Article Text
PDF
Statistics from Altmetric.com
We read with interest the report by Bachmeyer et al on Mycobacterium xenopi pulmonary infection manifesting in an HIV infected patient after receiving highly active antiretroviral treatment (HAART).1 The diagnosis was made based on clinical, radiological, and histological findings of a granuloma in addition to one sputum specimen growing M xenopi. We think that the patient may meet the criteria set by the ATS for diagnosis and treatment of disease caused by non-tuberculous mycobacteria (NTM) but, according to this guideline, these recommendations fit best for M avium complex, M kansasii, and M abscessus. Too little is known about other NTM (such as M xenopi) and how applicable these criteria are to them.2 This case may be one of those situations where it is difficult to make a definitive diagnosis.
M xenopi is usually a non-pathogenic coloniser of airways that has occasionally been associated with nosocomial outbreaks related to growth in hospital hot water systems.3,4 A recent publication showed the incidence of M xenopi isolates in a large urban hospital and its pathogenicity to be low.5 Tuberculosis would have the same clinical/radiological presentation and would have improved with the same treatment given to the patient.6,7 The persistent negativity of tuberculin skin testing (TST) despite the increase in CD4 cell count cannot be used to exclude tuberculosis. TST has a high false negative rate even among non-HIV infected patients with confirmed tuberculosis.
While the management of this case would not have differed had the patient been treated as a presumed case of tuberculosis, it is important to keep in mind the need for contact investigation and appropriate public health interventions for tuberculosis cases.
References
Authors' reply
We thank Drs Salazar-Schicchi and Nachman for their interest in our paper and their valuable comments. However, we consider that Mycobacterium xenopi was responsible for the patient's disease despite the fact that the microbiological diagnosis was not “definitive”. Indeed, the criteria of the American Thoracic Society were not fulfilled.1 These criteria do not seem to be applicable to M xenopi in HIV infected patients, in whom two positive cultures of M xenopi and no other cause of symptoms have been proposed as criteria for the diagnosis.2 Our patient also did not fulfil these criteria. Indeed, we were concerned about the possible role of other pathogens—especially M tuberculosis—since coexistent pulmonary infections due to other pathogens had been reported.3 However, no other pathogens were found and a search for M tuberculosis in the three sputum samples and broncholoalveolar lavage fluid was negative on direct microscopic examination and culture. This is rare in cavitatory tuberculosis and makes this diagnosis unlikely.
Mycobacterium xenopi may be found in hospital water taps, hot water storage tanks, and contaminated bronchoscopes.4 Environmental contamination seemed unlikely since M xenopi was not isolated from samples in the microbiology laboratory during the period of management of our patient.
We conclude that M xenopi can be the cause of a lung disease in HIV infected patients that resembles tuberculosis and clinicians should not disregard the significance of this organism when isolated from respiratory specimen, even from only one.