Download Löwenstein-Jensen Media No Longer Necessary for Mycobacterial

Microbiology and Infectious Disease / ELIMINATION OF LÖWENSTEIN-JENSEN MEDIA FOR AFB CULTURE
Löwenstein-Jensen Media
No Longer Necessary for Mycobacterial Isolation
Susan E. Sharp, PhD, Maritza Lemes, Sandra G. Sierra, Anna Poniecka, MD,
and Robert J. Poppiti, Jr, MD
Key Words: Mycobacteria; Media; Isolation; Significance
Abstract
Isolation of mycobacteria on Löwenstein-Jensen
media (LJ) and in the BACTEC MB9000 (MB) system
was compared. Of 2,271 specimens, 317 were positive
for 331 mycobacteria isolated in 1 or both media. The
MB was positive in 238 isolates, and LJ was positive
for 239 isolates; 92 isolates were detected by MB only
and 93 by LJ only. Of the 331 isolates, 146 were
recovered by both media. MB recovered 38 of 38
Mycobacterium tuberculosis complex isolates, while LJ
recovered 23. MB recovered 94.1% (96/102) of
Mycobacterium avium complex isolates and LJ 69.6%
(71/102). The MB recovered 81% (65/80)
Mycobacterium fortuitum-chelonae isolates and LJ
68% (54/80). Of the remaining species, MB isolated 39,
while LJ isolated 91. Only 1 organism that was isolated
on LJ alone was medically significant based on medical
record review. The addition of LJ media to the MB9000
system is not warranted, as it causes clinically
irrelevant workload, increased expenditures for the
laboratory, and could cause the inappropriate treatment
of patients.
770
Am J Clin Pathol 2000;113:770-773
Specimens submitted for the detection of mycobacteria
historically have been cultured on solid agar–based media,
and for at least the last decade, broth media has been added to
increase the sensitivity of culture for the recovery of
mycobacteria. Indeed, several standard microbiologic references recommend that both solid and liquid media be used
for the successful cultivation of mycobacteria.1-5 However,
solid media such as Löwenstein-Jensen medium (LJ) and
Middlebrook 7H10 agar require incubation for up to 6 weeks
before identification methods can be used, and although LJ
medium recovers Mycobacterium tuberculosis well, it is not
reliable for the recovery of other species of mycobacteria.6 LJ
medium and Middlebrook 7H10 agar also are less sensitive
and less rapid in the recovery of mycobacteria compared with
broth media,7-9 and these media (especially LJ) are being
dropped from routine use in many clinical laboratories.10 It is
possible that these media may no longer be necessary as a
part of the standard processing protocol for mycobacterial
culture. In fact, the use of broth media followed by a DNA
probe often establishes an organism’s identification even
before growth usually can be seen on solid media.10
Many broth media formulations and systems have been
developed for the recovery of mycobacteria species. These
include the manual BACTEC 460TB radiometric instrument
and media, the manual biphasic Septi-Chek acid-fast bacillus
(AFB) system, the manual Mycobacterial Growth Indicator
Tube, the BACTEC MB9000 automated instrumentation (all
from Becton Dickinson Microbiology Systems, Cockeysville,
MD), the ESP-AFB automated system (DIFCO, Detroit, MI),
the MB/Alert automated system (Organon-Teknika Corporation, Durham, NC), and the Vital automated system
(bioMerieux Vitek, Hazelwood, MO). The BACTEC 460TB
system was the first improvement in the detection of
© American Society of Clinical Pathologists
Microbiology and Infectious Disease / ORIGINAL ARTICLE
mycobacteria to be developed in years, and it increased the
recovery and reduced the time to detection of mycobacterial
organisms.11 The BACTEC 460TB system plus solid media
became the “gold standard” for mycobacterial culture, the
system with which all other broth-based systems have been
compared.
The MB9000 automated mycobacterial detection instrument is a fluorescence-based continuous monitoring detection system that uses a modified Middlebrook 7H9 broth in
conjunction with a supplement containing antibiotics to
suppress the growth of contaminating microorganisms. The
MB9000 system has compared favorably with the BACTEC
460TB system1 and the Septi-Chek AFB biphasic system12,13
for the detection of mycobacteria; however, the use of solid
media in conjunction with this system is still recommended.
The present study was undertaken to determine whether the
addition of LJ media to the MB9000 system is necessary for
the cultivation of clinical specimens for mycobacteria.
at 37°C in a 5% carbon dioxide atmosphere and were examined visually for growth every day for the first week and then
twice per week thereafter for a total of 8 weeks.
