JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1987, p. 1428-14320095-1137/87/081428-05$02.00/0Copyright © 1987, American Society for Microbiology

Simplified Acetylcysteine-Alkali Digestion-DecontaminationProcedure for Isolation of Mycobacteria from Clinical Specimens

SAMUEL RATNAM,l* FLORENCE A. STEAD,' AND MARY HOWES2Newfoundland and Labrador Public Health Laboratories, St. John's, Newfoundland AIB 372,1 and

Laboratory Services Branch, Ministry of Health, Toronto, Ontario MSW JRS,2 Canada

Received 13 January 1987/Accepted 7 May 1987

During our attempts to reevaluate and improve mycobacteriology laboratory procedures, we found thatisolation of mycobacteria from undecontaminated specimens through the use of selective media was unreward-ing and that specimen decontamination was an essential step. A reevaluation of the N-acetyl-L-cysteine-sodiumhydroxide (NALC-NaOH) decontamination procedure (G. P. Kubica, W. E. Dye, M. L. Cohn, and G.Middlebrook, Am. Rev. Respir. Dis. 87:775-779, 1963) indicated that this method could be simplified andshortened without affecting the isolation rate or growth and without increasing the contamination rate. Themodified NALC-NaOH procedure effectively combines the decontamination step with the concentration step:specimens are mixed with the NALC-NaOH solution on a Vortex mixer and, without any waiting or additionof buffer or water to the digested specimens, centrifuged at 3,000 x g for 15 min, and the sediment is suspendedin phosphate buffer (pH 5.3). The modified method'is simpler, faster, and safer than the original procedure,and the reduced manipulation is also likely to minimize the chance for cross-contamination in sequentialspecimen processing. The total alkali exposure time in the modified NALC-NaOH method should be kept to aminimum; if necessary or feasible, the concentration of NaOH in the digestion solution may be reduced fromthe usual strength of 2% to about 1.5% to compensate for the longer alkali exposure time in the modifiedmethod to maintain the isolation rate or to improve it.

Decontamination of clinical specimens such as sputumand urine is an important and critical step in the isolation ofmycobacteria, and the mildest decontamination procedurewhich provides sufficient control of contaminants is likely toyield the best results (9). In diagnostic laboratory practice,the routinely used decontamination methods have certaindisadvantages: they kill many mycobacteria, the proceduresare tedious and time-consuming, and there is an increasedrisk of specimen cross-contamination as well as occupationalhazard. An alternative to specimen decontamination is theuse of selective culture media, and their successful applica-tion in the isolation of mycobacteria from undecontaminatedclinical specimens has been reported (5, 8). We investigatedthe efficacy of this approach in a reference laboratory settingin which a large number of mailed-in specimens are proc-essed.The N-acetyl-L-cysteine-sodium hydroxide (NALC-

NaOH) method of Kubica et al. (2, 3) not only is consideredto be one of the most satisfactory digestion-decontaminationprocedures but also is the most widely recommended andused technique for the isolation of mycobacteria from clini-cal specimens (9, 10). In the NALC-NaOH method, speci-mens are digested and decontaminated in a single step duringa 15- to 20-min treatment period. This step is followed by theaddition of phosphate buffer (PB) or water to fill the centri-fuge tube within 1 to 2 cm of the top, presumably to dilutethe alkali and to lower the viscosity of the specimens.

Mycobacteria are concentrated by centrifugation, and thesediment is resuspended in a minimal volume of bovinealbumin solution (0.2%, pH 6.8) and used for cultures andsmears. A chance observation in the St. John's laboratoryindicated that the omission of the PB or water addition todigested specimens prior to centrifugation did not have a

significant effect on the growth or isolation rate of the

* Corresponding author.

mycobacteria. In a preliminary study, we demonstrated thatthis modification of the NALC-NaOH method in combina-tion with a reduction of the recommended treatment time of15 to 20 min had no adverse effects on the isolation ofmycobacteria (S. Ratnam and F. Stead, Abstr. Annu. Meet.Am. Soc. Microbiol. 1983, C218, p. 348). Since the above-described modifications improved, simplified, and shortenedthe NALC-NaOH procedure, we carried out additionallaboratory studies and field trials of the modified protocol indiagnostic laboratory settings, the description and results ofwhich form the main subject of this report.

