We recently managed 4 patients with cystic fibrosis who had acquired Burkholderia pseudomallei infection after exposure in a region of endemicity. Person-to-person transmission between 2 siblings may have occurred; otherwise, the evidence suggests that cystic fibrosis may increase the likelihood of infection with this organism, and patients should be warned of this possibility and cautioned to avoid high-risk activities.
Cystic fibrosis (CF) is the most common fatal genetic disorder among white persons, with death usually associated with bacterial infection of the lung. Burkholderia cepacia is a well-recognized pathogen in CF lung disease and is associated with more rapid deterioration of respiratory function than is CF without B. cepacia infection . Burkholderia pseudomallei is the cause of melioidosis and is indigenous to certain tropical areas of the world, such as Thailand and Northern Australia, where it is a common cause of severe community-acquired pneumonia [2, 3]. There have recently been 2 separate case reports of isolation of B. pseudomallei from the sputum samples of patients with CF after the patients had traveled to areas of endemicity [4, 5]. We describe 4 more patients with CF who had B. pseudomallei isolated from sputum samples, and we discuss the significance of this association.
Case reports. Patient 1 lived in Darwin, Australia, until the age of 5 years, when his family migrated to New Zealand, a country with no indigenous B. pseudomallei. When he was 8.5 years of age, B. pseudomallei was first isolated from a sputum sample. The isolate was identified by full biochemical testing, and the identification was confirmed by 16S rRNA gene sequencing. The isolate was susceptible to ceftazidime, trimethoprim-sulfamethoxazole (TMP-SMZ), amoxicillin-clavulanate, doxycycline, piperacillin, and meropenem, as determined by use of agar dilution and Etest methods.
The patient remained clinically stable until the age of 9 years, when he had acute deterioration in pulmonary function. His condition responded well to treatment with intravenously administered ceftazidime and tobramycin (TMP-SMZ had been withdrawn after the patient developed a rash), although the patient had a relapse when parenteral therapy was discontinued. He was given further treatment with ceftazidime and meropenem for 17 weeks, but he remained clinically unstable, and B. pseudomallei persisted in his sputum. CT of the chest revealed widespread mild bronchiectasis and mucous plugging. Oral TMP-SMZ was reintroduced after desensitization. Three months later, the patient had further pulmonary deterioration, and B. pseudomallei was isolated from blood cultures. The isolate was now resistant to ceftazidime, piperacillin, TMP-SMZ, and doxycycline, and it had intermediate susceptibility to amoxicillin-clavulanate; however, it remained susceptible to meropenem, to which the infection slowly responded.
Patient 2 is the younger sister of patient 1 by 20 months. When she was 7.5 years of age, chest radiography revealed a nodular opacity in the right lower lobe. B. pseudomallei was first isolated from a sputum sample 9 months after the first isolate had been recovered from her brother; the isolates had the same susceptibility pattern. Oral TMP-SMZ therapy was commenced, but bone marrow suppression necessitated a change to doxycycline. The patient remains healthy, but B. pseudomallei persists in her sputum.
Patient 3 is a 10-year-old boy who lives in Darwin and who had previously been a playmate of patients 1 and 2. Patient 3 had B. pseudomallei isolated from a sputum sample when he was aged 9 years. The indirect hemagglutination antibody titer for B. pseudomallei was 1 : 1280. The patient was healthy, but he received treatment with ceftazidime and TMP-SMZ for 14 days, followed by TMP-SMZ for 3 months for presumed colonization or mild infection. Eighteen months after the completion of therapy, the patient remained healthy, and cultures have been negative for B. pseudomallei.
Identification of the isolates recovered from patients 1–3 was confirmed by the Menzies School of Health Research (Darwin) by use of in-house PCR and agglutination with B. pseudomallei-specific antisera, and the isolates were typed by PFGE. The isolates recovered from patients 1 and 2 were identical, and they differed from the isolate recovered from patient 3 (figure 1).
Patient 4 is a 38-year-old man who resides in New Zealand. While on vacation in Australia in March and April 2001, he spent 3 weeks driving from Perth to Darwin, camping, and swimming in rivers; on ⩾1 occasion, he dug his vehicle out of mud. In July 2001, he was admitted to a hospital in New Zealand with fever, productive cough, and decreased pulmonary function. Pseudomonas aeruginosa, Staphylococcus aureus, and B. pseudomallei were isolated from a sputum sample. Identification of the isolates was also confirmed by means of 16S rRNA gene sequencing. The patient was treated with intravenous ceftazidime, flucloxacillin, and tobramycin, but, 3 months later, B. pseudomallei was again isolated from a sputum sample.
Discussion. Published experience with B. pseudomallei in patients who have CF is very limited. In addition to the 2 aforementioned recently published case reports [4, 5], a report by Dance et al.  described a 20-year-old man with CF who had B. pseudomallei infection , and a conference abstract mentioned 3 patients, all of whom died . We report 4 more cases seen by us within a relatively short time, and we have also become aware of other patients with CF who are infected with this organism (2001 World Melioidosis Congress, personal communication).
Clinical and environmental isolates of B. pseudomallei in Northern Australia are diverse , but the isolates recovered from patients 1 and 2 were identical and were rather uncommon strains (figure 1). Although person-to-person transmission of B. pseudomallei occurs very rarely [8, 9], we believe that it may have occurred between these siblings. We speculate that human transmission may be more likely in the context of CF, chronic carriage of the organism, and residence in the same household as an infected person, akin to the well-documented means of spread of B. cepacia among patients with CF. An alternative explanation is that acquisition of the identical strain by the siblings occurred at the same time and from the same unidentified source. This would then suggest that CF markedly increases the risk of acquisition of B. pseudomallei. Patients 1–3 each had lived in the Northern Territory of Australia for some years, and they may have had a number of exposures to B. pseudomallei; however, patient 4 had only a relatively short exposure period, albeit a period in which there was activity involving mud and rivers—known reservoirs of the organism.
Recognized predisposing conditions for melioidosis include diabetes, excessive alcohol consumption, chronic lung disease (other than CF), and renal disease. The paucity of reports of CF and B. pseudomallei infection in the past may be because few individuals with CF live in areas of endemicity. The relative contribution of other factors (e.g., exposure to water) is unknown, but available information indicates that patients with CF who travel to regions where B. pseudomallei is endemic should be warned of the potential risk and cautioned to avoid activities that involve aerosols or percutaneous exposure from mud and/or dirty water. The difficulty in treating infection with B. pseudomallei is a function of the intrinsic resistance of the organism and is further complicated in these cases by the environment of the CF lung. We favor the use of ⩾2 antibiotics (preferably ceftazidime and TMP-SMZ) as initial therapy.
We thank Mark Mayo and Daniel Gal of the Menzies School of Health Research (Darwin, Australia) for typing the isolates.