Isolates of acid-fast bacilli were identified as M tuberculosis complex (MTB) or Mycobacterium avium complex
(MAC) by DNA-RNA hybridization (Gen-Probe, San Diego,
CA) or were grouped into 1 of the following categories:
Mycobacterium fortuitum-chelonae complex (MFC),
pigmented mycobacterial species, or nonpigmented
mycobacterial species that were not MTB or MAC. Cultures
were evaluated to determine which isolates were recovered
by both the MB9000 system and LJ solid media and which
were recovered by only 1 of these methods. Medical record
reviews were performed for patients who had a specimen in
which a mycobacterial isolate was recovered only on LJ
solid media. An isolate was considered clinically relevant if
the attending physician indicated this in the medical record,
based therapy on the report of the organism, or both.
Materials and Methods
Results
All clinical specimens submitted for AFB culture from
June 1996 through May 1997 were included in the study, and
all specimens were processed within 24 hours after receipt in
the laboratory. Standard N-acetyl- L -cysteine sodium
hydroxide digestion (final concentration, 1% NaOH), decontamination, and concentration methods were used for all
specimens from nonsterile sources.3 These specimens were
concentrated by centrifugation at 2400g for 15 minutes after
being diluted with an equal volume of sterile saline. Tissue
specimens were not decontaminated, but ground in sterile
0.9% sodium chloride before centrifugation. The exception
to this procedure was with intestinal biopsy specimens,
which were decontaminated as previously described for
nonsterile specimens. The concentrated specimen was used
to make 2 smear preparations and then suspended in 0.2%
bovine albumin to a final volume of 1.5 mL. One milliliter of
this suspension was added to the MB9000 bottle and 0.1 mL
to an LJ slant. Two milliliters of freshly reconstituted
PANTA/F solution (consisting of polymyxin B, amphotericin
B, nalidixic acid, trimethoprim, and azlocillin) was added to
each MB9000 bottle, and patient samples were inoculated
within 2 hours of the addition of PANTA/F. MB9000
cultures were held 6 weeks in the MB9000 instrument and
monitored every 10 minutes by the instrument for microbial
growth. The MB9000 automated system uses a modified
Middlebrook 7H9 broth and a silicon rubber disk attached to
the inside base of the bottle that is impregnated with a ruthenium metal complex. This complex can detect decreases in
the dissolved oxygen in the medium resulting from microorganism metabolism and growth. LJ slants were held 8 weeks
A total of 2,271 clinical specimens were cultured in
both the MB9000 and on LJ. Specimens included in the
study were respiratory, blood, body fluid, tissue, and stool.
Of 2,271 specimens, 317 (14.0%) were positive for 331
mycobacterial isolates in 1 or both media ❚Table 1❚. The
MB9000 was positive for 238 (71.9%) of 331 isolates, while
LJ was positive for 239 (72.2%) of 331 isolates. The same
146 mycobacterial species were recovered in both systems,
while 92 isolates were recovered by MB9000 only and 93
isolates by LJ only. Table 1 shows the recovery of isolates by
each method.
The MB9000 recovered 38 (100%) of the 38 MTB
isolates, while LJ recovered only 23 (61%) of 38 isolates. The
recovery of MAC isolates was 94.1% (96/102) for the
MB9000 and 69.6% for LJ (71/102). The MB9000 recovered
65 (81%) of 80 MFC isolates compared with LJ, which recovered 54 (68%) of 80. The remainder of the mycobacterial
isolates were pigmented species or nonpigmented species that
were not MTB or MAC by Gen-Probe analysis. Of this latter
group, the MB9000 isolated 39 species, while LJ isolated 91
(25 pigmented mycobacterial species in MB9000 vs 45 in LJ;
14 nonpigmented mycobacterial species [not MTB or MAC]
in MB9000 vs 46 in LJ). In addition, isolates detected by each
system alone, MB9000 and LJ respectively, were 31 vs 6 for
MAC, 26 vs 15 for MFC, 12 vs 32 for pigmented mycobacterial species, and 8 vs 40 for nonpigmented mycobacterial
species (not MTB or MAC).
Medical records were available for review for 80 of 93
patients for whom isolates were recovered only by LJ media.
