- Email: [email protected]
Direct antimicrobial drug susceptibility testing of Mycobacterium tuberculosis by the radiometric method
41
DIAGN MICROBIOL INFECT DIS 1988;10:41-48
Direct Antimicrobial Drug Susceptibility Testing of Mycobacterium tuberculosis by the Radiometric Method Joseph P. Libonati, Charles E. Stager, James R. Davis, and Salman H. Siddiqi
Direct-drug-susceptibility tests were performed on clinical specimens positive for acid-fast bacilli by either Ziehl-Neelsen or fluorochrame staining. The results of conventional agar dilution (Vestal, 1975) and a modified radiometric (BACTEC) method were compared. A total of 580 smear-positive specimens were tested by the BACTEC method at three separate sites. Three hundred and seventy-seven of these were culture positive for M. tuberculosis, and 343 (91%) yielded acceptable direct-susceptibility-test results. We used the conventional method to determine that 343 of 519 smear-positive specimens were culture positive for M. tuberculosis, and 212 (62%) produced acceptable results within 3 wks. Conventional results were reported in 3-4 wks, while the time required to obtain results with the BACTEC method ranged from 5 to 21 days (average 11.5 days). Results indicate that the radiometric method provides reportable results more frequently with time savings as compared to the conventional method.
INTRODUCTION
Drug susceptibility testing of Mycobacterium tuberculosis (MTB) is generally performed after the organism has been isolated. Because this indirect-susceptibilitytest approach requires several additional weeks for isolation, a direct-susceptibilitytest procedure is desirable {Vestal, 1975}. Currently, direct agar dilution susceptibility testing in which the inoculum is a digested and decontaminated clinical specimen k n o w n to be positive for acid-fastbacteria {AFB) can provide results 3-4 w k earlier than the indirect test (Gangadharam, 1984; Vestal, 1975). The introduction of rapid radiometric techniques has reduced the time required for drug susceptibility testing of MTB from 3 wk to less than 6 days (Roberts et al., 1983; Siddiqi et al., 1981, 1984, 1985; Vincke et al., 1982}. However, all current literature citations of radiometric-drug-susceptibility testing described the indirect method and required an average of 10-16 days for isolation of MTB prior to susceptibility testing. This report presents the results of direct-susceptibility testing using the radiometric technique and evaluates the efficacy and turnaround time w h e n compared to the conventional agar dilution procedure.
From the Maryland Department of Health and Mental Hygiene, Baltimore,Maryland {J.P.L.}, Harris County Hospital District,Baylor Collegeof Medicine Houston, Texas {C.E.S.,J.R.D.},and Johnston Laboratories,Research and Development, Towson, Maryland {S.H.S.). Address reprintrequests to: Dr. Joseph P. Libonati,Division of Microbiology,Department of Health and Mental Hygiene, P.O. Box 2355, Baltimore,M D 21203. Received October 8, 1987; revised and accepted April 22, 1988. © 1988 Elsevier Science Publishing Co., Inc. 52 Vanderbilt Avenue, New York, NY 10017
0732-8803/88/$03.50
42
J.P. Libonati et al.
MATERIALS A N D METHODS
This work is comprised of three separate studies. The first study was a joint effort of the Maryland Department of Health and Mental Hygiene (DHMH) and Johnston Laboratories (JL), Research and Development department. Specimens were processed, cultured, and a conventional direct-susceptibility test was performed at the DHMH laboratory. Portions of the smear-positive specimen concentrates were provided to JL, where radiometric-direct-susceptibility tests were performed. Later, a second study was initiated in which the conventional and radiometric work was performed at the DHMH laboratories. The third study was carried out at Jefferson Davis Hospital and Baylor College of Medicine, Houston, Texas. Clinical specimens were digested and decontaminated with N-acetyl-L-cysteineNaOH according to established procedures (Vestal, 1975). After centrifugation, the sediment was suspended in water, and smears were prepared and stained by the fluorochrome (FL) and Ziehl-Neelsen (ZN) procedures. The smears were examined for numbers of AFB per field after FL and ZN staining. If there were more than 10 AFB per field, it was classified as "numerous," while 1-10 AFB per field was "few." Less than one AFB per field was considered "rare." The results were arranged into three categories: 1) high, in which both FL and ZN smears had numerous AFB; 2) medium, with numerous/few, numerous/rare, or few/few by FL and ZN stains, respectively; and 3) low, in which smears had numerous/negative, few/rare, rare/rare, or rare/negative AFB by FL and ZN, respectively. Conventional agar dilution susceptibility testing was performed on all acid fast smear-positive specimens, by using the 1% proportion method described by the Centers for Disease Control (Vestal, 1975). Middlebrook 7H10/7H11 agar medium was used for susceptibility testing, and plates were examined after 2, 3, and 4 wk incubation. The modified radiometric procedure used is shown in Figure 1. All specimens positive for AFB were used for direct-susceptibility testing and were inoculated without any dilution. Middlebrook 7H12 broth (BACTEC 12A, Johnston Laboratories, Towson, MD) was used. Lyophilized drugs (Johnston Laboratories, Towson, MD) were rehydrated and added prior to the susceptibility testing as recommended. The concentration of streptomycin (SM) was the same for all three study sites, 4 ~g/ ml(radiometric) and 10 p.g/ml(conventional). For isoniazid (INH), 0.2 p.g/ml was used for both radiometric and conventional in study 1 and 2, while in study 3, 1 ~g/ml was used in both methods. Rifampin (RIF) concentration was 2 p.g/ml radiometric and 1 p~g/ml for conventional in study 1 and 2. One p.g/ml of RIF was employed in study 3 for both methods. Ethambutol (EMB) was 10 p.g/ml in both methodologies for study 1 and 2, while in study 3, only 3 p.g/ml was used in radiometric and 5p.g/ ml in the conventional methods. The specimen were diluted 1:10 with Diluting Fluid (Johnston Laboratories, Towson, MD), and 0.1 ml was inoculated into a drug-free control vial; the vials containing drug and 12A medium were inoculated with 0.1 ml of the specimen. An antimicrobial mixture (PACT, Johnston Laboratories, Towson, MD} was added to each of the 12A vials to suppress the growth of any contaminants. The addition of PACT yielded final concentrations of polymyxin B, 50 units; amphotericin B, 5 ~g; carbenicillin, 25 ~g; and trimethoprim, 2.5 p.g per ml of 12A medium. The inoculated 12A vials were incubated at 37°C -+ 1°C and tested every 2-3 days on the BACTEC 460. The susceptibility test vials were read for a maximum of 3 wk. When the control GI was 20 or more, the results were considered reportable and were interpreted in the same manner as in the indirect method (Roberts et al., 1983; Snider et al., 1981). When the change in GI from the previous day (AGI) was greater than that of the control, the organism was judged resistant to the drug. If the AGI was less, then the isolate was susceptible.
