Azlocillin

An Emerging Pathogen or a Product of Improved Laboratory Methods?

Study objectives: To investigate the dramatic rise in number of Mycobacterium xenopi isolates identified in our mycobacteriology laboratory, and to determine if this increase was due to emerging clinical pathology or to changes in culture technique.

Design: Retrospective chart and laboratory review.

Setting: University-affiliated tertiary-care city hospital.

Patients: Eighty-one patients with a single culture positive for M xenopi from 1975 to 1994 (period 1), and 47 patients with two or more cultures positive from 1994 to 1998 (period 2).

Interventions: The Bellevue mycobacteriology laboratory changed the culture medium from solid Lowenstein-Jensen medium (used from 1975 to 1990) to the Septi-Check AFB System (Becton-Dickinson; Glencoe, MD; used from 1991 to 1994), to the Mycobacteria Growth Indication Tube (MGIT; Becton-Dickinson; used from 1995 to 1998).

Measurements and results: We recovered 29 M xenopi isolates from 1975 to 1990, 12 isolates from 1991 to 1994, and 381 isolates from 1995 to 1998. We subsequently identified and reviewed the medical records of all 81 patients who were culture positive for M xenopi from 1975 to 1994 (period 1), and 46 patients who had two or more isolates culture positive for M xenopi from 1995 to 1998 (period 2). For period 1, 75% of the subjects were male, 80% were minority, and at least 43% were HIV positive. Only one patient had clinical M xenopi lung disease during this period. For period 2, 79% of the subjects were male, 83% were minority, and at least 58% were HIV positive; two additional patients were identified who had clinical M xenopi lung disease.

Conclusions: The dramatic increase in M xenopi isolates noted in our hospital was due to a more sensitive laboratory isolation technique, rather than a true increase in clinical disease. Other hospitals utilizing MGIT systems for mycobacterial recovery should interpret positive M xenopi cultures with caution. (CHEST 2000; 118:1365-1370)

Key words: clinical review; laboratory methods; Mycobacterium xenopi; nontuberculous mycobacteria

Abbreviations: ATS = American Thoracic Society; MAI = Mycobacterium avium-intracellulare; MGIT = Mycobacteria Growth Indication Tube; NTM = nontuberculous mycobacteria

Mycobacterium xenopi is a slow-growing scotochromogen[1] that was first isolated from skin granulomas of the Xenopus laevis in 1959.[2] It was first reported in the sputum of a man with COPD in 1965.[3] It grows at increased temperatures (42 [degrees] C) and has been reported to contaminate hot-water systems.[4] Although it is one of the most common nontuberculosis mycobacterial pathogens in parts of England,[5] its isolation in the United States was relatively uncommon.[6]

At Bellevue Hospital Center, we observed a striking increase in M xenopi isolates: from 29 isolates (1975 to 1990) to 12 isolates (1990 to 1994), to 381 isolates (1995 to 1998). It is not clear, however, whether this represents a true increase in incidence or if it is a result of improved culture media, which has changed significantly in recent years. At Bellevue between 1975 and 1991, Lowenstein-Jensen and Middlebrook solid media were employed exclusively. From 1991 to September 1994, the laboratory used the Septi-Check AFB System (Becton-Dickinson; Glencoe, MD) biphasic agar/broth media, which for the first time added a Middlebrook 7H9 broth to the routine culture protocol. In October 1994, the laboratory switched to the Mycobacteria Growth Indicator Tube (MGIT; Becton-Dickinson), which incorporated a modified and enriched Middlebrook 7H9 broth with a fluorescent detection system. In order to assess whether increased detection of M xenopi represented colonization or an actual increase in the incidence of infection, we undertook a chart review of all patients with one or more positive cultures for M xenopi from January 1975 to October 1995 (period 1), followed by a chart review of those patients with two or more positive cultures for M xenopi from November 1995 to January 1998 (period 2).