MATERIALS AND METHODSThe efficacy of isolating mycobacteria from undecontam-

inated specimens through the use of selective media wasdetermined by processing routinely submitted clinical spec-imens as previously described (8). Specimens were liquefiedin NALC and centrifuged; one portion of the resuspendedsediment was directly cultured on selective media, and asecond portion was decontaminated by the original NALC-NaOH method (10) and then cultured on both selective andstandard media.

Studies relating to modifications of the NALC-NaOHmethod were carried out in two stages: (i) the initial labora-tory analysis series with sputum and urine specimens seededwith mycobacterial strains and (ii) the field trial series withroutinely submitted clinical specimens. The test organismsfor the initial series included Mycobacterium tuberculosis(21 strains), M. bovis (3 strains), M. kansasii (3 strains), M.marinum (3 strains), M. scrofulaceum (7 strains), M. xenopi(3 strains), M. gordonae (3 strains), M.flavescens (3 strains),M. avium (5 strains), M. intracellulare (8 strains), M. triviale(3 strains), M. nonchromogenicum (3 strains), M. fortuitum(11 strains), M. chelonae (M. chelonei) (17 strains), M.smegmatis (3 strains), and M. phlei (3 strains). Mycobacte-rial suspensions and sputum and urine specimens were

1428

Vol. 25, No. 8

on April 21, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

NALC-NaOH METHOD FOR ISOLATION OF MYCOBACTERIA 1429

prepared, and the test organisms were seeded into sets oftubes containing 10 ml of the specimens as previouslydescribed (7). The seeded sputum and urine specimens were

processed in triplicate by various modifications of theNALC-NaOH procedure, with a set of specimens processedby the original NALC-NaOH method (10) serving as a

control. Modifications of the treatment procedure includedthe omission of the PB or water addition to digested speci-mens prior to centrifugation in combination with the follow-ing parameters: decreased concentration of NaOH in thedigestant solution from the usual 2% to 1.5 and 1%; additionof up to 2 volumes of NALC-NaOH digestant to specimens;blending of specimen-digestant mixtures on a Vortex mixerfor 15, 30, or 60 s; reduced decontamination times of 0, 2, 5,and 10 min; and resuspension of sediments in PB (0.067 M)ranging from pH 5 to 7. For the field trial series. sputumspecimens were liquefied with a small volume of NALCdigestant without NaOH, and all specimens were thoroughlymixed on a Vortex mixer, divided into two equal volumes,and processed simultaneously, one portion by a modificationof the NALC-NaOH procedure (see below) and the other bythe original procedure (10). The field trials were indepen-dently carried out in two reference laboratories (St. John'sand Toronto).The modified NALC-NaOH treatment procedure is as

follows. (i) Up to 10-ml samples of specimens are placed in50-ml centrifuge tubes. No more than 24 tubes are processedat a time. (If <10 ml is used, the volume of a specimen maybe increased to this level with water). (ii) The tubes are linedup in order, and an equal volume of NALC-NaOH digestantsolution is added. (If specimen volumes are not alreadyuniform, the appropriate volumes of NALC-NaOH solutionmay be added in the following ratios: 3 ml of digestant to ca.2 ml of specimen, totalling 5 ml; 6 ml of digestant to ca. 4 mlof specimen, totalling 10 ml; or 7 to 9 ml of digestant to ca.6 to 8 ml of specimen. totalling 15 mi. Care should be takento balance the tubes in the centrifuge.) (iii) Each specimen ismixed on a Vortex mixer for about 15 s or until liquefied. (iv)The digested specimen is centrifuged at 2,500 to 3,000 x gfor 15 min. (v) AlU of the supernatant is decanted, and thesediment is resuspended in about 1 to 2 ml of PB (0.067 M;pH 5.3). Note that the total alkali exposure time for individ-ual specimens should be kept to a minimum, i.e., ca. 30 min,when 24 specimens are processed. Also, the NaOH concen-

tration in the solution may be reduced from the usual 2%7 to1.5% if necessary or feasible.The NALC-NaOH procedure (10) or its modifications (see

above) were used throughout to decontaminate specimens,and no more than 24 specimens were processed at a time.Specimens and reagents were brought to room temperature(24 to 25°C), and all procedures, including centrifugations,were carried out at this temperature; centrifugations were

done at 3,000 x g for 15 min. The inoculum size was keptuniform throughout (0.1 ml), and the culture media usedincluded selective 7H10 (S-7H10) medium (8), selectiveLowenstein-Jensen (S-LJ) medium (1), and standard 7H11and Lowenstein-Jensen (LJ) media (Difco Laboratories,Detroit, Mich.). Cultures were incubated at 35°C under 5 to10% CO, and examined weekly for contamination andgrowth of mycobacteria. Mycobacterial isolates were iden-tified by standard methods (10).