Direct Drug Susceptibility of M. tuberculosis
43
SPUTUM SPECIMEN
DIGEST & DECONTAMINATE
CENTRIFUGE
RESUSPEND SEDIMENT
AFB SMEAR FL/ZN
I
I
NEGATIVE (process for isolation
POSITIVE
I
I
RADIOMETRIC SUSCEPTIBILITY
CONVENTIONAL SUSCEPTIBILITY
I 1:10 DILUTION
I
O.1 ML
12A +PACT ~., .~ PRIMARY ISOLATION READ FOR MAXIMUM OF 6 WKS
0.1 ML
O.1 ML
O, ML
0.1 ML
O.1 ML
12A
12A + PACT +SM
12A +PACT +INH
12A +PACT +RIF
12A +PACT +EMB
+PACT (CONTROL )
\
¥ READ 3 TIMES WEEKLY MAXIMUM OF 21 DAYS CONTROL GI~20 IN ORDER TO INTERPRET RESULTS
G I > I O = POSITIVE
FIGURE 1. Combined isolation and susceptibility testing of smear-positive specimens.
)
44
J.P. L i b o n a t i et al.
T A B L E 1. R e p o r t a b l e R e s u l t s of D i r e c t - A n t i m i c r o b i a l - S u s c e p t i b i l i t y T e s t Radiometric
Study 1 Smear + Culture + Study 2 Smear + Culture + Study 3 Smear + Culture + Total Smear + Culture +
Conventional
No. tested
% reported
No. tested
% reported
203 115
48 84
142 81
39 68
252 165
60 92
252 165
36 54
125 97
75 97
125 97
54 69
580 377
59 91
519 343
41 62
W h e n t h e c o n t r o l v i a l d i d n o t y i e l d a GI of 20 or m o r e w i t h i n 21 days, r e s u l t s w e r e n o t r e p o r t e d , a n d t h e t e s t w a s d i s c a r d e d . C o r r e l a t i o n of s m e a r a n d c u l t u r e r e s u l t s with those obtained by the conventional method was analyzed.
RESULTS I n t h e first s t u d y , 203 s m e a r - p o s i t i v e s p e c i m e n s w e r e p r o c e s s e d for i s o l a t i o n a n d
T A B L E 2. C o m p a r i s o n of L e v e l of S m e a r P o s i t i v i t y a n d R e p o r t a b l e D i r e c t susceptibility Results with Radiometric Method (Culture-positive Specimens) Smear positivity
No. tested
No. results
Average time (days)
High Study 1 Study 2 Study 3
26 24 67
25 24 67
6.5 9.4 9.7
Total
117
116
8.9
Medium Study 1 Study 2 Study 3
37 45 8
33 40 8
10.3 10.9 12.5
Total
90
81
10.8
Low Study 1 Study 2 Study 3
52 96 22
39 88 19
13.8 14.1 13.4
Total
170
146
13.9
Total Study 1 Study 2 Study 3
115 165 97
97 152 94
10.7 12.5 10.6
Total
377
343
11.5
Direct Drug Susceptibility of M. t uberculosis
45
TABLE 3. Comparison of Direct-antimicrobial-susceptibility Test Results with Radiometric and Conventional Methods
Antimicrobial-susceptibilityResults
Antimicrobial
Susceptible by both methods
Susceptible conventional only
Susceptible radiometric only
Resistant by both methods
Streptomycin Study 1 Study 2 Study 3 Total
97 142 93 332
0 0 1 1
0 0 0 0
0 10 0 10
Isoniazid Study 1 Study 2 Study 3 Total
93 123 80 286
0 0 0 0
0 0 1 1
14 29 13 56
Study 1 Study 2 Study 3
93 131 90
0 0 0
0 0 0
4 21 4
Total
314
0
0
29
96 143 93
0 0 0
1 9a I
0 0 0
332
0
Rifampin
Ethambutol Study 1 Study 2 Study 3 Total
11
0
°Nine isolates fromthe same patient at different times during the study period. direct-drug-susceptibility testing. Some of the specimens could not be set up for direct susceptibility tests by the conventional method, and, thus, only 142 specimens were compared by this method with radiometric-susceptibility results. Since the specimen concentrates were transported from DHMH to JL after screening for AFB, the specimens were refrigerated overnight before radiometric-susceptibility testing was performed. In the second study, 252, and in the third study, 125 smear positive specimens were evaluated for direct-susceptibility tests by the conventional and the radiometric methods. Culture results of 580 specimens are summarized by study site in Table 1. There were 115 (79%) specimens positive for MTB in study 1,165 (87%) in study 2, and 97 (78%) in study 3. The percentage of reportable results was calculated for total smear-positive specimens and for those specimens that were culture positive for MTB only. Overall, 59% of the total specimens and 91% of the specimens that grew MTB yielded reportable drug susceptibility results by the radiometric method, while the percentage for conventional was 41% and 62%, respectively. Data on smear and culture positive specimens are further analyzed in Table 2. As the data indicate, highly smear-positive specimens yielded almost 100% reportable susceptibility results with an average of 8.9 days. As the smear positivity decreased, the average reporting time increased, and success in achieving reportable results declined. Within the data of the three sites combined, 90% of specimens with medium smear positivity yielded reportable results with an average of 10.8 days' reporting time. Specimens with low numbers of AFB on smear took an average of 14 days to
46
J.P. Libonati et al.