MATERIALS AND METHODS

A list of all patients with positive culture results for M xenopi from January 1, 1975, to October 1995, and two or more positive culture results from November 1995 to January 1998 was obtained from the mycobacteriology laboratory records of Bellevue Hospital. Medical records were reviewed for age, gender, ethnicity, chest radiograph, pathology specimens, HIV status, and clinical course.

Sputum samples using a nebulized saline solution induction technique or BAL samples were obtained. Specimens were digested and decontaminated with 2% sodium hydroxide, 2.9% sodium citrate, and 0.5% N-acetyl cysteine for 15 min, and an equal volume of 0.9% saline solution was added. The specimens were centrifuged at 3,000g for 15 min, decanted, and the sediment was used to inoculate media and prepare smears.

Between 1975 and 1991, Lowenstein-Jensen and Middlebrook 7H10 or 7H11 solid media were employed exclusively, and no liquid media were inoculated. From 1991 to October 1994, all specimens were inoculated into a sealed biphasic culture system (Septi-Check AFB System), which is a sealed quadriphasic culture system. This system consists of a lower chamber filled with liquid media (Middlebrook and Cohn 7H9), supplemented with a combination of oleic acid, albumin, dextrose, and catalase; and an antibiotic supplement containing amphotericin, azlocillin, colistin, and trimethoprim; and an upper chamber with a paddle with three areas coated with agar (Middlebrook and Cohn 7H10), an agar-based variant of Lowenstein-Jensen media, and chocolate agar. The inoculum was added to the liquid phase and incubated at 37 [degrees] C. The vials were examined daily for growth; if negative, they were inverted to seed the solid media in the upper chamber.

After October 1994, all cultures were inoculated onto solid medium and an MGIT. The MGITs contain Middlebrook and Cohn 7H9 broth supplemented with oleic acid, albumin, dextrose, and catalase; polymyxin, amphotericin B, nalidixic acid, trimethoprim, and azlocillin; and an indicator in the bottom of the tube that fluoresces as the level of oxygen in the tube decreases. MGITs are reported to permit the recovery of mycobacteria as rapidly as the BACTEC 460 (Becton-Dickinson) tuberculosis system. These were examined daily on an ultraviolet transilluminator to detect growth. Solid media were examined weekly for 8 weeks. Between the periods of October 1994 and September 1995, a remarkable 175 M xenopi cultures were identified. Laboratory protocol was subsequently changed. The culture time in the MGIT system was shortened from 8 weeks to 5 weeks. Two hundred six M xenopi cultures were identified from October 1995 to January 1998. Cultures showing signs of growth as turbidity in the liquid media, colonies on the solid media, or fluorescence in the MGITs were examined by performing a Kinyoun acid-fast stain, and any mycobacteria detected were identified by standard techniques. Twenty consecutive isolates identified as M xenopi using standard biochemical testing were sent to the New York City mycobacteriology laboratory for species confirmation via high-performance liquid chromatography. All 20 isolates were confirmed as M xenopi.

RESULTS

During period 1 (January 1975 to October 1995), we identified 81 patients (61 male and 20 female) with at least one M xenopi isolate. Specimens were recovered from sputum alone (60 eases), BAL (20 cases), and blood and sputum (1 case). Twenty-nine cultures of M xenopi were isolated prior to 1991, 12 from 1991 through 1993, and 40 from January 1994 to September 1995. Of the 81 patients, 33 were African American, 26 were Hispanic, 16 were white, and 6 were Asian, reflecting the ethnic distribution at Bellevue Hospital. Their mean ([+ or -] SD) age was 47 [+ or -] 14 years. The patients had significant comorbidity with a variety of primary diagnoses (Table 1). A positive HIV status was found in 43% of the group. M xenopi was isolated in 34 patients receiving treatment for Mycobacterium tuberculosis and was considered colonization. Six patients had several different nontuberculous mycobacteria (NTM) isolated, including four Mycobacterium avium-intracellulare (MAI), one Mycobacterium scrofulaceum, and one Mycobacterium gordonae. None of these six patients had clinical evidence of disease. The other patients with single isolates of M xenopi had no evidence of clinical NTM disease.