The record reviews showed that 1 of the 6 isolates of MAC
© American Society of Clinical Pathologists
Am J Clin Pathol 2000;113:770-773
771
Sharp et al / ELIMINATION OF LÖWENSTEIN-JENSEN MEDIA FOR AFB CULTURE
❚Table 1❚
Mycobacteria Detected by the BACTEC MB9000* and LJ Solid Media
Positive Results
Organism
MTB
MAC
MFC
Pigmented mycobacterial species
Mycobacterium species
Total
Total No. of Isolates
38
102
80
57
54
331
MB + LJ
23
65
39
13
6
146
MB Alone
15
31
26
12
8
92
Total
LJ Alone
0
6
15
32
40
93
MB
LJ
38
96
65
25
14
238
23
71
54
45
46
239
LJ, Löwenstein-Jensen; MAC, Mycobacterium avium complex; MB, BACTEC MB 9000; MFC, Mycobacterium fortuitum-chelonae complex; MTB, Mycobacterium tuberculosis
complex.
* From Becton Dickinson Microbiology Systems, Cockeysville, MD. Mycobacterium species are nonpigmented mycobacterial species, not MTB, and not MAC.
was relevant, but that this patient had a previously detected
MAC isolate (by MB9000). Both isolates were from respiratory specimens. Of the 5 remaining MAC isolates (2 respiratory, 2 stool, and 1 blood), none was considered significant
by the treating physician. Only 1 of 15 MFC organisms
isolated (12 respiratory, 3 stool) only by LJ was considered
clinically significant. In that case, the physician noted the
presence of a pneumonia-like illness and that an isolate of
MFC was isolated from the sputum of this patient. Furthermore, the patient was treated with cefoxitin, amikacin, and
clarithromycin and showed marked clinical improvement
after therapy. None of the pigmented mycobacterial species
or the nonpigmented mycobacterial isolates (not MTB or
MAC) were found to be clinically significant based on
medical record review.
Discussion
The MB9000 and LJ media had similar recovery rates
for mycobacteria from clinical specimens (238 vs 239,
respectively), and the number of isolates detected in each
medium alone was similar (92 vs 93). However, the MB9000
recovered 100% of the MTB isolates (38/38), while LJ
recovered only 23 of 38 isolates, and with the MAC
complex, the MB9000 recovered 96 (94.1%) of 102 isolates
compared with LJ, which recovered only 71 (69.6%) of 102.
Of the 6 isolates of MAC that were not detected by the
MB9000 system, 1 was detected in a previous specimen
from the same patient, while the remaining 5 were solitary
isolates, 2 from respiratory sources, 2 from stool specimens,
and 1 from blood. Colonization with MAC can occur in
immunocompetent and immunocompromised hosts11; hence,
their singular isolation often is considered irrelevant. As
MTB and MAC constitute the most frequently isolated clinically significant pathogens, and as the 6 isolates of MAC
recovered only on LJ did not have clinical significance for
the patients, the addition of LJ media contributed little to the
772
Am J Clin Pathol 2000;113:770-773
overall recovery of significant organisms from the patients
and nothing to clinical relevance.
MFC is also a ubiquitous organism that can be cultured
from soil, water, and dust and, when isolated from human
sources, usually represents respiratory tract colonization.
Thus, the clinical significance of the single isolates of MFC
in the present study is questionable. Studies have shown that
only 4% of all respiratory isolates of rapidly growing
mycobacteria are considered medically relevant and that
multiple positive specimens are necessary to establish a diagnosis of disease caused by these organisms.14 Of the 15 MFC
organisms not isolated by the MB9000, 13 were single respiratory isolates and 2 were single stool isolates. Only 1 of 15
isolates of MFC, although a single respiratory isolate, was
considered clinically relevant based on medical record
review and would have been missed if the MB9000 had been
used alone without the addition of LJ media. In fact, this was
the only isolate recovered from LJ alone in the entire study
that was potentially relevant to patient care.
The remainder of the mycobacterial isolates were
pigmented species or nonpigmented species that were not
MTB or MAC. LJ recovered 55 more of these isolates than did
the MB9000. Species identification of these organisms is
usually not performed in our facility except on special request.
No such requests were received for any of these cases during
the study period. Because these organisms are usually considered contaminants, the nearly 2.5 times higher isolation of
these organisms by LJ media resulted in a substantial amount
of clinically irrelevant work for the laboratory.
Although 1 isolate deemed by the treating physician as
clinically significant was missed by the MB9000 system, the
overwhelming majority (>99.7%) of relevant isolates were
detected by this method without the addition of LJ media. In
addition, the isolates recovered only from LJ media resulted
in the use of a substantial amount of unnecessary technical
labor by the laboratory and infection control department for
the isolation, reporting, and initial investigation. Moreover, if
a laboratory’s procedures were to fully identify all isolates of
© American Society of Clinical Pathologists
Microbiology and Infectious Disease / ORIGINAL ARTICLE
acid-fast bacilli, the isolates from LJ alone would cause an
additional financial burden for the facility. The reporting of
organisms isolated from LJ could also cause a physician to
begin therapy that may be unnecessary.
