elimination of organisms from blood than those in the chlor-
febrile patients was greater on each day of therapy in the amphenicol group. None of the 19 patients completing a
aztreonam group than in the chloramphenicol group (Table 2). course of aztreonam, compared with 7 of 22 patients receiving
This difference was significant beginning with the fifth day of chloramphenicol, had a positive blood culture after 24 h of
treatment. The difference in the persistence of fever in the therapy (Table 4) (P < 0.01). The same difference, although
groups was highly significant (P < 0.001). No patient in the less obvious, was seen for days 2 through 4 of therapy. Relapses
chloramphenicol group was febrile after the eighth day, while 8 did not occur in the aztreonam group. Stool cultures became
of 19 (42%) patients in the aztreonam group remained febrile negative more slowly than the blood cultures. At the end of
on the ninth day of therapy. Four patients of the aztreonam treatment (days 1 and 2 posttreatment), 2 of 15 (13%) patients
group (21%) remained with fever through 10 days of therapy, who completed their treatment in the aztreonam group and 1
when the study drug was changed to chloramphenicol. of 22 (5%) patients of the chloramphenicol group had had a
Twenty-two of 22 (100%) patients receiving chlorampheni- stool culture positive for Salmonella infection. One patient in
col were clinically cured, while 7 patients (32%) were declared each group had a stool culture positive for Salmonella infection
treatment failures in the group receiving aztreonam (Table 3). at the seventh day of follow-up, and one patient in the
One patient in the chloramphenicol group developed a bacte- chloramphenicol group was positive at the 14-day follow-up.
riologic and clinical relapse during the 8-week follow-up. Adverse experiences in either group were unusual and mild.
Patients receiving aztreonam appeared to have more rapid The reactions encountered in the aztreonam versus chloram-
phenicol groups, respectively, were as follows: paresthesia, 3 of
19 (16%) versus 0 of 22; acne, 0 of 19 versus 2 of 22 (9%);
TABLE 3. Clinical responses with aztreonam and chloramphenicol arthralgia, 0 of 19 versus 1 of 22 (5%); leukopenia, 4 of 19
typhoid fever therapy (21%) versus 6 of 22 (27%); increased serum glutamic oxal-
acetic transaminase, 4 of 19 (21 %) versus 0 of 22; and
No. with status/no. treated (%) in: increased alkaline phosphatase, 2 of 19 (10%) versus 0 of 22.
Patient's status Aztreonam Chloramphenicol No important changes or significant differences between treat-
group group ment groups in hematocrit or hemoglobin values were seen
Completed therapy 19/22 (86)- 22/22 (100) during therapy.
Was cured 15/22 (68)b 22/22 (100)
Was clinical failure 7/22 (32)b 0/22 (0)C DISCUSSION
Failed to complete triald 3/22 (14) 0/22 (0)
Had typhoid relapse 0/19 (0) 1/22 (5) During the last 20 years, reports of the occurrence of
a p, not significant. chloramphenicol-resistant S. typhi have occurred in virtually all
b Denominator includes the three patients that were dropped from analysis. parts of the world (1, 5, 6, 8). In a number of the reports,
P < 0.01 by Fisher's exact test. multiresistant typhoid strains have been identified against
d Two
patients were removed from the trial because of the occurrence of which none of the standard antityphoid drugs (including ampi-
gastrointestinal hemorrhage on day 1 for one patient and on day 4 for the second,
and one patient requested (on day 4) to be removed from the trial for unknown cillin and trimethoprim-sulfamethoxazole) were active. This
reasons. and the ineffectiveness of standard therapy (chloramphenicol)
VOL. 38? 1994 AZTREONAM VERSUS CHLORAMPHENICOL AGAINST TYPHOID FEVER 561

in preventing intestinal carriage of S. typhi or typhoid relapse trated within the intracellular tissues serving as the habitat of
make the search for effective antityphoid drugs essential. the infecting salmonellae.