(*) PPD = purified protein derivative.

There were 7 patients (8.6%) with multiple isolations of M xenopi. Only one patient met the American Thoracic Society (ATS) criteria[7] for pulmonary disease caused by M xenopi. Two were identified posthumously after dying of AIDS-related causes. Two others had COPD with bacterial pneumonia that responded to antibiotics, and one had pulmonary, tuberculosis and responded to routine antituberculous treatment. The last was a postpartum female patient with a normal chest radiograph. Since none of these four patients met the ATS criteria for NTM disease, it was concluded that the recovery of M xenopi most likely reflected colonization and not infection.

During period 2 (November 1995 to January 1998), we identified 47 patients with at least two M xenopi cultures: 37 male patients, 10 female patients, 23 African Americans, 14 Hispanics, 8 whites, and g Asians. Their mean age was 43 years, and Table 2 lists their significant comorbidities. Over 50% were HIV seropositive. We were able to obtain CD4+ cell counts on 26 of 27 patients in period 2, and the average was 99 cells/[micro]L (median, 30 cells/[micro]L; range, 0 to 580 cells/[micro]L). Most of the patients (n = 37; 80%) had abnormal chest radiographs (infiltrates, nodules, fibrotic changes, and cavities). Chest radiograph abnormalities were a common feature of M xenopi infection in the pre-HIV era and continued to predominate during the HIV epidemic.

Three patients received a diagnosis of M xenopi pulmonary disease, including one patient from period 1 and two patients from period 2. The first was a 70-year-old white woman with a history of COPD ([FEV.sub.1] of [is less than] 35% predicted), diabetes mellitus, and a positive result on tuberculin skin test in 1970. She presented with recurrent fever, chills, and myalgia for year. Her chest radiograph and CT scan from August 1994 (Fig 1, 2) showed a right-upper-lobe infiltrate suggestive of tuberculosis. Smears for acid-fast bacilli and cultures from two bronchoscopies were negative for M tubercolosis. She was given a trial of rifampin, ethambutol pyrazinamide, and isoniazid, but stopped after 1 month due to GI side effects and rash. In August 1995, a third bronchoscopy specimen was positive for acid-fast mycobacteria. The patient was started on antituberculous medications but developed a drug-induced lupus reaction. A right-upper lobectomy was performed, and the surgical specimen contained multiple necrotizing granulomas. Culture results were positive for M xenopi at that time from sputum, BAL, and lung tissue specimen. She completed a 1-year course of treatment with rifabutin, ethambutol, and clarithromycin, and remains well.

[Figures 1-2 ILLUSTRATION OMITTED]

Two additional patients with clinical M xenopi pulmonary disease were identified during period 2. The first individual was a 75-year-old male nursing home resident with severe COPD ([FEV.sub.1] of [is less than] 0.9 L). He had a positive sputum smear for acid-fast bacilli that consistently was determined to be an atypical mycobacterium but was ignored clinically for almost 10 years. Because of persistent fevers, he was evaluated at our University Hospital, where M xenopi was cultured from sputum repeatedly and a chest radiograph showed an upper-lobe infiltrate. Treatment with a macrolide and a fluoroquinolone resulted in gradual resolution of fevers and the chest radiograph infiltrate, with discharge back to the nursing home. The second individual was a 34-year-old HIV-positive woman with multiple cavities noted on her, scan (Fig 3). She had cough, fever, night sweats, and weight loss for several weeks. Her CD4+ count was 60 cells/[micro]L. Treatment with streptomycin, ethionamide, clarithromycin, and ciprofloxacin over 6 months resulted in clinical improvement. The streptomycin has been replaced by aerosolized amikacin given three times weekly.

