We studied 149 children and adolescents 3–17 years of age with clinical typhoid fever who were treated with either oral azithromycin (20 mg/kg per day; maximum dose, 1000 mg/day) or intravenous ceftriaxone (75 mg/day; maximum dose, 2.5 g/day) daily for 5 days. Blood and stool specimens were obtained for culture before the initiation of therapy and were repeated on days 4 and 8 of treatment. Isolation of Salmonella enterica serovar Typhi or S. enterica serovar Paratyphi from the initial culture was required for inclusion in the final analysis. S. Typhi was isolated from 68 patients, 32 of whom were receiving azithromycin. Cure was achieved in 30 (94%) of 32 patients in the azithromycin group and in 35 (97%) of 36 patients in the ceftriaxone group (P = NS). Mean time to clearance of bacteremia was longer in the azithromycin group than in the ceftriaxone group. No patient who received azithromycin had a relapse, compared with 6 patients who received ceftriaxone. A 5-day course of azithromycin was found to be an effective treatment for uncomplicated typhoid fever in children and adolescents.
Salmonella enterica serovar Typhi, the agent of typhoid fever, is estimated to cause ∼16 million cases of infection annually, predominantly in the developing world . Emergence of multidrug-resistant (MDR) S. Typhi has complicated therapy by limiting treatment options . Reports of infection with fluoroquinolone-resistant Salmonella species have raised concern that soon no available oral medication will exist to treat the infection [3, 4]. Although ceftriaxone and other third-generation cephalosporins are still highly effective against S. Typhi, they are considered to be less than ideal routine treatments [5–7]. In addition to the high cost and requirement of parenteral administration associated with ceftriaxone, S. Typhi isolates resistant to this drug have begun to appear . Therefore, other regimens are required for the treatment of typhoid fever.
Azithromycin, a member of the macrolide class of antibiotics, possesses many characteristics for effective and convenient treatment of typhoid fever, including in vitro activity against many enteric pathogens, excellent penetration into most tissues, and achievement of concentrations in macrophages and neutrophils that are >100-fold higher than concentrations in serum [9–11]. Previous studies have demonstrated that a 7-day treatment course of azithromycin was highly effective against uncomplicated typhoid fever in adults and children [12–14].
A recent report from Vietnam demonstrated that the duration of azithromycin therapy for uncomplicated typhoid fever in adults could be decreased to 5 days . The encouraging results from this trial prompted us to test whether a shorter treatment course could also be used in children and adolescents. We now report the results of a study of the efficacy of a 5-day course of azithromycin for treating uncomplicated typhoid fever in children and adolescents.
Patients, Materials, and Methods
Study site. Our study was conducted at the Abbassia Fever Hospital (Cairo, Egypt), a 1000-bed primary care and referral hospital for treating patients with infectious diseases who live throughout Egypt.
Study population. Individuals 3–17 years of age who were suspected of having typhoid fever were admitted to a single hospital ward where they were evaluated by a study physician. Patients with documented fever (a rectal temperature >38.0°C or an oral temperature >37.5°C) and ⩾2 of the following symptoms were eligible for enrollment in the study: abdominal tenderness, hepatomegaly, splenomegaly, and/or a coated tongue. Patients were excluded from the study if any of the following criteria were present: allergy to ceftriaxone or macrolides, major typhoid fever—associated complications, inability to swallow oral medication, significant underlying illness, pregnancy and/or lactation, and treatment within the past 4 days with an antibiotic that may be effective against S. Typhi.
Parents or guardians interested in enrolling their child received a detailed explanation of the study, and any questions they had were completely answered. Illiterate parents and guardians had the consent documents read to them, and a witness unassociated with the study was present during the consent process. After the explanations and answers were provided, informed consent was obtained, and subjects were randomized to a treatment group.
Ethics compliance. The study was performed in accordance with the ethical standards of the Institutional Review Board of the US Naval Medical Research Unit #3 (Cairo, Egypt) and under the guidelines of the US Department of Defense, which conform to the Helsinki Declaration of 1975, as revised in 1983.
Sample size requirements. The study was designed to test the therapeutic equivalence of ceftriaxone and azithromycin using the methods of Blackwelder . Nonequivalence was defined as a difference of ⩾20% in clinical cure rates between the 2 treatment groups, assuming that ceftriaxone would have a cure rate of 90%. Using a type I error rate of 0.05 and a type II error rate of 0.2, thirty-three evaluable subjects (i.e., those with S. Typhi isolates recovered from cultures of blood or stool specimens obtained at study enrollment) were needed in each treatment arm. Enrollment was stopped once the necessary sample size was reached.