obtain reportable results with an 86% success in the susceptibility test. As expected, there was some, but not significant, variation in the average reporting time and percentage of reportable results among the three studies. Overall, out of 377 smearand culture-positive specimens, 91% yielded final reportable results within an average of 11.5 days by the radiometric method. In Table 3, radiometric direct-susceptibility-test results are compared with those obtained by the conventional agar dilution method. In some cases the direct test with the conventional method did not yield reportable results, in which case the radiometric results were compared with the results of the indirect-conventional-susceptibility tests. With streptomycin, one culture was resistant by radiometric and susceptible by the conventional method. In total, there was 99.7% agreement with the susceptible and 100% with the resistant organisms. Testing with isoniazid yielded one discrepancy: resistant by conventional and susceptible by the radiometric method. Rifampin susceptibility tests produced 100% agreement between the two methods. The highest discrepancy rate was produced by ethambutol. There were 11 organisms resistant by the conventional method, all of which were susceptible by the radiometric method. Nine of these isolates were from the same patient, isolated at different times. All of the isolates susceptible to 10 ~.g EMB/ml on agar media were susceptible by the radiometric method. DISCUSSION
A direct-susceptibility test offers several theoretical advantages compared to indirect procedures (Gangadharam, 1984). In addition to the time saved, which can be 3-4 wk, other important factors are: a) the bacterial population being tested is directly from the patient and may be more representative of the actual population producing the infection, and b) there is less possibility of change in bacterial characteristics. Isolation and subculturing may result in selection of isolates with altered susceptibility. This phenomenon has recently been reported for the M. avium complex (Gruft, 1986). Direct-susceptibility testing also has some disadvantages. When all smear-positive specimens are set up for the direct test, there are a number of specimens that do not yield reportable results because of no growth or poor growth in the control, contamination, or the presence of mycobacteria other than M. tuberculosis (MOTT) bacilli. This would result in increased overhead cost. Moreover, an indirect test would have to be performed in such cases. A new procedure is valuable when it offers convenience, time savings, and greater success in achieving the desired results. The cost effectiveness of such a new technique will further improve if an early differentiation of MTB could be obtained, especially if such a test would eliminate unnecessary testing of MOTT bacilli. The BACTEC radiometric technique for isolation of mycobacteria, differentiation of MTB from MOTT bacilli, and indirect-susceptibility testing is well documented (Hawkins, 1986; Laszlo and Michaud, 1986; Middlebrook et al., 1977; Siddiqi et al., 1981; Siddiqi et al., 1985; Snider et al., 1981). The average time required for isolation and susceptibility testing is reported to be 18 days as compared to 38 days by the conventional method (Roberts et al., 1983). The single most important improvement offered by the radiometric method is reduction in time required to obtain results. The average time for the direct-radiometric-susceptibility procedure was 11.5 days, a time saving of about a week compared to the indirect radiometric method and about 10 days earlier when compared to the conventional procedure. The radiometric-direct-susceptibility procedure tests can be performed on all acid-
Direct Drug Susceptibility of M. tuberculosis
47
fast, smear-positive specimens without regard for number of organisms seen, and it is not necessary to dilute the inoculum as is recommended with the conventional method. The radiometric direct test, as the data indicate, yields a successful outcome more frequently than the conventional direct procedure (59% versus 41%). A large number of the specimens that did not yield a reportable direct-susceptibility result by radiometric testing contained MOTT bacilli. If only specimens that grew MTB are considered, the radiometric method produced successful results 91% of the time compared to 62% for the conventional method. The drug concentrations used in the radiometric method were those that were found to be equivalent with the concentrations routinely used in the conventional method (Siddiqi et al., 1985). In study 3, higher concentrations of INH and lower concentrations of EMB were used in the conventional method, and thus, the concentrations in the radiometric testing were adjusted accordingly. The correlation of susceptibility results obtained by the two methods was very high with SM, INH, and RIF. There were ten strains resistant to SM, 56 to INH, and 29 to RIF, and, with one exception, all results were in agreement. With EMB, however, there was a problem. Nine of 11 cultures, all from the same patient, were resistant by the conventional method and susceptible by the radiometric method. These results suggest that the EMB concentration (10 p~g/ml) is too high in the radiometric