[Figure 3 ILLUSTRATION OMITTED]

DISCUSSION

We have observed a dramatic increase in M xenopi isolates in the Bellevue Hospital mycobacteriology laboratory, from 29 isolates over a 15-year period to 381 isolates over a recent 3-year period. This dramatic increase was coincident with the institution of more-sensitive culture techniques in our mycobacteriology laboratory. Despite these 422 patients' isolates, we identified only three patients with clinical M xenopi pulmonary disease. Two of these patients had severe COPD, and the third patient had AIDS with multiple pulmonary cavities. The first patient also had cavitation in a fibrotic zone of the upper lobe.

Most NTMs are isolated from man-made environments (eg, tap water, food stuffs, soil, etc.), with different species displaying a large geographic difference in prevalence.[5-7] In the United States, MAI complex followed by Mycobacterium kansasii, are the most frequent NTM pathogens causing lung disease.[5,6] In Canada (Ontario), the United Kingdom, and in other parts of Europe, M xenopi ranks second overall to MAI as a cause of NTM lung disease. In Southeast England, for example, it is the most common NTM recovered in the laboratory.[5]

Nosocomial outbreaks[8] and pseudo-outbreaks[4] of infection caused by M xenopi have been reported. It is believed to be the result of organisms entering hospitals through municipal water mains, then multiplying in the hospital heating tanks where the water temperature is 43 [degrees] C to 45 [degrees] C, the optimal temperature for M xenopi to grow. M xenopi is an obligate thermophile that requires temperatures [is greater than or equal to] 28 [degrees] C to grow; it is also a slow grower that produces a characteristic yellow pigment. We cultured the tap water from multiple sources and none grew M xenopi despite several attempts. Bennett and colleagues[9] have reported M xenopi isolates resulting from contamination of bronchoscopes.

When detected in sputum specimens, M xenopi is generally considered a nonpathogenic colonizer of airways in patients with no immunologic compromise or pulmonary disease. When pathogenic, preexisting lung damage or an immunologic defect usually precedes the M xenopi infection. Chronic pulmonary disease is the most common clinical manifestation of M xenopi in the nonimmunocompromised patient.[8] Widespread dissemination[10] and Pott's disease[11] have also been reported. Sporadic outbreaks of M xenopi disease have been reported in the northeastern United States.[9,12,13] These reports show much higher disease prevalence than that reported here. The first report was believed to have been due to contamination of the water supply. Subsequent reports did not report any such contamination. However, these subsequent reports applied ATS criteria for NTM infections to define their disease cases. The ATS criteria[7] for identifying clinically significant NTM require repeated isolation from sputum or bronchial washings, and that other reasonable causes of disease have been excluded. Several articles have encouraged a much less stringent requirement to define M xenopi infection.[13-15] The large increase in numbers of new isolates allowed us to evaluate the validity of these recommendations. We extended our clinical chart review from the initial 7 patients with multiple M xenopi isolates to 46 patients in period 2.

The clinical presentation of NTM disease has changed since the advent of the HIV epidemic. Prior to AIDS, NTM disease was relatively uncommon, with MAI or M kanasii representing the major pathogens. Disease was primarily limited to the lungs, cervical lymph nodes, and skin, with rare dissemination. Pulmonary disease was found predominantly in elderly patients with predisposing lung conditions (eg, COPD and bronchiectasis). In contrast, 25 to 50% of immunocompromised hosts in the United States and Europe eventually become infected by NTM.[16,17] Most patients present with pulmonary disease primarily. Although M kansasii and MAI are still the major NTM agents, there are indications that other mycobacterial infections are increasing. Clinical M xenopi infection is still unusual, even in HIV-infected patients.

Because systems of NTM notification vary from country to country, the true incidence of NTM disease is difficult to determine. There are several reasons that the incidence of NTM disease may be increasing. First, there is a greater appreciation of the pathogenic potential of NTM in susceptible patients, which results in more specimens being sent for analysis, with a concomitant increase in positive cultures. Second, with the advent of AIDS and the increased use of chemotherapy, corticosteroids, and immunosuppressive drugs, the number of susceptible patients may be increasing. Third, man-made changes to the environment may have altered the risk of NTM infection. For example, hot-water systems provide better growth conditions for some species of NTM, and showers rather than baths may increase the chance of inhaling infectious droplets.