RESULTS

To determine the efficacy of isolation of mycobacteriafrom undecontaminated specimens through the use of selec-

tive media, we cultured a total of 2,100 consecutive speci-mens (sputum, 59Yc; urine, 27%; and other specimens, 14%)on S-7H10 and S-LJ media without prior decontamination.The culture results, as compared with those for decontami-nated specimens cultured on both selective and standardmedia, indicated that the former approach was unsatisfac-tory, owing to a significantly high rate of contamination, i.e.,39 versus 14.7%, respectively, for cultures on 7H media (P <0.01) (Table 1). The contamination rate was highest at 58%for undecontaminated specimens cultured on S-LJ medium,in contrast to 1 and 11%, respectively, for decontaminatedspecimens cultured on S-LJ and standard LJ media (data notshown).

In the laboratory analysis series carried out to modify theNALC-NaOH method, a total of 356 seeded specimens weretested, with the following results. The addition of PB orwater to digested specimens prior to centrifugation onlyslightly neutralized the alkali, and its omission did notsignificantly affect the recovery, growth rate, or colonycounts of most of the test organisms under controlledconditions. In fact, the addition of PB or water at timesresulted in a poor recovery, owing to a greater dissolution ofthe particulate matter in specimens; the particulate matterappears not only to aid in sedimentation but also to helpretain the bacteria in the sediment when the supernatant ispoured off following centrifugation. Alkali tolerance was nota consistent feature among the mycobacterial species andstrains tested, with the rapid growers being the most suscep-tible to a high pH. For this reason, the omission of PB wasdeleterious to these organisms, leading to decreased recov-ery rates because of the longer exposure to a higher pHduring centrifugation. Nevertheless, additional studies car-ried out with rapid growers indicated that by reducing oreliminating the standard treatment time of 15 to 20 minand/or by decreasing the concentration of NaOH in thedigestant solution from the usual strength of 2% to either 1.5or 1%, we could easily overcome the adverse effect (Fig. 1through 4). These modifications, in fact, resulted in a dra-matic increase in the recovery of mycobacteria that arerelatively more susceptible to alkali treatment (Fig. 1 to 4,panels c and d) but had no marked effect on the recovery ofother mycobacteria (Fig. 5 through 7). With 2% NaOH in thedigestant solution, reducing or eliminating the treatmenttime did not increase contamination when PB was not addedto digested specimens prior to centrifugation. (This, on theother hand, led to a slightly decreased recovery of somemycobacteria, e.g., Fig. la and b and 2a and b). Thecontamination rate was about the same even when lower

TABLE 1. Comparison of culture results for 2.100 specimensprocessed for mycobacteria without decontamination

and with decontamination"

Result for RResult for undecontaminated

decontaminated specimens cultured on S-7H10 agar Total no.specimens cultured No. No. No.

on 7Hll agar' contaminated positive negative

No contaminated 78 il 220 309 (14.7)No. positive 26 42 12 80 (3.8)No. negative 715 7 989 1.711 (81.5)

Total no. (%î) 819 (39.0) 60 (2.9) 1.221 (58.1) 2,100

Specimens were processed in parallel. and cultures were incubated for 8weeks.

Decontamination was achieved by the original NALC-NaOH procedure(10).