procedure, an observation previously reported (Hawkins, 1986; Siddiqi et al., 1985; Woodley, 1986). Based on these observations, it was felt that lower concentrations of EMB were required in radiometric liquid medium as compared to agar medium, in which some of the drug potency could be lost due to binding with the components of the medium. Recently, EMB concentrations in the radiometric method have been modified. EMB concentrations of 7.5 and 2.5 p~g/ml are currently recommended for the system. These concentrations are equivalent to 10 and 5 ~g/ml in 7H10 medium. This change is expected to minimize the false susceptibility to EMB observed with the radiometric method in this study. At present, there is no test available that can differentiate MTB from MOTT bacilli on AFB smears and, thus, direct susceptibility tests of MOTT bacilli, which may not be necessary, cannot be avoided. However, with the introduction of the BACTEC NAP test, an early differentiation of MTB from MOTT bacilli is possible. If a combined isolation and drug susceptibility test is performed, then reporting of the susceptibility test results can be done along with TB differentiation. The procedure for the radiometric-direct-susceptibility test was slightly different from that recommended for the indirect method. Since the number of AFB in the concentrated specimen was usually not high enough, a 1:10 dilution instead of a 1:100 dilution of the inoculum was used for the control. The results throughout the study were very clear, without a single borderline value, and a 1% calculation was not required. The schedule for determining the GI for direct susceptibility tests also differed slightly. The vials were tested every 2-3 days with no weekend testing. Moreover, in the direct test, the vials were tested on the BACTEC instrument for a maximum of 21 days, while in the indirect procedure, the results are completed within 4-10 days. A GI of 20 for the control vial was taken as a threshold for reporting the results, instead of a GI of 30, as recommended for the indirect procedure. In summary, the radiometric-direct-susceptibility test procedure offers some advantages over the conventional methods. Although a little over 40% of smear-positive specimens did not yield results, this technique did produce greater success than the conventional-direct method. Improved turnaround time is the most important feature of the radiometric direct susceptibility testing. In addition, such a test may reduce the possibility of changes in bacterial characteristics during the culturing and subculturing required in an indirect procedure. Earlier results can be of benefit to phy-
48
J.P. Libonati et al.
sicians i n the m a n a g e m e n t of tuberculosis infections, especially with the advent of 6-month (short-course) chemotherapy.
We thank Nancy Hooper, James Baker, Martin Carter, and Charles Hwangho for their skillful technical assistance.
REFERENCES Gangadharam PRJ (1984) Drug Resistance in Mycobacteria. Boca Raton, Florida: CRC Press, Inc., pp 55-110. Hawkins JE (1986) Nonweekend schedule for BACTEC drug susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol 23:934. Gruft H (1986) Direct susceptibility of M. avium complex. Tubercle 67:156. Laszlo A, Michaud R (1986) Primary isolation, preliminary identification, and drug susceptibility testing of mycobacteria (M. tuberculosis complex) by a rapid radiometric method. Bull Int Union Tuberc 59:185. Middlebrook G, Reggiardo Z, Tigertt WD (1977) Automatable radiometric detection of growth of Mycobacterium tuberculosis in selective media. Am Rev Bespir Dis 115:1067. Roberts GD, Goodman NL, Heifets L, Larsh HW, Lindner TH, McClatchy JK, McGinnis MR, Siddiqi SH, Wright P (1983) Evaluation of the BACTEC radiometric method for recovery of mycobacteria and drug susceptibility testing of Mycobacterium tuberculosis from acid-fast smear-positive specimens. J Clin Microbiol 18:689. Siddiqi SH, Libonati JP, Middlebrook G (1981) Evaluation of a rapid radiometric method for drug susceptibility testing of Mycabacterium tuberculosis. J Clin Microbiol 13:908. Siddiqi SH, Hwangbo CC, Silcox V, Good RC, Snider Jr., DE, Middlebrook G (1984) Rapid radiometric methods to detect and differentiate M. tuberculasis/M, bovis from other mycobacterial species. Am Rev Respir Dis 130:634. Siddiqi SH, Hawkins JE, Laszlo A (1985). Interlaboratory drug susceptibility testing of Mycobacterium tuberculosis by a radiometric procedure and two conventional methods. 1 Clin Microbiol 22:919. Snider Jr., DE, Good RC, Kilburn JO, Laskowski Jr., LF, Lusk RH, Marr JJ, Reggiardo Z, and Middlebrook G (1981) Rapid susceptibility testing of Mycobacterium tuberculosis. Am Rev Respir Dis 117:631. Vestal, AL (1975) Procedures for the isolation and identification of mycobacteria. Center for Disease Control, Atlanta: U.S. Department of Health, Education, and Welfare publication no. (CDC) 76-82301, pp 97-110. Vincke G, Yegers O, Vanachter H, Jenkins PA, Butzler JP (1982) Rapid susceptibility testing of Mycobacterium tuberculosis by a radiometric technique. ] Antimicrob Chemother 10:351. Woodley CL (1986) Evaluation of streptomycin and ethambutol concentrations for susceptibility testing of Mycobacterium tuberculosis by radiometric and conventional procedures. J Clin Microbiol 23:385.
