Mycobacterium celatum and Mycobacterium xenopi: Two Easily Confused Pathogens

Abstracts & Commentary

Synopsis: Mycobacterium celatum and Mycobacterium xenopi are two emerging pathogens that may be confused in the laboratory with each other and with other mycobacteria.

Sources: Zurawski CA, et al. Pneumonia and bacteremia due to Mycobacterium celatum masquerading as Mycobacterium xenopi in patients with AIDS: An underdiagnosed problem. J Clin Infect Dis 1997;24:140-143; Piersimoni C, et al. Isolation of Mycobacterium celatum from patients infected with human immunodeficiency virus. Clin Infect Dis 1997;24:144-147; Jiva TM, et al. Mycobacterium xenopi: Innocent bystander or emerging pathogen? Clin Infect Dis 1997; 24:226-232.

Zurawski et al report on two hiv-infected patients with Mycobacterium celatum infection. One patient, with a CD4+ lymphocyte count of 1/mm3 and a left lower lobe infiltrate, repeatedly had an organism identified by standard biochemical tests as Mycobacterium xenopi recovered from sputum over a period of four months. Subsequently, a blood culture also yielded and acid fast organism on culture, again identified by the local laboratory as M. xenopi. HPLC analysis, however, determined that this identification was incorrect; the organism was M. celatum. Treatment was begun, but the patient was lost to follow-up.

A 32-year-old AIDS patient (CD4+ lymphocyte count 10/mm3) developed fever and a right upper lobe cavitary lesion. Sputum and blood cultures grew M. celatum, identified by HPLC. A later blood culture yielded both this organism and an organism of the Mycobacterium avium complex (MAC). Treatment was begun with clarithromycin and ethambutol, and all further blood and sputum cultures failed to yield mycobacteria.

Piersimoni et al describe seven HIV-infected subjects see over a two year period in northern Italy from whom M. celatum was recovered over a two-year period. Their mean CD4+ lymphocyte count was 19/mm3 (range, 1-66/mm3). Three patients had infection apparently limited to the lungs, two had pulmonary infection plus extra-pulmonary dissemination, and one had disseminated infection without apparent pulmonary involvement. In one patient, recovery of the organism from sputum was thought to represent colonization rather than infection.

Biochemical tests identified the isolates as MAC, while the Mycobacterium tuberculosis AccuProbe assay was weakly positive. Careful analysis of their mycolic acid HPLC profiles revealed that the organisms were, in fact, M. celatum. Four of the strains were demonstrated to represent M. celatum type 1 by 16s rRNA gene fragment sequencing. The organisms were highly resistant to rifampin but susceptible to most other drugs tested, including clarithromycin. Five patients were treated with varying antimycobacterial regimens, and cultures obtained 2-9 months later were negative in four. One subject relapsed after nine months.

Jiva et al in Rochester, New York, report having recovered, mostly from respiratory sites, M. xenopi from 103 patients over a two-year period. Only nine men and one woman were believed to have disease caused by this organism, rather than colonization. Five of the 10 with infection were HIV-infected; their disease was limited to the lungs in two (CD4 counts 340 and 168/mm3) and disseminated in three (CD4 counts < 10, 10 and 50/mm3). The other five were older (age, 51-70 years) individuals with underlying chronic obstructive pulmonary disease (COPD) and infection limited to the lungs, with cavitary lesions in four. Most subjects appeared to respond to multidrug therapy.

Half the colonized patients had underlying COPD, and 11 colonized patients had active or prior pulmonary tuberculosis. Most of the remainder of the colonized patients were HIV-infected, with 95% having AIDS and a mean CD4+ lymphocyte count of 114/mm3 (range, 0-340/mm3).

In this study, the organism was identified based on growth characteristics, colonial morphology, and biochemical tests alone. Isolates from two patients were confirmed as M. xenopi at the National Jewish Center for Immunology and Respiratory Medicine, but they do not indicate what methodology was used for this purpose is not indicated.


M. xenopi is a slow-growing scotochromogenic acid-fast thermophile with a temperature optimum of 43°C. M. celatum is also slow-growing but is nonphotochromogenic, and it is reported to grow best at 37°C but to also grow at temperatures as high as 45°C. Based upon widely used biochemical tests, M. celatum appears biochemically indistinguishable from MAC and M. xenopi and to have a closely related, but distinguishable, mycolic acid pattern to the latter. Both organisms are potentially pathogenic.

In an editorial accompanying the paper by Jiva et al, Wolinsky indicates that M. xenopi "is not difficult to identify in the laboratory" because of its colonial morphology (Wolinsky E. Clin Infect Dis 1997;24:233-234). However, he does not specifically discuss possible confusion with M. celatum (even though all three of the articles reviewed here, as well as the editorial, appear in the same issue of Clinical Infectious Diseases). The isolates reported by Jiva et al were identified as M. xenopi on the basis of their growth characteristics as well as upon biochemical tests. The experience of Jurawski et al suggests, however, that not all clinical mycobacteriology laboratories are sufficiently facile to warrant this optimism. The clinical laboratory initially used by the former group misidentified their isolates of M. celatum as M. xenopi, and they claim that, "Genomic analysis or HPLC is necessary to distinguish the two species." They indicate that M. celatum, a slowly growing nonphotochromogen resembles both MAC and M. xenopi biochemically and with regard to colonial morphology.

Another reason to suspect the identification of the isolates of Jiva et al as M. xenopi is the fact that they were recovered from January 1, 1991, to December 31, 1993, while M. celatum was first described in 1993 (Butler WR, et al. Int J Syst Bacteriol 1993;43:539-548). Thus, they were likely unaware of the existence of this potentially masquerading organism at the time of the identification of their isolates. This caveat also applies to their prior report of two of the HIV-infected patients included in their current paper (Jacoby HM, et al. Clin Infect Dis 1995;20:1399-1401).

In addition to potential confusion with M. xenopi, M. celatum may also be confused with both MAC and M. tuberculosis, as occurred in the laboratory used initially by Piersimoni et al. Confusion of M. celatum type 1 with M. tuberculosis is the result of weak cross reactivity in the AccuProbe hybridization assay. In fact, it has been suggested by Piersimoni et al that this cross reactivity may be used diagnostically: "a weakly positive reaction in testing of nontuberculous mycobacteria with the M. tuberculosis complex AccuProbe strongly suggests the identification of M. celatum."

On the other hand, Piersimoni et al state that "M. celatum should be suspected when a mycobacterium, behaving in conventional tests as MAC or M. xenopi, fails hybridization with probes specific for MAC."

The distinction of M. celatum from MAC, from which it may be indistinguishable biochemically, may be made on the basis of either genomic analysis or HPLC analysis of mycolic acids.

The correct identification of these organisms is of importance to the patient because of the varying susceptibility patterns of the isolates. The susceptibility pattern of M. celatum, which appears to regularly be highly resistant to rifampin, appears to resemble that of MAC, while M. xenopi is often susceptible to a larger number of standard antimycobacterial drugs.