VOL. 25, 1987

on April 21, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

J. CLIN. MICROBIOL.1430 RATNAM ET AL.

'4

FIG. 1-7. Comparison of mycobacterial cultures obtained with the original and modified NALC-NaOH treatment methods. Sputumspecimens containing mycobacteria were digested and decontaminated by the original or modified NALC-NaOH method and cultured onS-7H10 and S-LJ media. FIG. 1 and 3. M. fortuitum. FIG. 2 and 4. M. chelonae. FIG. 5. M. scrofulaceum. FIG. 6. M. avium-M.intracellulare complex. FIG. 7. M. tuberculosis. (a) Treatment with original NALC-NaOH method. (b) Treatment with modifiedNALC-NaOH method with 2% NaOH. (c) Same as panel b but with 1.5% NaOH. (d) Same as panel b but with 1.0% NaOH.

Cq

on April 21, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

NALC-NaOH METHOD FOR ISOLATION OF MYCOBACTERIA

TABLE 2. Effect of NaOH concentration in the digestantsolution on the contamination rate of mycobacterial

cultures on different mediaa

No. of cultures lost completglyb withindicated NALC-NaOH method

Medium Original with Modified with NaOH at:

2% NaOH 2% 1.5% 1.0%

S-LJ 3 2 3 5S-7H10 3 3 4 5Standard LJ 12 il 14 44Standard 7H11 18 16 28 61

a Specimens were simultaneously digested, decontaminated, and cultured.A total of 100 sets of cultures were used.

b Because of excessive contamination within 8 weeks of incubation.

concentrations of NaOH (1 to 1.5%) were used to treatspecimens as described above, provided that the cultureswere grown on the selective media. Cultures of the samespecimens grown on nonselective standard media, however,were heavily contaminated, especially at 1% NaOH (Table2). Neither increasing the volume of the digestant-decon-taminant solution in relation to the specimen nor extendingthe specimen-digestant solution mixing time influenced theculture results. When the sediment was resuspended in pH 5to 5.5 PB instead of the pH 6.8 solution of the originalmethod (10), a less alkaline inoculum pH of 7 to 8 wasgenerally obtained.On the basis of the results of the laboratory analysis

series, the NALC-NaOH procedure was modified as de-scribed above and evaluated in comparison with the originalprocedure in the field trial series. A total of 199 specimens,most of which were sputum, from patients previously knownto be positive for mycobacteria were processed. The resultsindicated no significant differences in the isolation or con-tamination rates between the original and modified methods(P > 0.05), and the similarity of the two methods was evidentfrom the sum of cultures yielding concordant results (Table3). Further, cultures yielding discordant results were equallydistributed between both methods, indicating the occurrenceof only random errors (Table 3). In addition to the concor-dant results, the growth rates and colony counts were aboutthe same for both methods. The few specimens which werefound to be positive by either one method or the other hadscant growth, suggesting that the difference was most likelya sampling artifact. The species breakdown of the 154isolates obtained during the field trial was as follows: M.tuberculosis, 107; M. avium-M. intracellulare complex, 20;M. xenopi, 11; M. kansasùi, 4; M. fortuitum, 4; M. chelonae,3; M. gordonae, 3; and M. scrofulaceum, 2. The overallfindings in this series of the study were identical between thetwo testing laboratories.

Subsequent to the field trial, over 13,000 routine clinicalspecimens have been processed in the St. John's laboratoryby the modified NALC-NaOH method with 1.5% NaOH.The isolation and contamination rates have remained atabout the same levels, i.e., approximately 3 and 4%, respec-tively, as in previous years when the original NALC-NaOHmethod was in use.

DISCUSSION

The specimen decontamination process used to facilitatethe isolation of mycobacteria also kills mycobacteria, and

the percentage of organisms killed varies depending on themethod used and the mycobacterial species present in thespecimens. The NALC-NaOH procedure of Kubica et al. (2,3) uses 2% NaOH as the decontaminant (final specimenconcentration of 1%) and is a relatively mild decontamina-tion method. However, even with this procedure, the loss ofmycobacteria could be as much as 104 depending on thespecies present (unpublished observation). Many mycobac-terial species are not as plentiful in clinical specimens as isM. tuberculosis, and these are also generally more suscep-tible to alkali treatment than is M. tuberculosis. Therefore,direct culturing of specimens on selective media withoutprior decontamination is highly desirable. Our attempts touse this approach proved to be unsatisfactory, particularlyfor mailed-in specimens. These results reiterated the impor-tance of specimen decontamination, without which the con-tamination rate reached unacceptably high levels and lead toa loss of cultures and poor isolation rates regardless of theselective medium used. It should be mentioned that theworkers who originally developed the selective agar mediumalso found the medium to be of benefit for direct culturing ofonly lightly contaminated specimens (6). Although we havefound that S-7H10 and S-LJ media are inhibitory to certainspecies and strains of mycobacteria, we consider them to behighly useful and a valuable adjunct to nonselective media inculturing decontaminated specimens as previously described(4).The modified NALC-NaOH procedure eliminates two