Randomization and treatment. Patients were stratified into 3 age groups (3–6 years, 7–11 years, and 12–17 years) to control for age bias. A random-number generator was used to determine treatment assignments within blocks of 8 patients. Sequentially numbered sealed envelopes containing the name of the study drug were then created. Treatment assignments were made by opening the lowest numbered envelope that had yet to be used in the patient's age group. Before randomization, neither the patient nor the study physician were aware of the drug that would be administered in each case.
After assignment, patients were treated once daily for 5 days with either oral azithromycin suspension (20 mg/kg/day; maximum dose, 1000 mg/day) or intravenous ceftriaxone (75 mg/day; maximum dose, 2.5 g/day) in an open-label format. All doses of both study medications were administered in the hospital by the nursing staff.
During hospitalization, vital signs (including body temperature) were measured every 8 h, and a clinical examination based on a structured form was performed daily. All patients were asked to return to the hospital 1 month after discharge for a final examination or, if they became sick before the scheduled follow-up visit, to return immediately.
Procedures. Patients were hospitalized during the entire treatment period and for 3 days after therapy was completed. Before initiation of antibiotic therapy, stool and urine samples and 2 blood specimens were obtained, and routine bacterial cultures were performed. Additional baseline laboratory testing included urinalysis, complete blood cell count, and serum chemistry analysis.
Blood cultures were performed for all patients after they received 3 days of antibiotics and again 3 days after antibiotics were discontinued (day 8 of the study). If results of baseline tests were abnormal, stool and urine cultures and urine analysis were repeated 3 days after initiation of antibiotics. A second complete blood cell count and serum chemistry analysis were performed on day 8 for all patients. A stool specimen was obtained for culture for all patients 1 month after initiation of therapy.
Any patient with fever or other symptoms suggestive of typhoid fever ⩽1 month after receiving therapy was evaluated and had a blood culture performed to determine whether there had been a relapse of typhoid fever. Other samples were obtained for laboratory analysis as clinically indicated.
Specimen processing. Standard clinical methods were used to culture the blood, stool, and urine specimens. All blood cultures were blindly subcultured after 1, 7, and 14 days of incubation and whenever the broth appeared to be cloudy. Stool specimens were plated on MacConkey agar and Salmonella-Shigella agar. Bacterial colonies in blood, stool, or urine cultures that were suggestive of Salmonella species were further evaluated using standard methods to confirm their identification .
Testing of S. Typhi isolates for susceptibility to ciprofloxacin, chloramphenicol, ampicillin, and trimethoprim-sulfamethoxazole was performed using the Kirby-Bauer method, and susceptibility to azithromycin and ceftriaxone was determined using Etest strips (AB Biodisk) . NCCLS guidelines were used to assign susceptibility and resistance breakpoints for each antibiotic tested .
Definitions. Clinical cure was defined as the resolution of all typhoid-related symptoms within 7 days of initiating antibiotic therapy. Clinical failure was defined as the persistence of ⩾2 typhoid-related symptoms or signs present at study entry or as the development of a typhoid-related complication. Clinical improvement was defined as partial resolution of illness. A microbiological cure was defined as a sterile blood culture on day 8 (3 days after discontinuation of antibiotic therapy). Microbiological failure was defined as an S. Typhi—positive blood culture on day 8. Patients with S. Typhi in their blood 3 days after initiating antibiotic therapy were said to have persistent bacteremia. Clinical relapse was defined as recurrence of fever and clinical features of typhoid within 30 days of completing therapy, along with isolation of S. Typhi from the blood.
Statistical analysis. The Mantel-Haenszel χ2 test of proportions was used to determine significant differences in cure rates between treatment groups. For groups with <5 events in a cell, Fisher's exact test was used. SAS, version 8 (SAS Institute), was used for all statistical analyses. For all statistical testing, the single-tailed level of statistical significance was set at P = .05.
One hundred twenty-eight patients (83 males) with clinical typhoid fever were enrolled in the study. Of these, 68 patients (32 of whom were in the azithromycin group) had S. Typhi isolated from cultures of blood or stool specimens obtained at enrollment and constituted the treatment group. Analysis of demographic characteristics and results of pretreatment laboratory tests revealed no statistically significant differences between patients treated with ceftriaxone and those treated with azithromycin (table 1).
Both antibiotic therapies were highly effective (table 2). Mean time (±SD) to defervescence was 4.5 ± 1.9 days for patients who received azithromycin and 3.6 ± 1.6 days for patients who received ceftriaxone (P = NS). Clinical cure was achieved in 30 (94%) of 32 patients treated with azithromycin and in 35 (97%) of 36 patients treated with ceftriaxone (P = .5). Both clinical failures in the azithromycin group were due to mild gastrointestinal symptoms that resolved 7 days after treatment was started without any additional treatment. Persistent fever caused the single treatment failure in the ceftriaxone group, but the fever resolved without additional therapy.