Among the new drugs being examined are those with in vitro Recent studies indicate that aztreonam shows a high degree
activity against S. typhi and S. paratyphi, including broad- of in vitro activity against enteric bacterial pathogens (8, 16).
spectrum cephalosporins, the new fluoroquinolones, and az- The drug has been shown, furthermore, to be effective in the
treonam (2, 6, 12, 15). Two recent reports suggested that treatment of bacterial enterocolitis of U.S. travelers to Mexico
aztreonam might be of value in the treatment of typhoid fever when given orally (4). Further studies to examine the effect of
(7, 18). One of these studies was with children treated in a aztreonam on other forms of bacterial enteric infection are
randomized clinical trial in Mexico City (18), and the second indicated. Considering its rapid clearance from the blood-
was an open trial with adults treated in Egypt (7). In the study streams of patients with typhoid fever, it is suggested that the
in Mexico, 36 children received either aztreonam or chloram- value of the drug in treating nontyphoid Salmonella infections,
phenicol. Clinical cure was seen in 18 of 18 patients randomized such as Salmonella enteritidis gastroenteritis, with or without
to aztreonam. These children were given 150 mg/kg/day, sug- bacteremia, and extraintestinal infection, including central
gesting that perhaps a higher dose of aztreonam might be more nervous system infection in young infants and systemic salmo-
effective. In the study carried out in Egypt, only four patients nellosis in patients infected with the human immunodeficiency
were treated with aztreonam (same dose as in the present trial) virus, be determined.
after undergoing an initial trial with chloramphenicol. One of
the strains was resistant to chloramphenicol in the Egypt study.
The present study was a controlled randomized trial com- ACKNOWLEDGMENT
paring aztreonam with chloramphenicol. The study confirmed This study was supported by a grant from Bristol-Myers Squibb,
the value of bone marrow aspirate cultures in establishing the Princeton, N.J.
diagnosis of typhoid fever. Among 44 patients with invasive
disease due to chloramphenicol-susceptible typhoid Salmo- REFERENCES
nella organisms, bone marrow cultures were positive in 43 of 44
(98%) patients, blood cultures were positive in 31 of 44 (70%) 1. Anderson, E. S., and H. R. Smith. 1972. Chloramphenicol resis-
tance in the typhoid bacillus. Br. Med. J. 3:329-331.
patients, bile cultures by the string test (Entero-Test) were 2. Cherubin, C. E., R. H. K. Eng, S. M. Smith, and E. J. C. Goldstein.
positive in 23 of 37 (62%) patients, and stool cultures were 1988. Cephalosporin therapy for salmonellosis: questions of effi-
positive in 15 of 43 (35%) patients. Despite a high degree of cacy and cross resistance with ampicillin. Arch. Intern. Med.
susceptibility of all strains to aztreonam (MIC for 90% of the 146:2149-2152.
strains, 0.25 [Lg/ml), there was a prolonged period of fever in 3. Christie, A. B. 1977. Infectious diseases. Churchill Livingstone,
the aztreonam-treated patients, with an average of 2 days Inc., Edinburgh.
longer than was seen in the chloramphenicol-treated subjects; 4. DuPont, H. L., C. D. Ericsson, J. J. Mathewson, F. J. de la Cabada,
in addition, the percentage of febrile patients through treat- and D. A. Conrad. 1992. Oral aztreonam, a poorly absorbed yet
ment was higher in the aztreonam group than in the chloram- effective therapy for bacterial diarrhea in U.S. travelers to Mexico.
JAMA 267:1932-1935.
phenicol group. Four of 19 (21%) patients receiving aztreo- 5. Edelman, R., and M. M. Levine. 1986. Summary of an interna-
nam, versus none of 22 receiving chloramphenicol, remained tional workshop on typhoid fever. Rev. Infect. Dis. 8:329-349.
febrile for 10 days of therapy and were declared treatment 6. Eykyn, S. J., and H. Williams. 1987. Treatment of multiresistant
failures (P = 0.04). Two patients in the aztreonam group were Salmonella typhi with oral ciprofloxacin. Lancet ii:1407-1408.
excluded because of intestinal bleeding. If the three patients 7. Farid, Z., N. I. Girgis, M. Kamal, E. Bishay, and M. E. Kilpatrick.
prematurely dropped from the study after being randomized to 1990. Successful aztreonam treatment of acute typhoid fever after
receive aztreonam are included in the analysis, the rate of chloramphenicol failure. Scand. J. Infect. Dis. 22:505-506.