steps in the original procedure, i.e., the 15- to 20-mindecontamination time following the digestion of specimensand the addition of PB or water to digested specimens priorto centrifugation. It also uses a more acidic buffer than wasoriginally recommended to resuspend centrifuged sediment;the use of a reduced concentration of NaOH is optional. Inessence, the modified method effectively combines the de-contamination step with the concentration step and is sim-pler, faster, and safer than the original method. The additionof PB or water to digested specimens prior to centrifugationis a time-consuming step. More importantly, since the con-tents are being blended on a Vortex mixer in the precedingdigestion step, opening tubes to add the diluent often allowsspecimens to drip from the cap and/or run down on the outersurface of the tube; the major safety concern is the use ofsuch tubes for the next step, i.e., centrifugation. The modi-fied method not only eliminates this potentially hazardousprocedure but also saves technologists' time and is likely tominimize the chance for cross-contamination in sequentialspecimen processing. It was proposed that the addition ofPB to digested specimens resulted in the increased isolation

TABLE 3. Comparison of culture results for 199 routine clinicalspecimens processed for mycobacteria by the original and

modified NALC-NaOH methods

Result with modified NALC-NaOHResult with original method with 2% NaOH Total no.NALC-NaOH

No. No. No. (%)method

contaminated positive negative

No. contaminated 5 2 0 7 (3.5)No. positive 1 142 6 149 (74.9)No. negative 2 3 38 43 (21.6)

Total no. (%) 8 (4) 147 (73.9) 44 (22.1) 199

a Specimens were processed in parallel and cultured on standard LJ, S-LJ,and S-7H10 media.

b Complete loss of cultures on all media within 8 weeks of incubation.

VOL. 25, 1987 1431

on April 21, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

1432 RATNAM ET AL.

of mycobacteria and, in general, larger colonies (3). Weobserved that the isolation rate, growth rate, and colonycount were about the same between the modified and stan-dard methods, suggesting that under controlled conditions,the addition of PB or water to the digested specimensperhaps has only a minimal effect on the culture outcome.During the course of a related study (7), we also observedthat the addition of PB or water to specimens did notsignificantly increase the sedimentation rate of mycobacteriaduring centrifugation (unpublished observation). The omis-sion of PB or water necessitates that the specimen-digestantvolume be adjusted to obtain uniform levels for balancing inthe centrifuge, and this can be easily accomplished. The lossof mycobacteria occurring during the total alkali exposuretime of 30 min or so in the modified method appears to be notsignificantly greater than the loss occurring during the stan-dard alkali treatment time of 15 to 20 min. This may be dueto the fact that the greatest killing of mycobacteria occursduring the first few minutes of exposure to the alkali and,with most mycobacterial species, the extended exposureresults in only a minimal additional loss (unpublished obser-vation). Although the addition of PB to digested specimensonly slightly neutralizes the alkali, it should be emphasizedthat the omission of PB or water allows for the continuedexposure of mycobacteria to a higher pH during centrifuga-tion, which could be detrimental to at least some mycobac-terial species. To counteract any adverse effects, the treat-ment should be carried out as fast as practical until thesediment is resuspended in an acidic buffer. For example, inour laboratory, technologists use both hands and two Vortexmixers simultaneously to blend specimen-digestant mix-tures, permitting the mixing of 8 to 10 specimens per minute.A reduced concentration of NaOH in the digestant solutionmay also be considered as a means to compensate for theincreased exposure to the alkali in the modified method.There is no doubt that treatment of specimens with reducedconcentrations of NaOH results in a dramatic increase in therecovery of some mycobacterial species that are relativelymore susceptible to the alkali. This, however, may be offsetby an increase in the contamination rate, especially whencultures are obtained on nonselective media.On the basis of our experience, we recommend the mod-