Microbiological cure was achieved in every patient treated with azithromycin and in 35 (97%) of 36 patients treated with ceftriaxone (P = .5). The single patient who did not respond to therapy was clinically healthy and, after receiving a second course of antibiotic therapy (chloramphenicol), achieved a complete cure, including sterilization of the blood.
Antibiotic susceptibility testing was performed on all the S. Typhi isolates, with >90% demonstrating susceptibility to all antibiotics tested. According to NCCLS guidelines, no isolate was determined to be resistant to either ceftriaxone or ciprofloxacin, 1 isolate was resistant to trimethoprim-sulfamethoxazole, 2 were resistant to chloramphenicol, and 3 were resistant to ampicillin. Only 1 isolate was MDR, with resistance to ampicillin, chloramphenicol, and trimethoprim-sulfamethoxazole. For azithromycin, the MIC90 was 6 µg/mL, and 7 isolates had an MIC of ⩾8 µg/mL, classifying them as resistant. Four of the “resistant” isolates were from patients randomized to receive azithromycin, and all of these patients achieved clinical and microbiological cure without deviating from the original treatment regimen.
Among patients treated with ceftriaxone, none had S. Typhi recovered from blood cultures after having received 3 days of therapy. In contrast, 12 patients (37.5%) treated with azithromycin still had S. Typhi recovered from their blood 3 days after initiation of antibiotic therapy (P = .0001). Results of antibiotic susceptibility tests revealed that all 12 isolates were susceptible to azithromycin. Despite the persistent bacteremia, all 12 patients had clinically improved by the time samples were obtained for culture, and all were asymptomatic by the time the culture result was known. Cultures of blood samples that were obtained on the day that the previous culture was found to be positive were sterile in every case.
After hospital discharge, 6 patients from the ceftriaxone group returned before their scheduled 1-month follow-up visit because of recurrence of typhoid fever—related symptoms. Cultures were performed, and 5 patients again had S. Typhi recovered from their blood, which indicated a relapse of infection. All 5 patients were treated with a second course of antibiotics, with resolution of their symptoms and sterile blood cultures after completion of the treatment regimen. None of the 5 isolates had developed resistance to any antibiotic tested, including ceftriaxone. No relapses occurred in the azithromycin group.
Fifty-seven of the 68 culture-positive patients (25 of 32 azithromycin recipients and 32 of 36 ceftriaxone recipients; P = .2) returned for the scheduled follow-up evaluation 1 month after completion of treatment, and, with the exception of the 6 described above, all were healthy at this visit. In addition, no child was found to have S. Typhi in stool specimens at the 1-month follow-up visit.
No serious adverse events occurred. Of the minor adverse events, gastrointestinal symptoms were the most common in both treatment groups (table 3). Vomiting occurred more frequently among patients treated with azithromycin (11 patients) than among those treated with ceftriaxone (7 patients), but the difference between the groups was not significant (P = .2). Vomiting was typically mild and transient and did not require treatment or alteration of the antibiotic therapy regimen. Diarrhea was the most common adverse event in patients treated with ceftriaxone, occurring in 15 patients, compared with 10 in the azithromycin group. Diarrhea also did not require treatment or alteration of the antibiotic therapy regimen.
At the end of treatment, mild, asymptomatic, increased aspartate aminotransferase levels were present in 4 patients (2 of whom were in the azithromycin group), and an additional 7 patients (5 of whom had received ceftriaxone therapy) had mild, asymptomatic increases in alanine aminotransferase levels. Thrombocytosis (defined as >500,000 platelets/mm3) was present in 7 patients in each treatment group at the time of hospital discharge, but all patients were asymptomatic, with normalization of laboratory results in the absence of treatment.
In the present study, a 5-day course of oral azithromycin was found to be highly effective and appeared to be therapeutically equivalent to intravenous ceftriaxone for the treatment of uncomplicated typhoid fever in children and adolescents. Patients treated with azithromycin were febrile for a longer period than were those treated with ceftriaxone (4.5 vs. 3.6 days). However, the difference was neither statistically significant nor clinically relevant. In addition, the durations of fever in both treatment groups were within the time frames (3.5–8 days) reported in previous trials on the treatment of typhoid fever [14, 15, 20]. Both study medications also were effective in clearing fecal carriage of S. Typhi, with no child having this pathogen isolated from culture of stool specimens obtained after antibiotic therapy was initiated.