Because today’s methods for culturing mycobacteria
are so far advanced compared with the days when only solid
media were used, it is time to reevaluate the new systems for
their ability to stand alone as acceptable culture methods.
The present study has shown that at least 1 of the new automated methods, the MB9000, can be relied on for the appropriate cultivation of these organisms. We submit that the
routine use of LJ media with the MB9000 automated system
for the culture and detection of medically relevant mycobacteria is not clinically warranted, is not cost-effective, and
could cause the inappropriate treatment of patients.
From the Arkadi M. Rywlin Department of Pathology and
Laboratory Medicine, Mount Sinai Medical Center, Miami Beach,
FL.
Presented in part as a poster at the annual meeting of the
American Society for Microbiology, Atlanta, GA, May 20, 1998.
Address reprint requests to Dr Sharp: Microbiology and
Immunology, Mount Sinai Medical Center, 4300 Alton Rd, Miami
Beach, FL 33140.
7. Morgan M, Reves RR, Wilson ML, et al. Comparison of
BACTEC 12B vs solid media for the recovery of
Mycobacterium avium complex from blood cultures in AIDS
patients. Diagn Microbiol Infect Dis. 1997;28:45-48.
8. Stager CE, Libonate JP, Siddiqi SH, et al. Role of solid
media when used in conjunction with the BACTEC system
for mycobacterial isolation and identification. J Clin
Microbiol. 1991;29:154-157.
9. Wilson ML, Stone BL, Hildred MV, et al. Comparison of
recovery rates for mycobacteria from BACTEC 12B vials,
Middlebrook 7H11–selective 7H11 biplates, and
Löwenstein-Jensen slants in a public health microbiology
laboratory. J Clin Microbiol. 1995;33:2516-2518.
10. Koneman EW, Allen SD, Janda WM, et al. Mycobacteria.
In: Koneman EW, Allen SD, Janda WM, et al, eds. Color
Atlas of Diagnostic Microbiology. Philadelphia, PA:
Lippincott; 1997:906.
11. Heifets LB. Rapid automated methods (BACTEC systems)
in clinical mycobacteriology. Semin Respir Infect. 1986;1:242249.
12. Sharp, SE, Lemes M, Erlich SS, et al. A comparison of the
Bactec 9000MB system and the Septi-Chek AFB system for
the detection of mycobacteria. Diagn Microbiol Infect Dis.
1997;28:69-74.
13. Van Griethuysen AL, Jansz AR, Buiting AGM. Comparison
of fluorescent BACTEC 9000 MB system, Septi-chek AFB
system, and Löwenstein-Jensen medium for detection of
mycobacteria. J Clin Microbiol. 1996;34:2391-2394.
14. Wallace RJ Jr. Diagnostic and therapeutic considerations in
patients with pulmonary disease due to the rapidly growing
mycobacteria. Semin Respir Infect. 1986;1:230-233.
References
1. Baron EJ, Peterson LR, Finegold SM. Mycobacteria. In:
Baron EJ, Peterson LR, Finegold SM, eds. Diagnostic
Microbiology. St Louis, MO: Mosby; 1994:603-605.
2. Master RN. Mycobacteriology. In: Eisenberg H, ed. Clinical
Microbiological Procedures Handbook. Washington, DC: ASM
Press; 1992:3.6.3.
3. Nolte FS, Metchock B. Mycobacterium. In: Murray P, Baron
EJ, Pfaller MA, et al, eds. Manual of Clinical Microbiology.
6th ed. Washington, DC: ASM Press; 1995:400-437.
4. Tenover FC, Crawford JT, Huebner RE, et. al. The
resurgence of tuberculosis: is your laboratory ready? J Clin
Microbiol. 1993;31:767-770.
5. Woods GL, Witsbsky FG. Mycobacterial testing in clinical
laboratories that participate in the College of American
Pathologists’ mycobacteriology E survey: results of a 1993
questionnaire. J Clin Microbiol. 1995;33:407-412.
6. Metchock B, Nolte FS, Wallace RJ Jr. Mycobacterium. In:
Murray P, Baron EJ, Pfaller MA, et al. Manual of Clinical
Microbiology, 7th ed. Washington, DC: ASM Press;
1999:399-437.
© American Society of Clinical Pathologists
Am J Clin Pathol 2000;113:770-773
773