ified NALC-NaOH method with 1.5% NaOH for the diges-tion and decontamination of specimens in the isolation ofmycobacteria. Ideally, the maximum alkali exposure timeshould not exceed 30 min per specimen when 24 or morespecimens are processed, and the minimum should be about25 min when fewer specimens are processed. Selectivemedia should be used in conjunction with nonselectivemedia for culturing for maximum isolation with minimumcontamination. The routine protocol in the St. John's Lab-oratory consists of the use of S-LJ, standard LJ, andselective 7H10 media for culturing.We carried out extensive trials with seeded sputum and

urine specimens in standardizing the modified NALC-NaOH

method. Nevertheless, experimental studies may not accu-rately simulate natural conditions. Although the field trialsincluded a large number of M. tuberculosis specimens, theywere rather limited in terms of other commonly encounteredmycobacterial pathogens. Therefore, while we recommendthe modified NALC-NaOH procedure for routine applica-tion, we suggest a cautious approach. Parallel comparisonsof the modified and standard methods may be worthwhileunder individual settings, particularly with specimens con-taining rapid growers such as M. chelonae and M. fortuitum,before implementation of the modified NALC-NaOHmethod as a routine procedure.

ACKNOWLEDGMENTS

This study was supported by a grant from the NewfoundlandLung Association.We thank R. W. Butler for support; W. A. Black, Division of

Laboratories, Ministry of Health, Vancouver, British Columbia,Canada, J. M. S. Dixon, Provincial Laboratory of Public Health,Edmonton, Alberta, Canada, G. Hammond, Cadham ProvincialLaboratory, Winnipeg, Manitoba, Canada, and S. M. Helbecque,Laboratory Centre for Disease Control, Ottawa, Ontario, Canada,for providing mycobacterial cultures; and L. Summers and P. Kellyfor secretarial assistance.

LITERATURE CITED1. Gruft, H. 1971. Isolation of acid-fast bacilli from contaminated

specimens. Health Lab. Sci. 8:79-82.2. Kubica, G. P., W. E. Dye, M. L. Cohn, and G. Middlebrook.

1963. Sputum digestion and decontamination with N-acetyl-L-cysteine-sodium hydroxide for culture of mycobacteria. Am.Rev. Respir. Dis. 87:775-779.

3. Kubica, G. P., A. J. Kaufmann, and W. E. Dye. 1964. Commentson the use of the new mucolytic agent, N-acetyl-L-cysteine, asa sputum digestant for the isolation of mycobacteria. Am. Rev.Respir. Dis. 89:284-286.

4. McClatchy, J. K., R. F. Waggoner, W. Kanes, M. S. Cernich,and T. L. Bolton. 1976. Isolation of mycobacteria from clinicalspecimens by use of selective 7H11 medium. Am. J. Clin.Pathol. 65:412-415.

5. Mitchison, D. A., B. W. Allen, L. Carrol, J. M. Dickinson, andV. R. Aber. 1972. A selective oleic acid albumin agar mediumfor tubercle bacilli. J. Med. Microbiol. 5:165-175.

6. Mitchison, D. A., B. W. Allen, and R. A. Lambert. 1973.Selective media in the isolation of tubercle bacilli from tissues.J. Clin. Pathol. 26:250-252.

7. Ratnam, S., and S. B. March. 1986. Effect of relative centrifugalforce and centrifugation time on sedimentation of mycobacteriain clinical specimens. J. Clin. Microbiol. 23:582-585.

8. Rothlauf, M. V., G. L. Brown, and E. B. Blair. 1981. Isolation ofmycobacteria from undecontaminated specimens with selective7H10 medium. J. Clin. Microbiol. 13:76-79.

9. Sommers, H. M., and R. C. Good. 1985. Mycobacterium, p.216-248. In E. H. Lennette, A. Balows, W. J. Hausler, Jr., andH. J. Shadomy (ed.), Manual of clinical microbiology, 4th ed.American Society for Microbiology, Washington, D.C.

10. Vestal, A. L. 1975. Procedures for the isolation and identifica-tion of mycobacteria. Center for Disease Control, Atlanta.

J. CLIN. MICROBIOL.

on April 21, 2020 by guest

http://jcm.asm

.org/D

ownloaded from