DIAGN MICROBIOL INFECT DIS 1988;10:41-48
Direct Antimicrobial Drug Susceptibility Testing of Mycobacterium tuberculosis by the Radiometric Method Joseph P. Libonati, Charles E. Stager, James R. Davis, and Salman H. Siddiqi
Direct-drug-susceptibility tests were performed on clinical specimens positive for acid-fast bacilli by either Ziehl-Neelsen or fluorochrame staining. The results of conventional agar dilution (Vestal, 1975) and a modified radiometric (BACTEC) method were compared. A total of 580 smear-positive specimens were tested by the BACTEC method at three separate sites. Three hundred and seventy-seven of these were culture positive for M. tuberculosis, and 343 (91%) yielded acceptable direct-susceptibility-test results. We used the conventional method to determine that 343 of 519 smear-positive specimens were culture positive for M. tuberculosis, and 212 (62%) produced acceptable results within 3 wks. Conventional results were reported in 3-4 wks, while the time required to obtain results with the BACTEC method ranged from 5 to 21 days (average 11.5 days). Results indicate that the radiometric method provides reportable results more frequently with time savings as compared to the conventional method.
INTRODUCTION
Drug susceptibility testing of Mycobacterium tuberculosis (MTB) is generally performed after the organism has been isolated. Because this indirect-susceptibilitytest approach requires several additional weeks for isolation, a direct-susceptibilitytest procedure is desirable {Vestal, 1975}. Currently, direct agar dilution susceptibility testing in which the inoculum is a digested and decontaminated clinical specimen k n o w n to be positive for acid-fastbacteria {AFB) can provide results 3-4 w k earlier than the indirect test (Gangadharam, 1984; Vestal, 1975). The introduction of rapid radiometric techniques has reduced the time required for drug susceptibility testing of MTB from 3 wk to less than 6 days (Roberts et al., 1983; Siddiqi et al., 1981, 1984, 1985; Vincke et al., 1982}. However, all current literature citations of radiometric-drug-susceptibility testing described the indirect method and required an average of 10-16 days for isolation of MTB prior to susceptibility testing. This report presents the results of direct-susceptibility testing using the radiometric technique and evaluates the efficacy and turnaround time w h e n compared to the conventional agar dilution procedure.
From the Maryland Department of Health and Mental Hygiene, Baltimore,Maryland {J.P.L.}, Harris County Hospital District,Baylor Collegeof Medicine Houston, Texas {C.E.S.,J.R.D.},and Johnston Laboratories,Research and Development, Towson, Maryland {S.H.S.). Address reprintrequests to: Dr. Joseph P. Libonati,Division of Microbiology,Department of Health and Mental Hygiene, P.O. Box 2355, Baltimore,M D 21203. Received October 8, 1987; revised and accepted April 22, 1988. © 1988 Elsevier Science Publishing Co., Inc. 52 Vanderbilt Avenue, New York, NY 10017
0732-8803/88/$03.50
42
J.P. Libonati et al.
MATERIALS A N D METHODS
This work is comprised of three separate studies. The first study was a joint effort of the Maryland Department of Health and Mental Hygiene (DHMH) and Johnston Laboratories (JL), Research and Development department. Specimens were processed, cultured, and a conventional direct-susceptibility test was performed at the DHMH laboratory. Portions of the smear-positive specimen concentrates were provided to JL, where radiometric-direct-susceptibility tests were performed. Later, a second study was initiated in which the conventional and radiometric work was performed at the DHMH laboratories. The third study was carried out at Jefferson Davis Hospital and Baylor College of Medicine, Houston, Texas. Clinical specimens were digested and decontaminated with N-acetyl-L-cysteineNaOH according to established procedures (Vestal, 1975). After centrifugation, the sediment was suspended in water, and smears were prepared and stained by the fluorochrome (FL) and Ziehl-Neelsen (ZN) procedures. The smears were examined for numbers of AFB per field after FL and ZN staining. If there were more than 10 AFB per field, it was classified as "numerous," while 1-10 AFB per field was "few." Less than one AFB per field was considered "rare." The results were arranged into three categories: 1) high, in which both FL and ZN smears had numerous AFB; 2) medium, with numerous/few, numerous/rare, or few/few by FL and ZN stains, respectively; and 3) low, in which smears had numerous/negative, few/rare, rare/rare, or rare/negative AFB by FL and ZN, respectively. Conventional agar dilution susceptibility testing was performed on all acid fast smear-positive specimens, by using the 1% proportion method described by the Centers for Disease Control (Vestal, 1975). Middlebrook 7H10/7H11 agar medium was used for susceptibility testing, and plates were examined after 2, 3, and 4 wk incubation. The modified radiometric procedure used is shown in Figure 1. All specimens positive for AFB were used for direct-susceptibility testing and were inoculated without any dilution. Middlebrook 7H12 broth (BACTEC 12A, Johnston Laboratories, Towson, MD) was used. Lyophilized drugs (Johnston Laboratories, Towson, MD) were rehydrated and added prior to the susceptibility testing as recommended. The concentration of streptomycin (SM) was the same for all three study sites, 4 ~g/ ml(radiometric) and 10 p.g/ml(conventional). For isoniazid (INH), 0.2 p.g/ml was used for both radiometric and conventional in study 1 and 2, while in study 3, 1 ~g/ml was used in both methods. Rifampin (RIF) concentration was 2 p.g/ml radiometric and 1 p~g/ml for conventional in study 1 and 2. One p.g/ml of RIF was employed in study 3 for both methods. Ethambutol (EMB) was 10 p.g/ml in both methodologies for study 1 and 2, while in study 3, only 3 p.g/ml was used in radiometric and 5p.g/ ml in the conventional methods. The specimen were diluted 1:10 with Diluting Fluid (Johnston Laboratories, Towson, MD), and 0.1 ml was inoculated into a drug-free control vial; the vials containing drug and 12A medium were inoculated with 0.1 ml of the specimen. An antimicrobial mixture (PACT, Johnston Laboratories, Towson, MD} was added to each of the 12A vials to suppress the growth of any contaminants. The addition of PACT yielded final concentrations of polymyxin B, 50 units; amphotericin B, 5 ~g; carbenicillin, 25 ~g; and trimethoprim, 2.5 p.g per ml of 12A medium. The inoculated 12A vials were incubated at 37°C -+ 1°C and tested every 2-3 days on the BACTEC 460. The susceptibility test vials were read for a maximum of 3 wk. When the control GI was 20 or more, the results were considered reportable and were interpreted in the same manner as in the indirect method (Roberts et al., 1983; Snider et al., 1981). When the change in GI from the previous day (AGI) was greater than that of the control, the organism was judged resistant to the drug. If the AGI was less, then the isolate was susceptible.