In our case, we found that the dramatic surge in M xenopi isolates identified at Bellevue Hospital was predominately due to more-sensitive laboratory techniques and was not truly the harbinger of an emerging pathogen. The change from Lowenstein-Jensen solid media to the MGIT system resulted in detection of [O.sub.2] consumption (respiration rather than growth) of the mycobacteria, likely increasing the sensitivity of detection. All positive MGITs were subsequently subcultured at 42 [degrees] C, which selects for and enhances recovery of M xenopi. We contacted the New York City mycobacteriology laboratory that routinely processes multiple specimens from throughout New York City and also uses the MGIT system as their culture system. In addition, the laboratory routinely identifies all its isolates using high-pressure liquid chromatography. They too found a large increase in M xenopi isolates as well as a large increase in M gordonae and MAI. Our mycobacteriology lab has noted an increase in MAI but not in M gordonae. Lastly, a recent study on various mycobacterial recovery systems reported the impressive ability of the MGIT system to detect M xenopi in comparison to other detection systems.[18] Other hospitals utilizing MGIT systems for mycobacterial recovery should interpret positive M xenopi cultures with caution when deciding which patients require treatment. Our findings support the current ATS recommendations on identification of NTM infection

Given the large numbers of infected patients reported by others, we were surprised by our low numbers of true M xenopi disease. Our discrepant findings may be partially explained by the effect of recently introduced highly active antiretroviral therapy. This therapy may reconstitute the immune system and thus prevent disease appearance or progression. Additionally, the current prophylaxis and treatment regimens used against MAI also have substantial activity against M xenopi. This may have decreased the incidence of observed disease in the immune-compromised population.

Treatment of M xenopi disease may present a problem because of its variable in vitro susceptibility to antimycobacterial agents.[19,20] Therapy always should be guided by organism sensitivities. The addition of macrolides and quinolones to the available treatment regimens should help reduce toxicity and improve compliance.[21] In the event of treatment failure or relapse, surgical treatment needs to be considered.[22] We believe that treatment should be completed even after surgical intervention.

ACKNOWLEDGMENT: The authors thank Drs. Claudia Plottel, Karen Hoover, and Fred Bevelacqua for sharing information on their patients; the New York City mycobacteriology laboratory for sharing its data concerning the changing isolate profiles; and Jessie A. Pierre for editorial assistance.

REFERENCES

[1] Timpe A, Runyon EH. The relationship of "atypical" acid-fast bacteria to human disease: a preliminary report. J Lab Clin Med 1954; 44:202-209

[2] Schwabacher H. A strain of mycobacterium isolated from skin lesions of a cold-blooded animal, Xenopus laevis, and its relation to atypical acid-fast bacilli. J Hyg (Lond) 1959; 57:57-67

[3] Marks J, Schwabacher H. Infection due to Mycobacterium xenopi. BMJ 1965; 1:32-33

[4] Sniadack DH, Ostroff SM, Karlix MA, et al. A nosocomial pseudo-outbreak of Mycobacterium xenopi due to contaminated potable water supply: lessons in prevention. Infect Control Hosp Epidemiol 1993; 14:636-641

[5] Yates MD, Grange JM, Colins CH. The nature of mycobacterial disease in southeast England, 1974-84. J Epidemiol Community Health 1986; 40:295-300

[6] O'Brien RJ, Geiter LJ, Snider DE Jr. The epidemiology of nontuberculous mycobacterial disease in the United States. Am Rev Respir Dis 1987; 135:1007-1014

[7] American Thoracic Society. Diagnosis and treatment of disease caused by nontuberculous mycobacterium. Am J Respir Crit Care Med 1997; 156(suppl):S1-S25

[8] Costrini AM, Mahler DA, Gross WM, et al. Clinical and roentgenographic features of nosocomial pulmonary disease due to Mycobacterium xenopi. Am Rev Respir Dis 1981; 123:104-109