The finding that 12 (38%) of 32 patients treated with azithromycin remained bacteremic 3 days after starting treatment was highly unexpected and surprising. Combining the results of 2 previous typhoid fever treatment trials, 69 patients with blood culture—proven typhoid fever were treated with azithromycin, and only 2 remained bacteremic 4 days after starting antibiotic therapy [13, 14]. In the current study, because the duration of treatment was shortened from 7 to 5 days, the on-therapy blood samples were obtained for culture on day 3, one day sooner than treatment was ended in our previous trials [13, 14]. The fact that blood samples were obtained for culture earlier in the present study is postulated to be the reason for the apparent, rather than actual, prolongation of bacteremia. The rate of persistent bacteremia found in the current study, however, is not inconsistent with the findings of a study by Islam et al. , in which 65% of patients with typhoid fever who received chloramphenicol therapy still had S. Typhi cultured from their blood after 3 days of therapy, compared with 0 of 28 patients who were treated with ceftriaxone. It is thought that more-rapid sterilization of the blood after receipt of ceftriaxone is associated with high serum levels resulting from a 75-mg/kg intravenous dose, compared with serum levels after an orally administered dose of azithromycin. In addition, antimicrobial drugs with sites of action at the cell wall tend to kill faster than those with a ribosomal site of action. Despite the continued presence of bacteria in their blood, all patients achieved clinical improvement, none required alteration of therapy, and there was no effect on the final clinical or microbiological outcome.
Although bacteremia may have persisted longer, no patient treated with azithromycin had either a suspected or microbiologically confirmed relapse of typhoid fever. However, 6 (19%) of 32 patients who received ceftriaxone therapy and were available for follow-up observation at day 30 had suspected relapses, and 5 of these 6 had microbiologically confirmed relapses. These findings are consistent with reports in the literature of relapse rates of 5%–15% for patients treated with ceftriaxone [14, 20–22]. Although the sample size in the present study is small, combining these results with those of our previous trials [13, 14], along with those of studies by Chinh et al.  and Butler et al. , nearly 200 patients with blood culture–confirmed typhoid fever have been treated with azithromycin, and no relapses have occurred. The concentration of azithromycin within cells and its secretion into the biliary tree, in conjunction with the long half-life of the drug, likely explain why relapses have not occurred when treating a principally intracellular infection such as typhoid fever [14, 15].
In the present study, both medications were well tolerated by the majority of patients. The most-common adverse events were associated with gastrointestinal symptoms, which included vomiting and diarrhea. These events principally occurred within the first day or two of treatment and did not require therapy or alteration of the treatment regimen. Although this cannot be proven, it is likely that many of the gastrointestinal events were associated with the underlying disease and not with the treatment.
On the basis of our experience and that of others, MDR S. Typhi was becoming a common clinical problem and was a major reason for performing the present study. Thus, it was an unexpected finding that, in the present study, only a single isolate of S. Typhi was MDR. Before 1988, MDR S. Typhi were almost nonexistent in Egypt, but once resistance developed, there was a steady increase in prevalence until 1993, when 100% of the isolates recovered at our institution exhibited the MDR phenotype . Since then, there has again been a decrease in the prevalence of S. Typhi with MDR in Egypt, but, even as recently as 3 years ago, 11 (17%) of the 64 S. Typhi from our institution were MDR . This decrease in the prevalence of MDR S. Typhi has also been described in India and Bangladesh [25, 26]. The cause of the change in prevalence of MDR S. Typhi is uncertain but has been postulated to be the result of bacteria losing their resistance plasmids to antibiotics that are uncommon in the environment .
Another interesting finding was that in vitro resistance to azithromycin did not correlate well with its in vivo effectiveness against typhoid fever. This is possibly because susceptibility testing is based on serum drug levels, whereas for typhoid fever, a major mechanism of action is thought to be intracellular killing, in which the azithromycin levels may be 100-fold greater than serum levels [9–11].
Including the present trial, 5 studies have now demonstrated the effectiveness of azithromycin for the treatment of uncomplicated typhoid fever in children, adolescents, and adults [12–14, 23]. In each of the studies, clinical and microbiological cure rates have exceeded 90% without any serious adverse events or relapses of typhoid fever.
The results of the present study aptly demonstrate that a 5-day course of azithromycin (a dosage of 20 mg/kg per day, with a maximum dose of 1000 mg/day) is effective against uncomplicated typhoid fever in children and adolescents. The once-daily administration of azithromycin, combined with the short duration of therapy, may improve compliance and ease treatment of typhoid fever. Future research directions include testing whether the dosage of azithromycin can be decreased to 10 mg/kg per day, which was previously shown to be effective in a 7-day treatment course, and/or whether the duration of therapy can be further shortened to minimize costs while maintaining efficacy.