Direct Drug Susceptibility of M. tuberculosis
43
SPUTUM SPECIMEN
DIGEST & DECONTAMINATE
CENTRIFUGE
RESUSPEND SEDIMENT
AFB SMEAR FL/ZN
I
I
NEGATIVE (process for isolation
POSITIVE
I
I
RADIOMETRIC SUSCEPTIBILITY
CONVENTIONAL SUSCEPTIBILITY
I 1:10 DILUTION
I
O.1 ML
12A +PACT ~., .~ PRIMARY ISOLATION READ FOR MAXIMUM OF 6 WKS
0.1 ML
O.1 ML
O, ML
0.1 ML
O.1 ML
12A
12A + PACT +SM
12A +PACT +INH
12A +PACT +RIF
12A +PACT +EMB
+PACT (CONTROL )
\
¥ READ 3 TIMES WEEKLY MAXIMUM OF 21 DAYS CONTROL GI~20 IN ORDER TO INTERPRET RESULTS
G I > I O = POSITIVE
FIGURE 1. Combined isolation and susceptibility testing of smear-positive specimens.
)
44
J.P. L i b o n a t i et al.
T A B L E 1. R e p o r t a b l e R e s u l t s of D i r e c t - A n t i m i c r o b i a l - S u s c e p t i b i l i t y T e s t Radiometric
Study 1 Smear + Culture + Study 2 Smear + Culture + Study 3 Smear + Culture + Total Smear + Culture +
Conventional
No. tested
% reported
No. tested
% reported
203 115
48 84
142 81
39 68
252 165
60 92
252 165
36 54
125 97
75 97
125 97
54 69
580 377
59 91
519 343
41 62
W h e n t h e c o n t r o l v i a l d i d n o t y i e l d a GI of 20 or m o r e w i t h i n 21 days, r e s u l t s w e r e n o t r e p o r t e d , a n d t h e t e s t w a s d i s c a r d e d . C o r r e l a t i o n of s m e a r a n d c u l t u r e r e s u l t s with those obtained by the conventional method was analyzed.
RESULTS I n t h e first s t u d y , 203 s m e a r - p o s i t i v e s p e c i m e n s w e r e p r o c e s s e d for i s o l a t i o n a n d
T A B L E 2. C o m p a r i s o n of L e v e l of S m e a r P o s i t i v i t y a n d R e p o r t a b l e D i r e c t susceptibility Results with Radiometric Method (Culture-positive Specimens) Smear positivity
No. tested
No. results
Average time (days)
High Study 1 Study 2 Study 3
26 24 67
25 24 67
6.5 9.4 9.7
Total
117
116
8.9
Medium Study 1 Study 2 Study 3
37 45 8
33 40 8
10.3 10.9 12.5
Total
90
81
10.8
Low Study 1 Study 2 Study 3
52 96 22
39 88 19
13.8 14.1 13.4
Total
170
146
13.9
Total Study 1 Study 2 Study 3
115 165 97
97 152 94
10.7 12.5 10.6
Total
377
343
11.5
Direct Drug Susceptibility of M. t uberculosis
45
TABLE 3. Comparison of Direct-antimicrobial-susceptibility Test Results with Radiometric and Conventional Methods
Antimicrobial-susceptibilityResults
Antimicrobial
Susceptible by both methods
Susceptible conventional only
Susceptible radiometric only
Resistant by both methods
Streptomycin Study 1 Study 2 Study 3 Total
97 142 93 332
0 0 1 1
0 0 0 0
0 10 0 10
Isoniazid Study 1 Study 2 Study 3 Total
93 123 80 286
0 0 0 0
0 0 1 1
14 29 13 56
Study 1 Study 2 Study 3
93 131 90
0 0 0
0 0 0
4 21 4
Total
314
0
0
29
96 143 93
0 0 0
1 9a I
0 0 0
332
0
Rifampin
Ethambutol Study 1 Study 2 Study 3 Total
11
0
°Nine isolates fromthe same patient at different times during the study period. direct-drug-susceptibility testing. Some of the specimens could not be set up for direct susceptibility tests by the conventional method, and, thus, only 142 specimens were compared by this method with radiometric-susceptibility results. Since the specimen concentrates were transported from DHMH to JL after screening for AFB, the specimens were refrigerated overnight before radiometric-susceptibility testing was performed. In the second study, 252, and in the third study, 125 smear positive specimens were evaluated for direct-susceptibility tests by the conventional and the radiometric methods. Culture results of 580 specimens are summarized by study site in Table 1. There were 115 (79%) specimens positive for MTB in study 1,165 (87%) in study 2, and 97 (78%) in study 3. The percentage of reportable results was calculated for total smear-positive specimens and for those specimens that were culture positive for MTB only. Overall, 59% of the total specimens and 91% of the specimens that grew MTB yielded reportable drug susceptibility results by the radiometric method, while the percentage for conventional was 41% and 62%, respectively. Data on smear and culture positive specimens are further analyzed in Table 2. As the data indicate, highly smear-positive specimens yielded almost 100% reportable susceptibility results with an average of 8.9 days. As the smear positivity decreased, the average reporting time increased, and success in achieving reportable results declined. Within the data of the three sites combined, 90% of specimens with medium smear positivity yielded reportable results with an average of 10.8 days' reporting time. Specimens with low numbers of AFB on smear took an average of 14 days to
46
J.P. Libonati et al.