[9] Bennett SN, Peterson DE, Johnson DR, et al. Bronchoscopy associated Mycobacterium xenopi pseudoinfections. Am J Respir Crit Care Med 1994; 150:245-250

[10] Weinberg JR, Dootson G, Gertner D, et al. Disseminated Mycobacterium xenopi infection. Lancet 1985; 1:1033-1034

[11] Miller NC, Perkins MD, Richardson WJ, et al. Pott's disease caused by Mycobacterium xenopi: case report and review. Clin Infect Dis 1994; 19:1024-1028

[12] EI-Solh AA, Nopper J, Abul-Khoudoud MR, et al. Clinical and radiographic manifestations of uncommon pulmonary nontuberculous mycobacterial disease in AIDS patients. Chest 1998; 114:138-145

[13] Tajuddin MJ, Helen MJ, Jacoby LA, et al. Mycobacterium xenopi: innocent bystander or emerging pathogen? Clin Infect Dis 1997; 24:226-232

[14] Stauffer F, Bankier AA, Strasser G, et al. Mycobacteria other than tuberculosis with an emphasis on Mycobacterium xenopi in clinical specimens from AIDS patients at the University Hospital of Vienna from 1989 to 1996. Wien Klin Wochenschr 1999; 111:56-58

[15] Juffermans NP, Verbon A, Danner SA, et al. Mycobacterium xenopi in HIV-infected patients: an emerging pathogen. AIDS 1998; 12:1661-1666

[16] Eng RHK, Forrester C, Smith SM, et al. Mycobacterium xenopi infection in a patient with acquired immunodeficiency syndrome. Chest 1984; 86:145-147

[17] Ausina V, Barrio J, Luquin M, et al. Mycobacterium xenopi infections in the acquired immunodeficiency syndrome. Ann Intern Med 1988; 109:927-928

[18] Alcaide F, Benitez MA, Escriba JM, et al. Evaluation of the BACTEC MGIT 960 and the MB/BacT Systems for the recovery of mycobacteria from clinical specimens for species identification by DNA AccuProbe. J Clin Microbiol 2000; 38:398-401

[19] Baugnee PE, Pouthier F, Delaunois L. Pulmonary mycobacteriosis due to Mycobacterium xenopi: in-vitro sensitivity to classical antitubercular agents and clinical development. Acta Clin Belg 1996; 51:19-27

[20] Tortoli E, Simonetti MT. Radiometric susceptibility testing of Mycobacterium xenopi. J Chemother 1995; 7:114-117

[21] Banks J, Hunter AM, Campbell A, et al. Pulmonary infection with Mycobacterium xenopi: review of treatment and response. Thorax 1984; 39:376-382

[22] Parrot RG, Grosset JH. Post-surgical outcome of 57 patients with Mycobacterium xenopi pulmonary infection. Tubercle 1988; 69:47-55

Vincent Donnabella, MD; John Salazar-Schicchi, MD; Stanley Bonk, BS; Bruce Hanna, PhD; and William N. Rom, MD, MPH, FCCP

(*) From the Bellevue Chest and Mycobacteriology Services (Drs. Donnabella, Salazar-Schicchi, and Rom), Division of Pulmonary and Critical Care Medicine, and Departments of Medicine, Environmental Medicine, and Pathology (Mr. Bonk and Dr. Hanna), New York University School of Medicine, New York, NY. Supported by National Institutes of Health grant No. M01RR00096.

Manuscript received November 9, 1999; revision accepted June 20, 2000.

Correspondence to: Vincent Donnabella, MD, Bellevue Chest and Mycobacteriology Services, Division of Pulmonary and Critical Care Medicine, New York University School of Medicine, 550 1st Ave, New York, NY 10016; e-mail: donnav01@popmail.med.nyu.edu

COPYRIGHT 2000 American College of Chest Physicians
COPYRIGHT 2001 Gale Group