failure for aztreonam increases to 7 of 22 (32%) (P < 0.01). 8. Goldstein, F. W., J. C. Chumpitaz, J. M. Guevara, B. Papadopou-
Aztreonam appeared to have a more rapid effect on elimi- lou, J. F. Acar, and J. F. Vieu. 1986. Plasmid-mediated resistance
to multiple antibiotics in Salmonella typhi. J. Infect. Dis. 153:261-
nation of the infecting Salmonella organisms from the blood- 266.
stream than chloramphenicol did. In 100% of the patients 9. Gotuzzo, E., L. Benavente, J. Guerra, C. Maguifia, and C. Carrillo.
receiving aztreonam, the infecting S. typhi organisms were 1988. Relapse in typhoid fever, abstr. 496, p. 44. Proceedings of the
eradicated from blood cultures during the first day of treat- International Congress for Infectious Diseases, Rio de Janeiro,
ment. This is in contrast to a culture positivity rate for day 1 Brazil. International Society for Infectious Diseases.
blood cultures of 32% for the chloramphenicol-treated pa- 10. Guerra, J., L. Benavente, E. Gotuzzo, 0. Grados, and N. Bravo.
tients (P < 0.01). This may have been even more important, 1987. Predictive value of stoolculture and bileculture post-treat-
considering the higher rate of pretreatment blood culture ment of typhoid fever (TF) in the identification of relapse and
positivity in the aztreonam group than in the chloramphenicol chronic carrier state (CC), abstr. 1169. Program Abstr. 27th
Intersci. Conf. Antimicrob. Agents Chemother. American Society
group (86 versus 54%, respectively; P < 0.05). Interestingly, for Microbiology, Washington, D.C.
the four patients declared treatment failures, with fever per- 11. Herzog, C. H., and A. M. Geddes. 1982. Chloramphenicol in the
sisting throughout the 10 days of aztreonam treatment, had treatment of enteric fever. Trans. R. Soc. Trop. Med. Hyg.
consistently negative blood cultures. 26:848-849.
Comparing the present study with previous clinical trials 12. Islam, A., T. Butler, S. K. Nath, N. H. Alam, K. Stoeckel, H. B.
carried out at this institution, the rate of cure for aztreonam Houser, and A. L. Smith. 1988. Randomized treatment of patients
was lower than that seen for standard therapy, but the bacte- with typhoid fever by using ceftriaxone or chloramphenicol. J.
riologic response to aztreonam was superior. One possible Infect. Dis. 158:742-747.
13. Kumate, J. 1977. Manual de infectologia. Hospital Infantil de
explanation for the findings of the various studies is that Mexico, Mexico City.
aztreonam is able to clear Salmonella organisms from the 14. National Committee for Clinical Laboratory Standards. 1990.
blood of bacteremic patients, but it is less effective in eradica- Performance standards for antimicrobial disk susceptibility tests,
tion of the organism from the intracellular location character- 4th ed. Approved standard M2-A4. National Committee for
istic of the later stages of typhoid infection. We speculate that Clinical Laboratory Standards, Villanova, Pa.
aztreonam may not have adequately penetrated and concen- 15. Pape, J., H. Gerdes, L. Oriol, and W. D. Johnson, Jr. 1986.
562 GOTUZZO ET AL. ANTIMICROB. AGENTS CHEMOTHER.

Typhoid fever: successful therapy with cefoperazone. J. Infect. Dis. teric fever. N. Engl. J. Med. 278:171-176.
153:272-276. 18. Tanaka-Kido, J., L. Ortega, and J. I. Santos. 1990. Comparative
16. Paradelis, A. G., A. G. Stathopoulos, J. Tselentis, K. Delidou, A. efficacies of aztreonam and chloramphenicol in children with
Stravrakis, and L. G. Crassaris. 1985. Comparative antibacterial typhoid fever. Pediatr. Infect. Dis. J. 9:44-48.
activity of aztreonam and other antimicrobial agents against gram 19. Woodward, T. E., J. F. Smadel, H. L. Ley, Jr., R. Green, and D. S.
negative microorganisms. Chemioterapia 4(Suppl. 1):14-22. Mankikar. 1948. Preliminary report on the beneficial effect of
17. Robertson, P., M. F. Abdel Wahab, and F. 0. Raasch. 1968. chloromycetin in the treatment of typhoid fever. Ann. Intern.
Evaluation of chloramphenicol and ampicillin in Salmonella en- Med. 29:131-134.