obtain reportable results with an 86% success in the susceptibility test. As expected, there was some, but not significant, variation in the average reporting time and percentage of reportable results among the three studies. Overall, out of 377 smearand culture-positive specimens, 91% yielded final reportable results within an average of 11.5 days by the radiometric method. In Table 3, radiometric direct-susceptibility-test results are compared with those obtained by the conventional agar dilution method. In some cases the direct test with the conventional method did not yield reportable results, in which case the radiometric results were compared with the results of the indirect-conventional-susceptibility tests. With streptomycin, one culture was resistant by radiometric and susceptible by the conventional method. In total, there was 99.7% agreement with the susceptible and 100% with the resistant organisms. Testing with isoniazid yielded one discrepancy: resistant by conventional and susceptible by the radiometric method. Rifampin susceptibility tests produced 100% agreement between the two methods. The highest discrepancy rate was produced by ethambutol. There were 11 organisms resistant by the conventional method, all of which were susceptible by the radiometric method. Nine of these isolates were from the same patient, isolated at different times. All of the isolates susceptible to 10 ~.g EMB/ml on agar media were susceptible by the radiometric method. DISCUSSION
A direct-susceptibility test offers several theoretical advantages compared to indirect procedures (Gangadharam, 1984). In addition to the time saved, which can be 3-4 wk, other important factors are: a) the bacterial population being tested is directly from the patient and may be more representative of the actual population producing the infection, and b) there is less possibility of change in bacterial characteristics. Isolation and subculturing may result in selection of isolates with altered susceptibility. This phenomenon has recently been reported for the M. avium complex (Gruft, 1986). Direct-susceptibility testing also has some disadvantages. When all smear-positive specimens are set up for the direct test, there are a number of specimens that do not yield reportable results because of no growth or poor growth in the control, contamination, or the presence of mycobacteria other than M. tuberculosis (MOTT) bacilli. This would result in increased overhead cost. Moreover, an indirect test would have to be performed in such cases. A new procedure is valuable when it offers convenience, time savings, and greater success in achieving the desired results. The cost effectiveness of such a new technique will further improve if an early differentiation of MTB could be obtained, especially if such a test would eliminate unnecessary testing of MOTT bacilli. The BACTEC radiometric technique for isolation of mycobacteria, differentiation of MTB from MOTT bacilli, and indirect-susceptibility testing is well documented (Hawkins, 1986; Laszlo and Michaud, 1986; Middlebrook et al., 1977; Siddiqi et al., 1981; Siddiqi et al., 1985; Snider et al., 1981). The average time required for isolation and susceptibility testing is reported to be 18 days as compared to 38 days by the conventional method (Roberts et al., 1983). The single most important improvement offered by the radiometric method is reduction in time required to obtain results. The average time for the direct-radiometric-susceptibility procedure was 11.5 days, a time saving of about a week compared to the indirect radiometric method and about 10 days earlier when compared to the conventional procedure. The radiometric-direct-susceptibility procedure tests can be performed on all acid-
Direct Drug Susceptibility of M. tuberculosis
47
fast, smear-positive specimens without regard for number of organisms seen, and it is not necessary to dilute the inoculum as is recommended with the conventional method. The radiometric direct test, as the data indicate, yields a successful outcome more frequently than the conventional direct procedure (59% versus 41%). A large number of the specimens that did not yield a reportable direct-susceptibility result by radiometric testing contained MOTT bacilli. If only specimens that grew MTB are considered, the radiometric method produced successful results 91% of the time compared to 62% for the conventional method. The drug concentrations used in the radiometric method were those that were found to be equivalent with the concentrations routinely used in the conventional method (Siddiqi et al., 1985). In study 3, higher concentrations of INH and lower concentrations of EMB were used in the conventional method, and thus, the concentrations in the radiometric testing were adjusted accordingly. The correlation of susceptibility results obtained by the two methods was very high with SM, INH, and RIF. There were ten strains resistant to SM, 56 to INH, and 29 to RIF, and, with one exception, all results were in agreement. With EMB, however, there was a problem. Nine of 11 cultures, all from the same patient, were resistant by the conventional method and susceptible by the radiometric method. These results suggest that the EMB concentration (10 p~g/ml) is too high in the radiometric