Melioidosis, an infection caused by the bacterium Burkholderia pseudomallei, is a major cause of community-acquired septicaemia in northeast Thailand. Common manifestations include cavitating pneumonia, hepatic and splenic abscesses, and soft tissue and joint infections. Despite improvements in diagnostic procedures and treatment, the mortality of severe melioidosis remains unacceptably high - approximately 35% with currently used antibiotics (ceftazidime or co-amoxiclav). There is clear evidence that antibiotics can affect mortality; the use of ceftazidime rather than previous regimens (doxycycline + chloramphenicol + co-trimoxazole) led to a 50% reduction in mortality from 80% to 35%. However, the mortality in the first 48 hours has not been altered by any treatment regimen. A key question is whether alternative antibiotics could improve early outcome. The hypothesis tested is that meropenem is superior to ceftazidime in terms of mortality for the treatment of melioidosis.

Mortality rate of patients with severe melioidosis is still unacceptably high. Response to high dose parenteral ceftazidime treatment in survivors is also slow, as median time to abatement of fever is approximately 9 days. B. pseudomallei is susceptible to ceftazidime, imipenem, co-amoxiclav (Augmentin®), piperacillin and doxycycline, but unlike most other pseudomonads it is resistant to aminoglycosides, apart from kanamycin which has borderline activity. The fluoroquinolone compounds also have borderline activity. Two large published in-vitro studies have shown that the carbapenem group are the most active antibiotics against B. pseudomallei, with an MIC90 of 0.5 or 1.0 mg/L, and an MBC90 of 1 mg/L. We have tested the susceptibility to meropenem of 100 recently isolated strains of B. pseudomallei, all of which were assessed as susceptible (MIC90 = 0.5 mg/L; range 0.125-1 mg/L). Furthermore, 13 isolates in our collection assessed as resistant to ceftazidime were susceptible to meropenem. Using time-kill kinetic studies, ceftazidime did not show "significant" bactericidal activity whereas meropenem was bactericidal (99.9% kill) within 6 hours. Previous treatment trials have demonstrated the importance of the choice of antibiotic at the time of presentation. A study that compared a four-drug combination of chloramphenicol, doxycycline, and trimethoprimsulfamethoxazole (TMP-SMX) with ceftazidime alone demonstrated a 50% reduction in the mortality rate from 80% to 35%. Several previous randomized controlled trials have been conducted to determine whether the administration of alternative antimicrobial drugs are associated with further improvements in outcome. A comparison of TMP-SMX plus ceftazidime versus ceftazidime alone demonstrated that the addition of TMPSMX did not reduce the acute mortality rate. A previous study comparing ceftazidime and imipenem/cilastatin in the treatment of severe melioidosis was performed in Ubon Ratchathani between 1994 and 1997. This showed that "treatment failure" rate (a potentially subjective endpoint in this open-labelled trial) in the imipenem/cilastatin group was lower than in the ceftazidime group. Endotoxin release, believed to be important to the pathogenesis of severe sepsis, was also lower in the imipenem group than the ceftazidime group. No difference in mortality was observed, but this study was underpowered following early termination due to a lack of imipenem supply from the manufacturer. As a result, ceftazidime has remained the treatment of choice for melioidosis, but the question remains as to whether a carbapenem drug would be more effective. A second, sufficiently powered clinical trial would address this important question.

Inclusion criteria (all criteria must be satisfied)

A. Community acquired sepsis, and melioidosis is suspected:

Suspected melioidosis (12): all of the following are defined as 'clinically probable' melioidosis

• A history of frequent contact with soil or surface water in the endemic area
• At least one of the following risk factors: diabetes mellitus, chronic renal failure or renal calculi, thalassaemia, aplastic anaemia or steroid abuse
• An illness compatible with melioidosis, including the presence of sepsis, acute pneumonia, acute pyelonephritis, septic arthritis, parotid disease or skin or soft tissue infection, or
• An evidence of intra-abdominal suppuration (hepatic or splenic abscesses) regardless of risk factors or exposure history

Sepsis: defined as patients who have Systemic Inflammatory Response Syndrome (SIRS) - two or more of the following, clinically ascribed to infection:

• Fever: temperature >38°C or <36°C

• Tachycardia: heart rate >90 beats/min

• Tachypnoea:

  1. Respiratory rate >20 breaths/minute; or
  2. PaCO2 <32 mmHg; or
  3. Mechanical ventilation

• White cell count >12,000 cells/mL or <4,000 cells/mL or >10% band forms B. Age > 14 years. C. Need hospitalisation and intravenous antibiotic administration. D. Willingness to participate in the study and written, informed consent obtained from the patient.

Exclusion Criteria (any one of the following):

A. Pregnant or lactating women. B. Known hypersensitivity to meropenem or ceftazidime. C. Previous isolate with known resistance to ceftazidime or meropenem. D. Patients not expected to remain in hospital for treatment. E. Patients with community-acquired sepsis with cultures positive for other organisms.

F. Patients treated with antibiotics active against B. pseudomallei (including ceftazidime, amoxicillin-clavulanate, meropenem) for this episode for greater than 24 hours.