A middle-aged woman from Bangladesh sought treatment for acute lymphoblastic leukemia (ALL) in Singapore. Her diagnosis was first made in a hospital in Dhaka (the capital of Bangladesh), where she stayed for 10 days prior to her transfer here. She had presented with fever, muscle aches and increasing fatigue, and had received antibiotics for unremitting fever and neutropenia.

In Singapore, she received chemotherapy for ALL (hyperCVAD #1a) 5 days after admission, when the initial septic workup was negative and her diagnosis of ALL was confirmed. However, she spiked a fever on Day 10 of her hospitalisation in Singapore and became more ill (but was not in septic shock). She was profoundly neutropenic at this point in time. She was prescribed intravenous (IV) piperacillin/tazobactam – a broad-spectrum penicillin – according to the local protocol for febrile neutropenia, but continued to have fever the next day. Her blood cultures were reported positive for Gram-negative bacteria after only one day.

Question: What would be the appropriate antibiotic regimen at this point in time?

[Updated 17th January 2015]

The standard protocol would have been to stop the piperacillin/tazobactam, and to switch over to a carbapenem (IV imipenem). This is in view of the high rates of extended-spectrum beta-lactamase producing Enterobacteriaceae in Singaporean hospitals, especially among patients with hematological disorders and febrile neutropenia,  causing bloodstream infections (as described in the Singapore Medical Journal in 2012).

However, Dhaka is one of the places on the Indian subcontinent where NDM-1-producing Enterobacteriaceae has been found in the hospitals. It would have been prudent to isolate and screen all such patients for the presence of carbapenemase-producing Enterobacteriaceae upon transfer into a local hospital, although of course such insights were not known then. This patient did indeed develop NDM-1-producing Escherichia coli bacteremia and eventually died in Singapore in November 2010, as described by Dr Esther Chan and her colleagues in the Singapore Medical Journal in 2011.

It is not known currently what is the best treatment regimen for serious infections caused by NDM-1-producing bacteria. In a case such as this, it would seem logical that empirical treatment with another antibiotic (other than the carbapenem) active against NDM-1-producing Enterobacteriaceae be prescribed if the patient was a known carrier of such bacteria.