procedure, an observation previously reported (Hawkins, 1986; Siddiqi et al., 1985; Woodley, 1986). Based on these observations, it was felt that lower concentrations of EMB were required in radiometric liquid medium as compared to agar medium, in which some of the drug potency could be lost due to binding with the components of the medium. Recently, EMB concentrations in the radiometric method have been modified. EMB concentrations of 7.5 and 2.5 p~g/ml are currently recommended for the system. These concentrations are equivalent to 10 and 5 ~g/ml in 7H10 medium. This change is expected to minimize the false susceptibility to EMB observed with the radiometric method in this study. At present, there is no test available that can differentiate MTB from MOTT bacilli on AFB smears and, thus, direct susceptibility tests of MOTT bacilli, which may not be necessary, cannot be avoided. However, with the introduction of the BACTEC NAP test, an early differentiation of MTB from MOTT bacilli is possible. If a combined isolation and drug susceptibility test is performed, then reporting of the susceptibility test results can be done along with TB differentiation. The procedure for the radiometric-direct-susceptibility test was slightly different from that recommended for the indirect method. Since the number of AFB in the concentrated specimen was usually not high enough, a 1:10 dilution instead of a 1:100 dilution of the inoculum was used for the control. The results throughout the study were very clear, without a single borderline value, and a 1% calculation was not required. The schedule for determining the GI for direct susceptibility tests also differed slightly. The vials were tested every 2-3 days with no weekend testing. Moreover, in the direct test, the vials were tested on the BACTEC instrument for a maximum of 21 days, while in the indirect procedure, the results are completed within 4-10 days. A GI of 20 for the control vial was taken as a threshold for reporting the results, instead of a GI of 30, as recommended for the indirect procedure. In summary, the radiometric-direct-susceptibility test procedure offers some advantages over the conventional methods. Although a little over 40% of smear-positive specimens did not yield results, this technique did produce greater success than the conventional-direct method. Improved turnaround time is the most important feature of the radiometric direct susceptibility testing. In addition, such a test may reduce the possibility of changes in bacterial characteristics during the culturing and subculturing required in an indirect procedure. Earlier results can be of benefit to phy-
48
J.P. Libonati et al.
sicians i n the m a n a g e m e n t of tuberculosis infections, especially with the advent of 6-month (short-course) chemotherapy.
We thank Nancy Hooper, James Baker, Martin Carter, and Charles Hwangho for their skillful technical assistance.
REFERENCES Gangadharam PRJ (1984) Drug Resistance in Mycobacteria. Boca Raton, Florida: CRC Press, Inc., pp 55-110. Hawkins JE (1986) Nonweekend schedule for BACTEC drug susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol 23:934. Gruft H (1986) Direct susceptibility of M. avium complex. Tubercle 67:156. Laszlo A, Michaud R (1986) Primary isolation, preliminary identification, and drug susceptibility testing of mycobacteria (M. tuberculosis complex) by a rapid radiometric method. Bull Int Union Tuberc 59:185. Middlebrook G, Reggiardo Z, Tigertt WD (1977) Automatable radiometric detection of growth of Mycobacterium tuberculosis in selective media. Am Rev Bespir Dis 115:1067. Roberts GD, Goodman NL, Heifets L, Larsh HW, Lindner TH, McClatchy JK, McGinnis MR, Siddiqi SH, Wright P (1983) Evaluation of the BACTEC radiometric method for recovery of mycobacteria and drug susceptibility testing of Mycobacterium tuberculosis from acid-fast smear-positive specimens. J Clin Microbiol 18:689. Siddiqi SH, Libonati JP, Middlebrook G (1981) Evaluation of a rapid radiometric method for drug susceptibility testing of Mycabacterium tuberculosis. J Clin Microbiol 13:908. Siddiqi SH, Hwangbo CC, Silcox V, Good RC, Snider Jr., DE, Middlebrook G (1984) Rapid radiometric methods to detect and differentiate M. tuberculasis/M, bovis from other mycobacterial species. Am Rev Respir Dis 130:634. Siddiqi SH, Hawkins JE, Laszlo A (1985). Interlaboratory drug susceptibility testing of Mycobacterium tuberculosis by a radiometric procedure and two conventional methods. 1 Clin Microbiol 22:919. Snider Jr., DE, Good RC, Kilburn JO, Laskowski Jr., LF, Lusk RH, Marr JJ, Reggiardo Z, and Middlebrook G (1981) Rapid susceptibility testing of Mycobacterium tuberculosis. Am Rev Respir Dis 117:631. Vestal, AL (1975) Procedures for the isolation and identification of mycobacteria. Center for Disease Control, Atlanta: U.S. Department of Health, Education, and Welfare publication no. (CDC) 76-82301, pp 97-110. Vincke G, Yegers O, Vanachter H, Jenkins PA, Butzler JP (1982) Rapid susceptibility testing of Mycobacterium tuberculosis by a radiometric technique. ] Antimicrob Chemother 10:351. Woodley CL (1986) Evaluation of streptomycin and ethambutol concentrations for susceptibility testing of Mycobacterium tuberculosis by radiometric and conventional procedures. J Clin Microbiol 23:385.